Winning Chess Middlegames - Sokolov.pdf 4r2jr

Ivan Sokolov

Winning Chess Middlegames An Essential Guide to Pawn Structures

© 2008 New I n Chess Published by New In Chess, Alkmaar, The Netherlands www.newinchess.com All rights reser ved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission from the publisher. Cover design: Steven Boland Front cover photograph: Evgeny Atarov Supervisor: Peter Boel Proofreading: Rene Olthof Production: Anton Schermer

New In Chess 2 0 0 8

ISBN-13: 978-90-5691-264-2

Foreword

Contents by Michael Adams Foreword by Michael Adams .......................... 7 Introduction ..............................' .......... 9 Chapter 1 .......................................... 11 Doubled Pawns: 12 essential structures

Chapter 2 .......................................... 85 Isolated Pawns: 10 essential structures

Chapter 3 .........................................179 Hanging Pawns: 4 essential structures

Chapter 4......................................... 221 Pawn Majority in the Centre: 7 essential structures

Index of Players ................................ .. 283 .

List of Games...................................... 285

5

I think the first tournament that I played with Ivan Sokolov was the Oakham Young Masters (yes, we were young once) back in 1 9 8 8 . We then started to bump into each other at tournaments on a regular basis, became friendly and often socialised to­ gether following our games, sharing some glasses of wine. Our discussions were many and varied and some of them were even about chess. I one particular conversation where I proposed a certain plan in a mid­ dlegame position; Ivan looked a little confused, his eyebrows started twitching, and he responded 'yes, but this is just a normal position' . In his chess education, he had broken down structures into various typical situations and analysed these. The Eng­ lish school of chess had a slightly more chaotic approach. In this book, Ivan shares the fruits of his labours, which are invaluable tools for any player. This structured approach is quite instructive as often players choose their opening repertoire according to quite haphazard criteria, randomly picking up lines without giving serious consideration to how the resulting middlegames would suit them. To resolve this problem a bit of retro-analysis can prove helpful. By studying the resulting middlegames and determining your strengths and weaknesses, it is possible to go back to the earlier stages of the game and set out your stall accordingly. The book also gives interesting pointers about the direction in which modern chess is heading. The role of the computer in opening preparation has become all-important and players' homework can progress right from the starting position until the end of the game. Reading the book, I was very happy that I was not caught in Ivan's preparation for his game against Krasenkow featured on page 2 6 5 ! I expect his opponent also thinks himself lucky that he diverged at an early stage. Although the general advice will be especially helpful to club players, other topics covered will certainly be useful to players of a greater strength. Undoubtedly they will make note of the original opening suggestions , as I certainly have, but also enjoy the in-depth annotations to some excellent games. Club players shouldn't be dis7

Winning Chess Middlegames

heartened if the number of variations sometimes seems a bit daunting. As you can see from Ivan's own games featured here and elsewhere, aggressive, challenging play al­ ways makes it difficult under the rigours of a practical game to find the best defence. The level of defensive play that is possible at home in your study with some helpful suggestions from your silicon friend is rarely replicated in the heat of battle. There are also a few ofIvan's losses in this book, I know from commiserating with him afterwards that some of them were quite painful, and he can't have enjoyed re­ visiting them for publication in this book. But an essential part of improvement, from which Ivan does not shirk here, is analysing what has gone wrong and being objective about not only the assessment of the position reached from an opening but your understanding of the resulting structures. After all , most main opening systems are completely viable but your mastery of them may not be. Ivan's introduction struck a chord with me in this respect, as I suspect it does with most other players , as I personally sometimes persisted with unsuited opening sys­ tems too long. I would j ustify my choices to myself, on the grounds that objectively I had a good position, but the situations arising did not suit my style and I was then never able to get a good grasp of them. Everyone must decide for themselves which systems best serve them but this book will certainly be a great asset in making these sometimes hard choices. Of course armed with the knowledge gained here you will be considerably better versed in all of them. I hope you enjoy reading it as much as I did. Michael Adam s, October

8

2008

Introduction I started working on this book in March, 2 0 0 7 . As usually happens, due to different reasons, it took me longer than my publisher and me had planned and I finished my work in August 2 0 0 8 . In this book I wanted to explore the - in my opinion - four most important types of pawn structure in chess. Quite a number of books on pawn structures have been published, and one may rightly wonder what makes this book different. Well, I have tried, as much as possible, to 1 . systematize the thematic plans used and give clear explanations of them, and 2 . incorporate the ideas of the featured opening variation into the pawn struc­ ture that ensues. The latter is actually quite important. In the pre-computer era players normally pol­ ished their opening repertoire over the years, and even though opening preparation did not go nearly as far as today, years of theoretical and practical experience brush­ ing up one's repertoire would normally result in a reasonably good strategic under­ standing of the positions arising from the openings played. In the past I S years, the involvement of computer programs and databases has made it considerably easier to prepare a particular variation for a particular oppo­ nent. However, thorough study and good strategic understanding of the positions still remains a must in order to capitalize successfully on your opening preparation. I still watching one of Anatoly Karpov's post-mortems, when he had won from some initially inferior Ruy Lopez with black. His opponent, slightly annoyed, remarked: 'Here, after the opening, you were definitely worse' , to which the 1 2th World Champion calmly replied: 'Yes , but soon I was better' . Indeed, Karpov has won from quite a number of inferior positions (his encoun­ ters with Garry Kasparov included) , due to his superior strategic understanding of the openings he was playing. Kasparov has won many Najdorfs and King's Indians not only because he had the best novelties, but because he fundamentally understood those positions better than his opponents. On the other hand he was too stubborn to it that the Berlin Variation of the Ruy Lopez was not 'his cup of tea' , which ulti­ mately cost him his World Championship title against Vladimir Kramnik in 2 0 0 0 . Kramnik, on the other hand, being devastating i n Catalan-type systems with white and Meran Slavs with black, at some stage started to opt for sharp Sicilians with white and King's Indians with black. That adventure did not last very long. Nowadays he is a merciless killing machine with his Catalans again, squeezing out the smallest of microscopic advantages, while the King's Indian with black is a long-forgotten voyage. 9

Winning Chess Middlegames

If such mistakes are committed by the world's very best, then what are we to expect from lesser gods? Throughout my own career, I have also scored reasonably well in the positions I understood and paid the price for being too stubborn to stay away from position types that did not suit me. So the reasons why I have tried in this book to incorporate the strategic middlegame ideas and the games which I view as important into the four different types of pawn structure discussed in this book, were : 1 . to provide a complete guide for the club player; 2 . through a process of serious analysis of the material in this book, to also give the club player a reasonably accurate feeling as to which particular positions suit him and which do not; and 3 . to give the club player who takes his time for a thorough study of this book, new strategic and also practical opening knowledge, after which he will defi­ nitely see a clear improvement in his results. In the introductions to the four different chapters, I will further explain the distin­ guishing types of position, games and variations featured. I hope that, apart from trying to improve his chess skills, the reader will also simply enjoy studying the games selected in this book. Ivan Sokolov, August

2008

Chapter 1 Doubled Pawns Introduction In this chapter, I will try to make structures with doubled pawns easier �o understand , . and analyse, by systematizing them into 1 2 standard positions. The 1 2 Structure dI­ agrams I have connected with the commented games that follow, reflect those stan­ dard positions, which are reached the most frequently. As we shall see, a vast majority of structures with doubled pawns arise from the various lines with 4.e3 or with 4.a3 (the Samisch Variation) of the Nimzo-Iridian Defence. Apart from the fact that it is good to understand these positions in general, for those who play l .d2 -d4 with white, or the Nimzo-Indian with black, the com­ mented games below will have the additional practical value that they will clearly improve your opening knowledge, an improvement that can be immediately imple­ mented in tournament play. 1.1

(Game 1 - Sokolov -Johansen) shows a rel­ atively rarely reached type of position. I have included this game primarily in order to improve the reader's gen­ eral understanding. In practice it does not often happen in the Nimzo-Indian that White gets doubled pawns while Black keeps his bishop pair. Structure 1 . 1

(Game 2 - Sokolov-Winants) and Struc­ (Game 3 - Gligoric-Nikolic) show what has been for many years the main line of the Nimzo-Indian. Anyone who attempts to fundamentally improve his chess skills needs to analyse these positions thoroughly. In the games related to Structures 1 . 2 and 1 . 3 I have tried to explain the pros and cons of these positions , which are difficult to play for both sides. Structure 1 .2

ture 1 .3

10

11

Chapter

Winning Chess Middlegames

(Game 4 - Bronstein-Najdorf and Game 5 - Sky-Hiibner) and Structure I.S (Game 6 - Kotov­ Keres) deal with the positions arising from the Samisch Variation. They are essential for an understanding of this variation, which was very popular S O years ago, and played by the world's greatest players of that time. Structure 1 .4

(Game 7 Keres-Sky) explains why for White it does not have to be bad to lose his c3 pawn in positions with a full centre and doubled pawns on c3 and c4 . This idea is in use in a few Nimzo-Indian lines, and useful to and understand. Structure 1 .6

1.6

(Game 1 4 - Bronstein-Simagin) shows an original strategic idea by Bronstein, played almost S O years ago and still very viable. Structure 1 .8

DOll bled Pawns

1.8

Structure 1 . 9 (Game 1 5 - Botvinnik-Chekhover and Game 1 6 - Kuzubov-Van der Wiel) shows an important strategic idea for White. He does not mind making the centre static, seemingly isolating his c4 pawn weakness even more, by exchanging his d4 pawn, in order to open the d-file and gain an important outpost on the central dS -square. An idea which was beautifully executed by former World Champion Mikhail Botvinnik 7 0 years ago (!) and still highly topical.

-

Structure 1 . 7 (Game 8 - Gligoric-Ivkov, Game 9 Vyzhmanavin-Beliavsky, and , with a mobile centre : Game 1 0 - Z.Polgar-Sokolov, Game 1 1 - Sokolov-Bologan, Game 1 2 - Sokolov -Dizdarevic and Game 1 3 Radjabov-Anand) shows a plan that is often seen in a dif­ ferent type of position with doubled pawns in a full-cen­ tre Nimzo-Indian : Black targets (and often wins) White's weak c4 pawn, but by doing this, he gets his knight tem­ porarily or permanently stranded on the edge of the board, i.e. on the a s -square. As you will see from the games analysed, most of the time White gets plenty of initiative on the kingside to compensate for the loss of his c4 pawn, but the game remains double-edged and White has to be energetic and accurate in developing and exe­ cuting his kingside attack.

12

1.4

1:

Structure 1 . 1 0 (Game 1 7 - Topalov-Aronian) deals with an idea similar to the one demonstrated in Structure 1 . 9 , with the difference that Black's e-pawn and White's f-pawn have disappeared here and White has to rely more on his bishop pair.

(Game 1 8 - Kaidanov-Onischuk) shows a beautiful positional exchange sacrifice idea in the Samisch Variation of the Nimzo-Indian. Structure 1 . 1 1

Structure 1 . 1 2 (Game 1 9 - Kasparov-Ivanchuk) deals with pOSitions arising from the English Opening where White has doubled c-pawns. Mostly White also takes cen­ tral control, but his d-pawn has not yet been pushed to d4 (which brings some clear advantages here) . Further­ more, White's f-pawn has been exchanged for and Black's e-pawn, so that the f-file is open for White's rook.

13

Winning Chess Middlegames

Chapter

Structure 1 . 1

Black has two bishops instead of �

+

tb

In the vast majority ofpositions with a full centre and doubled pawns, the side with the doubled pawns (usually White) has a bishop pair versus bishop + knight. How­ ever, if instead of bishop + knight Black also had two bishops, would it benefit him or would it (in a closed position) make it more difficult for him to manoeuvre his pieces? An interesting question. In the following game (which went rather wrong for me) I got the opportunity to find out some pros and cons. NI 1 3 . 6 (E44)

GAME 1

Ivan Sokolov Darryl Johansen Manila 01 1992 (2) 1.

d2-d4

2.

c2-c4

e7-e6

3.

ttJb1 -c3

�fS-b4

ttJgS-f6

4.

e2-e3

b7-b6

5.

ttJg 1 -e2

ttJf6-e41?

6.

f2-f3 !?

ttJe4xc3

7.

b2xc3

�b4-e7

S.

e3-e4

Deciding to advance the central pawn and then develop the knight further. In Shirov-Adams, Las Palmas 1 9 94, White opted for a different set-up : S .ttJg3 ttJc6 9.�d3 �a6 1 0 . 0-0 ttJa5 1 1 .'iVe2 c6 (in the event of 1 1 . . . d6 I guess Shirov's idea was to push 1 2 .f4) 1 2 .f4 d5 1 3 .cxd5 �xd3 1 4 . 'ifxd3 exd5 1 5 .f5 h5 ! with a complex game. Note that with his last move (typical for these positions) Black disturbed White's attacking plans. S. 9.

ttJbS-c6 ttJe2-g3

�cS-a6

1 o.

�f1 -d3

ttJc6-a5

11.

'iVd 1 -e2

d 7-d6

1 2.

0-0

'iYdS-d 7

Black follows the regular Samisch Nimzo plan, as if there was a knight on f6 and the e7 bishop did not exist. With a black bishop on e 7 instead of a knight 14

on f6 , White misses the important tempo move e4-e5 with which he often develops an initiative. Black could also have decided to harass the g 3 knight immediately by playing 1 2 . . . h5 ? - however, with his king in the middle, this would give White the op­ portunity to take immediate action and sacrifice a piece with 1 3 .f4! h4 1 4.ttJh5 ! g6 1 5 .ttJg7 + �fS 1 6 .ttJxe6+ fxe6 1 7 .f5 (White has opened up the black king's position and charges with a direct mat­ ing attack, while Black's bishop and knight are stranded on the a-file) 1 7 . . . 'litg 7 (or 1 7 . . . exf5 I S .exf5 �f6 1 9 .�g5 ! �g7 (on 1 9 . . . �xg5 , 20 .fxg6+ �f6 2 1 .'iYe6 wins) 20.�xf6+ 'llVxf6 2 1 .fxg6 and White wins) I S .fxg6 lifS ( I S . . . £Lf6 1 9 .e5) 1 9 .�d2 and White wins in the attack ( 1 9 . . . �f6 20 .e5) .

1 3.

1:

Doubled Pawns

game. White has to sense this and seize the initiative, and should not be afraid to sacrifice material and take risks. If White misses such opportunities and continues playing 'regular moves' , like I did in this game, then he will often get outplayed positionally, lose the game without firing a shot and wonder throughout a sleepless night how he ended up with such a lousy position af­ ter making all those 'normal' moves. In the game I definitely saw the possi­ bilities associated with 1 3 .f4! , but play­ ing White against some under-25 00 Australian guy, I thought 'regular' moves should suffice, and the win should arrive without any risks in­ volved. This is perhaps a reasonable way of thinking when you play some Cata­ lan line with white, but not in this type of Nimzo. White is about to learn this lesson soon. 1 3.

...

1 4.

�f1 -e1 ?!

h7-h5!

�a1 - b 1 ?!

Typically in such positions, White has to use his extra space and his advantage in development to work out an initia­ tive, before his structural deficiencies will start to be felt. 1 3 .f4! was a good and energetic way to start: 1 3 . . . 'iYa4 1 4.f5 .itxc4 1 5 .fxe6 fxe6 1 6 . ttJh5 � g S 1 7 . ttJf4 0 - 0 - 0 I S .ttJxe 6 ! and White i s better. I n such positions, most of the time the critical moment arrives rather early in the

One more 'regular' , timid move and Black is already slightly better. It was high time - and White's last chance - to realize the need for energetic action and opt for 1 4.f4! and the consequential piece sacrifice. Not an easy decision, but nevertheless this was the principled way to proceed: 1 4 . . . h4 1 5 .ttJh5 g6 15

1 6 .ttJg 7 + �f8 1 7 .ttJxe 6 + 'iYxe6 1 8 .d5 ( 1 8 .f5 gxf5 1 9 . 1lxf5 is also possible) 1 8 . . . 'iYd7 (or 1 8 . . . 'iYf6 1 9 .e5 'iYg7 20 .�e3 with compensation) 1 9 .f5 gxf5 20 .11xf5 with a strong attack. Black's �a6 and ttJa5 are again stranded on the wrong side of the board. 1 4.

�

h5-h4

1 5.

ttJg 3-f1

c7-c5

White has missed his opportunities to take the initiative and Black, due to his better pawn structure, now enjoys a slight but lasting advantage 1 6.

�c1 -e3

1laS-cS

1 7.

ttJf1 -d2

e6-e5

1 S.

f3-f4

Trying to remain active and keep some central tension. In general this is a good strategy - however, the position does not offer the same opportunities as a few moves earlier. Black now has firm control of the central squares and White's action will soon result in more pawn weaknesses . I t was better t o accept that things have not gone White's way and close the centre with 1 8 .d5 iVa4 1 9 .f4 �f6 20 .f5 , and even though Black has achieved his strategic objectives, a draw is the most likely outcome. 1 S.

...

1 9.

d4xe5

1�

d 6xe5

20.

ttJd2-f3

21 .

f4xe5

16

'iVd 7-e6

29.

g2xh3

l:IcS-eS

30.

cJtg 1 -g 2

ttJc6xe5

31 .

l:td 7xa7?!

21 .

...

�f6-e7!

Looking at my horrible pawn structure and slowly becoming aware of the long and difficult defence that lay ahead, I could not help but wonder about the speed of my positional collapse: in a mere 20 moves, playing White against an opponent I had never heard of and not having made any clear mistake - ex­ cept that my whole concept was a posi­ tional blunder.

31 . 32.

1:f1 -f2

32.Jde l does not help after 32 . . . Jda8 3 3 .�xa8 �xa8 3 4.lde2 Jda3 . ...

33.

ttJf3-g5

l:reSxe4

It seems as if White has created some counterplay, but the tactics are working for Black. 34.

...

35.

cJtg 2-g3

ttJc4-e3+!

Or 3 5 .�g l .tId5 . 35.

...

.l::re 5-f5!

Winning a piece. 36 .

.r!f2-e2

3 6 . Jdxf5 ttJxf5 + 3 7 . �g4 (3 7 . �f4 ttJh6) 3 7 . . . �xfl 3 8 .ldxfl ttJh6 + .

ttJe5-c4!

32.

Doubled Pawns

33 ... l:re5! 34.ttJxf7

To suffer in the rook ending after 3 1 .ttJxe5 �xe5 32.l::t x a7 I:txe4 was a better choice, but choosing between two such evils is a difficult task. Also, with time-trouble approaching, I thought that with knights on the board my chances of a swindle should be reasonable.

36.

ttJe3-f1 +

37.

cJtg3-g2

l:lfSxf7

3S.

Ila 7xf7

�gSxf7

39.

�g2-g1

cJtf7-f6

The black knight is temporarily corral­ led, but White soon runs out of moves. 40.

�g1 -g2

b6-b5

41 .

�g2-g1

c5-c4

42.

cJtg 1 -g 2

�f6-g5

'ue1 -f1

�a6xc4

43.

�g2-g 1

cJtg5-h4

23.

�e3-g5

�c4xd3

44.

cJtg 1 -g 2

l:If5-f4

24.

'iVe2xd3

ttJa5-c6

45.

a2-a3

g7-g6

22.

25.

�g5xe7

'iYe6xe7

26.

Ilb1 -d1

0-0

27.

'iYd 3-d7

'iVe7xd7

2S.

lld1 xd 7

And White reSigned. A painful defeat and a positional lesson that I have taken to heart.

Structure 1 . 2

�e7-f6 !

Maintaining central tension was per­ haps better, but Black would keep the upper hand after 1 9. ttJf3 h3 ! - but not 1 9 . . . cxd4 ? 20 . cxd4 exd4 2 1 .�xd4 �xd4+ 22.ttJxd4 and White gets what he wants. He will develop an initiative, since 22 . . . �xc4? 23 .�xc4 ttJxc4 loses to 24.�ec l ttJa5 25 .ttJf5 0 - 0 26 .'iYg4.

1:

Chapter

Winning Chess Middlegames

2S.

...

h4-h3!

Inflicting the final damage to White, rendering his pawn weaknesses defi nitely irreparable.

Nimzo-Hiibner Variation - the mobile pawn centre

17

Chapter

Winning Chess Middlegames

NI 1 4. 8 (E4 1 )

GAME 2

Ivan Sokolov Lue Winants Netherlands tt 1994/95 (9) 1.

d2-d4

ttJgS-f6

2.

c2-c4

e7-e6

3.

ttJb1 -c3

.,ifS-b4

4.

e2-e3

c7-c5

5.

ttJg 1 -f3

ttJbS-c6

6.

.,if1 - d 3

.,ib4xc3+

7.

b2xc3

d7-d6

White sacrifices his central pawn. It is, however, very risky for Black to accept it, since in this open position, White's lead in development and strong bishop pair are worth considerably more than the small material investment. The fol­ lowing game is a good example : 9 . . . cxd4 1 0 .cxd4 exd4 l 1 .exd4 ttJxd4 1 2.�e l + ttJe6 1 3 .�a3 0-0 1 4.ttJb3 'iYd7

1 9 .�b 2! (White trusts his calculations. Also good was 1 9 .�e4 'iYxc4 20.�b2 and the black position should soon collapse) 1 9 . . . 'iYxg2+ 20 .�xg2 ttJf4+ 2 1 .�g l ttJxhS 22.�d l ! (this is the posi­ tion White had to be sure about when he played 1 9 .�b2. Black cannot escape the deadly d-file pin and will lose material) 22 .. Jlf8 23 .�a3 �d8 24.�fS ttJhf6 2S .�e7 �e8 26.�xf6 ttJxf6 27 .�xe8+ ttJxe8 28 .lld8 and White won in Portisch-Timman, Wijk aan Zee 1 9 7 8 . 9.

.. .

0-0

18

S.

0-0

9.

ttJf3-d2

e6-e5

1 0.

l:ta 1 -b1

Taking an immediate decision in the centre with 1 0 .dS ttJe 7 1 1 .'iYc2, with the idea of f4, looks attractive at first sight, but Black has a strong antidote in 1 1 . . . g 6 ! 1 2.f4 exf4 1 3 .exf4 �fS and it will take White a long time to activate his dark-squared bishop - if he ever succeeds. 1 0.

analysis diagram

h2-h3

I am trying to keep the pawn structure flexible (as explained above) and pre­ pare the typical f2-f4 push which, if executed at the right moment, almost invariably works in White 's favour.

analysis diagram

1 S . IIe3 ! (a good, multifunctional move. This rook is ready for the kingside attack, while White now also threatens to take the d6 pawn. The im­ mediate 1 S .�xd 6 ? would have been wrong due to 1 s . . . lld8 1 6 .cS ttJxcS) I S ... 'iYc6 (if Black hangs on to his pawn with 1 S .. J�d8 , White gets a strong attack with 1 6 .�b2) 1 6 .�xd6 �d8 1 7 .�eS ttJd7 1 8 . 'iYhS ! h6

Doubled Pawns

(in principle this is a good, multi­ functional move, preparing a rook transfer to the kingside and allowing a bishop retreat to b 1 . It was, however, probably better to start with lle 1 , when White has the knight transfer ttJf1 -e3 fS at hand) 1 3 . . . ttJg6 1 4.1:le l .!df8 1 S .ttJf1 �g4? (with the white rook on b2 ready for a kingside transfer, this 'weakness provocation ' plays into White's hands; 1 S . . . h6 was a better move) 1 6 .f3 �d7 1 7 .g3 h6 1 8 . �f2 lib8 1 9 .ttJe3 a6 20.ttJfS �xfS 2 1 .exfS ttJe7 22.g4 and White firmly took the initiative on the kingside and went on to win in Bareev-Short, Geneva 1 9 9 6 . 11.

The Hubner Variation, for many years considered one of the main variations of the Nimzo-Indian. Black's counter­ play is not immediately obvious (con­ trary to the Samisch Nimzo, here White's potentially weak c4 pawn can­ not easily be attacked) , so he has to play constructive moves, exerting cen­ tral pressure and waiting for White to make a decision about his pawn centre. Once White pushes his pawns and the central structure becomes fixed, Black should be able to shuffle his pieces and find targets in the white camp. White should, for his part, remain as flexible as possible, keeping central tension and delaying any pawn push until the mo­ ment when a central blockade works in his favour.

1:

...

b7-b6

Black can also increase the central pres­ sure with 1 0 . . . �e8 and force White to take an immediate decision. This looks like a good plan, but it costs two tempi, because most of the time the rook has to return to f8 later on. The following game is a good example of the way the play may then develop : 1 0 . . J:te8 l 1 .dS ttJe 7 1 2.e4 b6 1 3 .llb2.

11.

.,icS-d7?

Allowing White to execute his idea. After 1 1 . . .!:le 8 1 2.dS ttJe7 (the tactical solution with 1 2 . . . e4? 1 3 .ttJxe4 ttJxe4 1 4.dxc6 ttJxc3 would backfire terribly after 1 S . 'iYhS ! g 6 1 6 . 'iYh6 ttJxb 1 1 7 .�b2 f6 1 8 .�xg 6 hxg 6 1 9 .'iYxg 6 + �f8 20 .�xf6) 1 3 . e4 h6 1 4.1::le 1 bIf8 1 S . ttJf1 , the play would develop simi1arly to the previous comment with White having a small plus. 19

Chapter

Winning Chess Middlegames

1 l . . . h6! was a good waiting move. Such useful moves , anticipating the oppo­ nent's actions, are of extreme impor­ tance in these positions. If White carries on with his plan, Black gets good counterplay after 1 2 .f4 exd4 1 3 .cxd4 dS , and if White decides not to push f4 , it is not easy for him to find useful moves, while in the event of 1 3 .dS CiJe 7 1 4.e4, Black is two tempi up compared to the previous comments. 1 2.

f2-f4!

e5xf4

The same counter plan 1 2 . . . exd4 1 3 .cxd4 dS ( 1 3 . . . cxd4 1 4.e4 leads to thematic positions where after �e 1 , CiJf3 and �b2 , White regains the d4 pawn , retaining the better play ; 1 3 . . . CiJb4 would allow the exchange sacrifice 1 4JIxb4 cxb4 I S . e4 and White develops a strong attack) does not work now because of 1 4 . dxcS bxcS I S . cxdS CiJxdS 1 6 . �xh 7 + �xh 7 1 7 .'ifhS + �g8 1 8 .'ifxdS and White is a sound pawn up. 1 3.

l:If1 xf4

14.

'iYd 1 -f3

'iYd8-e7

I had every reason to be happy with the outcome of the opening. White has a strong pawn centre, which has re­ tained its mobility (important ! ) , and the open f-file can be used to increase the pressure. Black has no counterplay. 20

1 4.

...

1:ta8-c8

1 5.

ttJd2-f1

ttJfS-e8

1 S.

ttJf1 -g3

g7-gS

1 7. 1 8.

ilLc1 -d2

ttJe8-g7

l:tb1 -f1

I wanted to bring all my pieces into at­ tacking positions on the kingside im­ mediately, getting everything ready for the execution. Also strong was 1 8 . 11f6 CiJe8 (or 1 8 . . . �e6 1 9 . 1:rfl CiJe 8 2 0 .CiJhS ! ; 2 0 . CiJe4 ? allows counterplay with 20 . . . dS) 1 9 .CiJe4 CiJaS 2 0 J�fl and White is poised for an annihilating attack. 1 8.

...

f7-f5

The pressure along the f-file forces Black to weaken his kingside structure. Taking on d4 with 1 8 . . . cxd4 first would not help after 1 9 .exd4! (not 1 9 .cxd4? fS and now the d4 pawn hangs with tempo, so White is deprived of the e 3 -e4 break) 1 9 . . . fS 2 0 . lIe l 'iff7 2 l . CiJe4! fxe4 (2 l . .J:rce8 2 2 .CiJxd6 l:rxe 1 + 2 3 .�xe 1 'if e6 24. 'if dS wins; 2 l . .. 'iVe7 2 2 .CiJcS iVfl 2 3 .CiJxd7 'iYxd 7 24.'ifdS + �h8 2 S .�f2 , doubling on the e-file, while the bishop pair would be very strong here) 2 2 .�xe4 'iVe7 2 3 . �d S + CiJe6 2 4 .�xe 6 + �xe 6 2 S . �xf8 + 'iYxf8 2 6 .'ifxf8 + �xf8 2 7 .nxe6 and White is a clear pawn up.

1 9.

White is very much ready to open up the position. The execution will be swift. 1 9.

...

ttJg7-eS

Or 1 9 . . . CiJeS 2 0 . dxeS dxeS 2 l . �xfS gxfS 2 2 .exfS with a terrible attack. 20.

e4xf5

ttJeSxf4

21 .

ilLd2xf4

c5xd4

22.

'iYf3-d5+

�g8-g 7

On 22 . . . 'iVfl , 2 3 .�xd6 wins. 23.

ilLf4xdS

'iYe7-e3+

24.

Wg 1 -h 2

a.f8-fS

1:

Doubled Pawns

2 4 . . . 1lfl would not save Black after 2 S .�e4 �e8 2 6 .�f4 'iYxc3 2 7 .fxg6 . 25.

ilLd S-f4

'iYe3-e8

2S.

f5xgS

h 7xgS

27.

ilLf4-hS+

Black resigned. After my opponent allowed me to exe­ cute my strategic idea with f2 -f4, it was rather easy, but note that after 1 l . . . h6 ! instead of 1 l . . . �d7 ?, a complex strate­ gic battle would have ensued.

Structure 1 . 3

Nimzo-Hiibner Variation - fixed pawn centre

In Nimzo-Indian positions with doubled pawns and a full centre, it is extremely im­ portant for White not to push his central pawns too quickly, which would block the position and take out the dynamic element. This is the kind of mistake that is easy to make and a lot ofa world-class players have suffered painful losses as a result. In the 1980s, Yugoslav grandmaster Svetozar Gligoric, in general an expert in classical-type positions, lost a lot of games in the Hubner Variation with white, all of them following a very similar pattern. He would push his central pawns too quickly, gaining space but eliminating the dynamic element in the centre. As a result, Black got enough time to shuffle his pieces around and improve his position, gradually out­ playing White. The fact that a player of Gligoric's stature committed this mistake so many times, underlines once more the difficulty of coping with the strategic complex­ ity here. Thefollowing game is one of several good examples.

e3-e4!

21

Chapter

Winning Chess Middlegames

NI 1 4. 3 (E4 1 )

GAME 3

Svetozar Gligoric Predrag Nikolic Novi Sad 1982 (5) ttJg8-f6

1.

d2-d4

2.

c2-c4

e7-e6

3.

ttJb1 -c3

itf8-b4 c7-c5

4.

e2-e3

5.

itf1 -d 3

ttJb8-c6

6.

ttJg 1 -f3

itb4xc3+

7.

b2xc3

d 7-d6

8.

e3-e4

Marking the beginning of a strategy I've never liked. Many strong players have opted for this way of immediately seiz­ ing space and it is difficult to be critical of this , but I feel that keeping the ten­ sion in the centre and delaying the push of the pawns is a much more appropri­ ate strategy - see the previous game. One interesting plan to occasionally try here is 8 .'iVe2 eS 9 .ttJd2 .

analysis diagram 22

A) At first sight it is not easy to under­ stand the point behind White's idea af­ ter the simple 9 . . . 0 - 0 . However, this logical move would justify White's idea: 1 0 .dS ttJe 7 1 1 .e4. The black king is already committed to the kingside and therefore it is considerably easier than usual for White to develop a kingside attack here. A I ) In the case of 1 1 . .. �as ? White, as will be discussed in Game 7 , sacri­ fices his c3 pawn, developing a deadly attack after 1 2 . ttJb3 ! 'iYxc3 + 1 3 .�d2 'iYb2 1 4.0-0 ttJg6 l S .g3 ! 'iYa3 1 6 .f4 l:Ie8 1 7 .fS ttJf8 1 8 .g4 ttJ6d7 1 9 .9S with a dream scenario for White and a night­ mare for Black, Knaak-Spiridonov, Rubinstein Memorial, Polanica Zdroj 1 979; A 2 ) 1 1 . . . ttJ g 6 1 2 .g3 �h3 1 3 .ttJf1 �d7 1 4.ttJe3 .

1 8 .exfS e4 (a standard motif here) 1 9.fxe4 ttJeS and Black is better) 1 7 . . . 'iYd8 1 8 .ttJfS ttJh7 1 9 .�e3 'iff6 2 0 .'iVd2 �xfs 2 1 .exfS ttJe7 2 2 .g4 hxg4 2 3 .fxg4 and White soon won in Timman-Knaak, Skopj e Olympiad 1 972. B) Correct is 9 . . . cxd4 ! 1 0 .cxd4 0 - 0 and now the position i s approximately equal after 1 1 .dS ttJb4 1 2 .�b 1 ttJa6 with . . . ttJcS to follow; C) On the other hand, the attempt at counterplay with 9 . . . 'ife 7 ? 1 0 .d S e4 is not advisable, since White is better after 1 1 .ttJxe4 ttJxe4 1 2 .dxc6 bxc6 1 3 . 0 - 0 fS 1 4.f3 ttJf6 1 5 .e4 0 - 0

1:

Doubled Pawns

As early as move 9 , a position has been reached with fixed strategic features, where in the long run the knights should be superior to the bishops. Even though White has more space, he runs the risk that his position soon becomes impossible to improve, while Black is just beginning to carry out his plans. 1 0.

.l:ta1 - b1

White has tried many different plans here and we will show some of them: A) A classic example is 1 0 .ttJh4 h6 1 1 .f4

analysis diagram

analysis diagram

White has a strong kingside attack, sim­ ply due to the fact that Black has castled too early. If Black had his king on e8 and the possibility to castle queenside here, the assessment of the position would have been entirely different. 1 4 . . . ttJ e 7 1 5 . J:r g I ! h 5 1 6 . f3 ttJ g 6 1 7 .�c2 ! (good execution is always im­ portant. Faulty would have been the im­ mediate 1 7 .ttJfS ? due to 1 7 . . . �xfS

1 6 .exfS ! (correctly steering into an endgame where Black's d6 and cS pawns will be targeted by White's dark-squared bishop. Black would be doing fine in case of 1 6 .�gS h6 1 7 .�h4 gS) 1 6 . . . 'ifxe2 1 7 .�xe2 �xfS 1 8 .�d 1 l:Ife8 1 9 .�f1 �ad8 2 0 .�f4 dS (now Black's cS and dS pawns are rather weak) 2 1 . �d2 h6 2 2 .nad 1 l:Ie 7 2 3 .�f2 �e6 24.�e3 �f7 2 S .�g l ttJd7 2 6 .�b2 and Black's position soon col­ lapsed in Portisch-Csom, Hungarian Championship, Budapest 1 9 7 1 . 8.

...

e6-e5

9.

d4-d5

ttJc6-e7

analysis diagram

1 1 . . . ttJg6 ! (this was a revolutionary novelty at the time) 1 2 .ttJxg6 fxg6 and Black had an excellent game in the 5 th match game Sky-Fischer, Reykj avik 1 972; B ) Another interesting option i s 1 0 .ttJd2 , but compared to the com­ ments given after 8 .'ife2 eS 9 .ttJd2 0 - 0 23

Chapter

Winning Chess Middlegames

1 0 .d5 CiJe 7 1 1 .e4, the situation is dif­ ferent. The position is sharp and com­ plex: 1 0 .. .'iVa5 1 1 .'ifb3 (White wants to play 1 2 .'�'b5 , exchanging queens, improving his pawn structure and forc­ ing Black to castle; another option is 1 1 .'ifc2 , when Black can decide to cas­ tle queenside and enter an unclear game after 1 1 . . . h5 1 2 .h4 ild7 1 3 . g3 0-0-0 , as in Gligoric-Larsen, Bugojno 1 9 S 0) 1 1 . . . 0 - 0 1 2 . 0 - 0 CiJh5 1 3 .g3 ilh3 1 4Jie l "flic7 (the immediate 1 4 . . .f5 is favourable for White after 1 5 .exf5 CiJxf5 1 6 . "flic 2 g 6 1 7 Jib 1 ) 1 5 . "flid I g 6 (White is better in case of 1 5 . . . CiJf6 1 6 .CiJfl h6 1 7 .CiJe3 "flics I S .f3 CiJh7 1 9 .94 ! h5 2 0 . �h l ) 1 6 .CiJfl 'ifd 7 (in case of 1 6 . . . f5 Black will feel the effects of missing his dark-squared bishop af­ ter 1 7 .ilh6 l::t f7 I S .exf5 ilxf5 1 9 .ile2 CiJf6 2 0 .g4 ild 7 2 1 . CiJg 3 ) 1 7 .ilh6 CiJg 7 I S .f4 ilxfl ? (the source of Black's trou­ bles. Better was l S . . . �hS ! 1 9 .fxe5 dxe5 followed by 2 0 . . . CiJgS , with an unclear game) 1 9 J:txfl 'iYh3

C) 1 0 . 0 - 0 has also often been tried. Black is doing fine after 1 0 . . . CiJg 6 I 1 .CiJe l 0-0 1 2 .g3 ilh3 1 3 .CiJg2 CiJeS , for example : 1 4.�e 1 f5 1 5 .exf5 ilxf5

1 0.

...

11.

h2-h4

h 7-h6

Trying to either gain space or force Black to push . . . h6-h5 and allow the white bishop to g5 .

analysis diagram

2 0 .ilg5 ! (this idea Black had probably missed) 2 0 . .. f6 2 1 .fxe5 ! fxg5 2 2 .exd6 iVd7 2 3 .dxe 7 'iYxe7 2 4.'ifd2 VieS 2 5 .ilc2 , with d5-d6 to follow, and with a pawn up White was clearly better in Gligoric-Timman, Tilburg 1 9 7 7 ; 24

11.

...

1 2.

ttJf3-h2

'iWd8-c7

Doubled Pawns

Preparing 1 3 .h5 (1 2 .h5 ? ilg4) , forcing Black to make a decision. 1 2.

...

h6-h5!

Good decision. As we will see later, the activity connected with the white bishop on g5 will be of a temporary na­ ture, and Black will soon be calling the shots on the kingside. 1 3.

ttJh 2-f1

�c8-d7

14.

ttJf1 -g3

0-0-0

1 5.

�c1 -g5

23.

...

�b8-a8!

24.

'iWb3-a3

ttJd7-b8

25.

a4-a5

ttJb8-a6

Black's play has been very instructive. All potential white activity has been eliminated, while the attack on the white monarch is about to start any time now, and will be impossible to stop. White is completely lost.

analysis diagram

1 6 .f4? (this 'active' move does not help White's cause. Better was 1 6 .ilxf5 with an approximately equal game) 1 6 . . .ilxd3 1 7 .'iYxd3 Vid7 I S .�b l CiJf6 1 9 .�b2 �ae S 2 0 . �fl 'iYh3 (Black has taken over and White is in trouble) 2 1 .f5 CiJg4 2 2 .CiJe 1 e4 ! 2 3 .'iYe2 CiJ6e5 2 4 . 'ifxe4 nf7 2 5 . 'iYg 2 ? "flixg 2 + 2 6 . �xg2 CiJxc4 and Black soon won in Gligoric-Timman, Bled/Portoroz 1 9 7 9 .

1:

1 5.

...

ttJe7-g8!

26.

�d3-c2

1:1f8-g8

27.

�c2-a4

g 7-g5

28.

h4xg5

f6xg5

29.

�a4-b5

ttJa6-b8

The start of a very instructive ma­ noeuvring phase. Black is going to dis­ entangle his pieces and launch a strong attack on the kingside, while the white pieces, who are piled up on the queenside, will be unable to create any serious threat. J::td 8-f8!

1 6.

0-0

1 7.

tlf1 -e1

�c8-b8

1 8.

l:Ie1 -e2

�d7-g4!

1 9.

f2-f3

�g4-c8

Now Black is ready to start kingside ac­ tion. 20.

J:te2-b2

ttJf6-d7

21 .

'i¥d 1 -b3

f7-f6

22.

�g5-e3

ttJgB-e 7

23.

a2-a4

If 2 3 .CiJf5 , then 23 .. J�tf7 with . . . g 7 -g6 to follow, with plans similar to the game.

A sorry Sight. On the queenside White is not able to create a Single threat, while on the other side of the board the battle is lost. During - and after! - the game the white player must have thought: 'I had my two beautiful bish­ ops, more space, even something which looked like an initiative, I did not blun­ der anything and still I lost a horrible game. How could it go so wrong ? ' 25

Chapter

Winning Chess Middlegames

White could safely resign here. His last moves are agony.

30.

ttJg3-f1

gS-g4

31 .

f3-f4

eSxf4

32.

�e3xf4

ttJe7-g6

34.

e4-eS

g4-g3

33.

nb2-f2

hS-h4

3S.

.t:tf2-f3

ttJg 6xf4

36.

I:tf3xf4

d 6xeS

37.

l:!:f4-f6

h4-h3

3S.

as-a6

h3xg2 �cSxb7

39.

a6xb7+

40.

nb1 -a 1

g2xf1 'iY+

41 .

llf6xf1

a 7-a6

42.

�f1 -f6

l::t h S-h1 +

White resigned.

Structure 1 . 4

S.

...

d 7-d6?!

Later it was established that it's better for Black to immediately start active counterplay with 8 . . . b6 and . . . �a6 , . . . ttJa5 t o follow. The relative tempo loss by 8 . . . d6 combined with the plan that follows, will prove to be fatal. White is ready to advance towards the black king and immediate counterplay - mostly related to the capture of the c4 pawn is needed to achieve a dynamiC balance. 9.

e3-e4

1:

Doubled Pawns

necessary to sense them and realize that the consequence of a strategic error is, most of the time, irreparable. 1 2.

f4-fS!

1 3.

fS-f6 !

e6-eS

ttJf6-eS

More in the spirit of Black's previous move would have been 9 . . . e 5 . 1 0.

0-0

11.

f2-f4

b7-b6

A thematic idea, worth ing­ and likely to have been missed by Najdorf. 1 3.

...

�gS-hS

1 3 . . . ttJxf6 steps into a deadly pin after 1 4.�g5 with ttJg3 -ttJh5 to follow.

11. Nimzo-Samisch structure - black pawn o n c5

NI 1 8 . 1 3 (E2 9 )

GAME 4

David Bronstein Miguel Najdorf Budapest Candidates' Tournament 1950 (5)

26

1.

d 2-d4

ttJgS-f6

2.

c2-c4

e 7-e6

3.

ttJb1 -c3

�fS-b4

4.

a2-a3

�b4xc3+

S.

b2xc3

c7-cS

6.

e2-e3

ttJbS-c6

7.

�f1 - d 3

0-0

S.

ttJg 1 -e2

. ..

�cS-a6?

Crucial mistake. By now it is rather evi­ dent that Black's planned counterplay will not come in time. It is rather puz­ zling that a player of Najdorf's stature did not realize the danger he was facing and did not see the necessity to stop White's kingside advance with 1 l . . . f5 ! . After the text, the white attack develops quickly and devastatingly. It is worth noting that the critical - and ultimate - mistake occurred as early as move 1 1 in this game. As I mentioned before, even though these positions seem to be rather closed, critical moments often occur very early on. It is

1 4.

d4-dS

1 S.

ttJe2-g3

ttJc6-aS

White is undisturbedly launching a mating attack, while Black's �a6 and ttJa5 are stuck on the wrong side of the board (capturing the c4 pawn would be rather irrelevant here) . A typical Nimzo gone terribly wrong for Black. 1 S.

...

g7xf6

On 1 5 . . . �xc4, probably the easiest way to win is 1 6 .�xc4 ttJxc4 1 7 .fxg 7 + ttJxg 7 1 8 .�h6 f6 1 9 .�a2 ! a6 2 o J laf2 lIa7 2 l .ttJh5 ttJxh5 2 2 .'ifxh5 �ff7 2 3 . tIxf6 . 1 6.

ttJg3-fS

�a6-cS

1 7.

'iY d 1 -hS

�cSxfS

1 S.

e4xfS

!:rfS-gS

1 9.

r!f1 -f3

1:1gS-g7

20.

�c1 -h6

Irg 7-gS

21 .

�f3- h 3

1 -0

There is no defence against 2 2 .�f8 . 27

Chapter

Winning Chess Middlegames

NI 1 9 . 7 (E2 4)

GAME S

Boris Sky Robert Hiibner Turin 1982 ( 1) 1.

d 2-d4

2.

c2-c4

e7-eS

3.

tLlb1 -c3

�fS-b4

4.

a2-a3

�b4xc3+

5.

b2xc3

c7-c5

S.

f2-f3

tLlbS-cS

tLlgS-fS

Nowadays, 6 . . . dS is, for some reason, considered to be best here. The text, however, leads to a complicated strate­ gic battle where Black has good chances to play for a win - contrary to 6 . . . dS lines, where every mediocre booked-up player can draw with white against a strong grandmaster. White's c4 weak­ ness remains a target, while at the same time he does not have an obvious way to efficiently push his central pawns. 7.

e2-e4

played between the same opponents in Bugojno 1 9 8 2 . That game continued 8 . . . b6 9 .�d3 and now: A) 9 . . . ttJaS 1 0 .ttJh3 �a6 1 1 .�e2 �d7

d 7-dS!

A clever move. Black temporarily stops e4-eS and also does not allow White to develop his knight to f2 via h3, since 8. ttJh3 fails to 8 . . . eS 9 .dS �xh3 1 0 .dxc6 �c8, with a good game for Black.

tating tactics on the other side of the board) 1 7 . . . gxh6 1 8 . �c2 �c6 ( 1 8 .. .'ifxc4Ioses to 1 9 .'ii' d 2) 1 9 .ttJf6+ �e7 2 0 J �fd l �xc4 2 1 .iVf2 �dS 2 2 .�h4 with a crushing attack; A3) 1 6 . . . 0-0 1 7 .�g S ttJg6 1 8 .ttJd6 . B) In the game Botvinnik-Furman, training match 1 9 6 0, Black came up with an interesting plan : 9 . . . 0 - 0 1 0 .ttJe2 ttJe8 1 1 . 0 - 0 ttJaS ! ? (planning . .. �a6 . Black wants to force the white knight to g 3 . Generally, White should be happy with this development, but Black has his reasons) 1 2 .ttJg3

analysis diagram

1 2 .eS ! (White takes action because he was, I assume, worried about 1 2 . . . 'lWa4) 1 2 . . . dxeS 1 3 .dxe S ttJg8 1 4.0-0 ttJe7 ( 1 4 . . . 'iYa4 I S .ttJgS ttJe 7 1 6 .ttJe4 or 1 S . . . �xc4 1 6 .ttJe4 looks good for White) . Thus far Sky­ Hubner. Now it seems to me that Black has problems after I S .ttJg S ! (instead of 1 S J �ad l ? ! ) 1 S . .. iYa4 1 6.ttJe4 and now: A I ) 1 6 . . . ttJfS ? is wrong due to 1 7 . g4 ttJxe3 1 8 .ttJd6+ �e 7 1 9 .'lWxe3 ; A2) In the case of 1 6 . . . ttJc8 ? White has a tactical stroke :

1 0.

1:

Doubled Pawns

f3-f4

White does not have a convenient way to defend his c4 pawn, so he opts for a central pawn push, hoping to create threats. White's activity is not going to produce any compensation for the pawn, so it is safe to say that as early as on move 1 0 White is already in serious trouble. So, white players beware ! If you play these systems you take a positional risk often bigger than you've bargained for! 1 0.

...

11.

e4-e5

tLlfS-eS

1 2.

�c1 -e3

�cS-aS

1 3.

�f1 -d3

�aSxc4

0-0

Apart from the fact that he has lost a pawn for absolutely nothing, White also has a positional problem: Black controls a lot of light squares and the white bishop is not well placed on e3 . 1 4.

0-0

1 4.

...

analysis diagram

1 2 .. .fS ! 1 3 .exfS exfS 1 4.ttJhS ttJc7 I S .iVe2 �a6 1 6 .g4! ? (with all Black's minor pieces on the queenside, White is anxious to begin kingside action) 1 6 . . . 'ii' e 8 ! 1 7 .�ae 1 'iff7 1 8 .dS 1:lae8 and Black was better. a. 9.

�

b7-bS

tLle2-g3

tLlcS-a5 f7-f5!

A good move, preventing White from creating a mess.

S.

tLlg 1 -e2?!

I do not like this move here. 8 .�e3 , protecting the d4 pawn in order to first move a bishop to d3 and then develop the knight, looks more logical and was 28

1 5.

e5xfS

tLleSxfS

1 S.

'ifd1 -e2

.ic4xd3

1 7.

'iVe2xd3

'iVdS-d 7

analysis diagram

1 S.

l:ta1 -e1

l:IaS-eS

1 7 .�h6 ! (as shown several times in this book, with the black pieces piled up along the a-file, White often has devas-

1 9.

�e3-c1

tLla5-cS

20.

f4-f5

Trying to organize some sort of play. 29

Chapter

Winning Chess Middlegames

2S.

NI 1 8 . 1 (E24)

ttJf5- h4

GAME 6

20.

...

eSxf5

21 .

ttJg3xf5

l:reSxe1

Alexander Kotov

22.

I.:tf1 xe1

c5xd4

23.

c3xd4

Paul Keres Budapest Candidates' 1950 ( 14)

2S .

.

..

1.

c2-c4

ttJgS-fS

2.

d2-d4

e7-eS

3.

ttJb1 -c3

�fS-b4

4.

a2-a3

�b4xc3+

5.

b2xc3

ttJbS-cS

S.

f2-f3

b7-bS

7.

e2-e4

�cS-aS

�d 7-a4!

Looks like a good moment for White to resign. 23 .

.

..

dS-d5!

The black knight will soon jump to e4 and that will be the end of White's 'initiative ' . 24.

'iYd3- h 3

�gS-hS

29.

.tIf3xfS

'iVa4xd4+

30.

�c1 -e3

'iYd4xfS ttJe4xg5

31 .

g4-g5

32.

'ifh3-d7

ttJg5-e4

33.

�g1 -g2

ttJe4-c5 ttJc5-eS

25.

l';le1 -f1

ttJfS-e4

34.

'iYd 7-c7

2S.

I.:tf1 -f3

llfS-fS

35.

'ifc7-bS+ ttJe7-gS

27.

g2-g4

Desperation. 27.

..

.

ttJcS-e7

3S.

ttJh4-f3

'iHfS-gS+

37.

�g2-h3

'iYgS-e4

And White resigned.

Structure 1 . 5

Nimzo-Samisch structure - black pawn on c7

30

Both sides have clear objectives. White is gaining space, while Black is going to capture the white pawn on c4. Given the fact that no pawn weakness has been created in his kingside camp, he is not afraid of White's pawns advancing there. S.

e4-e5

White can also pin the f6 knight with 8 .�g5 h6 9 .�h4 ttJa5 and now : A) 1 0 .e5 forces Black to advance his kingside pawns, which could in this case turn out to be rather favourable for the second player, since his king will be safe after castling queenside. A good ex­ ample is 1 0 . . . g5 1 1 . �f2 ttJh5 ! ( l 1 . . . ttJh 7 allows White to take the ini­ tiative with 1 2 .f4!) 1 2 .h4 f5 ! 1 3 .'ifa4 (Black is firmly in control after 1 3 .c5 �xfl 1 4 .�xfl �e 7 ! 1 5 .ttJe2 'iVg8) 1 3 . . . g4! 1 4 .fxg4 fxg4 1 5 .'iYc2 'iVe7

1:

Doubled Pawns

1 6 .'iVe4 �b 7 1 7 .'iYxg4 'iYf7 and Black quickly obtained a winning attack and won easily in Bannik-Petrosian, Soviet Union 1 9 6 1 ; B) 1 0 .'iYa4 'iYc8 ! Black steps away from the pin and is ready to remind White of his weak c4 pawn with the 'iVb 7 -'iV c6 transfer. The following short game demonstrates the complexity of the struggle ahead: 1 1 .�d3 (to me it seems better for White to advance his central pawns with 1 1 .�d 1 'iVb 7 1 2 .e5 ttJh5 1 3 .d5 or 1 1 .ttJh3 'iYb7 1 2 .d5 c6 1 3 .e5 ttJh7 1 4.d6) 1 1 . .. 'iVb7 1 2 .e5 ttJd5 ! (Black correctly takes advantage of the pin on the a6 -fl diagonal) 1 3 .cxd5 �xd3 1 4J::td 1 �c4 1 5 .dxe6 fxe6 1 6 .'iVc2

analysis diagram

1 6...0-0? I t i s hard to believe, but this logical move turns out to be the crucial mis­ take. White now quickly executes a mating attack. 1 6 . . . �f7 ! (covering the g6 -square ! ) 1 7 .ttJh3 g 5 ! (not allowing ttJf4 I 'if g6) 1 8 .�g3 �b3 was the way to proceed. Black wins the exchange and, presum­ ably, soon the game. The problem of his king will be easily solved with . . . �af8 , . . . �e8 -�d8 and the black king runs to safety on the queenside. 31

Chapter

Winning Chess Middlegames

1 7 .'iVg 6 + ! . At first sight the white at­ tack does not look that serious yet. Wrong ! It is quite possible that Black's position can no longer be held . 1 7 . . . nf7 1 8 .�f6 'iVc6 ( 1 8 . . . tbc6 does not save Black in view of 1 9 .h4 ! tbe7 2 0 .�xe 7 �xe 7 2 1 .tbh3 with tbgS to follow, and the attack continues) 1 9 .h4! �a6 2 0 .tbh3 ! and the knight travels to g 5 , forcing immediate resig­ nation. The way White executes this mating attack is thematic, instructive and definitely worth ing - it could easily happen in one of your own games ! 2 0 . . . 'iVxc3 + 2 1 .�f2 tbc4 2 2 . tb g 5 and Black resigned in Timman-Brinck Claussen , Wijk aan Zee II 1 9 7 1 . 8.

...

9.

ttJg 1 -h 3

ttJf6-g8

The regular square for the knight in these positions. It can later j ump to gS (or via f2 to e4) or the bishop on gS . 9.

...

ttJc6-a5

Targeting the only weakness in the white camp. 1 0.

'iWd 1 -a4

1 O .�gS was definitely a possibility, and then, after 1 0 . . . tbe 7 (taking the pawn with 1 0 . . . f6 is very risky and probably not good: 1 1 .�h4 �xc4 1 2 .�xc4 tbxc4 1 3 .'ife2 bS 1 4.f4 gives White plenty of compensation) , 1 1 . 'ii a 4. 1 0.

...

ttJg8-e 7

Black's regular plan, to bring the queen to c6, does not work now, for example : 1 0 . . . 'ii c 8 1 1 . �d3 'ii b 7 1 2 .�e4! c 6 1 3 .�b l ! tbxc4 1 4.�d3 tb a S I S . c4. White has lost a rather irrelevant pawn and won a few tempi to develop his initiative. 11.

32

�f1 -d3

1 2 .tbgS looks interesting, but White does not have enough after 1 2 . . . h6 1 3 .tbh7 (or 1 3 .�h7 + �h8 1 4.h4 �xc4 1 5 .�b 1 d6 ! (always an important move) 1 6 .'iYc2 tbfS 1 7 . g4 tbb3 ! 1 8 .gxfS ( 1 8 .�b 2 g6 1 9 . 9xfS tbxa l 20 .�xa l exfS is better for Black) 1 8 . . . tbxa 1 1 9 . 'ife4 exfS 2 0 . 'ifxfS g 6 and Black fends off the attack) 1 3 . . J ie8 1 4.�xh6 gxh6 1 5 .tbf6+ �g 7 . 11.

...

0-0

To castle or not to castle ? That is a diffi­ cult decision here. Castling prepares . . . d6, with counterplay that is unpleas­ ant for White, hitting the latter's vul­ nerable centre. On the other hand, White is ready to attack and the king on g8 could prove to be an easy target. I believe that with optimal play Black will be better, and White's centre will be tar­ geted and, eventually, left in ruins. However, the position is very complex and even a player of Keres's class soon goes astray and at the critical moment does not feel the danger, resulting in disaster. 1 1 . . . h6 ! was probably best . After 1 2 .tbf4 0 - 0 1 3 .h4 (or 1 3 .tbhS d6 ; 1 3 .�b 1 also does not work in view of 1 3 . . . �xc4 1 4.'iVc2 tbfS ) 1 3 . . . d6 ! (a losing mistake is 1 3 . . . d S ? 1 4.�b I ! and White gets a devastating attack: 14 ... �xc4 I S .'ii c 2 tbfS 1 6 .g4 �b3 1 7 .'iff2 tbe7 1 8 .tbhS �h8 1 9 .�xh6 ! gxh6 2 0 .'iVe3 tbg8 2 1 .tbf6 Gutman­ Levchenkov, Riga 1 9 7 6) 1 4 .�b l dxeS I S .'ifc2 tbfs and Black should win. The white centre is ruined and his attacking potential is not sufficient, for instance : 1 6 .g4 (or 1 6 .dxe S �xc4 1 7 .g4 �b3 ) 1 6 . . . exf4 1 7 . gxfS exfS 1 8 .'iVxfS g6 19 .'iVxf4 tbb3 . 1 2.

�c1 -g5

1 2.

...

13.

�g5-h4

1 5.

'iYa4-c2

1 5.

...

1:

Doubled Pawns

h 7-h6 ttJe7-g 6?

The critical and final mistake. Immedi­ ately I S . . . tbfS ? would lose to 1 6 . g4, but Black had an excellent, hard-to-find resource in 1 5 . . .fS ! ! 1 6.exf6 tbfS 1 7 .g4 'iYxf6 1 8 . gxfS exfS 1 9 .tbxgS nae8 + ! (not 1 9 . . . hxg S ? 2 0 . � g 1 ) 2 0 . �f2 ( 2 0 .�d l tbxc4) 2 0 . . . �xc4 2 1 Jig 1 �h8 . The white pieces are coordinating poorly and Black seems to be better. 16. 1 3.

...

d7-d5?!

A strategic mistake, hitting the wrong pawn and leaving White's central pawn chain intact. 1 3 . . . d6 ! was again the right way to attack the white centre : 1 4.�b l dxeS I S .dxeS ( 1 S .'iVc2 fS) I S . . . gS ! 1 6 .'ifc2 ( 1 6 .tbxgS does not help : 1 6 . . . tbfS ! 1 7 .�xfS exfS 1 8 .nd l 'iVe 7 and Black wins) 1 6 . . . tbfS 1 7 . g4 tbxc4 ! 1 8 .gxfS tbxeS and Black wins. Note that the variations are very sharp and there is a rather thin line between winning and losing. 1 4.

1 6.

...

g5xh4

1 6 . . . 'ife8 1 7 .tbhS ! . 1 7.

ttJf4xg6

1 8.

ttJg 6-h8!

l;U8-e8

�d3-b1 !

Keres had underestimated this move. Thanks to his unchallenged central pawn chain White has time to attack. 1 4.

ttJh 3-f4 !

A star move !

...

g 7-g5

1 4 . . . �xc4 I S .'ifc2 g6 1 6 .�f6 looks terrible for Black.

Once given the chance, White conducts the attack beautifully. On the other hand, as we see so many times in such posi­ tions when Black's strategy has failed, 33

Chapter

Winning Che s s Mi ddlegames

the �a6 and lLJa5 are useless pieces, stuck on the wrong side of the board. 1 S.

...

1 9.

'iVc2-h7+ �gS-fS

l:ieS-e7

20 . . . �xc4 would prevent White from castling, but would not change the out­ come after 2 l .f5 exf5 2 2 .�xf5 , with 2 3 .'ifxh6 + to follow. 21 .

f4-f5

eSxf5

22.

0-0

�aS-cS

23.

�b1 xf5

�cSxf5

24.

l:tf1 xf5

�fS-eS

25.

�f5xf7

�eS-d7

2S.

'iVh7-f5+

�d 7-cS

27.

'iYf5-fS+

�cS -d7

2S.

e5-eS+

�d 7-cS

28 . . . Wd6 Ioses the queen after 2 9 J�xe 7 'iYxe 7 3 0 .lLJf7 + . 20.

f3-f4!

Simple and strong. After 20 .�g6 fxg6 2 1 . lLJxg 6 + � e 8 22 . lLJ xe 7 'iYxe 7 2 3 . 'if g 8 + iYf8 2 4 . 'iYxe 6 + 'iYe 7 2 5 .'iYg8+ 'iYf8 2 6 .'iYxd5 �d8 Black could still fight. 20.

...

ttJa5xc4

29.

l:if7xe7

30.

ge7xc7+!

'iVdSxhS

A nice execution till the very end. 30.

...

�cS-b5

3 0 . . . �xc7 3 1 .'JiJie7 + �c8 3 2 J:tfl wins. a7-a5

31 .

'iYfS-e7

32.

'iYe7-d7+ �b5-aS

33.

�a 1 - b1

Doubled Pawns

dered, played a role, so I found it difficult to understand that being a sound pawn up I had had to fight to hold a draw. Later I analysed the game with Ljubojevic (for the younger generation: in the pe­ riod of 1978-1988, Ljubomir Ljubojevic was one of the world's leading players), who opined that taking the c3 pawn was a serious strategic mistake, because according to him I had helped my opponent by giving him more space to move his pieces around and by opening the diagonalfor his dark-squared bishop. 'Anyway, you are not the first one to take pawns in such positions and later realize that White gets a strong, lasting initiative that's very difficult to cope with. Even a player of 5ky's stature once suffered a terrible defeat agains t Keres in their match, misjudging those risks. The pawn on c4 is always worth taking, because you threaten to exchange the light-squared bishops and at the same time the a1-h8 diago­ nal remains closed for the dark-squared one. Taking the pawn on c3 only opens the di­ agonal and helps White', Ljubo was quick to explain. In 1992 I did not have a compu ter, so the above-mentioned game was not easy to find. This explanation, however, seemed logical and I ed it. Around the end of 1993 I bought my first laptop and after installing the database, one day Ljubo 's comments resurfaced in my memory. So, I found the game he had mentioned. Well, he was very right. Take a look for yourself.

1 -0

NI 1 2 . 1 0 (E43 )

Structure 1 .6

GAME 7

Paul Keres Boris Sky Riga Candidates' match 1965 (8)

7.

Classical Nimzo full centre-doubled pawns structure - White sacrifices his c3 pawn

A long time ago, in Oviedo 1992, I was playing Black in a typical Nimzo-Indian with doubled pawns and a closed centre agains t some 2450 player. At some stage there was a tactical swindle and I won his c3 pawn. To my surprise, my opponent obtained nice play, getting ample compensation for the pawn, and I was lucky to get away with a draw. The fact that his c3 pawn was not voluntarily sacrificed but obviously blun-

0-01

Nowadays this is a well-known con­ cept and according to the theory it is ' the best move in this position. In 1 9 6 5 : good judgment by Keres.

1.

d 2-d4

2.

c2-c4

e7-eS

7.

...

�b4xc3

3.

ttJ b 1 -c3

�fS-b4

S.

b2xc3

ttJe4xc3?

4.

e2-e3

b7-bS

5.

�f1 -d3

�cS-b7

S.

ttJg 1 -f3

ttJfS-e4

ttJgS-fS

Sky believes that, due to the tactical resource we will see, this capture leads to a forced draw. It was much better to forget about that pawn and castle. 9.

34

1:

'iYd 1 -c2

�b7xf3

1 0.

g2xf3

'iVdS-g5+

11.

�g 1 - h 1

'iYg5-h5

The only move and the resource Black has been counting on. 1 2.

�f1 -g1

Naturally White is in no mood to allow perpetual check after 1 2 .'iYxc3 'iYxf3 + . 1 2.

...

'iYh5xf3+

1 3.

l::r g 1 -g2

f7-f5 35

Chapter

Winning Chess M i ddlegames

Fortunately for Black, his c3 knight is not immediately threatened, so he has time to defend it . This probably con­ vinced Sky that White had no more than a draw. The text is the best defence. Black loses his knight after 1 3 . . . h5 ? 1 4.�b2 h4 1 5 . �g 1 , while 1 3 . . . d5 does not help either: 1 4.cxd5 ! exd5 ( 1 4 . . . ttJxd 5 ? 1 5 .�e2 'iYf6 1 6 .e4 with a crushing at­ tack) 1 5 .�a3 ttJe4 1 6 .�xe4 'iYxe4 1 7 .'iYxc 7 ttJd7 1 8 . g l g6 1 9 J ic l .tId8 20 .'iYxa 7 and White has regained his sacrificed material, while the black king remains stuck in the centre.

14 .

.ic1 -a3!

Rejecting the draw that would result from 1 4.'iYxc3 'iYd l + 1 5 .11g 1 �f3 + . Keres correctly judges that with all his pieces active and diagonals and files open, his initiative is worth more than two pawns. 1 4.

...

1 5.

l::t a 1 -f1 !

ttJc3-e4

Preparing to drive the queen away with 1 6 .�e2 and to start pushing the pawns (f3 , d5 , e4) , gaining space. 1 5 . <Jig I ? would not be effective, be­ cause after 1 5 . . . �f7 1 6 .�e2 'iYh3 1 7 .f3 ttJf6 1 8 .d5 ttJa6 , the king on g l would be taking the square away from the a 1 rook, so most of the tactical possibilities would not be working. 36

1 5.

...

l:rh8-g8

Black has a number of possibilities, but no good defence, in this difficult posi­ tion: A) 1 5 . . . ttJc6 , to try to castle queen­ side and connect the rooks, would not work: 1 6 .d5 ! (White has to be quick) and now: A I ) Putting the knight on the edge of the board almost never helps : 1 6 . . . ttJa5 1 7 .�e2 'iYh3 1 8 .dxe6 dxe6 1 9 . .tId 1 and the black king remains in the mid­ dle, since 1 9 .. J:td8 would lose to 2 0 . f3 and after the e4 knight has been chased away, White would take on g 7 ; A2) 1 6 . . . ttJe5 1 7 .�e2 'iYh3 1 8 .dxe6 ! (precise play is needed. Wrong would be 1 8 .f3 ? ttJf6 1 9 .dxe6 dxe6 2 0 .'iYa4+ �fl and White does not have enough compensation) 1 8 . . . dxe 6 1 9 .'iYa4+ <Jifl (after 1 9 . . . ttJd7 White would win a piece with the not so difficult, but still elegant 2 0 .Itd l 0-0-0 ( 2 0 . . . ttJf6 runs into a nice mate : 2 1 .'iYc 6 ! 0-0-0 2 2 J :rxd 7 ! ttJxd7 2 3 .�d6) 2 1 .�xd7 ! �xd 7 2 2 .'iYc6 �b8 2 3 . 'iYxd 7 ) 2 0 .'iYb5 ! ttJc5 (other moves lose imme­ diately : 20 . . . c5 2 1 .�b 2 , or 20 . . . ttJg4 2 1 .iYd 7 + �g6 2 2 .�xe6+ ttJf6 2 3 .e4) ttJg6 2 2 . 'iYxc5 bxc 5 2 1 .�xc 5 2 3 .'iYxc 7 + �f6 24Jid l and White has regained his sacrificed material, the ex­ posed black king remains a target and

White's ed c-pawn will be qUick. White has a winning advantage. B) 1 5 . . . �fl is an attempt to gain a tempo compared to the game. However, with correct play by White, it does not help either: 1 6 .�e2 ! (again precise play is required. In the event of 1 6 .d5 ? ! Black would defend with 1 6 . . . ttJa6 ! 1 7 .�xe4 'iYxe4 1 8 .'iYc3 g6 1 9 .�b2 �hg8 ! 2 0 .'iVf6 + (or 2 0 . f3 �h4) 20 . . . �e8 and the white attack runs out of steam) 1 6 . . . 'iYh3 1 7 .f3 ttJf6 1 8 .dS ttJa6 1 9 .dxe6+ dxe6 2 0 .c5 ! and now: ax b 6 2 1 . cx b 6 B 1 ) 2 0 . . . ttJ b 8 (2 1 . . . cxb6 2 2 . 'iYc 7 + ttJbd 7 2 3 J:td l nhd8 2 4 .�b2) 2 2 . 'iVxc 7 + ttJbd7 23 J:rd l :thd8 24.�b2 g6 2 5 .�xf6
.id3-e2

'iYf3-h3

1 7.

f2-f3

ttJe4-f6

1:

Doubled Pawns

With his king stuck firmly in the centre, and the l:la8 and ttJ b8 out of play, Black is helpless against the advance of the white pawns. White plays energetically and accurately until the very end. 1 8.

...

�e8-f7

1 9.

e3-e4!

c7-c5

Trying to close one of the diagonals. White's pawn advance, however, con­ tinues undisturbed. Black is completely lost. 20.

.ia3-b2

f5-f4

21 .

e4-e5

ttJf6-h5

22.

�h1 -g1

g 7-g6

23.

l:!:g2-g4 !

Trapping the queen. 23.

...

ng8-d8

Other moves also lose : 2 3 . . . ttJa6 24.'iVa4; or 2 3 ... ttJg7 24. nxf4+ ttJf5 2 5 .'iVe4. 24.

�e2-d3!

Precision till the end. The immediate 24JH2 would allow 24 . . . ttJg 7 , when Black would at least save his queen. Now the threat of 2 S .�xg6+ forces the rook back and does not allow the knight to return to g 7 .

1 8.

d4-d5!

24.

...

25.

l:!:f1 -f2

1:Id8-g8

Powerless against 2 6 .�f1 , Black re­ Signed. A terrible defeat for Spas sky in his prime. 37

Chapter

Winning Chess Middlegames

Structure 1 . 7

1 2.

1 6.

tbd2-f1 !

Exploiting the fact that the pawn on c4 is indirectly protected (due to the check on a4) in order to transfer the knight to e3 , which is almost always an excellent square for it in these positions. From here the knight protects the pawn on c4 while at the same time targeting the fS -square. 1 2.

...

1 3.

tbf1 -e3

'iWd8-d7

GAME 8

Svetozar Gligoric Borislav Ivkov Bled 196 1 ( 10) 1.

d2-d4

tbg8-f6

2.

c2-c4

e7-e6

d4-dS

tbc6-aS

The pawn on c4 is not likely to fall, while the knight will be stuck on as for a long time, so 1 0 . . . ttJe 7 was a better and more logical move. 11.

tbf3-d2

�c8-a6

1 7.

1 3.

...

1 4.

1 4.

...

�f8-b4

1 S.

0-0

S.

b2xc3

c7-cS

1 6.

�d3-e2

6.

e2-e3

tb b8-c6

7.

�f1 -d3

b7-b6

...

d 7-d6

9.

e3-e4

e6-eS

1 8.

�g4xhS

1 9.

tbe3-fS!

� h8xhS

hS-h4!

Fixing the white kingside and prepar­ ing . . . ttJhS -ttJf4.

�b4xc3+

8.

'iWd7-e7

In case of 1 4 . f3 , Black continues 1 4 . . . h4! with ttJhS to follow, as in the game.

tb b 1 -c3

8 .e4! is a good way to take advantage of Black's inaccurate move order.

...

1 7 . . � d8 ?! walks away from the tempo-winning ttJfS , but now the queen is removed from the e-file and this allows White to take the initiative with 1 8 . �xh S l:ixhS 1 9 .f4 ! exf4 2 o . ldxf4 'iVe7 (or 20 . . . Wf8 2 1 .lda2 ! ) 2 1 .!la2 ! and White doubles his rooks on the f-file, causing Black a lot of trou­ ble.

h2-h3

a2-a3

tbg 1 -f3

h 7-hS!

Planning ... ttJg4 in order to exchange the e3 knight and step up the pressure on c4.

3.

8.

�e2-g4!

.

4.

It is better to castle first.

38

1 0.

g 7-g6?!

White begins a series of energetic moves, crushing Black's position in the end.

Classical Nimzo-Indian doubled pawns with fixed pawn centre - Black has a stranded knight on a5

NI 1 9 .4 (E2 6)

Doubled Pawns

This natural move allows Gligoric to develop a direct attack on his oppo­ nent's maj esty with a lot of precision and power. Immediately 1 6 . . . ttJf4 ! was necessary. With it, Black would achieve his strate­ gic goals, since the knight cannot be kicked off of f4 , for instance : 1 7 .�g4 'iVd8 1 8 . �h2 (preparing g 2 - g 3 ) 1 8 . . . �c8 ! . 1 7.

Black always has to be careful not to get stuck with his knight on as and his bishop on a6, attacking the c4 pawn which cannot be taken, while White is taking action in the centre or on the kings ide, using his attacking potential, resulting from the absence of two pieces that are being inactive on the edge of the board. When Black puts his knight on as and his bishop on a6, he'd better make sure that he either takes the pawn on c4, or seriously disturbs White's piece coordination, otherwise the absence of those two pieces could be felt dearly on the other side of the board! Ex­ changing pieces on the kings ide often does not help, because the absence of the pieces stuck on the a-file will be felt even more. Thefollowing two games are good examples.

...

1:

tbf6-hS

1 6 .ttJfS is another option, trying to col­ lect the h4 pawn. However, Black has a good game after 1 6 . . . g6 ( 1 6 . . . �xc4 1 7 . ttJxd6 + ! 'iYxd6 1 8 . �xc4 ttJxc4 1 9 .'iVa4+ We7 2 0 .'iVxc4 f6 2 1 .a4 as 22 .ldb 1 is better for White) 1 7. ttJxh4 �xc4 1 8 . �xc4 ttJxc4 1 9 . a4 'iYe7 20.ttJf3 ttJf4.

Highlighting the drawback of putting the queen on e 7 . 1 9.

...

20.

f2-f4!

'ii' e7-f8

Black does not get a break. 39

Chapter

Winning Chess Middlegames

Excellent play. The other way to sacri­ fice the piece was tempting , but far from clear : 2 1 .fxe S gxfS 2 2 .�gS (or 2 2 . 'iYa4+ �d8 2 3 . �g S + �c8 24. l::r xfS dxe S 2 S .�f6 �g8 2 6 . �xe S 'iYd6 2 7 J:te7 �xc4) 2 2 . . . �c 8 ! 2 3 .exfS dxeS 24.�f6 �h6 ! . 21 .

...

�eS-dS

25 .

26. 20 .

...

1:Ih5-hS?

20 . . . l::r h 7 , to protect the 7 th rank, may look more logical, but it has its draw­ backs as well. On h7 the rook would be badly placed after 2 1 . 'iYa4+ �d8 2 2 . ttJe3 ! (it is also possible to sacrifice a piece with 2 2 .fxeS gxfS 2 3 .�gS + �c8 24. �xfS �xc4 2 S .e6 f6 ! - the only move, otherwise 2 6 .iVd7 + - 2 6 .�xf6 and White definitely has compensation, but things are far from clear) 22 . . . 'ife8 (22 . . . exf4 does not help on of 2 3 .tLJg4 and the white attack contin­ ues) 2 3 .'iYxe8 + �xe8 24 .ttJg4! (the drawback of 2 0 . . . l:Ih 7 ) 2 4 . . . 1:1h8 2 S .fxeS dxeS and White has a pleasant choice between taking the exchange with 2 6 .�h6 ttJxc4 2 7 .�g 7 or win­ ning a clear pawn after 2 6 .ttJxeS ttJxc4 2 7 .ttJxf7 lIf8 2 8 .ttJgS llxfl + 2 9 . �xfl . Black needed strong nerves and good calculation to find the best move : 20 . . . �xc4 ! ! , holding the game after 2 1 . fxe S dxe S ! (not 2 1 . . . �xfl ? 2 2 .ttJxd6+ �d7 and now, not so clear is 2 3 .'ifxfl �xeS 24.ttJxf7 llxe4; but after 2 3 .�f4! �a6 24.ttJxf7 ! , or also the immediate 2 3 . ttJxf7 , White has a win­ ning attack) 2 2 .'ifa4+ �d8 2 3 .�h6 ! �xh6 ! 24. ttJxh6 �xfl 2 S J �xfl 'iVxh6 2 6 .l:rxf7 'iYe3 + and Black saves his skin with perpetual check. 21 .

40

�d 1 -a4+!

...

ifc7-c6?

Black misses his chance to stay in the game with 2s . . . iVb 7 ! , since after the seemingly crushing 2 6 Jhd l (or 2 6 .e6 fxe6 2 7 . �ad 1 'if c6 ! ) , Black has the cool 26 . . . �xc4 ! 2 7 .d 7 + 'ft;c7 2 8 .e6 �xe6 ! and it seems that White has no more than a draw by perpetual check with 29 .�f4+ �d8 3 0 .�g S + �c7 .

ttJf5xd 6 ! !

'iYfSxd6

23.

f4xe5

ifd6-c7

The barricades are stormed. 23 . . . 'ifxeS loses to 24J:txf7 'ife8 (or 24 . . . �c8 2 S .�f4) 2 S .�g S + �c8 2 6 .'iYxe8+ nxe8 2 7 .d6 ! . 24.

�c1 -g5+? !

24.�f4! would have finished the game in style : 24 . . . 'ifb 7 2 S .e 6 ! fxe6 2 6 .�eS 'ifd7 (or 26 . . . l'ie8 2 7 J lad l with 2 8 .dxe6 to follow) 2 7 . 'iYd l ! .l::t e 8 2 8 J�f7 ! and White wins. 24.

...

25.

d5-d6

�dS-cS

Doubled Pawns

S.

ttJg1 -e2

d 7-d6

9.

0-0

0-0

'iYa4xc6+ ttJa5xc6 ttJc6xe5

27.

l:f.f1 xf7

2S.

l:f.f7-c7+

�cS-bS

29.

�g5-f6

l:IhS-eS

30.

l:lc7-e7!

1 0.

The smoke has cleared. White gets his piece back, and reaches an easily winning endgame. 22.

1:

..

31 .

d6xe7

WbS-b7

32.

�f6xe5

�a6xc4

33.

lla 1 -d 1

l::[ a S-eS

34.

�e5-f6

�c4-b5

35.

.t!d 1 -dS

Wb7-a6

36.

�f6xh4

�a6-a5

37.

�h4-f6

Wa5-a4

3S.

h3-h4

c5-c4

.

1 0.

:aeSxe7

30.

39.

g2-g4

Wa4xa3

40.

h4-h5

g6xh5

41 .

g4xh5

1 -0

NI 1 9 .4 (E2 6)

GAME 9

Alexey Vyzhmanavin Alexander Beliavsky Lvov ch-URS 1984 (7) ttJgS-f6

1.

d2-d4

2.

c2-c4

e7-e6

3.

ttJb1 -c3

�fS-b4

4.

a2-a3

�b4xc3+

5.

b2xc3

c7-c5

6.

e2-e3

ttJbS-c6

7.

�f1 -d3

e6-e5

The main move, 7 . . . 0 - 0 , is better here.

e3-e4!

A well-known plan. White keeps his central pawn structure flexible and is ready to temporarily sacrifice a pawn. ...

b7-b6

Taking the pawn with 1 0 . . . cxd4 I l .cxd4 exd4 is a strategically risky de­ cision. The a 1 -h8 diagonal is opened for White 's dark- squared bishop, Black's d6 pawn is weak, White has good control in the centre and after some time he will regain the d4 pawn, obtaining the better play. A good illustration of how the play may continue is the blitz game Tal-Sosonko, Brussels 1 9 8 7 : 1 2 . �b2 J:te8 1 3 .f3 'ifb6 1 4.ldb l �cS I S .�h l �e6 1 6 .Ilc l ttJd7 1 7 .�b 1 (the d4 pawn is now go­ ing to fall, so Black has to try to get some of the other white pawns) 1 7 . . . 'ifb6 1 8 .�a l ttJdeS (it seems that Black is going after the wrong pawn. Better was 1 8 . . . 'ifa6 1 9 .ttJxd4 'ifxa3 2 0 .�d3 ttJxd4 2 1 .�xd4 and White should have compensation, but no more) 1 9 .ttJxd4 �xc4 2 0 .tLJxc6 ! bxc6 2 1 .�e l nad8 2 2 .f4 �b 3 2 3 .'ife2 ttJg6 24.'iYg4! �f2 2 S .1dfl 'iYh4 2 6 .'ifxh4 ttJxh4 2 7 .fS ! and the knight is trapped. 11.

d4-d5

ttJc6-a5?!

41

Chapter

Winning Chess Middlegames

The knight will stay here, doing nothing, for quite some time. Maybe I have too much of an aversion for this kind of moves, but 1 1 . . . CiJe7 looks like an al­ most automatic reply to me in such po­ sitions. 1 2.

ttJe2-g3

1 3.

a3-a4

1 4.

h2-h3

1:.aS-bS 1:US-eS

Getting ready for f2-f4. Imagine Black had played 1 1 . .. CiJe 7 ; in that case 1 4 . . . CiJg6 would have been a simple so­ lution here. Now, in order to properly control the e 5 -square, Black has to al­ low the white knight on f5 . 1 4.

...

ttJf6-d7

1 5.

ttJg3-f5

ttJd7-fS

16.

�d 1 -g4

ttJfS-g6

21 .

...

l:rf7-fS

22.

l:ia1 -e1

ttJg 6-e7

Exchanging off the annoying knight on f5 . White will, however, start his pawn avalanche with f2 -f4, not afraid to sac­ rifice a pawn (or more) in the process, because with the black knight stranded on as , White is practically a piece up. 23.

ttJf5xe7

tieSxe7

24.

f2-f41

'iYdS-eS

25.

'iYh5-e2

�c1 -g51

Depriving the black queen of the f6-square. In case of the immediate 1 7 .�e3 , Black would have good play after 1 7 . . . h6 ! 1 8 .g3 'iYf6 . 1 7.

...

f7-f6

1 S.

�g5-e3

l:rbS-b7

1 9.

g2-g3

1:.b7-f7

20.

�g 1 -h 2

�gS-hS

21 .

'i¥g4-h5

White is slowly getting ready for the pawn advance on the kingside, while Black has no adequate counterplay to keep White busy. 42

26.

f4-f51

�cS-d7

27.

h3-h4

�d7xa4

2S.

g3-g4

The pawns will advance and White's dark-squared bishop will later be sacri­ ficed - a standard sacrifice in these po­ sitions - to clear the way for the further advance of the foot soldiers. Black is helpless. 25 .

1 7.

'iYa 2 + 3 3 . �h 1 White soon delivers mate) 3 2 . �xg 7 + � g 8 3 3 . � e 5 + II x g 1 3 4 . 'iY x g 1 + � f7 3 5 . 'iY g 7 + mate) 3 1 . �h6 ! Wins) 2 7 .IIg 1 (White has enough compensation also after 2 7 .'iYh5 , with the following possible vanatIon : 2 7 . . . CiJxc4 2 8 . e 5 g 6 2 9 .'iYh6 II g 7 3 0 . exf6 ! IIxf6 3 1 .�xc4 'iYxc4 3 2 .�d4! cxd4 3 3 . II e 8 + IIg8 3 4. II e 7 and White wins) 27 . . . �fe8 2 8 . 'iYh5 , with a strong attack worth much more than a pawn.

...

l:re7-f7

Black has to choose between two strate­ gic options here : A) to take the a4 pawn and hope for queenside counterplay, allowing White to advance his kingside pawns , or B) to take on f4 , allowing the central pawn break, but eliminating the king­ side pawn push. It's a difficult decision, and it's impos­ sible to calculate over the board. Black finally goes for option A. Option B would not have solved his problems either, because in most of the critical lines , he is practically a piece down , for example : 2 5 . . . exf4 2 6 . gxf4 'iYxa4 (in case of 2 6 . . . f5 White would get a strong attack with 2 7 . e 5 ! dxe 5 2 8 . fxe5 IIxe 5 2 9 .'iff2 'ifxa4 3 0 . IIg 1 ! and now 3 0 . . . 'iYd 7 (or 3 0 . . J:lfe8 3 l .�h6 ! .J:lxe 1 (on 3 l . .. gxh6 3 2 .'iYh4

2S.

...

b6-b5

29.

c4xb5

c5-c4

30.

�d3-b1

�eSxb5

31 .

g4-g5

nf7- b7

32.

'li'e2-h5

'li'b5-eS

33.

g5-g61

h7-h6

Now White has to get ready to sacrifice his bishop on h6 . 34.

l:tf1 -g 1

�hS-gS

35.

l:re1 -e2

ttJa5-b3

36.

l:te2-g2

ttJb3-c5

1:

Doubled Pawns

37.

�e3xh61

g7xh6

3S.

'i¥h5xh6

J:[b7-g 7

39.

h4-h5

The road has been cleared for the white foot soldiers to decide the battle. iVeS-b5

39. 40.

�h6-e3

l:tfS-bS

41 .

h5-h6

'li'b5-d7

Or 4 l . . . .J:l gb 7 42 . g 7 . 42.

h6xg 7

Black reSigned. When a wrong strategic decision is taken at the start, then any player (world-class included) risks losing the game without firing a bullet. Stranded

black

knight

on

the

queenside - White has a mobile pawn centre

I would be a liar if I were to claim that piling up minor pieces on the queenside and being slaughtered on the kingside, is a disaster which I have never experienced. The following game is a good example of how things can go wrong.

NI 1 8 . 1 3 (E2 9)

GAME 1 0

Zsuzsa Polgar Ivan Sokolov Pardubice 1 994 (4) ttJgS-f6

1.

d 2-d4

2.

c2-c4

e7-e6

3.

ttJb1 -c3

�fS-b4

4.

a2-a 3

�b4xc3+

5.

b2xc3

c7-c5

6.

e2-e3

ttJbS-c6

7.

�f1 -d3

0-0

S.

ttJg 1 -e2

b7-b6

9.

e3-e4

ttJf6-eS

1 0.

0-0

�cS-a6

11.

f2-f4

f7-f5

43

Chapter

Winning Chess Middlegames

Taking on d4 first does not solve Black's problems : 1 5 . . . cxd4 1 6 .cxd4 ttJa5 1 7 . c5 �xd3 1 8 .'iYxd3 ttJdc4 1 9 .d5 ! ttJxe3 (or 1 9 . . J dc8 2 0 .�f2 ! bxc 5 2 1 .�ac l ) 2 0 .'iYxe 3 'iYf6 ( 2 0 . . . exd5 ? 2 1 .l:tad 1 only helps White) 2 1 . �ad 1 and due to his better pawn structure and safer king, White is better.

So far a well-known position, and one of the main lines of the Samisch Nimzo-Indian. 1 2.

tbe2-g3

g 7-g6

1 3.

iLc1 -e3

tbe8-d61?

This variation owes a considerable deal of its popularity to the efforts of Artur Yusupov. A number of times Artur tried this variation as White, also against 1 2 th World Champion Anatoly Karpov. Karpov scored well in those encounters. This variation came as a slight surprise to me, so I was trying to those Yusupov-Karpov games. I did re­ member that in one of the games there was a pawn exchange in the centre with 1 3 . . . cxd4 1 4 .cxd4 d5 , and Black ob­ tained an equal game (Yusupov-Karpov, Linares 1 9 9 3 ) , while in the other game between the same opponents, a differ­ ent plan was tried: the c4 pawn was im­ mediately attacked - a plan which ap­ pealed to me more at the time. 1 4.

e4xf5

1 4.dxc S ttJxc4 is good for Black. The immediate 1 4.'iVe2 does not promise anything : 1 4 .. .fxe4! 1 5 .ttJxe4 ttJxe4 1 6 .�xe4 d5 1 7 .'iV g4 (Black is better after 1 7 .�d3 cxd4 1 8 .cxd4 ( 1 8 .�xd4? does not work due to 1 8 ... ttJxd4 1 9 .cxd4 dxc4 20 .'ifxe6+ �g7 ) 1 8 . . .�xc4 1 9 .�xc4 dxc4 2 0 .'iYxc4 'iYd5 ) 1 7 . . . dxe4 1 8 .'iYxe6+ �f7 1 9 .'iYxc6 �xc4. 44

1 4.

...

'iYd 1 -e2!

The correct reaction, taking advantage of Black's mix-up. There is no possibil­ ity to take on e4 any longer, so note that the main source of Black's counterplay has disappeared! (see the comment to 1 4.exf5 ) White is firmly in control. Black's position is already critical. Somewhere around here I started to re­ alize that I was definitely not following any of Karpov's games. Things are soon going to deteriorate further. 1 5.

...

d4xc5

1 7.

iLe3-d4

tbd6xc4

g6xf5?

If you have not done your homework properly or you have forgotten your preparation, then it is better to look for the moves over the board instead of try­ ing to '' the theory - or your preparation. Firmly believing I was still in one of the Karpov games, the only question I asked myself was with which pawn to recapture (taking with the e-pawn was better, by the way) . 1 4 . . . ttJxc4! was played in Game 3 of the Yusupov-Karpov Candidates' Match, London 1 9 8 9 , and after 1 5 .�xc4 �xc4 1 6 .fxg6 �xfl 1 7 .'iYh5 'iVe7 1 8 .l:ixfl hxg6 1 9 .'iYxg 6 + fig 7 2 0 .'iYd3 cxd4 2 1 .cxd4 'iYh7 ! White did not have enough compensation for the sacrificed exchange. Black consolidated and went on to win. 1 5.

1 6.

tbc6-a5

1:

Doubled Pawns

safe to say that Black is completely lost here. I wondering at the time how it was possible to end up in a completely lost position in a mere 2 0 moves, more than half o f which were established theory, without making any obvious mistake. Well, the mistake on the 1 4th move ( 1 4 . . . gxf5 ?) is indeed not an obvious one, but it was a very serious strategic error which allowed White to easily achieve her strategic goals and deter­ mined the course of the game. 22.

23.

�a6-b7

...

2 2 . . . ttJc6 ? is 2 3 .ttJxf5 +- .

a

blunder

due

to

�f1 -f2

Defending her only potential weakness. There was also nothing wrong with the simple 2 3 .a4, keeping the pawn. 23.

...

24.

.:t.b1 -e1

tbc4xa3

White has three minor pieces (bishop pair and knight) well-coordinated and ready to attack, while Black has his three minor pieces piled up on the edge of the board and completely out of play. 1 7.

...

b6xc5

1 8.

�d4xc5

�f8-f6

1 9.

�a 1 - b1

nf6-g6

20.

iLc5-d4

lla8-c8

a7 is a pawn Black would prefer to keep, but it is difficult to give good advice here. White is well-coordinated and ready to deliver a lethal blow. In the event of 2 0 . . . 'iYc7 ? , 2 1 . ttJxf5 ! qUickly decides : 2 1 . . . exf5 2 2 . �xf5 ttJc6 (22 . . . ttJd6 ? loses immediately to 23 .'iYe5) 2 3 .�xg6 hxg6 24.f5 and White wins. 21 .

�d4xa7

22 .

.ta7-d4

'iVd8-c7

White is a sound pawn up, while Black has not solved any of his problems. It is

I guess White decided to trust me and did not fully explore the consequences of 24.�b6 ! , which would have won be­ cause Black's tactics do not work, for in­ stance : 24 . . . 'iYb8 (24 . . . 'iVc6 2 5 .�xa5 ttJxb 1 2 6 .�xb 1 �a8 2 7 .�b4 �a 1 2 8 . 'iYb2 'iVa4 29 .c4 with 3 0 .�c3 to follow) 2 5 . l:ib4 ! ( 2 5 .�xa5 ttJxb 1 2 6.�xb l �xg2 2 7 .
chances to survive) 2 5 . . . tt:J 5 c4 (after 2 5 . . . tt:J 3 c4? 2 6 .kxc4 Black loses a piece without any tricks) 2 6.kxc4 (White can also decide not to bother about Black 's tricks and win in a direct attack after 2 6 .kd4 'iVc7 2 7 .kxf5 ! exf5 2 8 . tt:Jxf5 ) 2 6 . . J �xc4 2 7 J lxc4 ka6 . Now i t seems that Black has managed to create a mess after all. However, White has 28 .ka 7 ! (a move difficult to see from a distance) 2 8 . . . �b l + (or 28 . . . 'iYxa 7 2 9 . �a4) 29 . .tt f l 'ifb7 30 .llc7 ! , winning. 24.

...

llcS-fS

In the case of 24 . . . tt:J 3 c4 the standard sacrifice 2 5 .kxf5 ! exf5 2 6 . tt:Jxf5 qUickly decides. 2S.

'iVe2- h S !

White is now attacking the black king with all his pieces. As for the defenders - well, I'd say that it would not be easy to find worse squares on the entire board for the black knights. 2S.

...

Wf7 3 1 . 'iVxd 7 +) 2 9 .kxf5 exf5 3 0 J:re 7 h6 3 1 . 'iYh5 ! �c6 3 2 . !'f. e 8 + Wf7 3 3 . .!dfe2 and the only way for Black to prevent the immediate mate is to give a full queen. 26.

...

'iVc6-bS

The bishop cannot be taken, since after 26 . . . fxe4 White has 2 7 .'iVe5 ! (the point behind the piece sac) 2 7 . . . !:Iff6 2 8 . tt:Jxe4 winning, while the ugly 2 6 . . . d5 ? would shut off the only source of Black's counterplay, the a8-h 1 diago­ nal, and would lose immediately after 2 7 .kxf5 . 27.

�e4xfS!

�d3-e4 !

46

bIe2-a2

l!gS-aS

47.

l':tg6xd6+ �d4-eS

36.

l::t cS-hS

ttJc6xd4

4S.

l:td6-d 1

�e4xfS

37.

c3xd4

�b7-dS

49.

�d1 -f1

.l:!f4xf1 +

3 7 . . . ke4 does not help due to 38 . .!dh6 + ! �d5 (or 38 . . . �e 7 3 9 .g4 and the white pawns decide) 39 .l:th3 . 3S.

l::t h S-h6+

�d6-c7

39.

l:Ia2-a1

�dS-e4

40.

g2-g4

49 . . . !!g4+ 5 0 J:tg3 +- . SO.

wg 1 xf1

S1 .

.tia3-g3!

�eS-f6

.idfSxfS

2S.

ttJg3xfS

'iVbSxfS

29.

"iY hSxfS

e6xfS

30.

�e1 -eS+! �gS-f7

31 .

�eS-eS

The black king remains cut off forever.

White is rolling her pawns forward, while the pin along the a-file prevents Black from organizing any kind of counterplay.

The smoke has cleared. It will still take Black quite some time to coordinate his pieces , while White's ed pawns on the kingside will decide. 26.

Doubled Pawns

3S.

On 2 7 . . . exf5 ? , 2 8 .I:Ie 7 wins.

'iVc7-c6

Probably even stronger was 2 6 . tt:Je4! and White wins in the attack: 2 6 . . . fxe4 (or 26 .. Jlf7 2 7 .tt:Jg5) 2 7 .kxe4 !:If5 2 8 .'iYh4 ! iVc8 (White also wins in the event of 2 8 . . . 'iYb5 2 9 .kxf5 'iYxf5 3 O . 'iYd 8 + �f7 3 1 . 'iYxd 7 + or 28 . . . 'iYxe4 2 9 . I:rxe4 kxe4 3 0 .�d8+

1:

Chapter

Winning Chess M i ddlegames

31 .

...

ttJaS-c6

32.

�eSxfS+

�f7-e6

33.

l::tf S-cS

3 3 . .!dh5 also wins. The text, however, underlines all the misery caused by the knight being on a3 . 33 .

...

1:g 6-gS

34.

�f2-e2+

�e6-d6

40 .

...

d7-d6

41 .

f4-fS

<;t>c7-c6

42.

llh6xh7

�c6-dS

43.

Ilh7-h3

kIaS-gS

44.

bIa1 xa3

44.

... l:Ih3-g3

...

�fS-g6

h2-h4

1 -0

The win is very simple, for example: 5 2 ... �f5 5 3 .Wfl \t>f6 54. �f3 \t>f5 5 5 J�g5+
At last some good news. After twenty moves I have managed to get rid of my terrible knight! 4S.

S1 . S2.

NI 24. 1 3 (E3 2)

GAME 1 1

I van Sokolov

Viktor Bologan Sarajevo 2 0 04 (4)

l:tgSxg4+

Also sufficient was 45 .�fl ! ? �xd4 46.f6 htg 2 + (46 . . . .!df4+ 47 Jlhf3 +- ) 47.\t>fl Il g 6 48.,lda4+ \t> e 5 49 Jih5 + �f5 5 0 J�a5 + d5 5 1 ..!dxd 5 + �xd5 5 2 . .!dxf5 + +- .

ttJgS-f6

1.

d 2-d4

2.

c2-c4

e7-e6

3.

ttJb1 -c3

�fS-b4

4.

�d 1 -c2

0-0

S.

e2-e4

d7-d6 �b4xc3+

4S.

...

1:g4-f4

6.

a2-a3

46.

J::t g 3-g6!

<;t>dSxd4

7.

b2xc3

e6-eS

S.

�f1 -d3

ttJbS-c6

9.

ttJg 1 -e2

b7-b6

46 . . . �xf5 (Hecht) 4 7 . Idf6 48 . .!dg3 a:f4 49 Jlg5 +- .

l:ig4+

47

Cha p ter

Winning Chess Mi ddlegames

In the 4.'it'c2 Nimzo, this kind of pawn structure normally does not appear. Here, in this sideline starting with S .e4 d6 , the pawn structure resembles the Samisch Nimzo and most of the rules apply. 1 0.

0-0

�c8-a6

11.

f2-f4

tbf6-d7

An important zwischenzug. 1 S . . . dxeS ?? would be a terrible blunder that simply loses a piece after 1 6. 'iY d 1 . 1 6.

nd3-d5

1 7.

na 1 - b1

tba5-b3

White is 2 l .itxf4 ? 'iVh4 2 2 .g3 'iVg4, while after 2 l .�d4 cS ! 2 2 . l:rxd7 'iVe8 White has to return the piece with 2 3 .'iVc6 cxd4 24.cxd4 �d8 2 S Jic7) 2 l . . . 'iVe8 ! . The white queen on a4 is not defended and due to the e8-a4 pin , Black gets his material back, obtaining the better game after 2 2 .ttJxf4 ttJeS (not 22 .. J �i d 8 ? ? 2 3 . ttJxd3 cxd 3 24.l:1d4) 2 3 .�bd l l::I d 8 24 .itcS I;Ixd 7 . 1 8.

tbe2-g3

Heading for its optimal square : fS .

1 7.

White is trying to keep his pawn centre flexible, while Black wants to target the c4 weakness. 1 2.

I:rf1 -f3!?

The introduction to an interesting con­ cept. In my game against Anatoly Kar­ pov, Hoogeveen 2 0 0 3 , 1 2 .ite3 was played. 1 2.

...

...

'iVd8-e7

d6xe5

tbg3-f5

'ife7-e6

20.

'ifc2-d 1 !

With this tempo White is hoping to win the eS pawn.

Doubled Pawns

'iYb6 + 2 7 . 'iYn 'iYxc 7 2 8 .�h6 + ! �xh6 2 9 . 'iVh4+ � g 7 3 0 .ttJfS + gxfS 3 1 .'iVg S + . Winning tries could get White in trouble : 2 6 .itgS h6 ! , or 2 6 .iVd l �ae8 2 7 .itgS h6 2 8 .tDxg6 �xg 6 ! ; not 28 . . . 'iVxg 6 2 9 .ite 7 . 21 .

nd5xe5

J::ta 8-d8!

An important tempo. 22.

'ifd 1 -e2

23.

'iYe2xc4!

'ife6-d7

This pawn has to be eliminated, among other things to undermine the position of the b3 knight. A mistake would be 2 3 .�e3 ? 'iVd3 with an advantage for Black. 23 .

...

'ifd 7-d 1 +

24.

'if c4-f1

'ifd 1 -c2

25.

�c1 -e3

25 .

...

tbc6-a5 20.

c4-c5

Trying to make use of the fact that the bishop can be recaptured with the rook, when pressure along the d-file will be exerted. Opting for the text, I underesti­ mated Black's defensive resources. Very much in the spirit of the position was 1 3 .ttJg3 ! itxc4 1 4.itxc4 ttJxc4 l S .ttJfS and White gets an initiative on the kingside which definitely compensates for the sacrificed pawn. �a6xd3

1 3.

...

1 4.

nf3xd3

b6xc5

1 5.

d4xe5

c5-c4!

48

...

1 9.

With the inclusion of the last two moves things have changed and 1 7 . . . dxeS ! was now possible : 1 8 .'it'd 1 (not 1 8 . btxb 3 ? cxb 3 1 9 .'it'd2 c6 2 0 .�xd7 �b6 + 2 l . �fl �ad8 and the black b-pawn would decide the out­ come) 1 8 . . . ttJbcS 1 9 . ite3 ttJd3 ! ( 1 9 . . . 'it'e 7 ? 2 0 .itxcS ttJxcS 2 l .J:lxeS is clearly better for White) 2 0 .'it'a4.

Black targets White's only weakness. 1 3.

1 8.

1:

analysis diagram

Both players had seen this pOSItIOn, both had concluded that White wins a piece and stopped exploring further. However, tactics now start to work for Black: 20 . . . exf4! 2 1 .�xd7 (bad for

...

tbd 7-f6

Black had a tactical option in 20 . . . exf4 ! ( 2 0 . . J:lad8 ? i s wrong after the simple 2 1 .fxeS , and Black has stepped into the pin for no reason at all) 2 l . 'it' g4! (the white pieces would be extremely clumsy after 2 1 .l::tx d7 ? 'it'xe4 and Black gets back his piece, with a winning ad­ vantage after 2 2 .ttJe 7 + �h8 2 3 .�b2 tDxc 1 ! 24.'iYxc l I;Iae8 2 S .ttJdS �e 6 ! ) 2 1 . . . g6 2 2 . 1::rxd 7 'iYxe4 2 3 .ttJe 7 + �g 7 24.'iYxf4 and now Black saves the game with 24 . . . 'iVe 1 +! 2 S .'iYfl 'iYe 6 ! and White has nothing better than to force a draw by perpetual check after 2 6 . I;Ixc 7

tbf6xe4?

Critical moment. Black is moving his second knight away from his king. Even though it seems Black is taking over the initiative, very soon White will develop a direct attack against the black king while Black's knights will be stranded on the wrong side of the board. It was necessary for Black to anticipate the danger and find 2 5 . . . ttJd2 ! (taking a pawn with 2 5 . . . 'iVxc3 does not help due to 2 6 .'iYe l ! ) . Now, a long variation peters out in a dead-drawn rook en­ ding : 2 6 .bk l 'iYb2 2 7 J lb5 ! 'iVxa3 49

Chapter

Winning Chess Middlegames

2 8 .i.cS 'iYa6 2 9 J�a l 4:Jxfl ! 3 0 . kixa6 �d l ! 3 1 .kixf6 ! 4:Jd2 + ! 3 2 . \t>f2 4:Jxe4+ 3 3 .\t>f3 4:Jxf6 3 4.i.xf8 �xf8 3 s .kib8+ 4:Je8 3 6 .4:Jxg 7 \t>xg 7 3 7 .ldxe8 kid3 + 3 8 . �g4 l:ixc3 3 9 Jh 8 l:ia3 40 .llc8 . 26.

tecting the first rank b y eliminating the checks) 34 . . . 4:Jd4 3 S .�g4 4:Je6 3 6 .h4 ! 'iVc 1 3 7 .hxg S + 4:JxgS 3 8 . \t>h2 ! (free­ ing the bishop from the pin) 38 . . . kid3 3 9 .i.b4.

In a lot of variations , having the pawn on as will be to Black's disadvantage. 4.

...

S.

e2-e41

ttJgS-f6

1:

Doubled Pawns

centre and keeps it flexible, achieving a very favourable version of the Nimzo­ Indian.

l:tb1 -e1 !

White has moved all his pieces to the side of the board where the action is. 26.

...

27.

'iYf1 -f3!

ttJe4xc3

Focusing on the black king. I was no lon­ ger interested in settling for a small ad­ vantage after 2 7 .'iYf2 'iYxf2 + 2 8 .i.xf2 . 27.

...

l::!: d S-d1

2S.

�e3-f2

g 7-g6?

Allowing a quick execution. However, Black's pieces (queen and two knights) are far away from their monarch and it is difficult to defend here. 29.

ttJfS-h6+!

30.

f4-fS!

WgS-g 7

White has a winning attack. 30.

...

Wg 7xh6

30 . . . 4:Jd2 loses to the attractive 3 1 .f6 + �h8 3 2 .4:Jxf7 + ! \t> g 8 3 3 .4:Jh6+ �h8 3 4.'iYe3 4:Jde4 3 S . lde 8 ! Jd.xe8 3 6 .f7 kied8 3 7 .'iYd4+ � l xd4 3 8 .i.xd4 + . 31 .

fS-f6

The escape route for the black king is cut off. 31 .

...

32.

�f2xe1 �e1 xc3

g6-gS

Missing the best execution: 3 3 .h4! 4:Jd4 34.'iVg4 kig8 3 S . �xgS . 33.

...

ttJb3-d2

The best defence was 33 . . . �d8 . How­ ever, White wins after 34.i.e I ! (pro50

'iYf3-e31

34 .

...

�c2-c1 +

3S.

Wg 1 -f2

iYc1 -f1 +

36.

Wf2-g3

'iYf1 -f4+ gSxf4+

37.

'iYe3xf4

3S.

Wg3xf4

ttJd2-c4

39.

l:teS-cS

ttJc4xa3

40.

Wf4-fS

.::tf S-dS

41 .

�c3-eS

1 -0

White soon delivers mate, while the black knight ends its career at an unde­ sirable square.

l:td1 xe1 +

White delivers a forced mate after 3 1 . . . gS 3 2 . 'iVh3 + \t>g6 3 3 .kixg S + ! \t>xgS 34 .i.h4+ ! �hS 3 S .i.g3 + \t>gS 3 6 .'iYh4+ \t>f5 3 7 .�f4+ \t>g6 3 8 .�g4+ <Jtxf6 3 9 .i.h4 mate. 33.

34.

The last difficult move. Black soon runs out of checks.

VO 2 2 . 1 2 (A40)

GAME 1 2

I van Sokolov

Emir Dizdarevic Saraj evo 1 998 (9) 1.

d2-d4

e7-e6

2.

c2-c4

.ifS-b4+

3.

�c1 -d2

a 7-aS

4.

ttJb1 -c31

Steering the game into Nimzo-Indian positions where the inclusion of the moves i.d2 as is going to favour White.

Taking immediate advantage of the in­ clusion of the moves 3 . i.d2 as . An­ other interesting option is 5 .i.g S , when the extra move 3 . . . aS is probably working against Black. S.

...

d 7-d6

7.

a2-a31

S.

b2xc3

�b4xc3

It is rather obvious that here the extra move . . . a7 -as is useless. S.

...

c7-cS

9.

ttJg 1 -e2

ttJbS-c6

1 0.

0-0

0-0

11.

f2-f31

Taking the pawn with S . . . i.xc3 6 .i.xc3 4:Jxe4 only helps White, since after 7 .iVg4 fS (after 7 . . . 4:Jxc3 ? 8 .iVxg 7 �f8 9.bxc3 , with the black king stuck in the centre, White is much better) 8 .'iYxg 7 'iYf6 9 .'iYxf6 4:Jxf6 1 0 .dS , his bishop pair and better pawn structure ensure White a clear advantage. 6.

.if1 -d3

e6-eS

It was not the first time I tried this set-up as White. In my game against Michael Adams , Wijk aan Zee 1 9 9 6 , play continued 6 . . . cS 7 .dS (also an in­ teresting idea, similar to this game, was 7 . a3 i.xc 3 8 . bxc 3 ! 4:Jc6 9 . 4:J e 2 ) 7 . . . i.xc3 8 . i.xc3 exdS 9.exdS 0 - 0 1 0 .4:Je2 4:Jbd7 1 1 . 0-0 l:i e 8 1 2 .4:Jg3 and White was clearly better. With the text Black would like to provoke White into making a decision in the centre. How­ ever, this is not going to happen. With his next move White strengthens his

It is important for White to keep his centre flexible and push f4 at a favour­ able moment. White also prepares 1 2 .i.e3 . As usual in such positions , White has no reason to worry about Black taking on d4 , as already explained in the notes to Vyzhmanavin-Beliavsky. The immediate 1 1 .f2 -f4 ? would have been a mistake, since after 1 1 . . . exd4 1 2 .cxd4 cxd4 Black would transfer his 51

Winning Chess Middlegames

f6 knight either via g4 to e3 or via d7 to cS , making use of the weakness just cre­ ated in the white camp. 11.

...

tL'lf6-d7?!

Taking on d4 would now produce a very different result: 1 1 . . . exd4 1 2 .cxd4 cxd4 (exchanging a pair of knights does not change much. White is also better after 1 2 . . .ttJxd4 1 3 .ltJxd4 cxd4 1 4.�gS 'iYe 7 l S .l::f.b l ) 1 3 .�gS 'iVe 7 1 4.f4 ( 1 4.l:rb l is also good) 1 4 . . . h6 I S .�h4 'if e6 1 6 .ltJg3 and White gets a strong initia­ tive worth much more than a pawn. Black's plan to transfer his c6 knight to b3 is an attempt to justify the a-pawn push. However, it is not going to work, since White will be quick with his kingside attack. It was better to leave the knight on f6 and play a few useful moves, like . . . l::f. e 8 and . . . h6 . 1 2.

�d2-e3

a5-a4

1 3.

lla 1 -b 1

J:1a8-a6

Chapter

(on 1 6 . . . l:le8 , rather strong i s 1 7 .cS ! dxcS 1 8 .ltJfS �b6 1 9 .eS) 1 7 .ltJbS . 1 5.

tL'le2-g3

tL'la5-b3

1 6.

.Rd3 -c2

�f8-e8

1 7.

tL'lg3-h5!

Black's knight is rather irrelevant on b3 and he does not have any meaningful counterplay. On the other hand, the white onslaught on the kingside is soon going to be impossible to ward off.

1 9.

'ifd1 -g4!

tL'lf8-g6

The only move. In case of 1 9 . . . g6 White obtains a mating attack with 2 0 .�g S ! gxhS 2 1 .'iYg3 'iVd7 2 2 .�f6 + ltJg6 2 3 .'iVg S dS 24.eS . 20.

c3xd4

21 .

.Re3-f2

c5xd4

The pin along the g-file is terrible for Black, so White has enough time to prepare the execution. 21 .

...

'ifd8-e7

2 1 . .. ltJd2 loses to 2 2 .�xd4! ltJxfl 2 3 . brxfl f6 24.'iY g3 . 22 .

.Rc2xb3!

Now, 2 2 . 'iYg3 would offer Black some chances of survival after 2 2 . . .'iVe S ! 2 3 . fxg6 'iYxhS 2 4 . gxf7 + �xf7 2 S .�xb3 axb3 2 6 .�xd4 'ifg6.

1 7.

tL'ld 7-f8?

The final mistake. Now the white attack becomes impossible to stop. Though not an easy move to play, the best defence was 1 7 . . . g 6 ! 1 8 .ltJg3 cxd4 1 9 . cxd4 exd4 2 0 .�xd4 ltJxd4 2 1 .'ifxd4 ltJcs and Black would still be in the game. 1 8.

f4-f5!

22.

...

23.

nb1 xb3

a4xb3

As the last attacking piece, the second rook s White's attacking forces, forCing resignation. 23.

...

tra6-c6

Or 23 . . .f6 24.'iYd l ! ltJeS 2 S .!:f.g3 g6 2 6 .fxg6 hxg6 27 .�xd4 and White wins.

e5xd4

Probably best under the circumstances. Black loses immediately after 1 8 . . . f6 ? 1 9 .'iVg4 'iYe 7 2 0 .�h6 g6 2 1 .'iYg3 . 1 4.

f3-f4!

1:

Dou b led Pawns

The following game is a good example of a strategic and tactical battle in posi­ tions with doubled pawns and a full centre. Anand does everything contrary to the general set of strategic rules I have been trying to explain and to the examples I have given so far. Black opts for the inferior opening vari­ ation, then voluntarily strands his knight on as without haVing any realis­ tic chances to ever take on c4. White, meanwhile, gets his flexible pawn cen­ tre ready to roll and has well-developed pieces to it, without having to make any concessions from his side. At first Sight it looks as if, with all those trumps, White should be having a size­ able advantage (as indeed he had) . Black then comes up with a rather inge­ nious strategic plan, starting with 1 2 . . . gS , which is worth studying and ing. However, the black plan involves a weakening of his king and with White's advantage in space as well as development, the position should logically contain a tactical blow. The problem with such tactical solutions is that they must be detected and executed at once. Being just one move late will allow the opponent to coordinate his pieces and the possibility is lost forever. Even though this was a rapid game, it was well played, rather complex and an excellent learning example.

The time has come to start action in the centre/ on the kingside. 1 4.

...

S2

NI 1 9 . 3 (E2 6)

tL'lc6-a5

Black carries on with his plan as if nothing had happened. However, there are no really satisfactory options and it all boils down to a choice between evils. White is also clearly better after 1 4 . . . exd4 I S .cxd4 cxd4 1 6 .ltJxd4 ltJcs

GAME 1 3

24.

'iYg4-d 1 !

tL'lg 6-e5

Teimour Radjabov

25.

�f2xd4

�g8-h8

26.

f5-f6

1 -0

Viswanathan Anand M onaco (rapid) 2 0 0 7 (2)

Black is mated soon, for instance: 2 6 .f6 gxf6 2 7 .ltJxf6 IIg8 28 .'ifhS I1g7 29 .:g3 'iff8 3 0 .'iYh6 , with ltJh7 to follow.

tL'lg8-f6

1.

d2-d4

2.

c2-c4

e7-e6

3.

tL'l b 1 -c3

.Rf8-b4

S3

Chapter

Winning Chess Mi ddlegames

4.

a2-a3

�b4xc3+

5.

b2xc3

c7-c5

S.

e2-e3

d 7-dS?!

It is difficult to explain why Vishy re­ jected the usual 6 . . . ttJc6 and opted for the inferior text. White now gets a very good version of the well-known Samisch Variation of the Nimzo-In­ dian. 7.

�f1 -d3

S.

tLlg 1 -e2

9. 1 0.

b7-bS?! e3-e4

tLlcS-a5

0-0

Pinning the knight immediately with 1 O .�g5 is also an option. 0-0

1 0. 11.

54

�c1 -g5

11.

...

1 2.

�g5-h4

h 7-hS

well-developed and coordinated, so the time to act was NOW! Black would have been in trouble after 1 5 .dxc5 ! dxc5 1 6 .fxg5 hxg5 1 7 .exf5 exf5 . Now the position is open and White has to use this opportunity with l S .�e5 ! (otherwise Black plays I S . . . f4) 1 S . . . ttJc6 (bringing the knight back into play and attacking the bishop) 1 9 . ttJg 3 ! .

tLlbS-cS

Black's next two moves are rather sur­ prising. First he loses time playing . . . b6 and then he strands his knight with . . . ttJa5 , even though it is rather clear that there will be no time to take the c4 pawn . Meanwhile, White fol­ lows the standard procedure, improv­ ing his position and getting ready to strike. Had I not known the names of the play­ ers , I would have thought that the black player must be a complete patzer. How­ ever, in chess as in life things are often not the way they seem. S.

Here, the standard idea 1 1 .f4 is defi­ nitely an option. White should not fear sacrificing his c4 pawn; after 1 1 . . . �a6 1 2 .f5 �xc4 (or 1 2 . . . e5 1 3 .�g5 h6 1 4 .�h4) 1 3 .�g5 he will have good compensation.

Anand understands that things are de­ veloping very much White's way, so with the following sequence of moves he embarks on a risky strategy, expos­ ing his king but trying to shut in both of White's bishops. 1 2.

g 7-g5!

1 3.

�h4-g3

tLlfS -h5

1 4.

f2-f4

f7-f5!

Black stops the white advance and ham­ pers both the enemy bishops ' move­ ments. Even though with his two knights on as and h5 Black's strategy is highly risky, it is far from easy to find something clear for White. If he is to take advantage of his better develop­ ment and Black's potentially exposed king, it will have to be done either by a sacrifice or by a pawn breakthrough in the centre. And the sooner the better, before Black coordinates his pieces. 1 5.

'iYd 1 -c2?!

White has to act before Black gets orga­ nized. White is at the moment very

analysis diagram

The point of the previous move. Black is now going to feel the consequences of exposing his king : A) 1 9 . . . ttJxe 5 2 0 . 'iVxh5 ttJxd3 2 1 .�ad I ! (naturally White is not inter­ ested in the perpetual with 2 1 .'iVg6+) 2 1 ... 'iVf6 2 2 .ttJe4 ! (bringing in an extra attacking piece) 22 . . . 'iVg 7 2 3 . �xd3 ! fxe4 24. �xfS + �xfS 2 5 .�dS+ �e7 26 .'iVeS+ �f6 2 7 .'ifc6+ and White wins; B) 1 9 . . . 'iVeS 2 0 .�e4 ! (making maxi­ mum use of the exposed black king) 20 . . . ttJxg 3 (20 . . .fxe4 10ses to 2 1 .�xfS + 'iYxfS (or 2 1 . . . �xfS 2 2 .'iYd6 +) 22 .'iYd5 + iff7 2 3 .'ifxc6) 2 1 .�d5 + �e6 2 2 .�xg 3 and White has a winning advantage; C) 1 9 . . . ttJxg 3 (probably best under the circumstances. The other moves only underline the weakness of his kingside) 2 0 .�xg3 f4 2 1 .�e4 ! 'iYxd 1

1:

Doubled Pawns

2 2 .�axd 1 fxg3 (other moves do not solve the problems either: 2 2 . . . �g4 2 3 . � d 5 ; 2 2 . . . �b 7 2 3 . �d 7 fxg3 2 4 . �d 5 + �hS 2 5 . �xfS + �xfS 2 6 . �xb 7) 2 3 .�xc6 gxh2 + 24.�xh2 with a big advantage for White in the endgame. 1 5.

...

g5-g4!

Excellent move. White's g3 bishop is now shut off and tactical possibilities are eliminated. Black has positional control now. White missed the critical moment to act when he played the timid 1 5 .'iYc2 . 1 S.

l:ra1 -d 1

tLlh5-g7

1 7.

�g3-f2

'iVdS-c7

1 S.

tLle2-g3

h6-h5

1 9.

Mf1 -e1

VJlic7-f7

20.

e4-e5

With both of his bishops locked in, White wants to open up the position. Trying a different central break does not promise anything : 2 0 .d5 exd5 2 1 .cxd5 c4 2 2 .�fl h4 ! 2 3 . ttJxf5 ttJxf5 24.exf5 �xf5 and Black has a good game. d Sxe5

20. 21 .

:Ie1 xe5

tLla5-cS

22.

�e5-e1

'li'f7-c7

23.

d4-d5

tLlcS-dS!

As long as Black maintains the blockade c4 versus c5 and f4-f5 , it will be diffi­ cult for White to activate his bishops. 24.

tLlg3-e2

'iVc7-e7!

Not allowing 2 5 .�h4. 25.

tLle2-g3

'iYe7-fS

Black wants more than a draw, which could have been achieved by returning to c7 . 26.

tLlg3-f1

tLldS-f7

Black is about to take over, so White has to act energetically. Choosing simple moves will result in an inferior posi­ tion. 55

Chapter

Winning Chess Middlegames

wins i n case o f 3 S . . . \t> g 8 3 6 .nxe8+ tLlxe8 3 7 .'ife4 ! tLl c 7 3 8 . fS ..ta6 (3 8 .. J:lb8 3 9 .'ife 7 ! ..txfS 40 .'ifxc 7 �e8 4 1 . tLle 7 +) 39 .l::t d 7 ! 'iVc l + 40 . ..te 1 J::r e 8 4 1 .tLle 7 + llxe 7 42.l:txe 7 tLld6 43 .'ife5) 3 6 .tLle7 + �h8 37 .tLlg6+ �h7 and White has to agree to perpetual check, since 3 8 J�xe8 ? fails to 38 . . . 'iYxg6 l .

27.

dSxe6

Better was 2 7 .tLle3 ! eS 2 8 .d6! (freeing the dS-square for the knight) 2 8 . . . e4! (this tempo-winning move is logical and good. Wrong would be 2 8 . . . ..te 6 ? 29.fxeS tLlxeS 3 0 . ..tfl ! l::tad8 3 1 . tLldS ..txdS 3 2 . 11xdS l::r fe8 3 3 .d 7 ! l::t e 7 34.'iVd2 with a clear advantage for White) 2 9 .�xe4! fxe4 3 0 .tLldS 'iYxd6 3 1 ...th4 'iYe6 ! (the best defence, even though it does not look logical. Black puts his queen 'under the tempo' (Jdxe4) in order not to allow the centralization of the white queen ('ifxe4) after 3 1 . . . 'iYh6 ? 3 2 .tLle 7 + �h8 3 3 . 'iYxe4 �a6 (or 33 . . . ..te6 34.fS) 34.fS with a winning at­ tack) 3 2 .Jd.xe4 'iYh6 3 3 .tLle7+ �h8 3 4Jiee l ! (threatening 3 S .tLlg6+) 34 ... Me8 3 S .tLlg6+ (wrong is 3 S .fS ? due to 3S .. Jixe7 3 6 .11xe7 hfS)

3S.

...

tLlhS-g 7 !

Strong play. By bringing his hS knight to e8 Black releases his queen. 3 8 . . . tLle4 would not offer more than a drawish ending after 3 9 . ..txe4 'ifxc6 40.dxc6 fxe4 4 1 .fS �f6 4 2 . nd 6 + �xfS 43 J::tdS + \t>g6 44.IId7 �e8 45 J ixa7 e3 46 . ..te 1 �c8 47 . l::r e 7 IIxc6 4 8 J �xe 3 .

27.

...

ilLcSxe6

39.

a3-a4

tLlg7-eS!

2S.

tLlf1 -e3

l:!:aS-dS

40.

a4-aS

'iff6-b2!

Black has connected and coordinated all his pieces. White has missed his tactical opportunities and is now worse. 29.

�d 3-f1

l:tdSxd 1

30.

�e1 xd 1

hS-h4!

31 .

tLle3-dS

31 .

...

ilLe6xdS

c4xdS

tLlf7-d6

33.

�f1 -d3

tLlg7-hS

34.

g2-g3

l:US-eS

Possibly even stronger was 34 . . . c4 ! . I guess Anand didn't want to allow White's dark-squared bishop to d4. However, then its d3 colleague would be sent to fl and end up completely dead after 3 S . ..tfl �e8 3 6 . ..td4 'iYh6 3 7 .'if g2 tLlf6 . Black is much better.

tLle4xd 2

Black had a nice win with 44 . . . tLl8xd6 ! ! 4 S . l:ixd6 'iff3 4 6 .'iVd5 + �g7 4 7 . ..td4+ ( o r 4 7 . ..t e 1 'if e 3 +) 4 7 ... cxd4 4 8 . 'iYxd4+ \t>h7 . 4S.

�c6-d5+

4S .dxe 7 loses to 4S . . . 'iYf3 ! 46.'iYxb6 'iYd l . 4S .

...

�f7-f6

aSxb6

46.dxe 7 was the best defence, even though Black should win the resulting endgame after 46 . . . 'iV d 1 4 7 . 'if eS + \t>f7 4 8 . 'iYxfS + r:J;; xe 7 4 9 . 'iYh 7 + r:J;; d 8 S O . 'iYd3 + �c8 S 1 .'iYfS + �b8 S 2 . 'iYe S + �b 7 S 3 . 'iYd5 + �a7 S 4.'iYd7 + �a6 S S . 'iYb S + �b7 S 6 .'iYd 7 + tLlc7 S 7 . 'iYd3 bS ! S8 . ..te3 (5 8 .cxb S ? c4) S 8 . . . 'iYxfl + S 9 .�xfl tLlxfl 6 0 .�xfl bxc4 6 1 . ..txcS �c6.

46.

Hoping to open a diagonal for one of his bishops. 32.

...

DOllbled Pawns

4S . . . �e6 46.d7 is not clear: 46 . . . tLlf3 + 47 . \t> g 2 tLl c 7 4 8 . d8 'if (48 . 'ifxfS + \t>e 7 ) 48 . . . tLlxdS 49.cxdS and White has tactical chances.

The knight on d6 is defended, so the queen can enter the fray. 41 .

44 .

1:

41 .

...

'li'b2-b3!

42.

nd 1 -d2

a 7xb6

43.

�d3-f1

'iVdSxd 2

46.

...

l:te7-d7

47.

'iYd2-dS

tLleSxd 6

A ' move ' like 43 . \t>h2 loses to 43 . . :iYc3 44J::td l �aS ! with 4s . . . 11c7 to follow, trapping the white queen.

3S.

c3-c4



43.

...

36.

'iWc2-a41

JleS-e7

44.

dS-d6

37.

'ifa4-c6

h4xg3

3S.

h2xg3

tLld6-e4

Hoping for tactics. After 44.11e2 tLl8d6 Black has clearly improved his knights.

4S.

'iYdS-eS+

49.

'iVeS-dS+ �f7-e7

SO.

'iVdS-eS+ �e7-dS

S1 .

'iVeS-hS+ �dS-c7

analysis diagram

S2.

'iYhS-aS

3 S . . . �h7 ! ! (exposing his king to the dis­ covered check is the only defence. White

S3.

'iVaS-a7+ ttJd6-b7

56

'ifb3-b1

And White reSigned. 57

Chapter

Winning Chess Middlegames

Structure 1 . 8

1:

Doubled Pawns

�aSxd3

1 S. 1 7.

iVd1 xd 3

l:IcSxc5

1 S.

�c1 -e3

.l:1c5-d5

1 9.

'iVd3-aS

ttJa5-cS

20.

l:f1 -c1 !

White understands that he is not a pawn down, but rather virtually a piece up and so he makes ready to sacrifice material in order to force a break­ through.

Nimzo-Samisch structure - white pawn chain paralyses the black kingside

In these doubled pawns/full centre positions, Black has to be careful not to allow the white pawns too far. If this happens, then the black kingside could easily be paralysed and cut off from play. The following game is a good example. I am rather convinced that the modern treat­ ment of one of the topical Nimzo variations stems from this old game by David Bronstein.

NI 1 9 . 7 (E24)

GAME 1 4

ttJbS-cS

S.

David Bronstein

7.

e2-e4

0-0

Vladimir Simagin Moscow ch-URS 1 96 1 ( 1 2 )

S.

e4-e5

ttJfS-eS

9.

f3-f4

c5xd4

c3xd4

b7-bS

1.

c2-c4

ttJgS-fS

2.

d2-d4

e7-eS

3.

ttJb1 -c3

�fS-b4

4.

a2-a3

�b4xc3+

5.

b2xc3

c7-c5

S.

f2-f3

When I said 'modern treatment of this old Bronstein idea' , I meant the line 6.e3 ttJc6 7 .�d3 0-0 8.ttJe2 b6 9 . e4 ttJe8 1 0 . 0 - 0 �a6 1 1 .f4 f5 1 2 .d5 ttJa5 1 3 .e5 ! �xc4 1 4.�xc4 ttJxc4 1 5 .d6 and due to the fact that Black's kingside is paralysed and cut off from play, White has strong compensation for the pawn, as seen in Milov-J Polgar, Moscow, FIDE World Championship 2 0 0 1 . 58

1 0.

1 3.

d4-d5!

1 4.

d5-dS

ttJcS-a5

Like in the above-mentioned game Milov-Polgar, White's advanced pawns paralysing the black kingside are worth far more than a pawn, with the small difference that here White has an even better version because the knight is better placed on f3 than it was on e2 in the other example. 1 4.

...

.J:.aS-cS

1 5.

0-0

g 7-gS

11.

ttJg1 -f3

�cS-aS

�f1 -d3

f7-f5

The immediate 1 2 . . . ttJa5 ? ? fails to a well-known elementary school tactic : 1 3 .�xh7 + ! �xh 7 1 4 .ttJg 5 + �g6 (or 1 4 . . . �g8 1 5 .'lWh5 ) 1 5 .f5 + ( 1 5 .'lWg4 should win as well) 1 5 . . . exf5 1 6 . g4 and White soon wins.

...

ttJeS-g7

21 .

.i::r c 1 xcS !

:Id5-a5

The endgame after 2 1 . . . dxc6 2 2 .'ifxa 7 'lWa8 2 3 .'iVxa8 (not 2 3 .'lWxb6 �b8 24.iVc7 ? ? �b 7 and White has blun­ dered his queen: 2 5 .'lWxc6 1:Ib l +) 2 3 . . . 1:Ixa8 24.�xb 6 is difficult for Black.

Black understands the trouble he has got himself into and is no longer in such a hurry to collect the pawn. Putt­ ing the e8 knight on g7 (the only avail­ able square) and connecting the rooks becomes his chief priority.

Black is very much focused on collect­ ing the c4 pawn, allowing the white pawns to advance. Soon he will realize that with his kingside, i.e. his rook on f8 and his knight on e 8 , out of play, the extra pawn will not be worth much. 1 2.

20 .

22.

'iVaSxa5!

A nice positional queen sacrifice. In the resulting position the white rook will be devastatingly strong on the 7 th rank. On top of that, the black knight remains out of play, so it is fair to say that in real­ ity White is the one with the material advantage.

1 S.

c4-c5!

The more pieces White exchanges on the queenside, the more the absence of the e8 knight is going to be felt.

22 .

...

bSxa5

23.

RcS-c7

h 7-hS

Hoping to push . . . g6-g5 and create some sort of counterplay on the king­ side. 24.

I:ta 1 -b1

59

Chapter

Winning Chess M iddlegames

...

37.

.l:ib1 -c1

Doubled Pawns

Structure 1 . 9

rious threats. Furthermore, the black king is soon to become a prime target. 3S.

1:

'iVdS-d3

The other rook is ready to enter the 7 th rank. 24.

...

ttJg 7-eS

A sad story. The knight returns to eS , underlining the success of White's strategy. Black is without any kind of counterplay. 2S.

P.. c 7xa7

gS-gS

2S.

�e3-bS

'iVdS-bS

d7 is not the pawn Black likes to give, but 26 . . . 'iVcS runs into 2 7 .i!.. c 7 with b(bS to follow, and Black will lose his queen. 27.

:Ia 7xd 7

The pawn on d7 is gone and nothing has been solved. ...

'iVbS-cS

2S.

l:td7-e7

'iYcS-c2

29.

.t:Ib1 -f1

'iVc2-cS

30.

�bS-f2

ttJeS-g7

31 .

f4xgS !

27.

Now the time has come to collect this pawn. 31 .

...

hSxgS

32.

ttJf3xgS

�cS-dS

33.

ttJgS-f3

fS-f4

34.

'D.. e 7-c7

ttJg7-fS

The black knight is finally active, but at what price ! 3S.

�f1 -b1

3S.

h2-h3!

ttJfS-e3

Even with the knight on e3 Black does not have enough potential to create se60

Nimzo-Indian structure - White controls open d-file building on his control of d5

37.

...

ttJe3-dS

The attempt to create counterplay with 3 7 .. :if g6 does not work after 3 S .i!..xe3 fxe3 3 9 . ld l c4! (it's good that 3 6 .h3 ! has been played. Playing the other rook - 3 9 . � 7 c4 - may offer Black some chances to survive after 39 . . . 'if d3 4 0 . � d4 'iYxa3 ) 3 9 . . . �b l + (or 39 . . . 'iYd3 4o .11d4, and the rook ending is lost for Black after 39 .. Jixf3 40 Jlg4 'ifxg4 4 1 .hxg4 e2 42 . nc l ) 40.'it>h2 e2 4 1 .�g4+ 'it>hS 42.d7 e l 'if 43 .tDxe l �xe l 44.1dgS ! (the most exact - how­ ever, the prosaic 44.ldcS 'iYxe S + 4S .�g3 'iYd6 46 .�eS wins as well) 44 . . . 'iYe3 4S .!IcS 'iYf4+ 46.ng3 'it>h7 47 . neS and the d-pawn queens. 3S .

.l:Ic7-cS

White is converting his winning advan­ tage systematically and steady-handed. 3S.

...

ttJdS-e3

39.

.Q:cS-c3

'iVd3-bS

40.

..if2xe3

f4xe3

One of the strategic plans we do not see that often, but which is nevertheless impor­ tant in Nimzo-Indian type positions with doubled pawns and a full centre, is playing for the d5-square. White voluntarily exchanges on e5 or c5 and then uses the d5 out­ post for his knight or another piece. Taking control of the d-file is often part of the plan. The following game, played seventy years ago by the 6th World Champion, il­ lus trates this very well.

NI 2 S . 2 (E2 l )

GAME 1 5

Mikhail Botvinnik Vitaly Chekhover Leningrad ch-URS sf 1938 1.

d2-d4

ttJgS-fS

2.

c2-c4

e7-eS

3.

ttJb1 -c3

..ifS-b4

4.

ttJg 1 -f3

0-0

One of several possible moves here. The other continuations, more frequently seen nowadays, are 4 . . . cS and 4 . . . b 6 . S.

�c1 -gS

1 0.

...

h 7-hS

11.

�gS-h4

c7-cS

1 2.

na 1 -e 1 !

Preparing tDd2 and fl-[4. 1 2.

...

�cS-g4

Preventing White from achieving his desired set-up.

d 7-dS

S ... cS , challenging White in the centre, or S . . . dS , transposing to the Ragozin Variation, are better moves here.

41 .

nc3xe3

'iYbS-bS

S.

e2-e3

�d8-e7

42.

l:ic1 -c3

�gS-g7

7.

�f1 -e2

eS-eS

43.

�g 1 -h2

�g7-hS

S.

'iY d 1 -c2

:US-eS

44.

I:te3-e4

'iVbS-b2

9.

0-0

�b4xc3

and at the same time Black reSigned.

White has to recapture with the pawn; after 1 0 . 'iYxc3 ? Black has l O . . . tDe4.

1 0.

b2xc3

61

Chapter

W i nning Chess M iddlegames

1 3.

1 3.

...

'iVe7xf6

1 4.

'iVc2-e4

�g4xf3

Black exchanges his bishop in order to finish his development. Keeping the bishop with 1 4 ... �c8 with the idea to develop the knight to c6 and then chase away the white queen with . . .�fS does not work. White is in time with I S . ttJd2 ! ttJc6 1 6.f4, with a huge advantage. 1 S.

a blockade on d6 , closing the position, with the black knight soon becoming superior to the white bishop.

�h4xf6!

A strong and far from obvious strategic decision. White exchanges his well­ placed bishop, which was pinning the black knight , in order to take control of the light squares and the h l -a8 diago­ nal. A lot of strong players would not even consider exchanging the beautiful bishop for the stupid knight, which is pinned to boot. This decision shows the deep strategic understanding the 6th World Champion was famous for. The rest of the game, especially the execu­ tion phase, is quite impressive.

�e2xf3

ttJb8-c6

21 .

...

22.

�f3-g4

'iVf6-e7

The white rook will enter the 7 th rank. 22.

...

23.

�g4-fS!

'iVe7-b7

Taking away square h7 from the black king, creating a different mating tactic. 1 8.

23.

l::t d 1 -dS!

White builds on his beautiful outpost. 1 8.

...

b7- b6

The immediate 1 8 . . J:ixdS does not solve Black's problems , due to 1 9 .cxdS ttJe7 2 0 .d6 ! 'iYxd6 2 1 .'iYxb7 , with a clear advantage, since the white bishop is very much superior to the black knight here. 1 9.

l:i.f1 -d1

...

'iVb7-b8

2 3 . . . 'lW a8 was a better defence, though after 24.'iYh4 White maintains full domination. 24.

l::t d S-d7

l:re8-d8

Otherwise 2 S . 'lW dS .

h2-h3!

Improving his position, correctly antic­ ipating the course of events. 20.

...

2S.

1 6.

d4xcS!

1 6.

...

d6xcS

1 7.

l:.e1 -d1

l::ta 8-d8

62

c3-c4

b6-bS

31 .

c4xbS

ttJa3xbS

32.

e4-eS

as-a4

33.

f2-f4

ttJbS-d4

34.

�g1 -f2

g 7-gS

3S.

g2-g3

gSxf4

36.

g3xf4

Black is ive and helpless , unable to stop the advance of the white pawns. 36.

ttJd4-e6

37.

�f2-e3

cS-c4

38.

f4-fS

ttJe6-cS

39.

Ild 7-c7

ttJcS-d3

40.

eS-e6

f7xe6

41 .

fSxe6

l::tf 8-e8

42.

e6-e7+

g8-g7

43.

�dS-c6

1 -0

NI I S . 1 (E4 1 )

'lWe4xeS!

2S.

...

ttJaSxc4

26.

'iVeSxb8

l'ld8xb8

27.

�fS-e4!

Excellent execution. Improving the bishop, maintaining domination, is more important than snatching the a7 pawn. 21 .

a7-aS

29. 30.

l:rd8xdS

Making use of the tactical possibilities created with 2 3 .�fS ! .

Opening the d -file, though isolating his doubled pawn, is part of the plan started with 1 3 .�xf6 ! . Due to his con­ trol of the dS -square, White will take control of the file.

Doubled Pawns

In the following, more recent example, White brings his knight to the dS out­ post and after the knight swap recap­ tures with the pawn. In this case getting a ed pawn on dS improves White's position.

ttJc6-aS

Moving away from cxdS with tempo and thus getting ready to capture on dS . 20.

1:

l::t d 1 xdS !

Maintaining control of the d-file, with full domination. Improving his pawn structure with 2 1 .cxdS ? , which is in general a good idea, would be a serious strategic mistake here, because it allows

27.

...

ttJc4-a3

28.

�e4-dS

Mb8-f8

29.

e3-e4!

With full control ensured, the pawns can start to roll. This is a very systemati­ cally played game. A beautiful learning example.

GAME 1 6

Yury Kuzubov John van der Wiel Groningen Harmonie 2004 (6) 1.

d2 -d4

ttJg8-f6

2.

c2-c4

e7-e6

3.

ttJb1 -c3

�f8-b4

4.

e2-e3

During the process of writing this book I came across a game Wang Yue-Balogh, Taiyuan 2 0 0 7 , where White also exe­ cuted the strategic plan I am about to explain. That game continued 4.ttJf3 b6 S . �gS �b 7 6. e3 h6 7 . �h4 �xc3 + 8 .bxc3 d6 9. ttJd2 ttJbd7 1 0 .f3 'iYe7 1 1 .'iYa4 0-0 1 2 .e4 eS (so far all stan63

Chapter

Winning Chess Middlegames

dard theoretical moves) 1 3 .�d3 (it is obvious that the ideal square for the white knight is e3 and a logical follow up would be lLlfl -e3 and then lLlf5 or lLld5) 1 3 . . . .tHe8 1 4.0-0 (I have never understood why White castles here, in­ stead of continuing his plan with 1 4.lLlfl ! . The only drawback is that the white king temporarily stays in the middle. However, I do not see a way for Black to exploit this. If he takes action in the centre with 1 4 . . . exd4 1 5 .cxd4 c5 1 6 .lLle3 ! (a standard pawn sacrifice in these positions, shown before in this book. Closing the centre with 1 6 .d5 ? would allow Black to become unpleas­ antly active after 1 6 . . . 'iYe5 ! 1 7 .11d l 'iVc3 + 1 8 .�f2 lLle5) 1 6 . . . cxd4 1 7 .lLlf5 'iYe6 1 8 . 0 - 0 lLlc5 1 9 .'iYd l (White has full control, and he will retrieve the pawn with dividends) 1 9 . . . lLlh5 2 0 . lLlxd4 'iYe5 2 1 .�c2 lLlf4 2 2 .lLlf5 and White is much better) 1 4 . . . lLlf8 1 5 .�fd 1 lLlg6 1 6 .�f2 lLlf4 1 7 .�fl ! (the bishop has to be preserved. In case of the direct knight transfer with 1 7 .lLlfl , Black becomes active with 1 7 . . . lLl6h5 1 8 .lLle3 'iYg5) 1 7 . . . c5 1 8 .g3 (kicking back the knight) 1 8 ... lLle6

stands that the white knight will come to d5 , where it will have to be ex­ changed. As a result, White will get a ed pawn on d5 , so having the knight ready to jump to d6, blockading that pawn , will come in handy) 2 1 .lLlfl ! (the knight travels to its desti­ nation) 2 1 . .. �c6 2 2 .'iVc2 lLlb7 2 3 .lLle3 g6 24 .lLld5 �xd5 2 5 .cxd5 lLld6 2 6 .a4 and White was better. 4.

...

c7-c5

5.

ttJg 1 -f3

d 7-d6

6.

..tf1 -d3

..tb4xc3+

7.

b2xc3

'iY dS-e 7

An unusual move order, and not the best. It is normal to develop a knight to c6 before deciding where to put the queen. S.

plain about in these positions. With the pawn structure defined, moving the knights around is conSiderably easier than playing with the bishops. White's strategy is sound because his knight will reach the beautiful outpost on d5 . It is rather difficult for Black to tolerate the white knight there, so he will exchange it for one of his knights. White then recaptures with his c4 pawn and gets a ed pawn, while the a 1 -h8 diagonal will soon be opened for his dark-squared bishop. 11.

...

1 2.

ttJd 2-f1 !

Executing the idea. 1 2.

...

tbfS-e6

1 3.

ttJf1 -e3

ttJe6-f4

'iVd 1 -c2

...

1 4.

..td 3-f1 !

The black knight will soon be kicked out with g2-g3 . 1 4.

analysis diagram

1 9 .dxe5 ! (the right positional decision) 1 9 . . . dxe5 2 0 .�g2 lLld8 ! (Black under64

d4xe5!

d6xe5

1 0.

ttJf3-d2

ttJbS-d7

11.

e3-e4!

The idea introduced by 8. 'iY c2 becomes clear. For the time being, with the 9 . dxe 5 exchange White has defined the pawn situation in the centre, which in general is nothing for Black to com-

tbf6-g4

Otherwise 1 6 . g 3 and lLld5 . 1 6.

ttJe3-d5

tbf4xd5

1 7.

c4xd5

f7-f5

1 S.

..tf1 -b5+!

An important check. ..tcS-d7

1 S.

...

1 9.

..tb5xd7+ iY'e7xd 7

20.

e4xf5!

Making use of the fact that the d5 pawn is temporarily taboo : 2 0 . . . 'iYxd5 ? ? 2 1 .'iYa4+ .

e6-e5

9.

...

Dou b led Pawns

ttJd7-fS

With this move White shows that he has already decided on his strategic set-up. Otherwise he would have played 8 . 0 - 0 e5 9 .lLld2 . S.

1 5.

1:

...

h 7-h6

Preparing the ... lLlh3 jump after 1 5 .g 3 . In case o f the logical 1 4 . . . 0-0, White is clearly better after 1 5 . g 3 lLle6 ( 1 5 . . . lLlh3 ?? simply loses a piece after 1 6 .ttJf5 �xf5 1 7 .exf5 lLlf4 1 8 .gxf4 exf4+ 1 9 .'iYe2 'iYd7 2 0 .�xf4) 1 6 .ttJd5 . 1 5.

h2-h4

Not willing to allow 1 5. g 3 ttJh3 , since the position is rather unclear after 1 6 .ttJf5 �xf5 1 7 .exf5 lLlg5 1 8 .�g2 e4.

20 .

...

0-0

21 .

f2-f3

ttJg4-f6

22.

c3-c4

Now the ed pawn on d5 is de­ fended, while the a I -h8 diagonal is opened for the bishop. Last, but defi­ nitely not least, White also happens to be a pawn up. 22.

...

e5-e4

Hoping for a tactical solution to the problem. 23.

0-0

24 .

..tc1 -b2!

llaS-eS

This bishop now becomes a monster. 24.

...

e4xf3

25.

�f1 xf3

tbf6-g4

26.

�f3-f4!

Preparing the lethal blow. The immedi­ ate 2 6.f6 ? would not be effective due to 2 6 . . . 'iVd6 ! 2 7 . g3 ttJxf6 . 65

C hapter

Winning C hess Middlegames

31.

... I:ta 1 -e1 ?

Doubled Pawns

Structure 1 . 1 0

White slowly creates technical difficul­ ties for himself. 32.

1:

'iVg 7-gS

Much better is 3 2 .h5 ! with an easy win: 3 2 . . .' ifg 5 (or 3 2 . . . 'iYc2 3 3 .�xf6) 3 3 .�xf6 �xf6 34.l:rxf6 ttJxg2 3 5 .t1g 6 . 32.

2S .

...

hS-h5?

Black is not alert , allowing for his qUick annihilation. However, after retrieving the pawn with 26 . . . ttJe3 2 7 .'iYd3 �xf5 (or 2 7 . . . ttJxfS 2 8 .J:rafl ttJd6 2 9 .iYg6 and the white attack continues, whereas not taking the pawn and fishing for compensation with 2 7 . . .' iVd6 should not work after 2 8 .�f3 ttJg4 2 9 . g3) 2 8 J�e4 ! �xe4 29 .'ifxe4 ttJg4 3 0 . �e 1 h5 3 1 .d6 ! , White has a winning advan­ tage, since Black has no more than just a few checks : 3 1 . . .'ifxd6 3 2 .'iVxf5 'ifh2+ 3 3 . <Jtfl 'ifh l + 3 4 . �e2 'iYxg 2 + 3 S . �d I 'iVxb2 3 6 .l:Ie8 mate. 27.

...

tLle3-g4!

IntrodUCing some tactics, like 33 .gxe8 'iYb I + 34.'iffl 'iYxfl + 3 5 J:rxfl �xe 8 , with chances o f salvation. Still, a rela­ tively safe win for White should be 3 3 . �d l ttJe3 3 4JId2. 33.

�e1 -f1

34.

d5-dS?

fS-f5

After achieving a totally winning po­ sition, White is doing his best to spoil it.

Nirnzo-Indian structure - White combines control of the d- and f-file with the power of his bishops

During the process of writing this book I came across this very recent game and found Topalov's play as impressive as it was instructive.

EO 44. 8 (E2 I )

GAME 1 7

Veselin Topalov Levon Aronian Morelia/Linares 2 0 08 ( 1)

f5-fS!

Now it's immediately over.

1.

d2-d4

tLlg8-fS

g7xfS

2.

c2-c4

e7-eS

3.

tLlb1 -c3

kf8-b4

4.

tLlg 1 -f3

c 7 -c5

5.

g2-g3

27.

...

2 7 . . . ttJxf6 2 8 .�xf6 gxf6 2 9 .'ifg 6 + 'ifg7 3 o .' ifxhS is also lost. 28.

'iVc2 -gS+ �d 7-g 7

29.

'li'gSxh5

tLlg4-e3

30.

'iVh5-f3

�g8-h 7

31 .

�g1 -h1 ?

It is rather difficult to explain why White refrained from the simple 3 I .�xf6 , winning immediately. Now

34.

...

1d.f8-f7??

The moment he finally gets serious counterchances, Black blunders horri­ bly. After the normal 34 . . . 'iVxd6 3 5 .'iVxb 7 + .!:re7 3 6 .'�'f3 idfe8 Black would have a lot of tactics and be very much back in the game. 35.

J:If4xg4

1 -0

This variation used to be first Alekhine's and then Romanishin's favourite, but it gained real popularity thanks to Kasparov's efforts in one of his World Championship matches against Karpov. 5.

...

c5xd4

Another logical continuation is not to exchange any pawns in the centre and to exert immediate pressure along the as-e 1 diagonal with 5 . . . ttJe4 6 .'iVd3 'ifas . One of the stem games with that line was the first game of the 66

Kasparov-Karpov World Championship match in 1 9 8 5 . There followed : 7 .'ifxe4 �xc3 + 8 .�d2 �xd2+ 9.ttJxd2

analysis diagram

9 . . . 'ifb6 ? (a strange mistake coming from the 1 2th World Champion. Black is underdeveloped and he should bring his b8 knight into play. By the way, his queen is pinning White's d2 knight and is well placed on as . Modern theory is still trying to come up with a verdict in the line 9 . . . ttJc6 1 0 .dxc5 (or 1 0 .dS ttJd4 l I .�g2 ; the original 1 1 . �d 1 is also an option) 1 0 . . . b6 1 1 .�g2 �b7 ) I O .dxc5 'ifxb2 1 1 .�b I 'ifc3 (Black de­ cides, probably correctly, that taking a 67

Ch apter

Winning Chess Middleg ames

pawn with 1 1 . . . iYxa2 is too dangerous after all)

analysis diagram

1 2 . 'ifd3 ! . A nice, instructive concept often seen in such positions and worth ing. By exchanging the queens on d3 , White improves his pawn structure, while keeping all the pressure. In particular the weakness of b7 is annoying for Black, and he has terrible problems to finish his deve­ lopment. White's advantage is larger than it may seem at first Sight. I still be­ ing amazed (back in 1 98 S !) by the fact that an endgame virtuoso of Karpov's stature went down in this game, not be­ ing able to create any counterplay or put up any meaningful resistance to the young Kasparov: 1 2 . . . 'ifxd3 1 3 .exd3 4Ja6 1 4.d4 Jdb8 I S .�g2 r:i;e 7 1 6 . r:i;e2 l:rd8 (the attempt to solve the b7 weakness and at the same time exchange the ive c8 bishop would not work; after 1 6 . . . b6 1 7 .cxb6 axb6 1 8 .l:rb3 i.b7 1 9 .i.xb 7 !:txb 7 White exerts pressure on the weak b-pawn and obtains a winning position after 2 0 . l:[hb l l:[hb8 2 1 .cS ! bS 2 2 .4Je4 4Jc7 2 3 .a4! b4 24.4Jd6 l;Ia7 2 S . l:l:xb4) 1 7 . 4Je4 (another logical way was 1 7 . � b 3 d 6 1 8 . cxd 6 + !:txd6 68

1 9 . �e 3 b6 2 0 J�a3 ! with :Ib l to fol­ low. White has a massive advantage note the poor coordination of the black pieces) 1 7 . . . b6 1 8 .4Jd6 ! 4Jc7 (in the case of 1 8 . . . bxcS White can choose between simply taking an exchange with 1 9 .4Jxc8+ �dxc8 2 0 .i.b7 or playing for domination with 1 9 . 1::[xb8 4Jxb8 2 0 .dxcS) 1 9 .nb4 (White doubles his rooks on the b-file, a simple and power­ ful plan here) 1 9 . . . 4Je8 2 0 .4Jxe8 �xe8 2 1 .l:ihb 1 i.a6 (desperately trying to create some sort of counterplay in a lost position) 2 2 .�e3 dS 2 3 .cxd6 Ilbc8 (if 2 3 . . J �xd6 , 24.cS deCides) 24.r:i;d3 !Ixd 6 2 S . !:ta4 ! bS 2 6 . cxb S Iib8 27 .!:tab4 and White soon won. This game again confirms that when a positional set-up is fundamentally wrong, then even the very best - Kar­ pov in his prime - can lose without fir­ ing a Single shot. 6.

tbf3xd4

tbf6-e4

This is one of the main lines of the g 3 Nimzo. 7.

'iVd 1 -d 3

7.

...

�b4xc3+

The other frequently played line here is 7 . . . 'ifa s . For the sake of a better general understanding of this type of position I shall give a few instructive examples of how the play may proceed :

A) 8 .4Jb3 and now : A I ) White is better in the case of 8 . . . 'iffs 9 .'iVe3 4Jxc3 1 0 .bxc3 i.e 7 1 1 .i.g2 .

1:

Doubled P awns

9 . . . 4Je4 1 0 .iYxe4 �xd2 + 1 1 . 4Jxd2 0 - 0 1 2 .�g2 4J c 6 and now :

analysis diagram analysis diagram

With pressure along the d-file and on b7, while his doubled c-pawns are con­ trolling a lot of squares. It is important to note that doubled c-pawns are much better than they may seem in these po­ sitions. 1 1 . .. 4Ja6 1 2 . 0 - 0 0 - 0 1 3 Jid 1 'ifhS 1 4.�f3 �g6 l S .'ifd3 ! fS ( l S . . . 'ifxd3 1 6.exd3 would improve White's pawn structure) 1 6.i.f4 and Black has terri­ ble problems to finish his development : 1 6 . . . d6 1 7 .i.xd6 ! �d8 1 8 .i.xe 7 llxd3 1 9 J:txd3 �f7 2 0 .�a3 and Black's lIa8 and �c8 remain out of play. White soon won in Ljubojevic-Ermenkov, Malta Olympiad 1 9 8 0 ; A2) 8 . . . 4Jxc3 9.i.d2 (the position is about equal after 9 . 4JxaS 4Je4+ 1 0 . i.d2 i.xd2 + 1 1 . iYxd 2 4Jxd2 1 2 . �xd2 4Jc6 1 3 . 4Jb3 d6 ! (better than 1 3 . . . b6 1 4.i.g2 i.a6 (or 1 4 . . . i.b 7 l S . cS) I S J:tac 1 �c8 1 6 .cS ! r:i;e7 (not 1 6 . . . 4JeS 1 7 . cxb 6 4Jc4+ 1 8 J :txc4 .ixc4 1 9 . �b 7 ! and White wins) 1 7 . nc3 4JeS 1 8 . ldhc 1 and White was better in Wang Yue-Berg, Gibraltar 2008) 1 4.i.g2 i.d7 I S .!:thd 1 �e 7 )

A2 1 ) 1 3 .�e3 dS 1 4.0-0 d4 I S .�d3 eS 1 6 .a3 g6 1 7 . b4 'ifc 7 1 8 .cS . White has some initiative on the queenside and a small advantage, as in Cvitan-P. Nikolic, Sarajevo 1 9 8 7 ; A2 2 ) Putting the white queen on a more active - and more exposed - post with 1 3 .iYf4 gets White in trouble after 1 3 . . . eS ! 1 4. 'iYh4 d6 I S .a3 i.e6. White is pinned along the as -e 1 diagonal and has to spend a number of tempi to pre­ pare castling : 1 6 J ic 1 �ad8 1 7 .l:t.c3 dS 1 8 .4Jb3 �b6 1 9 .cxdS 1:IxdS ! and Black was better in Ljubojevic-Velimirovic, Yugoslav Championship, Vrbas 1 9 8 2 . B ) Another, less frequently played move here is 8 . 4Jc2 , involving a pawn sacrifice.

analysis diagram

69

C h apter

Winning Chess Middleg ames

8 . . . ttJxc3 (Black takes the pawn, which is a risky decision here. Correct was 8 . . . �xc3 + ! 9 .bxc3 ttJcS with approxi­ mate equality) 9 . ttJxb4 ttJxa2 1 0 J:txa2 'tWxb4+ 1 l .�d2 'ifb6 1 2 .�g2 (with a bishop pair and well-coordinated pieces in an open position, White has ample compensation for the sacrificed pawn) 1 2 . . . ttJc6 (it was better to bring the king into safety with 1 2 . . . 0 - 0 and, in case of 1 3 .b4, try to complete devel­ opment with 1 3 . . . d6 1 4. 0 - 0 ttJc6) 1 3. 'iY a3 (keeping the black king in the middle) 1 3 . . . ttJd4. Now White has to build on his initia­ tive and he correctly offers another pawn : 1 4. 0 - 0 ! .

analysis diagram

B 1 ) 1 4 . . . d6 ? ! I S .e3 ttJb3 1 6 .�c3 0 - 0 1 7 Jdd 1 (due t o the terrible weakness of d6 Black is lost here) 1 7 . . . ttJcS ( 1 7 . . Jld8 loses to 1 8 .cS ! ) 1 8 .b4 ttJd7 1 9 .1:iad2 and White had a massive ad­ vantage and soon won in Beliavsky­ Alburt, Soviet Championship, Lenin­ grad 1 9 74; B2) It was better for Black to be con­ sistent and snatch another pawn with 1 4 . . . ttJxe 2 + I S .�h 1 ttJd4 (it is obvious that White has enough compensation, but it's not easy to find more) 1 6 .�e3 eS 1 7 .l:td l d6 1 8 .'ifc3 70

9.

1 0.

...

d7-d6

�f1 -g2

e6-e5

analysis diagram

1 8 . . . 'ifc7 ! (an important move ; 1 8 . . . 0 - 0 ? loses to 1 9 J :txd4 ! exd4 2 0 .�xd4 'ifc7 2 1 .�xg 7) and now: B 2 1 ) The exchange sacrifice could still be tried, but the consequences are no longer clear: 1 9 . I:txd4 exd4 2 0 .�xd4 �e6 ! 2 l .b 3 f6 2 2 . �e2 �f7 and in order to keep the initiative, White probably has to sacrifice another exchange : 2 3 . ldxe 6 ! �xe 6 24.�dS + �e 7 2 S .'iVf3 . In this sharp position White should have enough compensa­ tion for his double exchange sacrifice, but probably not more. Other moves are not sufficient: 2 3 .'iff3 l:Ihe 8 2 4 .'ti'hS + Wg8 2 S . �e4 h6 2 6 .'ifg6 cbf8 ; or 2 3 .�e4 'ifd7 ! 24.'iff3 �g4 2 S .�dS + �f8 2 6 .'iff4 �xe2 2 7 .�xf6 �hS ! and Black wins. B22) 1 9 .�xd4 exd4 2 0 .�xd4 0 - 0 2 l . b3 I:re8 ! and Black should b e able to hold: 2 2 . 11xa7 l:ixa 7 2 3 .'ifxa7 'iVcs ; B3) 1 4 . . . 'ifb3 ?! i s not good: I S .'ifxb3 ttJxb3 1 6 .�c3 and now 1 6 .. .f6?? blun­ ders a piece: 1 7 .�a3 ttJcS 1 8 .b4. 8.

b2xc3

9.

'iVd3-f3

tDe4-c5

A position similar to the one in Ljubojevic-Ermenkov has been reached. White has pressure and the potential weakness of his doubled c-pawns is of no importance for the time being.

11.

'iff3-e3!?

A fresh Topalov idea, trying to improve on two other logical moves. A) 1 1 .ttJfs is one; after 1 1 . . . 'iYf6 1 2 .ttJe3 'iYxf3 there can follow: A 1 ) 1 3 .�xf3 �e6 1 4.�a3 ttJba6 1 S . 0-0-0 We7

analysis diagram

1 6 .�g2 ! (preparing a well-known plan to increase the pressure on Black's posi­ tion) 1 6 . . .l:!ad8 1 7 . f4 f6 1 8 . �hfI b6 1 9 . 94 ttJc7 2 0 .h4 and White had an initiative in Kramnik-Anand, Monaco blind 2 0 0 7 ; A2 ) Another interesting idea is 13 .exf3 ! ? - the reader should note that due to the good control of central squares and his pressure along the d-file White is in most of these lines not

1:

Doubled P awns

afraid to allow a doubling of his pawns : 1 3 . . . ttJba6 1 4. �a3 �e6 I S . 0 - 0 - 0 0 - 0 - 0 1 6 . f4 exf4 (the immediate 1 6 . . . fS !, achieving a similar aim but not allowing g-file pressure, looks like a better move) 1 7 . gxf4 fS 1 8 .h4 I;lhe8 1 9 .hS �c7 2 0 . 11hg l and due to his pressure along the d- and g-files and good control of the dS-square, White had an advantage in Ivanchuk-Asrian, Sochi 2 00 7 . B) Aronian also has experience on the white side of this line and he has tried l l .ttJb3 ttJba6 1 2 .�a3 'ifc 7 1 3 . �d l �e6 1 4. ttJxcS ttJxcS 1 S .�xcS dxcS (later on black players realized that creat­ ing a pawn structure similar to Game I S (Botvinnik-Chekhover) is not advisable and improved with I S . . . 'iVxcS ! 1 6 .�xb 7 �c8 1 7 .�dS �xdS 1 8 .'iYxdS 'iYxc4 with equality, Moiseenko­ Gyimesi, Maalot 2 0 0 8) 1 6 .0-0 ! (refrain­ ing from 1 6.�xb7 "iVxb 7 1 7 .�xb7 ldb8 1 8 .�dS We 7 1 9 .f4 I1b2 , after which a draw would be a likely outcome) 1 6 . . Jlb8 1 7 .'iYhS ! (an important move, achieving the strategic objective to build on his control of the dS -square) 1 7 . . . 0-0 ( 1 7 . . . �xc4? is bad due to 1 8 .'iVg4) 1 8 .�dS btfe8 1 9 .�xe6 �xe6 2 0 . �dS lde7 2 l .ldfd l �be8 2 2 .'iVg S .

analysis diagram

71

Chapter

Winning Chess Middlegames

White has d-file control and a dominant piece (this time his rook) excellently placed on the dS-square. Due to these two assets he had a long-term advantage and went on to win in Aronian-P.H.Niel­ sen, Turin Olympiad 2 0 0 6 . 11.

...

0-0

1 2.

ttJd4-b3

'iVd8-c7

to build on his dS-square dominance, like in Aronian-Nielsen; 1 6 . . . �h3 (try­ ing to prevent White from castling) 1 7 .f3 and this looks better for White. 1 3.

ttJb3xc5

d6xc5

1 4.

0-0

ttJb8-d 7

Not an easy moment for Black. The text allows Topalov to create the aforemen­ tioned favourable pawn structure (see Botvinnik-Chekhover or Aronian-Niel­ sen) . The other way to reach a similar pawn structure, but in a slightly differ­ ent version, was 1 2 . . . tLlba6, and now after 1 3 .�a3 'fic7 White has a few log­ ical plans at his disposal :

1 7.

1 5.

f2-f41

Well played . Immediate action is needed, so White opens up the position in order to activate his dark-squared bishop as well . The stereotypical I s Jid l does not promise anything, since Black gets a good game after I S . . . �b8 with . . . b 6 or . . . tLlf6 to follow. 1 5.

analysis diagram

A) 1 4.f4 l:ie8 l S . 0 - 0 �fS 1 6 .l:I ad 1 and White has pressure here ; B) A slightly different version with 1 4. 0-0 �e6 I S .f4 f6 looks OK for Black; C) After the well-known plan 1 4 .�xcS tLlxcs 1 s. tLlxcs dxcS (due to his d-file control and strong bishop White is better in the case of l S . . . 'fixcs 1 6 .'iYxcS dxcS 1 7 .0-0-0 ! fS 1 8 .�dS + �h8 1 9 .f4) 1 6 .JtdS (wrong is 1 6 . 0 - 0 ? �e6 1 7 .r:tab 1 llab8 and White has missed his chance to obtain a dominant piece on the dS outpost) and White tries 72

lld8 2 3 .�b4 and White is a sound pawn up.

...

e5xf4

Trying to keep the position relatively closed with I S . . . tLlb6 would not help. White is better after 1 6 .fS ! tLlxc4 (or 1 6 . . . f6 ? 1 7 . �a3 ) 1 7 .'iYe4 tLld6 ( 1 7 ... tLlb6 ? 1 8 .f6 gives White a crushing attack) 1 8 .'iYxeS �e8 1 9 .'iYdS . 1 6.

'iYe3-e71!

Domination. White has the initiative, but he has to play with a lot of power and precision. Black needs just a small break in order to develop his queenside and then White's seemingly strong ini­ tiative may evaporate. Topalov's plan is by far the best and the most exact, al­ though in my opinion White also had another way to get an advantage : 1 7 .�a3 n a 6 and now : A) Not convincing is 1 8 . 11e4. A I ) Now 1 8 . . . 'iYd6 is not good due to 1 9 JU1 'iYh6 and now: A l l ) 2 0 .'li'xh6 llxh6 2 1 .l:le7 b6 and now: A l I I ) It is never too late to get into trouble - even with a logical move : 22 Jld l ?

l::tf 1 xf4

Black has terrible trouble developing his queenside, something he will not manage to solve throughout the game. 1 6.

...

a 7-a5

Black is trying to bring the a8 rook into play and thus solve his queenside devel­ opment problems. 1 6 . . . tLleS looks logi­ cal , but does not help after 1 7 .�a3 tLlg6 1 8 J iffl �e6 1 9 .'iYxcS l:Ifc8 2 0 .'iYxc7 l::l x c7 2 1 .�d6 �xc4 2 2 .�xb7

analysis diagram

1:

Doubled Pawns

2 2 . . Jle6 ! ! 2 3 . Ildxd7 (2 3 . ttxe6 fxe6 2 4 J ld 6 tLleS) 23 . . . �xd7 24.I:rxd 7 ldxe2 and Black takes on a2 , pins White on the first rank and will push his own ed a-pawn very fast; A 1 1 2) 2 2 .�dS tLlf6 2 3 .�c 1 tLlxdS 24.cxdS ttd6 2 S .e4 f6 2 6 .Jte3 l:Id7 2 7 . �xd7 Jtxd7 2 8 . 11b 1 bS 2 9 .�xcS �e8 3 0 .!ie l �fS 3 1 .d6 �f7 and Black should survive. A 1 2) 2 0 .�xc S ! (this simple and di­ rect move promises more than line A l l ) 2 0 . . . tLlxcs 2 1 . 'iYxcS iYb6 2 2 .iYd4 Jte6 2 3 .cS . A2) 1 8 .. J::tf6 ! 1 9 .ne7 'iVd6 2 0 .�dS b 6 ! (neutralizing the a3 bishop) 2 1 J�d l Wh8 2 2 J:td3

analysis diagram

2 2 . . J ie6 ! 2 3 .nxe6 fxe 6 2 4.�xe6 'iVe7 (Black's pieces are coordinating well, while White 's pawn structure deficiencies are starting to be felt) 2 S .�xd 7 'iYxe 3 + 2 6 . �xe3 �xd7 and the position is about equal. B) 1 8 . �dS ! nf6 ! 1 9 . nafl (Black holds in the case of 1 9 .l:te4 'if d6 2 0 Jld 1 Wh8 2 1 . l:id2 (or 2 1 .�c 1 tLlb6 2 2 . �e 8 �d7 2 3 . �xf8 + iYxf8 ) 2 1 . .. 'iVb6 ! 2 2 .�b2 'iVd6 2 3 . .I:td2 iYb 6) 1 9 . . . 'iVd8 2 0 .�xcS l::t e 8 2 1 .'iVf2 tLlxcs 2 2 . iYxcS �xf4 (a blunder is 2 2 . . J �xe2 ? ? due to the simple 73

Chapter

Winning Chess Middlega mes

2 3 .�xf7 + �xf7 24.'iYhS +) 2 3 . laxf4 �e6 . Black is a pawn down, but due to White 's damaged pawn structure White's chances to win or Black's to draw should be estimated at about so-so.

l t should b e noted that these positions are considerably more difficult to play than they may seem. Playing over the board one often gets the impression that various moves should be good enough to maintain the initiative. How­ ever, due to the potential pawn struc­ ture problems and the fact that very of­ ten one wrong piece swap may cause White's initiative to evaporate entirely, playing those positions with white re­ quires energy and precision. 1 7.

'iYc7-e5

...

Black has to exchange queens , since 1 7 .. J�a6 ? ? loses to 1 8 .�dS +-.

22.

x:te4-e7

23.

�e3-f4

Even though the queens have been swapped and there is relatively limited material left on the board, Black is still completely paralysed and his pieces lack coordination. 23.

...

24.

�g 2-e4

�g6xh5

30.

...

a5-a4

3 0 . . J lf8 loses an exchange after 3 1 .�d6 nf6 3 2 .�e8 ! IIfS 3 3 .�e 7 ttJf8 34. l::r d 8 �b 7 3 S .�g 6 .

..

ttJd7-f6

31 .

.

32.

�h5xf7+

�g8xf7

33.

�c7xb6

�c8-a6

34.

�b6xc5

e6-e5

35.

a2-a3

�a6xc4

36.

�c5-b4

e5-e4

37.

�g1 -f1

1 -0

h 7-h5

24 .�dS ttJf6 2 S .e4 h4 2 6 .�c7 hxg3 2 7 .hxg 3 also looks winning. 24.

Dou b led Pawns

Black is so helpless that White is not even in a hurry to take the exchange. This looks like a good moment for Black to resign, For some reason, Aronian plays a few more useless moves.

b7-b6

Structure 1 . 1 1

1::l g 6-e6

25.

�e7xe6

f7xe6

1 8.

'iYe7xe5

ttJd 7xe5

26.

�f4-d6!

1::lf8-f6

1 9.

�c1 -e3

ttJe5-d7

27.

1::t b 1 -d 1

�g8-f7

The 'active ' attempt 1 9 . . . gS simply loses a pawn to 2 0 . �e4 f6 2 1 .�xc5 .

31 .

2 1 . . J :Id6 loses to 2 2 .�f4 J:Ig6 (or 22 .. Jdb6 2 3 .nxb6 ttJxb 6 24.�d6+-) 2 3 . IIe 7 , while in the case of 2 1 . .. �d8 22 .IId I ! still works , for instance : 22 . . . IIf8 2 3 Jle 7 bIe6 24J�xe6 fxe6 2 S .�xb7 �xb 7 2 6 .nxd7 �a6 27 .�xcS �fS 2 8 J�c7 �xc4 2 9 .�d4 �xe 2 3 0 .IIxg 7 + <;t>f8 3 1 . !:lxh 7 .

1:

2 7 . . . e S ? ? blunders an exchange after 2 8 .�e 7 .

Nlmzo structure - White has an open file and a doubled e-pawn

20.

nf4-e4 !

The rook gets t o the seventh rank, fully dominating events. 20.

2 0 . . J :Id8 2 2 .�xcS . 21 .

74

na8-a6

...

loses �a 1 -b1 !

to

2 1 .�d l

l:ia6-g6

I:ta6

28.

�d6-f4 !

�f7-g8

29.

�f4-c7

lH6-f7?

This blunders an exchange. 29 . . . e5 was a better defence, though after 3 0 .�d8 ! l::r f8 3 1 .�d5 + \t>h7 3 2 .�c7 l::r f6 3 3 .h4 Black remains completely tied up. 30.

In the following game White execu tes a beautiful positional sacrifice. I have to it that I have played this type of position with white in a number of games, and never thought of this kind of sacrifice at all. White sacrifices the exchange in order to achieve harmonious development and a long-term initiative, based on his powerful centre and strong light-squared bishop on a2, which will become a mons ter. The open f-file and the pressure on j7 perfectly fit in this strategy. Once made, the sacrifice looks very logical. However, it is far from easy to think up over the board (former World Champion Sky did not find it either) and very useful to , because it can definitely be used as a standard sacrifice in this type of position.

�e4-g61

75

Chapter

Winning Chess Middlegames

NI 1 8 . 1 0 (E2 8)

GAME 1 8

Gregory Kaidanov Alexander Onischuk Chicago 2 0 0 2 (7) 1.

d 2-d4

ttJgS-fS

2.

c2-c4

e 7-eS

3.

ttJb1 -c3

�fS-b4

4.

e2-e3

0-0

S.

a2-a3

�b4xc3+

S.

b2xc3

d 7-dS

A sideline with a different concept compared to the standard 6 . . . cS . 7.

ttJg 1 -e2

eS-eS

S.

ttJe2-g3

l:tf8-e8

9.

�f1 -d3

eS-e4

1 0.

iiLd3-b1

b 7-bS

11.

f2-f3

�c8-aS

Now Black's concept becomes clear. White has built a strong pawn centre and will have an open f-file, while Black gets a strong bishop on c4 which dis­ turbs White's development, controlling the fl -square and preventing kingside castling for White. 1 2.

f3xe4

�aSxc4

1 3.

'iYd 1 -f3

ttJb8-d7

1 4.

na 1 -a21!

In order to exert pressure along the f-file, White needs a rook there and this is a beautiful solution to the problem. If Black takes the exchange, White will get a monster bishop and will be able to 76

castle a s well, achieving harmonious development. Should Black decide to ignore the white rook (which was probably a better option) , then White transfers his rook from a 2 to f2 , exert­ ing pressure along the f-file. The above-mentioned game by Sky continued in a rather original way: 1 4.a4 �c8 1 S S�(d2 ! ? (looking for an original solution to the problem. It is interesting that a player of Sky's cal­ ibre and talent failed to find 1 4Jia2 ! ! , even though he was obviously looking for original solutions) 1 S . . .'iVa6 1 6 .h4 1:Iad8 1 7 .hS h6 1 8 .l:rh4 (solving the �h 1 problem; however, �a 1 , �b 1 and �c 1 remain out of play) 1 8 . . .t2Jh7 1 9 .�c2 ttJgS 2 0 .'�'f2 .t:te6 2 1. .t!g4 ldde8 2 2 .�a3 ttJf6 2 3 . r!xg S (a forced sacrifice) 2 3 . . . hxg S 24.'iYfS cS ? (it is rather difficult to imagine why Black refrained from 24 . . . iVaS , simply de­ fending the pawn, with a clear advan­ tage) 2 S .�xgS (now White gets an at­ tack) 2 S . . . cxd4 (better was 2 S . . . ttJh7 , organizing his defences. This probably leads to a draw after 2 6 .'iV g4 ttJf6 (2 6 . . . �h8 can be dangerous for Black since White gets an attack after 2 7 . ttJfS g6 2 8 . hxg6 fxg6 2 9 .r!h 1 ! ) 2 7 .'iYgS ttJh 7) 2 6 . ttJfS ! dxe 3 + (the correct de­ fence consists of 26 . . . dxc3 + ! 2 7 . �e 1 (2 7 .�c 1 ? ttJxhS 2 8 .'iVxhS �d3 ! -+ since there is no 2 9 . \i(b2 ! ; 2 7 . \t>xc3 ? 'iVaS + 2 8 .�b4 ttJxe4+ 2 9 .�xe4 'ifeS + 3 0 . c\t>xc4 'ifxa 1 3 1 .�dS .tIe4+ 3 2 . �xe4 l:txe4+ 3 3 . ttJ d4 ll e S +) 27 . . . ttJxhS 2 8 .'ii'xhS g6! 29 .�gS �fl ! 3 0 .ttJd4 a:eS 3 1 .'iVf6 �d3 3 2 . �f2 �hS 3 3 .�b 3 �c4 3 4.�b4! �xb3 3 S .�xc3 'iYd3 3 6 .ttJe6 ! (3 6 . ttJxb3 is a draw after 3 6 . . . l:rfS + ! 3 7 .exfS �xe 3 + 3 8 .\i(fl �f4+ 3 9 .\t>g 1 �e3 + 4 0 . �h2 iYh6+

with perpetual check) 36 . . . iYxc3 ! (3 6 . . Jl.h7 3 7 J::t h 1 ! ) 3 7 . iVxc3 �xe6 3 8 .'iVc7 �aS 3 9 . 'iYxd6 and due to the weak dark squares around the black king, White has attacking potential and is clearly better) 2 7 . �c 1 ! ttJxhS (the only way to prevent mate) 2 8 .'ii xhS �d3 29 .�b2 ! and White soon won in the attack, Sky-Uusi, Rostov 1 9 5 8 . 14.

...

1:

DOllbled Pawns

Now the d2 bishop has an open field. White is attacking with all his pieces. 20.

...

eSxd4

21 .

ttJfSxg 7 !

ttJgS-eS

.ic4xa 2?!

This gives White easy play. Ignoring the rook and keeping the strong bishop with 1 4 . . . cS was better. 1 S.

�b1 xa 2

ttJd7-f8

1 S.

0-0

ttJf8-gS

1 7.

iiLc1 -d2

Despite being an exchange up, Black is completely tied up and forced to wage a ive defence. 1 7.

...

'iVd8-d7

1 8.

ttJg 3-fS !

'iVd7-d8?!

Black hangs on to his material, though it was probably better to give back the exchange with 1 8 . . . !lxe4 ! 1 9 .ttJxg 7 tt:Jf4! (the only move) 2 0 . �b 1 dS 2 1 .exf4 \i(xg 7 2 2 .fS �g8 .

The best defence. 2 1 . . . �xg7 ? would lose quickly to 2 2 . �g S . 22.

'iYf3-h3!

2 2 .'ifg3 ? allows 2 2 . . . ttJxe4 with tempo. 22.

...

d4xc3

Now, in the event of 2 2 . . . ttJxe4, White gets a winning attack after 2 3 . ttJxe8 'iYxe8 24. �h6 . 23.

�d 2xc3

24.

'iVh3-g3+


The power of the bishops. Black is pinned all over and completely lost. 24.

...

2S.

�c3xfS+

ttJeS-gS

It was not necessary to hurry with tak­ ing on f6 . The knight on f6 cannot run away, so 2 S .h4 ! was better, winning easily. Compared to the game White would be more or less a tempo up.

1 9.

e4-eSI!

With this nice pawn sacrifice White ac­ tivates his d2 bishop. It is quite possible that Black thought White had no better moves than retreating with 1 9 .ttJg3 . 1 9.

...

20.

e3-e4

d SxeS

2S.

...

2S.

l:tf1 xfS

�g7xfS

27.

h2-h4!

l:Ie8-eS

28.

'iYg3-f3+



29.

h4-hS

'ifd8xfS

Black is forced to give his knight. 29.

...

f7-fS

30.

hSxgS

h 7xgS

31 .

'iVf3-d3

l::t a 8-e8

77

Winning Ch ess Middl egam es

Trying to build a fortress. 32.

'iVd3-c4

l:IeS-e7

33.

'iVc4-c6

l:Ie7xe4

34.

'ifc6xc 7 +

Chapt e r

34 .

...

lleS-e7

Black has a difficult position because the g6 pawn is also weak. It could be, however, that 34 . . J�4e 7 was a better try.

EO 1 . 1 0 (A2 9) Vasily Ivanchuk M oscow ch-URS 1988 (2) c2-c4

'ifc7-bS

�g7-h6

2.

ttJb1 -c3

e7-eS

36.

'ii' bS-fS+

h6-gS

3.

ttJg 1 -f3

ttJbS-c6

37.

�a2-f7

37.

�e7-c7

3S.

�fS-g7

llc7-c1 +

39.

\tJg 1 -f2

l:tc1 -c2+

40.

<;;tf2-f3

l:Ie4-f4+

41 .

\tJf3-e3

1 -0

Structure 1 . 1 2

4.

g2-g3

�fS-b4

S.

�f1 -g 2

0-0

6.

0-0

This variation of the English Opening was one of Kasparov's main weapons in his 1 9 8 7 World Championship match with Karpov, and has later remained one of his favourite lines. In this line Black takes on c3 , creating doubled c-pawns for White. However, White usually gets central domination. More­ over, his doubled pawns control central squares and are not easy for Black to target.

6. English Opening structure - White has a doubled c-pawn and an open f-file

White can easily get into trouble when he overestimates the dynamic possibilities as­ sociated with his pawn centre. Very often when it turns out that there is not enough compensation for the sacrificed c4 pawn (this is mos tly the pawn that falls), White is left with a ruined position and loses without a real fight. Since giving a pawn for dynamics based on a strong pawn centre is by definition an intuitive sacrifice, this is a tricky business and even the very best often get los t in the treacherous waters. Here is one good example.

_

Doubl ed Pawns

J:[ad8 with a better game for Black in Kasparov-Anand, Geneva rapid 1 9 9 6 ; B ) 9 . ttJd4 h6 1 O .dxe4 ttJxe4 1 1 .'iVc2.

ttJgS-f6

1.

3S.

The g6 weakness decides.

78

GAME 1 9

Garry Kasparov

1:

eS-e4

Another logical possibility for Black which is often tried here is 6 . . . �e8 7 .d3 .ixc3 8 .bxc3 e4. Play may con­ tinue as follows : A) Less convincing is 9 . dxe4 ttJxe4 1 0 .'iYc2 d6 1 1 .ttJd4 'fie 7 1 2 J�d l ltd7 1 3 .f3 ttJc5 1 4.e4 and White's central pawns become vulnerable : 1 4 . . . ttJa5 I S .'iYe2 f5 ! (creating an extra white weakness on e4) 1 6 .lta3 fxe4 1 7 .fxe4

analysis diagram

Note the pros and cons of White's dou­ bled c-pawn here : on the one hand, they control central squares and help White to build up central domination, while on the other hand, the c4 pawn can become a very easy target. A) 1 1 . . . ttJc5 ? is wrong due to 1 2 .lta3 and Black cannot develop ; while 1 1 . .. 'iVe 7 ! ? , trying to follow the same idea as in Kasparov-Anand, is perhaps an improvement for Black here, since after 1 2 . ttJ b 5 d6 1 3 . ttJxc 7 'iYxc 7 1 4 .ltxe4 ttJe5 he gets his pawn back and has a good game. B) 1 1 . . . d5 (Black opts for quick de­ velopment and therefore allows White to get rid of his weak c4 pawn) 1 2 .cxd5 'iYxd5 1 3 .e 3 ttJa5 1 4.f3 ttJd6 1 5 .e4 'iYc5 1 6 .lte3 ttJdc4 1 7 .ltf2 and due to his bishop pair and central domination, White was better in the 1 6th World Championship match game Kasparov­ Karpov, Sevilla 1 9 8 7 . 7.

ttJf3-gS

�b4xc3

S.

b2xc3

ttfS-e8

White now has to eliminate Black's e4 pawn in order to take control of the centre. Then the doubled c-pawns will 79

Winning Ch ess Middl egam es

actually be quite useful for White in many lines. 9.

9.

f2-f3

..

.

e4xf3

Chapt er

white pawn: 1 2 .ttJf4 'iVxc4 1 3 .e4 d6 (White gets good compensation for the pawn in the case of 1 3 o o .'iVxc3 1 4.Jdb l d 6 I S .�b2 'it'cS + 1 6 .Wh l )

analysis diagram

The alternative is to keep the f-file closed and create a slightly strange pawn struc­ A) 1 4.'it'd 3 ttJeS I S .'iVxc4 ttJxc4 ture with doubled white e-pawns with 1 6 .g4! l:tbs 1 7 .l:rd 1 b6 I S .gS ttJd7 9 o o .e3 1 0 .dxe3 (another logical solution 1 9 .ttJdS �b7 2 0 .f4 and White was is to keep the white pawn structure in­ better, though Black eventually man­ tact with 1 0 .d3 . In that case, however, aged to hold in Kasparov-Sadvakasov, White will remain slightly behind in de­ Astana 2 0 0 0 . velopment and Black will open the posi­ This i s definitely a safe way t o achieve tion with 1 0 o o .dS , with sharp play and an advantage. However, I think that mutual chances, for example : 1 1 . iYb3 White would get more out of his initia­ ttJaS 1 2 .'ifa3 c6 1 3 .cxdS cxdS 1 4.f4 tive with queens on the board, taking a ttJc6 1 s J�Ib l 'ifc7 ( 1 S o o .�g4! ? deserves more aggressive approach : attention) 1 6 .�b2 �g4 1 7 .c4 ! ? (look­ B) 1 4. �h l ! ? ttJeS (or 1 4 . . . 'ifxc3 ing for a tactical solution; other options 1 5 Jib 1 ) and now 1 5 .g4! should be were 1 7 .ttJf3 or 1 7 .�fe 1 ) 1 7 o o .dxc4 considered. Black has to take the pawn l S .�xf6 gxf6 1 9 .ttJe4 Wg7 , as seen in on c3 now (otherwise White will con­ Kasparov-Karpov, 2nd match game, tinue undisturbed) , after which White Sevilla 1 9 S 7 . ) gets a very strong attack. One of the I assume that Kasparov had analysed lines illustrating White's attacking po­ deeply and come to the conclusion that tential goes : 1 S o o . �xc3 1 6 .�d2 'iVcs the creation of a slightly strange pawn 1 7 . g S ttJfd7 I S .ttJdS ttJb6 1 9 J:1c l structure (after 1 0 .dxe3) would actually ttJec4 2 0 .�b4 'iYc6 2 1 .'iVd4 'it'd 7 be quite favourable for White. As we will 2 2 .�xc4 cS 2 3 .�xcS ! ttJxc4 24.�b4 see, Black will indeed find it difficult to ttJ b 6 2 S .�c3 fS 2 6 . gxf6 ttJxdS fight the advance of the white pawn: 2 7 .'ifxdS + 'iVe6 2 S .'iVd2 and White 1 0 o o .'iVe7 I I .ttJh3 'iYcs and Black gets wins : 2S . . . g6 2 9 .f4 'iffl 3 0 .fS �d 7 his pawn back, but he cannot find 3 1 . �g l etc. counterplay against the advance of the 1 0. ttJg5xf3 d 7-d5 80

Black has a temporary advantage in de­ velopment and decides, probably cor­ rectly, to open up the position. It would be a mistake for Black to stick to a 'reg­ ular' strategy, keeping White's doubled pawn on c4 and, in general, trying to play closed-type positions. White has the open f-file, his pawns in the centre (c3 , c4, d2 , e2) ensure central domina­ tion and prevent Black from getting any counterplay, while the doubled c-pawns are actually quite useful here. Even a player of Anatoly Karpov's stature - and in his prime - misjudged the dom­ ination White achieves, associated pri­ marily with his central pawn structure, and he lost an important game without getting the slightest chance of counter­ play This highly instructive game contin­ ued: 1 0 o o .iYe7 I I .e3 ttJeS 1 2 .ttJd4 ttJd3 (taking a pawn with 1 2 o o .ttJxc4 is not ad­ visable, since White gets a strong attack after 1 3 .e4 d6 1 4.d3 cS l S .ttJc2 , with 1 6 .�gS to follow) 1 3 .'iYe2 ttJxc l 14Jhxc l d6 l S J::tf4 c6 1 6.�cf1 .

1:

Doubl ed Pawns

A) If 1 6 . . . �e6 , White can decide that the time is ripe to go in for the kill with 1 7 . �xf6 gxf6 I S . �e4 'iYfs (if I S . . . �xc4 ? , 1 9 . �xh 7 + ! mates) 1 9 . 'iYhS 'ifg 7 2 0 J �f4; B) 1 6 o o .'iYeS 1 7 . 'iYd3 �d7 (probably a better defence was 1 7 . . . 1lfS) 1 S . ttJfS �xfS 1 9 .1d:xfS 'iYe6 2 0 .'iVd4 lie 7 2 1 .'iYh4 (also good for White was 2 1 . �xf6 gxf6 2 2 . .rtxf6 'iVeS 2 3 J lxd6 'iYxd4 24.cxd4 and again, the excellent central pawn formation c4, d2 , d4, e3 secures domination for White) 2 1 o o .ttJd7 2 2 .�h3 ! (now Black will be forced to give his fl pawn and allow the white rooks to enter) 2 2 o o .ttJfs 2 3 .l:ISf3 'iVeS 2 4 . d4 'iVe4 2 S . 'it'xe4 l:Ixe4 2 6 .11xfl J:Ixe3

analysis diagram

analysis diagram

Take note of White's excellent pawn structure (c3 , c4, d2 , e3 ) , his dominant knight on d4, and his f-file pressure, while Black does not have any ideas for counterplay. Also, again, the doubled c-pawns are very useful for White here.

2 7 .dS (White also wins with 2 7 Jlxb 7 Iixc3 2 S . l:Ic 7 ! (not clear is 2 S . l:Iffl ? ldxc4 2 9 . �xg 7 + �hS) 2 S . . . J:Id3 (2 S o o Jlxc4 loses a piece after 2 9 .l:Icfl ! with 3 0 .11fs and 3 1 .�e6 to follow) 2 9 J �xc6 l:Ixd4 3 0 . �g2 - the bishop gets transferred to dS and the black position will collapse) 2 7 . . . l:IaeS (or 2 7 o o .cxdS 2 S .�g2 ! ) 2 S J :rxb 7 (2 S .cS ! dxcS 2 9 . d6 wins more qUickly) 2 S . . . cxdS 29 .cxdS and White soon won in Kasparov-Karpov, 4th match game, Sevilla 1 9 S 7 . SI

Chapter

Winning Chess Middlegames

11.

d2-d4!

Correctly judging that the ensuing complications will be in White's favour. Black would have achieved his strategic aim after l 1 .cxdS 'iYxdS 1 2 .tLld4 'iYhS and Black takes the initiative.

11.

. ..

strong initiative. It is quite possible that Black had underestimated, or missed, White's 1 4th move. 12.

'iVd 1 -c2!

Putting further pressure on the central squares. 1 2.

...

1 3.

na1 -b1 !

d5xc4

Continuing to develop calmly. 1 3 . tLleS does not get White anywhere after 1 3 . . . tLlxeS 1 4.'iYxe4 tLlg6 l S .'iYdS 'iYe7 . 1 3.

...

f7-f5

If I 3 . . . f6 , then 1 4 .tLleS ! fxeS I S .i1Lxe4.

tLJf6-e4

One of two logical replies. The other was 1 1 . . . dxc4. The problem for Black is that White then pins the black knight with 1 2 .�g S and after 1 2 . . . h6 (if 1 2 . . . �d6 , 1 3 .tLld2 ! with a strong ini­ tiative; on the other hand, 1 3 .�xf6 'iYxf6 1 4.tLleS is not clear here because Black has 1 4 . . . 'iYh6 ! ) 1 3 .�xf6 'iYxf6 1 4.tLleS and now : A) If 1 4 . . . 'iYe6 , White gets a better endgame after some sharp complica­ tions : I S .i1Lxc6 bxc6 1 6 .I:txf7 cS 1 7 .e3 cxd4 1 8 .exd4 ..tb7

analysis diagram

82

1 9 .'iVb l ! �e4 (if 1 9 . . . ..tdS , 2 0 . l::t xc 7 ; while 1 9 . . . 'iYdS ?? i s a horrible blunder due to 2 0 J :txg 7 + �xg 7 2 1 .'iYg6+) 2 0 . 'iYxe4 'iYxf7 2 1 . 'iYxa8 .a:xa 8 2 2 .tLlxf7 �xf7 2 3 . l:tb 1 �e6 24.11bS �d6 2 S .a4 and due to his dominant rook, White is better. The more positional 1 s . tLlxc6 bxc6 1 6 .e4 does not promise an advantage, for example : 1 6 . . . �b8 1 7 .'iYhS (if 1 7 . I:tf2 , Black secures counterplay with 1 7 . . . cS) 1 7 . . Jlb2 1 8 .I:tf4 and White should have compensation, but not more. B) 1 4 . . . 'iYd6 , and now White has two acceptable continuations, but no clear road to an advantage:

analysis diagram

B 1 ) l S . tLlxc6 bxc 6 1 6 . e4 ..te6 1 7 .'iYa4 'iYd7 ( 1 7 . . . f6 ? is a blunder due to 1 8 .eS fxeS 1 9 .�xc6) 1 8 .�ab 1 with compensation for the pawn, but proba­ bly not more. B2) I s .tLlxf7 'iYe 7 1 6 . 'iYc 2 ! �e6 1 7 .�xc6 bxc6 1 8 . tLleS . The dominant white knight compensates for the light­ square weaknesses related to the ab­ sence of White's bishop, but the positi­ on remains sharp. It could well be that 1 1 . . . dxc4 is the way to improve on the game, since after the text White quickly gets a very

1 5.

1 4.

...

'iYd8-e7

If 14 . . . fxg4, White gets a strong initia­ tive after 1 5 .tLleS , and now: A) In case of l S . . . tLlxeS 1 6.�xe4 tLlg6 , the follOwing line illustrates White's attacking potential : 1 7 . .txg 6 hxg 6 1 8 .'iYxg6 'iYd7 1 9 .dS ! b6 (or 1 9 . . . 'iYe7 20.�a3 'iYe3 + 2 1 J:rf2) 2 0 .�a3 ! �b7 2 1 .e4 cS 22JHS ! .tIxe4 2 3 Jlbfl �xdS 24J:thS and Black is mated;

g4xf5

tLJe4-d6

If I S . . . �xfS : A) Possible, but rather unclear is 1 6 .tLleS tLlxeS ( 1 6 . . . tLlxc3 ? ? is a blun­ der due to 1 7 .'iYxc3 ..txb 1 1 8 .iYxc4+ �h8 1 9 .tLlf7 + �g8 2 0 .�g S ) 1 7 J:txfs tLlg4 1 8 .�f4; B) 1 6 .tLlgS g6 (or 1 6 . . . �g6 1 7 .tLlxe4 �xe4 1 8 .�xe4 'iYxe4 1 9 .'iVxe4 �xe4 2 0 Jdxb 7 l::t c 8 2 1 .ldf2) 1 7 .tLlxe4 �xe4 1 8 . �xe4 'iYxe4 1 9 . �xe4 �xe4 2 0 . l::l xb7 .

analysis diagram

g3-g4!

Black now realized that the central post of his e4 knight cannot be maintained after all.

Doubled Pawns

B) I S . . . tLld6 1 6 . tLlxc6 bxc6 1 7 .e4 with promising play. In general, after 1 4.g4 ! , Black has to make difficult decisions and he is often forced to choose between a rock and a hard place.

It may seem that Black has secured con­ trol of the light squares and has a good central post for his knight on e4. Kasparov's energetic play, right until the end of the game, is highly instructive. 14.

1:

White is clearly better in this endgame, for instance : 20 . . . tLle7 2 1 . nxc7 nxe2 2 2 .a4 tLldS 2 3 .!:txc4 ldc2 24.!:tcS tLlxc3 2 S .�h 1 and White wins , because Black cannot solve the problem of his pinned knight. 1 6.

tLJf3-g5!

1 7.

�g 2-d5+

'if e 7xe2 <;;t g 8-h8

1 8.

�c2xe2

ne8xe2

1 9.

itc1 -f4

The queen swap has not helped Black and White's mating attack continues with unabated vigour. 83

Winn i n g C h ess M i d d l eg a m e s

remaining on the board, Black still can­ not avoid being mated. 21 . l::t e2xe 1 22.

1:lf1 xe1

.!tc8-d7

23.

.i::t e 1 -e7

.!td7-c6

Chapter 2 Isolated Pawns I ntrod uction In Chapter 2 I have followed a similar structure as in the chapter on doubled pawns ,

1 9.

with the difference that here I have divided the material into two sections. The first

ttJc6-d8

1 9 . ttJ e 7

loses

. .

to

2 0 . �f3

section shows several successfully employed strategies to fight against the isolated

�xa2

pawn , while the second section is focused on plans to be implemented by the side

2 1 .�xd6 cxd6 2 2 . l:tbe l ; or 1 9 ... �xf5 2 0 . �xd6

�xb 1

2 1 . ttJ f7 +

that has the isolated pawn.

�g8

Isolated pawn structures are arguably the structures that arise from the most dif­

2 2 .ttJd8+ h8 2 3 .1df8 mate. 20.

�f4xd6

21 .

nb1 -e1 !

ferent openings (Tarrasch Defence and Semi-Tarrasch, Queen's Gambit Accepted,

c7xd6 24.

f5-f6!

Queen's Gambit Declined, Nimzo-Indian Defence, Meran Variation, Ragozin Varia­

Exchanging Black's only active piece

Finishing the game in style. There can

tion, Petroff Defence , etc.) and are therefore very important positions to understand,

and entering the black position with

follow: 24 . . . �xd 5 2 5 J�te 8 + �g8 2 6 .f7

regardless of the opening preferences a player may have. In the comments to the

the rook. It is quite amazing that with

ttJxf7 2 7 . tLlxf7 mate.

games, a club player will be explained not only the main strategic ideas , but he will

the relatively limited number of pieces

Black resigned.

also receive a lot of information about the openings played. If the reader takes enough time to study this chapter, he should be well-armed to meet a number of variations in practical tournament play.

A) Playing aga inst an isolated pawn Structure 2. 1 (Game 20 - Ivanchuk-Aronian) . Here a

2.1

classical isolated pawn-position is reached, where the side fighting against the isolated pawn has a good block­ ade while the side with the isolated pawn has seemingly enough counterplay to make a draw. With masterful play, Ivanchuk first quashes Aronian's counterplay and then exploits the weakness of the isolated pawn. Structure 2.2 (Game 2 1 - Kramnik-Illescas Cordoba) . In

one of the main variations of the Tarrasch Defence, fight­ ing against an isolated pawn, Kramnik employs a well-known strategy of creating a second weakness to tar­ get. When the knights are exchanged on c 6 , Black recap­ tures with his b7 pawn and now, instead of an isolated pawn on d5 , the newly-created weakness on c6 is the ob­

ject of White's attention. In our main game (Kramnik84

85

Chapt e r

Winning Chess Middl egam es

Illescas Cordoba) and other games featured in the com­ ments (Kasparov-Illescas Cordoba, Sokolov-Petrosian and Ljuboj evic-Gligoric) , this classical strategy is ex­ plained. (Game 22 - Jaracz-PH.Nielsen) . In this game another strategy is employed. Fighting against an isolated pawn , White at some stage exchanges a black knight on e4 and after Black recaptures with . . . dSxe4, a relative pawn symmetry is reached, where White's pieces are better placed for the ensuing actions. Structure 2.3

2.3

2:

Isola t ed Pawns

Structure 2.6 (Game 25 - Kasparov-Karpov and Game 2 6 - Sokolov-Schandorff) . Here, the same central exchange transaction occurs as in the two previous games (Petrosian-Sky and Sokolov-Cebalo) , with the differ­ ence that now White has a knight as a minor piece and Black a light-squared bishop. In general, the knight is su­ perior here. Most of the time, Black's light-squared bishop combines badly with his own pawn on dS - a light square -, and the dark-squared bishops are already exchanged. Structure 2 . 7 (Game 2 7 - Beliavsky-Illescas Cordoba) shows a beautiful, original rook manoeuvre, executed by Illescas Cordoba playing with an isolated pawn.

B) Playi ng with an isolated pawn Structure 2.4 (Game 2 3 - Petrosian -Sky) . In this beautiful classic World Championship encounter, Sky, playing with an isolated pawn, at some stage ex­ changes knights on d4, and after Petrosian recaptures e3xd4, a pawn symmetry in the centre (white d4/black dS) is reached, which drastically changes the strategic objectives. In general, given the fact that both sides have a light-squared bishop remaining (the dark-squared ones have been exchanged) , such a transaction should in gen­ eral favour White. However, with the beautiful knight manoeuvre 3 2 . . . ttJh7 ! , Spas sky brilliantly shows the defi­ ciencies of the white position and wins the game in style.

(Game 24 - Sokolov-Cebalo) . In my own game I employed a similar strategy; at some stage ex­ changing knights on dS , and after the forced recapture . . . e6xdS , a pawn symmetry in the centre (white d4/black dS) is reached , where the superiority of White's light-squared bishop to his black colleague (the dark ones being exchanged) , combined with White's control of the e S -square, along with the kingside attack that is developed , plays a crucial role. Structure 2.S

86

Structure 2 . 8 (Game 2 8 - Vyzhmanavin-Beliavsky and Game 29 - Sokolov-Nikolic) : White, playing with an iso­ lated pawn, executes a d4-dS pawn break, after which a few pieces are exchanged. When a pawn symmetry is left on the board (two vs two on the queenside, three vs three on the kingside) , White will have a dominantly-placed piece on the dS -square, causing trouble for Black.

Game 30 (Kramnik-Hiibner) shows a slightly unortho­ dox and uncommon idea : White sacrifices his central (isolated) pawn in order to eliminate Black's bishop pair and gain a few tempi to develop an initiative. Structure 2.9 (Game 3 1 - Kasparov-Timman) . Here and in the following games we analyse an important strategic idea for White, where he sacrifices his isolated pawn un­ der different circumstances (mostly by playing d4-dS) , either to disorganize Black's pieces or to take advantage (should Black recapture with a pawn) of a beautiful newly-created outpost for the knight on d4, as well as the open e-file for the white rook on e 1 , which creates vari­ ous sacrificial motifs around taking the black bishop on e7 with the rook. In our first game (Kasparov-Timman) , the purpose is to take advantage of the disorganizationof Black's pieces caused by the sacrifice.

87

Winning Chess Middlegames

(Game 3 2 - Petrosian-Sky, Game 3 3 - Kramnik-Anand 1 9 9 9 and Game 3 4 - Kramnik-Anand 2 0 0 1 ) . These games excellently show White's attacking potential should Black capture the pawn sacrificed on d5 with his e6 pawn. Structure 2 . 1 0

Game 3 5 (Illescas Cordoba-Short) again shows the power of the pieces after a central break with the isolated pawn, this time by Black with . . . d 5 -d4. An excellent learning example demonstrated by Short. In the com­ ments to this game and the analysis diagram (the game Karpov-Kir.Georgiev in the comments) , please observe the powerful strategy of the 1 2 th World Champion, showing that after a potential exchange of pawns on the queenside in a particular variation of the Tartakower Vari­ ation of the Orthodox Queen's Gambit, White's a4 and d4 pawns, which also seem weak, are not much of a rea­ son to worry, while Black's pawns on b6 and c6 could worry him for a long time to come.

Chapter

2.10

Isola ted Pawns

A) Playing against an isolated pawn Structure 2. 1

Pure blockade

Exploring isolated pawn positions, I would like to start with several instructive ex­ amples of play agains t the isolated pawn. The Tarrasch Defence is a perfect variation to study this type of position. As we will see from many examples in this book, former World Champion Vladimir Kramnik has played many excellent games against as well as with the isolated pawn. Given the fact that such positions occur frequently, it is worth spending a considerable amount of time on them to learn the ins and outs.

NI 2 7 . ? (D3 8)

GAME 2 0

Vasily I vanchuk Levon Aronian Morelia/Linares 2 0 0 7 (7) 1.

d2-d4

ttJg8-f6

2.

c2-c4

e7-e6

3.

ttJg 1 -f3

d 7-d5

4.

ttJb1 -c3

�f8-b4

S.

�c1 -gS

ttJb8-d7

6.

c4xd5

e6xdS

7.

'iUd 1 -c2

Starting somewhere in the beginning of the nineties, this move became more popular than the old main line 7 .e3 . 7.

...

c7-cS

The other, less played but rather inter­ esting continuation here is 7 . . . h6 and 88

2:

after 8 . �h4 g 5 ! (the correct reaction; in case of 8 . . . c5 White gets a better game after 9 .e3 c4 1 0 .�e2 'iVaS 1 1 . 0-0 ( 1 1 .ttJd2 0 - 0 1 2 .�f3 should also be better for White) 1 1 . . . �xc3 1 2 .bxc3 ttJe4 (here it is obvious that the inclu­ sion of 7 . . . h6 8 .�h4 favours White, as otherwise the bishop on g 5 would be attacked now) 1 3 .�fc 1 ttJb6. White now gave an instructive demonstration: 1 4.a4! �J5 1 5 .'iVb2 0 - 0 1 6 .�d 1 ! �fe8 1 7 .�a2 f6 1 8 .ttJd2 �d7 1 9 .tt:Jfl ! with f3 to follow, with a clear advantage for White in the second match game Kramnik-Lautier, Cannes 1 9 9 3 ) 9 . �g3 ttJe4 1 0 . ttJd2 (the natural 1 0 .e3 leads to very complicated play after 1 0 . . . h5 ! 89

Chapter

Winning Chess Middlegames

I I .h4 g4 1 2 .ttJgS ttJxg 3 1 3 .fxg3 'ife 7 1 4. 0 - 0 - 0 ! i.xc3 1 S .bxc3 ldh6 ! 1 6 .i.d3 f6 1 7 .ttJh7 , as seen in Sokolov-Rainfray, tt 2 0 0 3 ) 1 0 . . . ttJxg 3 I I .hxg3 ttJb6 1 2 .a3 i.f8 ! and Black had an equal game in Van Wely-Elianov, Foros 2 0 0 7 .

1 6 J:ihe 1 l:the8 and Black is better; B2) 9 . i.d2 0-0 1 0 .a3 i.xcs l l .e3 a6 1 2 .i.e2 i.d6 1 3 .0-0 ttJeS 1 4. llfd l i.e6 I S . l:: ac l �c8 1 6 .'iVb l ttJc4 1 7 .ttJd4 i.eS and due to his nicely developed pieces Black had a slight initiative in Aronian-Lautier, ICC Petrosian Memo­ rial 2 0 04. S.

...

9.

'iYc2xc3

.,ib4xc3+

White has to take with the queen, since after the pawn capture 9. bxc3 ?! comes 9 . . . 'ifaS with . . . ttJe4 to follow. 9. 1 0.

S.

a2-a3

It is already clear that a position with an isolated pawn is likely to arise. White could have forced this immediately with 8 .dxcS . A) Now, rather ive and not in the spirit of the position is 8 . . . ttJxcS , since after the virtually forced 9 .a3 i.xc3 + 1 0 .�xc3 ttJce4 1 1 . i.xf6 'iYxf6 1 2 .'iYxf6 ttJxf6 1 3 .e3 i.d7 1 4.ttJd4 White has a small but lasting advantage, pressing for a win without any risk; B) However, Black has good piece play to compensate for the pawn weak­ ness after 8 . . . h6 and now: B 1 ) Enterprising, but no good is 9 .i.xf6 'iYxf6 1 0 . 0- 0 - 0 i.xc3 1 l .'iVxc3 'iYxc3 + 1 2 .bxc3 ttJf6 1 3 .e4? ! dxe4 ( 1 3 . . . ttJxe4 would allow White to jus­ tify his idea after 1 4 . .t'1xd S i.e6 I S .i.bS + � e 7 1 6 . �eS ! ttJxf2 (or 1 6 . . . ttJxc3 1 7 .i.c4) 1 7 . :ahe l ttJg4 1 8 .� S e2 l::th c8 1 9 .c6 ! bxc6 2 0 .i.a6 .tIe8 2 l .ttJd4 and White is better) 1 4.i.b S + i.d 7 ( 1 4 . . . � e 7 I S . ttJd4 looks better for White) 1 S .ttJd4 O-O-O ! 90

...

h 7-h6

iiLg5xf6

Trying to keep the bishop with 1 0 .i.h4? is simply bad after 1 0 . . . gS 1 1 .i.g3 ttJe4. 1 0.

...

11.

e2-e3

'iV dSxf6

The other option is to play against an isolated pawn in the ending after 1 1 .'iYe3 + 'iVe 7 1 2 .'iYxe 7 + rJi;xe 7 1 3 .dxcS ttJxcS 1 4.l:k l . In the stem game Van Wely-Piket, Eindhoven ch-NED 1 9 9 3 , White could hope for an edge after 1 4 . . . ttJe4 1 S . e3 i.d 7 1 6 .i.d3 . The black knight, how­ ever, is better placed on e6 and with 1 4 . . . ttJe6 instead of 1 4 . . . ttJe4, control­ ling the d4-square, Black gains easy equality. 11.

0-0

1 2.

�f1 -e2

White plans to finish his development and then, at a favourable moment , to capture on cS , creating an isolated pawn. Forcing Black to make a decision in the centre with 1 2 .i.bS ? ! may look positionally sound, but Black gets ex­ cellent play on the queenside after 1 2 . . . c4 ! 1 3 .i.xd7 (if 1 3 . 0 - 0 ? , then 1 3 . . . ttJb6 with . . . a7 -a6 to follow, and the white bishop gets stuck) 1 3 . . . i.xd 7 1 4.ttJeS i.fS I S .b3 bS 1 6 . 0 - 0 iV e 6 and after 1 7 . . . f6 the pride of White's positi­ on, the eS knight, is kicked back, while Black has all the trumps. 1 2.

...

b7-b6

ClOSing the centre with 1 2 . . . c4 is now different, since White does not have to allow Black to advance his queens ide pawns. White's chances are to be pre­ ferred after 1 3 . 0 - 0 b6 (or 1 3 . . . l::t e 8 1 4.a4 b6 1 S .b 3 ) 1 4.b3 . 1 3.

0-0

.,icS-b7

In Yakovich-Sargissian , Moscow Aeroflot 2 0 0 7 , Black prepared to advance his queenside pawns with 1 3 . . . a6 in order, after 1 4 .llac 1 , to push 1 4 . . . c4. A principled battle ensued, both sides having their trumps: I S .ttJeS bS 1 6 .f4 'iYd6 1 7 .i.f3 i.b 7 . White has to be qUick with his kingside action, before Black starts rolling his pawns on the queenside. In such positions, which can arise from different openings, the choice is often a matter of taste. I have played plenty of similar positions and mostly preferred White. 1 8 .g4 �fe8 1 9 .1dce l ttJf8 2 0 .g S ! hxgS 2 1 .fxgS and White seized the initiative. 14.

J::tf 1 -c1

llaS-cS

Sensing that White plans to capture on cS , creating an isolated pawn, Black prepares counterplay along the c-file.

2:

Isola ted Pawns

1 5.

d4xc5

'ii'f6xc3

1 6.

�c1 xc3

�cSxc5

It seems that all four rooks are soon go­ ing to be exchanged along the c-file, re­ sulting in an easy draw for Black. But as we will soon see, the c-file is not that important here. Opting for two hanging pawns in the centre with 1 6 . . . bxc S ? would not be wise here, since White can undermine these pawns with the standard 1 7 .b4! c4 1 8 .ttJd4 and with a dominant knight and better pawn structure, White has a massive, probably winning advantage. 1 7.

J::tc 3-c1 !!

A beautiful move. White keeps the rooks on in order to target the weak isolated pawn on dS . For his part, Black cannot create any counterplay related to his control of the c-file. 1 7.

1:1fS-cS

1 S.

J::tc 1 - d 1

1 9.

�e2-b5!

l:tc5-c2

An important tempo. 1 9.

...

20.

J::ta 1 -b1

ttJd7-fS

On the next move, the black rook will be kicked out. 20.

Mc2-c7

21 .

�b5-a4

ttJfS-e6

22.

.,ia4-b3

�gS-fS

23.

h2-h3!

91

Chapter

Winning Chess Middlegames

...

�c7-c5

30 .

...

.l:tdS-cS?

3 0 . . . ttJcS looks like the best defence, but does not solve the problems, since after 3 1 . e4 ttJ b 3 3 2 . .txb 3 nxb 3 3 3 .exdS �d6 3 4 . ttJ d 3 1:tbS (or 34 .. . Iie8 3 S . ttJc 1 ! �bS 3 6 . ttJ a 2 ) 3 S . ttJf4 Black has a hard time ahead. 31 .

It seems that White will not be able to improve his position, and Black will achieve a draw after all. The technique that Ivanchuk now displays to convert his small advantage into a full point is an excellent learning example for ama­ teurs and grandmasters alike! 24.

�g 1 -h2!!

Stepping away from a check on c 1 , in order to double the rooks along the d-file. It is very important for White to keep all four rooks on the board.

Creation of a new weakness

ttJe1 -d3!

Threatening to trap the bS rook with 3 2.a4, so Black has to give a pawn. 31 .

...

d5-d4

32 .

.ia2xeS

�e7xeS

33.

ttJd3-f4+

�eS-e7

34.

lld 2xd4

White is now a sound pawn up, with the better position besides. The game is decided - the rest is agony for Black. 34.

...

.tl:cS-c7

TD 4. 1 6 (D3 4)

GAME 2 1

Vladimir Kramnik Miguel Illescas Cordoba Linares 1 994 (6) e7-eS

1.

d2-d4

2.

c2-c4

d 7-d5

3.

ttJg1 -f3

c7-c5

4.

c4xd5

eSxd5

5.

ttJb1 -c3

ttJbS-cS

S.

g2-g3

ttJgS-fS

35.

�d1 -d2

l:Ib5-c5

7.

..tf1 -g2

�fS-e7

3S.

e3-e4

l:rc5-c4

S.

0-0

0-0

37.

lld4-dS

l:Ic4-cS

9.

�c1 -g5

...

�fS-e7

25.

ll d 1 -d2

l:tc5-b5

3S.

e4-e5

llcS-c2

2S.

�b3-a2

l:tb5-c5

39.

lld 2xc2

l::t c7xc2

27.

ttJf3-e 1 !

a7-a5

40.

nd6xbS

�b7-cS

41 .

b2-b4

g 7-g5

9.

...

c5xd4

42.

ttJf4-h5

a5xb4

1 0.

ttJf3xd4

h 7-hS

43.

a3xb4

�cS-d5

11.

�g5-e3

l:IfS-eS

44.

ttJh5-g7

a.c2-e2

1 2.

lla 1 -c1

45.

ttJg 7-f5+

�e7-eS

4S.

ttJf5xhS

�d5-eS

24.

The attempt to become active with 2 7 . . . �c 1 would not help Black after 2 8 .Iixc l 1:txc l 2 9 .ttJd3 �a l 3 0 . ttJb4 d4 (the dS pawn will be lost anyhow) 3 1 .exd4 ttJf4 3 2 .dS �d6 3 3 . .tb3 1:te 1 34.ttJc6 ! . 2S.

llb1 -d1

l:tcS-dS

47.

llbS-b5

l:re2-b2

2 9.

�h2-g3

J:1c5-b5

4S.

llb5-bS+

�eS-d7

30.

f2-f3

49.

l:.bS-gS

1 -0

92

Isola ted Pawns

Structure 2 . 2

White wants to take the dS pawn on his own . Taking it immediately would result in a drawish endgame : 2 3 . .txd S ? ! �xd S 2 4 . �xdS llc l + 2 s J�1d l .tlxb l 2 6 J:lxb l l::t c 2 and due to his active rook, Black should hold. 23.

2:

One of the two main moves here, 9.dxcS .txcS 1 O .�gS being the other option.

This is one of many moves here, played very frequently in the beginning of the 1 9 80s and enjoying new popularity lately. It should be noted that the pawn structure can easily transform.

White may decide to take on c6 and play against the two connected pawns, targeting the newly-created weakness on c6 or, should Black decide to put his bishop on e 6 , White would take with the knight and then try to take profit from the bishop pair. 1 2.

...

�e7-fS

The other principled way to play this po­ sition is 1 2 . . . �g4 and after 1 3 .h3 �e6, White normally does not capture the bishop immediately (which would im­ prove Black's pawn structure) , but play some useful moves like 1 4.�h2 'ti'd7 93

Chapter

Winning Chess Middlegames

1 S . iYb3 (trying to play against the weak c-pawn with I S .tLlxc6 bxc6 1 6 .tLla4 does not bring anything : 1 6 . . . i�JS 1 7 .�cS �d8 ! 1 8 .�d4 tLle4, Ziiger­ Kasparov, Zurich 1 9 8 7 ) I S . . J:1�ac8 1 6.11fd 1 , with an advantage for White. In Van Wely -Magomedov, Yerevan Olympiad 1 9 9 6 , Black now blundered with 1 6 . . . tLleS ? ? , which White failed to punish with 1 7 . tLlxe 6 (he played 1 7 .'iYbS ? and was slightly better, but only drew in the end) 1 7 . . . fxe6 1 8. tLlxdS ! ! exdS 1 9 . 11xdS tLlxdS 2 0 .�xdS + �h8 2 1 .�e6, winning.

1 3.

tbc3-a4

White decides to remain focused on the dS weakness. The other way was to take on c6 in order to create and focus on a new weakness for Black. This used to be the main plan here - and it looks rather logical to me as well. I will give some examples : 1 3 .tLlxc6 bxc6 1 4.tLla4 and now: A) The attempt to create counterplay with 1 4 . . . 'iYaS ? does not work since the tactics favour White after 1 5 Jixc6 ! �d 7 1 6 .�d2 ! ( 1 6 . 11xf6 �xa4 1 7 .b3 gxf6 1 8 .bxa4 can also be considered) 1 6 . . . iYbS ( 1 6 . . . �b4 Ieads to a very dif­ ficult endgame for Black after 1 7 . ncS 'iYxa4 1 8 .'ifxa4 �xa4 1 9 .�xb4 llxe2 2 0 .�aS ! �d 7 2 l .�c 3 ) 1 7 Jlxf6 'iYxa4 94

(in case of 1 7 . . . gxf6 1 8 .tLlc3 'iYxb2 1 9 . tLlxdS White soon gets his minimal material investment back, obtaining a winning advantage) 1 8 .'iYxa4 �xa4 1 9 . 1:IfS and White is a sound pawn up ; B) 1 4 . . . �d7 I S .�cS �xcS 1 6 .tLlxcS . White's strategy is rather clear. The dark-squared bishops have been ex­ changed, the knight on cS is a strong, dominant piece and the black pawn on c6 is a newly-created weakness. Black has to create counterplay in order to achieve a dynamic balance. However, without the dark-squared bishops and with the white knight so well placed on cS , counterplay is not easy to find. It should also be mentioned that almost all further piece exchanges favour White : 1 6 . . . �g4 1 7 . .tIe 1 and now : B 1 ) The stem game in this line, Lj uboj evic-Gligoric , Bugojno 1 9 7 8 , continued 1 7 . . . 'li'b6 1 8 .'iVc2 lIad8 1 9 .h3 �c8 2 0 .b3 iYb8 2 l .e3 lle7 2 2 .ned l (with his firm control over the dark squares, White easily improves his position, while for Black counterplay is nowhere to be found) 22 . . . llde8 2 3 .l:!d4 tLlh7 24.iVd2 tLlgS 2 S .h4 tLle6 2 6 . tLlxe6 l:Ixe6 2 7 .'iYb4 'iYb6 2 8 .'iVcS and White continued to build on his dark-square domination and won easily; B2) 1 7 . . . 'iVaS I 8 .h3 and now:

analysis diagram

B2 1 ) 1 8 . . . �fS was played in Kasparov-Illescas Cordoba, Linares 1 9 94. The play by the former World Champion was very instructive. The game continued 1 9 .'iYd4 1:[ab8 2 0 .a3 'iYbS 2 l .b3 ! (cautious and strong. The immediate 2 1 . b4 would allow 2 1 . . . as ! , when White would have to be clever enough to force a drawish endgame with 2 2 .'iVf4! �g 6 2 3 .tLlb3 ! axb4 24. 4.Jd4 'iYb 7 2 S .axb4 (not 2 S .tLlxc6? b3 !) 2S . . . 'li'xb4 2 6 .tLl xc6 'ifxf4 27.gxf4 .tIb2) 2 l . . . tLle4 (2 l . . Jdxe2 ?? would be a terrible blunder, losing a piece after 2 2 . .tIxe2 'ifxe2 2 3 .'iYf4) 22 .b4! (great play. The situation has changed and White need not fear the . . . a7 -aS break any more) 2 2 . . . aS 23 .4.Jxe4 �xe4 2 4 . .tIcS 'ifb6 2 S .bxaS and White was a sound pawn up, soon converting it into a full point; B2 2) 1 8 . . . �hS is the latest attempt to improve on the previous examples and create the much-desired dynamic bal­ ance: 1 9 .tLld3 (with the black bishop on hS , the e 2 pawn is attacked and Kasparov's play could not be copied. However, the drawback of Black's strat­ egy is that with the bishop on hS , White has a possibility to transfer his knight to f4 with tempo) 1 9 . . . �g6 (the other, pro bably better option was 1 9 . . . 'iYxa2 ! 20 . .tIxc6 �g 6 ! 2 1 .tLlb4 (2 1 .4.Jf4 �e4 ! ) 2 l . .. 'iYxb2 2 2 .tLlxdS 4.JxdS 23 .�xdS �h7 24.e4 .tIac8 and a draw should be the most likely out­ come) 2 0 .tLlf4 and now: B2 2 1 ) 20 . . . 'li'xa 2 2 1 . tLlxg6 fxg 6 22Jlxc6 'iYxb2 i s better than i t looks , since after 2 3 .�xdS + tLlxdS 24.'iVxdS + �h8 ! (24 ... �h7 2 S .'iYd3) 2S .�xg6 as ! the a-pawn is very strong and Black should be able to draw;

2:

Isola ted Pawns

B 2 2 2) On the other hand, in case of a ive defence with 2 0 . . . iVb6 2 l .tLlxg6 fxg6 2 2 .b3 �e6 , White im­ proves his position with 2 3 .e3 nf8 24.l:Ie2 with .tIec2 to follow, and the black pawn weakness on c6 will start to become visible; B 2 2 3 ) 2 0 . . . �e4 2 1 . .tIxc6 �xg 2 2 2 . �xg2 d 4 (an attempt at dynamic play. Regaining the pawn with 22 . . . 'iVxa2 would lead to a better game for White after 2 3 .'li'a l 'iYb3 2 4.'ifa3 )

analysis diagram

2 3 . 'if c I ! (an important move) 23 . . . tLldS (Black continues his attempts to solve problems tactically. In the event of 2 3 . . . 'iYxa2 the idea behind White's previous move would become clear: 2 4.�c4! 'iYxc4 (24 . . . 'iYxb 2 ? ? loses to 2 S . nc7 nf8 2 6 .tLlg6) 2 S J�xc4 and Black remains stuck with his pawn weaknesses) 2 4 . tLld 3 ! tLlb4 (to 24 . . . �xa2 White again responds with 2 S .'ifc4) 2 S .tLlxb4 'ifxb4 (White is a pawn up, but Black has some temporary activity) 2 6 . bk4 �b 7 + 2 7 .�g l 'li'd 7 2 8 .�f4 .tIad8 2 9 . l:Id l ! (making use of a tactical motif) 29 . . . 'iYxh3 3 0 .lIcxd4 .tIxd4 3 1 .'iYxd4 a6 (3 1 . . . .tIxe2 ?? blun­ ders a rook after 3 2 .'iYd 8 + �h7 3 3 .'ifd3+) 3 2 . e3 and White was a sound pawn up. Later he exchanged the 95

Chapter

Winning Chess Middlegames

queens and won the rook endgame eas­ ily, Sokolov- T.1. Petrosian, European Club Cup, Kemer 2 0 0 7 . I have to it that in my preparation for this game, for my decision which variation to choose Kasparov's game played a crucial role. This decision­ making process is in no way connected to the pawn structure theme, but may be useful for readers to know. In the past ten years or so, my main weapon against the Tarrasch was the main line (like in the game) : 9 .�g S cxd4 1 0 .ttJxd4 h6 1 1 .�f4 (instead of 1 1 .�e3 ) , leading to a slightly better endgame for White. A few months be­ fore my game against Petrosian, I had a game against Akobian. In the slightly better endgame that resulted after 1 I .�f4, I never had any real winning chances and the game ended in a dull draw. Now, facing Petrosian, I knew the Tarrasch would appear on the board. I had only about two hours to prepare and wanted to play something new. Under those circumstances it is useful to check the games of the world's very best. Should the line chosen by them suit you and not seem too complicated to prepare in a few hours, PLAY IT. In my professional career this strategy has served me very well. 1 3.

...

tba4-c5

tbc6-a5

Not an easy choice. This knight looks clumsy at the edge of the board , but 96

1 5.

b2-b3

Limiting the scope of the as knight. lla8-c8

1 5. 1 6.

tbc5xd 7

'ifd8xd7

1 7.

'if d 1 -d 3

tba5-c6

Trying to play ' actively' with I 7 . . . ttJe4? would be bad after the simple 1 8 .Jdxc8 'iYxc8 ( I 8 . . . �xc8 loses a pawn without any compensation after 1 9 .�xe4 dxe4 2 0 .'li'xe4) I 9 J�c I 'iVd7 2 0 .'iVbS , with a winning advantage for White.

2 I .ttJb4 ! ) 2 1 .fxe3 bxc6, with at the very least good compensation. b7xc6

1 8. 1 9.

Jd.f1 -d1

20.

�e3-c5

'iWd7-b7

White now uses similar positional plans to those we have already seen under 1 3 .ttJxc6 bxc6 1 4.ttJa4. �f8xc5

20. 21 .

J::tc 1 xc5

'iWb7-e7

22.

l::t c 5-c2

'iWe7-a3

23.

e2-e3

Black is facing a very difficult defence. Due to his better pawn structure, White has a lasting advantage, while Black does not have dynamic activity to nearly compensate for the positional problems related to his weak pawns. It is important to note that transitions from an isolated pawn to two weak hanging pawns are a very common way to combat the isolated pawn. Contrary to parallel hanging pawns in the centre, such weak connected pawns are by def­ inition not mobile, so there are no tac­ tics that can be based on their mobility. 23.

...

l:le8-d8

24.

'ii d 3-f5

�c8-b8?1

Black prefers not to stay ive. A good idea in general ; however, it does not work here. It was better to opt for 24 . . . 'iVa6 or 24 . . . 'iYd6 . 25.

�c8-d7

The active I 3 . . . 'li'aS ? is bad due to I 4.ttJxc6 bxc6 I S . �xc 6 ! �d7 1 6 .�d2 ! , transposing to one of the previous comments - see I 3 .ttJc6 bc6 I 4.ttJa4 tvaS ?, while in the case of I 3 . . . ttJg4 White is better after 1 4.ttJxc6 bxc6 1 S .�d4. 1 4.

other moves also do not create the kind of play Black is hoping for. In general, Kramnik's plan is perhaps a sophisticated version of the already dis­ cussed 1 3 . ttJxc6 bxc6 1 4. ttJa 4, since in the current position the b7 pawn hangs, creating an extra problem for Black, while White keeps most of his posi­ tional trumps. In case you're wondering why I didn't play it myself (against Petrosian) - well, having limited time to prepare, I checked only Kasparov's White games against the Tarrasch!

1 8.

l:lc2xc6

26.

2:

Isola ted Pawns

�c6xf61

A rather thematic exchange sacrifice. 26.

...

g7xf6

Trying to be tricky with 2 6 . . . 'li'xb 3 ? would not help after 2 7 .'li'g4 h S 2 8 .'iVxhS gxf6 29 .�xdS , winning. 27.

�g 2xd5

.l::t d 8xd5

Black is obliged to return the exchange, either immediately or after 2 7 .. .r:J:ig 7 2 8 .'iVg4+ �f8 2 9 .'iVhS . 28.

� d 1 xd5

'iWa 2xb3

Black has ended up with a damaged pawn structure on the kingside, but there are not that many pieces left on the board and there is material equality, so at first sight it seems that Black should have reasonable drawing chances. If he managed to exchange ei­ ther queens or rooks, even losing his a -pawn in the process, he would reach the safety of a draw. It is rather possible that Illescas Cordoba aimed for this position when playing 24 . . . l:Ib8 , believing - with good reason - that he would have reasonable draw­ ing chances. Kramnik, however, does not allow exchanges and keeps the pressure on. Having to defend a diffi­ cult position, Illescas Cordoba ulti­ mately loses his way.

'iWa3xa2

tbd4xc6 1

Quite rightly, Kramnik now decides to change the pawn structure and focus on the newly-created weakness. The logical I 8 . lIfd I would reveal the drawback of I S .b3 after I 8 . . . �a3 ! I 9 Jib l �cS with good play for Black, since after 2 0 .ttJxc 6 ? ! he has a thematic exchange sacrifice in 2 0 . . . �xe3 ! (not 20 . . . �xe3

29.

'iWf5-g4+

�g8-h 7

30.

a.d5-d6

'iWb3- b 1 +

97

Chapter

Winning Chess Mi ddlegames

31 .

�g1 -g2

'iVb1 -g6

SL 8 . 5 (D45)

32.

'iYg4-f3

�h7-g7

Pawel Jaracz

33.

�d6-a6

nb8-bS?!

Peter Heine Nielsen Dresden Ech 2 0 0 7 (4)

Black loses his a-pawn, but the prob­ lems related to his exposed king re­ main. A better defence was 3 3 . . . �e8 ! 3 4J!a4 l::[ e 5 3 5 .h4 h5 , using the fact that either the queen or the rook ending with four versus three pawns on the same side, is a draw. 34.

na6xa 7

nbS-fS

3S.

'ii'f 3-e2

'ii' g 6-hS?

A tactical miscalculation, obviously thought up when he played 3 3 .. J:rb5 . 36.

g3-g4

'iYhS-g6?

An outright blunder. 36 . . . !:rg5 was the only move.

37.

l::t a 7xf7+!

38.

'iVe2-c4+

�g7xf7

GAME 2 2

1.

d2-d4

2.

c2-c4

e7-e6

3.

t2Jg1 -f3

d 7-dS

4.

t2Jb1 -c3

c7-c6

S.

e2-e3

t2Jb8-d7

6.

'iVd 1 -c2

�f8-d6

7.

b2-b3

0-0

8.

�f1 -e2

b7-b6

This is one of the most common posi­ tions in the 6. 'if c 2 Meran and has been one of the main lines over the years. All of the logical moves, 8 . . . l::[ e 8 , 8 . . . e5 , 8 . . . dxc4, 8 . . . 'iVe7 and 8 . . . a6, have been tried in hundreds of games. The game continuation also belongs to the main moves here. 9. 1 0.

Transferring into a favourable pawn symmetry

0-0

�c8-b 7

S:i.c1 -b2

c6-cS

Taking action in the centre. Black can also delay this decision and first play some useful moves, like 1 O . . . 'iVe 7 , 1 O . . . �e8 , or l o . J :k8 . .

In isolated pawn-positions, one of the sides often chooses the right moment to ex­ change a pair of minor pieces in the cen tre (mostly knights) in order to reach a favour­ able symmetry. Such structural transformations are very common and with the next few games I will try to give a few good examples.

98

11.

Isola ted Pawns

pawns in the centre, or an isolated pawn. 11.

...

1 2.

!1a 1 -d 1

1 2.

...

e6xdS

t2Jg8-f6

Black resigned. He ends up two pawns down : 38 . . . �f8 3 9 .'iVc8++- ; 3 8 . . St)e 7 3 9 .'iYe4++- ; or 3 8 . . . \t>g 7 3 9 . 'ifc 7 + �g8 40 .�c8 ++- .

Structure 2 . 3

2:

c4xdS

The strategic features of the position are becoming visible. It is rather clear that the current situation in the centre will transform to a position where Black will either have two parallel hanging

'iVd8-e7

In case of 1 2 . . . !:re8 1 3 .dxeS bxcS , a complex position with two hanging central pawns is reached. In 1 9 94 I had an interesting game of my own, which continued 1 4. iff5 ! (stepping up the pressure on Black's central pawns) 1 4 . . . 'iVe 7 and now : A) Now 1 5 .�b5 plays into Black's hands after 1 5 . . . ttJe5 ; B) Kramnik-Kasparov, 8 th blitz match game, Moscow 1 9 9 8 , produced inter­ esting and dynamic play after 1 5 . life 1 'iVe6 1 6 .ifb 1 h6 1 7 .�d3 ttJe5 ! 1 8 .�f5 ttJxf3 + 1 9 . gxf3 'if e 5 2 0 . f4 'if e8 2 1 .�h3 I:rd8 2 2 . .ig 2 'iVe6 with a com­ plicated game; C) 1 5 .�d3 and now : C l ) 1 5 . . . ttJe5 ? simply blunders a pawn after 1 6 . ttJxd5 ! ; C2) On the other hand 1 5 . . .'iie6 should definitely be considered; C3) 1 5 . . . �fe8 1 6 JHe l (with a black rook on e 8 , 1 6 .�bS ! ? , pinning the knight, was another option) 1 6 . . . �b8 1 7 .ttJe2 (with both white bishops working and the knights getting trans­ ferred to the kingside, threats will soon 99

Ch apter

Winning Chess Middleg ames

be created) 1 7 . . . g 6 1 8 .'ifgS c4 (Black tries to create threats of his own, but misses some tactical shots. Another op­ tion was to steer the game into the end­ game with 1 8 . . . tLJeS 1 9 .tLJxeS �xeS 2 0 .�xeS 'iYxeS 2 1 .'iYxeS .l::t xeS , though then White can target the black pawns and keep the advantage with 2 2 .lIc 1 as 2 3 Jdc2 , since 23 .. .!Ie7 fails to 24. lIec l Jdec7 2 S .tLJd4 ! )

1 3.

1 9 .�fS ! (the pins are becoming quite unpleasant for Black) 1 9 . . . tLJe4 ( 1 9 . . . h6 does not help, since after 2 0 .'iYxh6 gxfS 2 1 .'iYg S + �h8 2 2 .tLJed4 White gets a winning attack) 2 0 .'iYh6 ! c3 2 1 .�a3 ! (crucial move) 2 1 . .. 'iYxa3 2 2 .�xd 7 'iYxa2 ? (a blunder. Better was 22 . . . c2 , though after 2 3 .Ilc 1 lIed8 24.�xc8 �xc8 2 S .tLJfd4! 'iYxa2 2 6 .f3 Black's compensation is not sufficient) 2 3 .�xe8 J;Ixe8 24.l:Ic 1 (24.tLJxc3 ! was a direct win: 24 . . . 'iYxf2 + (24 . . . tLJxc3 2 S .tLJgS is a forced mate) 2 S .Wh l tLJxc3 2 6 .l:Id2 and the black queen is trapped) 24 . . . Iic8 2 S . tLJfd4 'ifaS 2 6 .f3 and White was clearly better in Sokolov-Lautier, Donner Memorial, Amsterdam 1 9 94. 1 3.

1 4.

tLJh4-f3

l:raS-cS

'iYc2-b1

a 7-aS

1 S.

a2-a4

llfS-eS

1 7.

'iYb1 -a1

...

bility. It was necessary to first secure the cS knight with 20 . . . aS ! and only then proceed with . . . tLJfe4, with better chances for Black. N ow White spots a new tactical motif that comes as a consequence of the possibility to change the pawn struc­ ture in the centre. Please note that generally, tactics created by a change in pawn structure are very easy to miss. 21 .

b3-b4!

tLJc5-d7

22.

.itf3xe4!

d5xe4

c5xd4!

1 S.

tLJf3xd4

tLJd 7-c5

1 9.

�e2-f3

�dS-e5

The bishop is well placed on eS , so the potential weakness created by White's 1 3 th move is rather irrelevant at the moment. 20.

tLJc3-e2

tLJfS-e4?

Now the pawn structure is symmetri­ cal, but all of White's pieces are work­ ing harmoniously, while Black's b 7 bishop i s hitting its own pawn. White now executes a very nice positional exchange sacrifice that is worth ing. 23.

tLJd4-f5!

gSxf5

24.

�d 1 xd 7

�e5xh2+

25.

\tlg1 xh2

'ife7xd 7

2S.

tLJe2-f4

For only an exchange White has full control, the black king is weak, his kingside pawn structure is destroyed, and the b7 bishop is reduced to a mere pawn. This standard move, though logical in itself, allows a beautiful tactical possi-

29.

Isol a ted P awns

\tlh2-g3!

Involving his last piece - the rook - in the attack, which now becomes devas­ tating. The game is decided. 29.

...

'iYb4-d2

30.

.l:t h 1 -h5

'iYd 2xf2+

31 .



The black monarch has been left on his own without a single piece defending him, while all of the white pieces are participating in the attack.

Black correctly decides that the time has come to change the pawn structure in the centre. He opts for the isolated pawn position, correctly judging that with his well-placed pieces he has nothing to fear.

tLJf3-h4

Provoking . . . g 7 -g6 in order to create threats along the a 1 -h8 diagonal. 1 00

g 7-gS

1 5.

1 7.

analysis diagram

...

2:

31 .

...

�eS-eS

32.

'iYa1 -d4!

'iYf2-d2

33.

llh5-g5+

'It>gS-fS

34.

tLJf4xeS+

f7xeS

35.

�fS-e7+!

The final stroke. �fSxe7

35 .

..

3S.

l:tg5-g7+

We7-eS

l:tg 7-gS+

�eS-e7

.

2S .

...

'ifd 7-e7

37.

27.

l:If1 -h 1 !

�cS-c2

3S.

'iYd4-g 7+

2S.

�b2-fS

'iYe7xb4

39.

l:lgS-dS+

1 -0

101

C h ap t e r

W i n n i n g C h es s M i d d l e g a m es

B) Playing with an isolated pawn

Structure 2 .4

Transferring to a favourable pawn symmetry in the centre

TD 4. 1 6 (D3 4)

GAME 2 3

Tigran V Petrosian Boris Sky Moscow World Championship 1 9 69 (4) 1.

c2-c4

e7-e6

2.

d2-d4

d7-d5

3.

lLlb1 -c3

c7-c5

4.

c4xd5

e6xd5

5.

lLlg 1 -f3

lLl bS-c6

6.

g2-g 3

lLlgS-f6

7.

.itf1 -g2

�fS-e7

S.

0-0

0-0

9.

.itc1 -g5

c5xd4

1 0.

lLlf3xd4

h 7-h6

11.

.itg5-e3

.itcS-g4

This is a less frequently played, but rather viable alternative to the more popular 1 1 . . .lle 8 . In the 1 9 69 World Championship match, 1 1 . . . iiLg4 was played no less than three times, and it served Sky well. It is not clear to me why this logical move has all but disap­ peared from grandmaster practice and 1 02

1 1 . . . �e8 has become an almost auto­ matic reply.

I S .ttJxc6 'ifxc6 1 6 .'iVxg4) I S .ttJxc6 bxc6 1 6 .ttJc3 ) 1 3 J�tad l and now: A I ) Black has to be careful not to blunder into a different kind of tactic. A 'logical' move like 1 3 . . Jlc8 runs into 1 4.ttJxdS ! ttJxdS 1 5 .ttJfS ! (the point. White gets his piece back, remaining a sound pawn up) I s . . JIc4 1 6 .'iVbS i.xfS 1 7 .lixdS a6 1 8 . 'iYxaS 'iYxaS 1 9 .1ixaS and White had a technically won endgame in Tal-Zhidkov, Soviet Championship, Baku 1 9 7 2 ; A2) 1 3 . . . ttJc4 1 4.i.c l 'ifc8 I S .'iVc2 (after 1 5 . ttJxdS ? ttJxdS 1 6 .i.xdS ttJ b6 1 7 .'iYb3 ttJxdS 1 8 .'iYxdS i.f6 Black gets strong compensation, worth at least a pawn) I S .. J�d8 1 6 .b3 ttJeS 1 7 .�b2 iYd7 1 8 .f3 ! �hS 1 9 .'iVfS ! and White was better. I think Black's play can be improved with 1 4 . . . ttJb6 ! , with a good game after I S .'ifc2 (or I S .'ifb3 l:Ic8 1 6 .h3 i.d7 1 7 .Wh2 �c4!) I S . . . 'iYd7 1 6 .f3 �hS , since now the idea to exchange queens with 1 7 .'iYfs fails to 1 7 . . . 'iYe8 ! and Black was better in Gen. Timoschenko­ Sturua, Pavlodar 1 9 8 2 ; B) 1 2 .h3 i s another standard move, played by Sky himself. Black kept dynamic piece play after 1 2 . . . i.e6 1 3 . l::k l (it i s good t o note that per defi­ nition Black is never worried about 1 3. ttJxe6 fxe6 , since the loss of the bishop pair is fully compensated for by the improvement of his pawn struc­ ture) 1 3 . . . 'if d7 1 4. �h2 ttJeS ! (other moves would allow White, like in Kramnik -Illescas Cordoba (Game 2 1 ) , to create a new pawn weakness in Black's camp : I S .ttJxc6 bxc6 1 6 .ttJa4) l S .'iYa4 'iYxa4 1 6 .ttJxa4 ttJc4 1 7 .�J4 gS! 1 8 .b3 �d7 ! in Sky-Kortchnoi, Soviet Championship, Moscow 1 9 5 7 .

2:

I s o l a t e d P a wns

1 2.

...

�g4-e6

1 3.

l:ta1 -c1

�fS-eS

1 4.

llf1 -e1

In the second game of the match, Petro­ sian had opted for a different plan and played 1 4.ttJbS , with the idea to transfer the knight to d4 and then to take on c6, creating a new pawn weakness to focus on. This well-known strategy, which worked well in Kramnik-Illescas Cor­ doba (Game 2 1 ) , did not work here, since Black quickly achieved dynamic play and took over the initiative, never allowing White to 'remind' him of his weak c6 pawn: 1 4 . . . ifd7 I S .ttJSd4 i.h3 1 6.ttJxc6 bxc6 1 7 .'iYd3 i.xg2 1 8 . �xg2 as ! 1 9 .nc2 a4 2 0 .ttJd2 (20 .ttJcS �xcS is a favourable swap for Black) 20 . . . ifb 7 2 1 .f3 ttJd7 ! - transferring the knight to eS and preparing . . . c6-cS at the same time. Black was better. 1 4.

...

1 5.

.ite3-c5

'iVdS-d7

Exchanging the dark-squared bishops is generally a good strategy here. 1 5.

...

�aS-cS

1 6.

.itc5xe7

'iYd 7xe7

1 7.

e2-e3

.

1 2.

lLld4-b3

White's idea is to first exert pressure on the dS pawn, and later play i.cs and ex­ change the dark-squared bishops. A) 1 2 .'ti'a4 is another logical contin­ uation, which was played in the 1 2th game of the match : 1 2 . . . ttJaS ! ( 1 2 . . . 'iYd7 ? looks like a standard move here, but it loses a pawn after 1 3 .i.xdS ! ttJxdS 1 4 .ttJxdS .td8 ( 1 4 . . . 'iYxdS

As a result of the exchange of the dark-squared bishops, White has good control of the dark squares in the centre (d4 and cS) and also enough space to 1 03

Chapter

W i n n i n g C h es s M i d d l e g a m es

manoeuvre his pieces. White should be slightly better here. 1 7....

l:re8-d8

1 8.

'iYd 1 -e2

1 8 .h3 ! , to prevent Black's active play, was better, securing White a small ad­ vantage. 18.

.

.

.

�e6-g4f

A good move, forcing White to choose between two unpleasant options : either to weaken his own pawn structure and close off the g2 bishop with f2 -f3 , or to move the queen to fl . The central pawn break 1 8 . . . d4, often an equalizer in these positions, does not work that well here : 1 9 .tDxd4 tDxd4 2 0 . exd4 l:txd4 2 1 .tDbS l:ixc l 2 2 .tDxd4! llxe 1 + 2 3 . 'if xe 1 and White is better. 1 9.

f2-f3

White will control the e4-square and later place his bishop on h3 , but this de­ cision compromises his pawn structure, even though, due to the fact that the dark-squared bishops are exchanged, it is conSiderably more difficult for Black to make use of this potential weakness. 1 9 .'iffl was a safer decision. 1 9.

...

�g4-f5

20.

l:rc1 -d 1

ttJc6-e5

White's pawn structure has been com­ promised (he has a potential weakness on e3 to worry about) , his bishop on g2 is temporarily shut off and cannot exert pressure on the dS pawn. His small ad­ vantage has vanished and he has to be careful to maintain the balance. Black, on the other hand, has easy piece play and the potential weakness of the isolani on dS does not play any role now. 21 .

ttJ b3-d4

�f5-g6

22.

�g2-h3

l::t c 8-c411

Keeping the cS -square free for the queen and also planning to put pressure 1 04

on the d4-square with . . . tDc6 . This plan has its drawbacks, however, as will soon become clear. 2 2 .. J:Ics was another logical option, when White can play 2 3 .�fS , with ap­ proximate equality.

since after 2 7 .exd4 l:le6 2 8 . gS ! , Black loses the exchange) 2 7 .�fl tDxd4 2 8 . exd4 l::r e 6 2 9 .tDcS l:Ixe 1 3 0 J:txe l 'iVc 7 , with approximate equality. 26.

...

�g6-h 7

Moving away from the potential tDf4 tempo. This pre-emptive type of strat­ egy (introduced a long time ago by Nimzowitsch) is often a good idea. 27.

�h3-g2

28.

ttJe2-g3

l:ld8-e8

By preparing tDdfS , White more or less forces Black to take on d4. Another so­ lution was 2 8 . tDf4.

23.

g3-g41

An excellent move that shows the draw­ backs of 2 2 . . J:tc4. White prepares 24.f4. In the event of the immediate 2 3 .f4 Black has 2 3 . . . �hS . In the case of 2 3 .'iffl iYcs , or 2 3 .�fS �xfS 24.4JxfS 'iV cS , Black keeps the initiative. 23.

...

l:Ic4-b4

Running away from the tempo-win­ ning f3 -f4 with 2 3 . . . tDc6 ? would be a terrible blunder due to 24.tDxdS . 24.

b2-b3

For some reason, White hesitates to fur­ ther pursue his plan started with the previous move. Better and more princi­ pled was 24.f4 ! tDc4 2 S .b3 4Jd6 2 6 .fS �h7 2 7 .�g2 and with the black bishop locked in on h7 and the rook on b4 starting to look clumsy, White is better. 24.

...

'iVe2-d2

26.

ttJc3-e2

...

ttJc6xd4

29.

e3xd4

�b6-e6

The situation in the centre has changed and an isolated pawn-structure has been transformed into a symmetrical one. Given the fact that the dark-squared bishops are exchanged, a symmetrical pawn structure d4/dS with only the light-squared bishops present, should, in principle, favour White. However, here Black has control over the e-file, he has a better-placed bishop and the white kingside can become vulnerable. 30. 31 .

l::t e 1 xe6

'iVe7xe6

llb4-b6

Heading for f4. Another move was 2 6 .tDa4 l:ta6 ( 2 6 . . . tDxd4? is a blunder,

I s o l a t e d P awns

This knight is heading for f4. 33.

'ifd2-f4

34.

�c1 -c51

ttJh 7-f8

This rook will remain here, doing noth­ ing, until the end of the game. 34.

�g 6-b1 1

35.

a 2-a4

ttJf8-g6

36.

'iYf4-d2

'i¥e6-f6

In the last few moves, the situation has changed drastically. Black has improved his pieces, while White has simply wasted a few tempi. 37.

\!tg1 -f2

38.

a4-a51

ttJg6-f4

White was definitely not aware of Black's direct threat. It was necessary to get the rook back into play with 3 8 . bt.c3 .

38 .

...

�b1 -d31

Black has a winning attack.

l:td1 -c1

39.

ttJg3-f5

3 9 .'iVc3 loses to 3 9 . . . 'iVh4 40 . �g l �xfl 4 1 .�c8 �bS . 39.

...

�f6-g51

Threatening 40 . . . tDh3 + . 3 9 . . . �xfS 40 .'iYxf4 gS ! 4 1 .'iYd2 �xg4 also wins, but the text is even better.

ttJe5-c6 1

Now the possibility for White to ad­ vance his f-pawn and shut out the black bishop has vanished forever. 25.

28 .

2:

31 .

...

�h7-g6 1 1

Preparing an excellent knight manoeu­ vre. 32.

�g2-f1

ttJf6-h71

40.

ttJf5-e3

'ifg5-h4+

41 .

\!tf2-g1

�d3xf1

White resigned. 42 . tDxfl (or 4 2 . �xfl 'iYh3 +) loses to 42 .. . l:Ie 2 . I t i s interesting t o note how easily a gi­ ant like Petrosian could go wrong in a seemingly symmetrical, equal position. 1 05

Chapter

Wi n n i ng C h ess M i d d l e g a m es

A game Kasparov-Tal, Brussels 1 9 8 7 , continued 1 0 .bxc3 0-0 1 1 .0 -0 ttJe7 ? ! (a plan that will not help Black. Better was 1 1 . . . h6) 1 2 .'iYc2 �d7 1 3 .�g5 ttJg6 1 4 .f4! (a typical plan , worth remem­ bering) 1 4 . . . h6 1 5 .�xf6 'iYxf6 1 6 .f5 CiJe7 1 7 .ttJg3 (the queen is extremely badly placed on f6) 1 7 . . . ttJc8 1 8 .tIf4 CiJd6 1 9 . iff2 .tIfe 8 2 0 . ttJh 5 'iVd8 2 1 .ttJxg 7 and White soon won) 1 0 . 0-0 0-0 1 1 .�c2 h6 1 2 .ifd3 (it was also possible to start with 1 2 .f3 ) 1 2 . . . l::l e 8 (another idea was to prevent White's subsequent pawn advance with 1 2 . . . ttJe7 1 3 .ttJg3 l:te8 , though White is better after 1 4.�d2 with 1 5 .�ae l to follow) 1 3 .f3 ttJe 7

Structure 2 . 5

NI 1 5 . 6 (E48) Ivan Sokolov

GAME 24

2:

I s o l a t e d P a wns

Black has taken a wrong strategic course at the start of the game and now loses without making any direct mistakes: 2 1 . . . �f8 2 2 .'iYd2 'iYh3 2 3 .�xg 6 ! (re­ moving the only defenders) 2 3 . . .fxg 6 24.�xf6 gxf6 2 5 .ttJf4 'iYxf3 2 6 .ttJxg6+ �f7 2 7 .ttJh4 (the game is decided) 2 7 . . .'ti'h5 2 8 .�g7 + �f8 2 9 .1dxc7 'iVxh4 3 0 .'iYg2 'iYh8 3 1 .lig7 'iVh5 3 2 .'iVg3 and Black resigned, Sokolov-Brunner, Oakham 1 9 8 8 . 9.

0-0

Around the end of 1 9 8 8 I played for the Yugoslav team in the Olympiad in Thessaloniki, where the game Knaak­ Christiansen caught my attention. This original game continued 9 .a3 �e7 1 0 .�c2 0-0 1 1 .'ii'd 3 g6

Miso Cebalo Pula ch-YUG 1 9 8 8 1.

d2-d4

tt:JgS-f6

2.

c2-c4

e 7-e6

3.

tt:Jb1 -c3

�fS-b4

4.

e2-e3

c7-cS

S.

�f1 -d3

tt:JbS-c6

6.

tt:Jg1 -e2

This variation I played a lot almost twenty years ago. In the resulting iso­ lated pawn-structures, White's kingside knight is placed on e 2 rather than, more classically, on f3 . Both plans have pros and cons. Based on my experience, however, I would say that in such classi­ cal isolated pawn-positions the white kingside knight is in principle better placed on f3 . 6.

cSxd4

7.

e3xd4

S.

c4xdS

d 7-dS

White has better chances of an advan­ tage if he takes on d5 himself, rather than allowing 8 . 0- 0 dxc4 9.�xc4 0-0. 1 06

analysis d i agram

S.

...

tt:Jf6xdS

One of the important strategic deci­ sions for Black. The other option was 8 . . . exd5 , creating a symmetrical pawn structure, which may seem like an easy way to equalize. Things are, however, far from simple. The white knight is now very well placed on e 2 (contrary to its colleague on f6) , allowing White to advance his kingside pawns, starting a serious attack on the kingside. One of my own games can serve as a good example : 9 �a3 �d6 (9 . . . �xc3 + is a wrong decision here, since White soon develops a strong kingside attack.

1 4.g4! (this is a typical pawn advance in these positions and useful for the reader to . White is advancing his pawns on the kingside, gaining space and soon developing a deadly attack. On the other hand, it is rather difficult for Black to create serious counterplay) 1 4 . . . ttJg6 1 5 .�h l �c7 1 6 .�d2 'iVd6 1 7 J:!f2 �d7 1 8 J:tg l (all White's pieces are now transferred to the kingside and ready to attack) 1 8 . . . 'iYe6 1 9 .g5 ! hxg5 20 .�xg5 �c6 2 1 J Hg2 . White has a terrible attack, while Black's counterplay is nowhere to be found.

analysis d i agram

1 2 .h4! ? (a very direct approach, use­ ful for the reader to take note of - it can also be used in similar positions ! . Black has to follow the old basic rule : a wing attack has to be countered in the centre ! ) 1 2 . . . e5 ! (an immediate, principled reaction involVing a sacri­ fice. The other possi bility is 1 2 . . . ttJxc3 1 3 . bxc3 e5 , when White gets compensation for the sacrificed pawn after 1 4 .iYg3 exd4 1 5 .h5 �d6 1 6 . �f4 �xf4 1 7 . ttJxf4) 1 3 . dx e 5 ttJxc3 1 4 . 'ti'xc3 � g 4 1 5 . f3 107

Chapter

W i n n i n g C h es s M i d d l eg am e s

ana lysis d i agram

I s . . J:k8 ! (Black's play is energetic, in the spirit of the position. His bridges have already been burned with 1 2 . . . eS ! and to keep the initiative, he must be ready to sacrifice more) 1 6 .fxg4 ttJxeS ! (excellent play. Black breaks open the white king's position) 1 7 .'iYb3 ! (White is shoWing good nerves in defence. Taking the knight is dangerous, for example : 1 7 .'iVxeS �e8 ! . White is two pieces up, but his king has nowhere to hide : 1 8 .'ifd4 �xh4+ 1 9 .�d l 'iYe7 2 0 .'ife4 l:tcd8 + 2 1 .�d2 'iYcS ! (less good is 2 1 . . . �gS 2 2 . 'iYxe 7 l:txe 7 2 3 .rJte l l:txd2 24.�d l �xb2 2 S .�f1 bS 2 6 J �h3 and White would disentan­ gle his pieces) 2 2 .b4 l'ixd2 + 2 3 .�xd2 'ifgS + 24.�d l .r:rxe4 2 S .�xe4 'iYxg4 and due to White's exposed king, Black has ample compensation for the small material sacrifice) 1 7 . . . �xh4+ 1 8 . g3

analysis d i agram

1 08

1 8 . . JIxc2 ? (in his attempt to keep the initiative, Black sacrifices one piece too many. Correct was 1 8 . . . �gS ! 1 9 .0 -0 (wrong is 1 9 .�f4? 'iVaS + ! 2 0 .ttJc3 ttJf3 + 2 1 .�f2 ttJd4 2 2 .'iYa4 'iYb6) 1 9 . . . �xc l 2 0 . 11axc l 'iYd2 2 1 .ttJf4 ttJxg4 2 2 . ttJg2 'ifd4+ 2 3 .<;t>h l ttJf2+ 24.rJth2 ttJg4+ 2 S . rJth l ttJf2 + and a draw could have been the outcome of this exciting game. Now it soon be­ comes obvious that the tables are turn­ ing in White's favour) 1 9 .'iYxc2 ttJd3 + 2 0 . �f1 'iYdS 2 1 .ttJg l ! (only move) 2 1 . . J:ie8 2 2 .�h6 ! (weaving a mating net around the black king) 2 2 . . . �xg3 2 3 Jld l 'iVxh l 24.'iVxd3 'ifh2 2 S .'iYe2 ! and White soon won. During the Olympiad in Thessaloniki the general opinion among grand­ masters had been formed that Christiansen's sacrifice was incorrect, so a couple of months later when I played in Wijk aan Zee against Van der Wiel (a reasonably good player at the time) I wanted to repeat Knaak's idea. However, after 9 . a3 Black played 9 . . . ttJxc3 1 0.bxc3 �e7 and now, for some rea­ son, instead of playing 1 1 .�c2 , trying after 1 1 . . . 0-0 to copy the above-men­ tioned game with 1 2 .'iVd3 g6 1 3 .h4 , I played 1 1 .�e4 and after 1 1 . . . 0-0 1 2 .'iYd3 h6 committed an incredible positional blunder playing the cof­ fee-house move 1 3 .f4? Black easily punished this 'aggressive' approach with 1 3 . . .'ifc7 1 4.tIb l ttJaS 1 5 . 0 -0 fS ! 1 6 .�f3 and now Black executed an in­ structive pawn sacrifice with 1 6 . . . 'i'ic4 ! 1 7 .'ifxc4 ttJxc4 1 8 .�xb 7 Rb8 1 9 .�xc8 l:ifxc8 . The dominant knight on c4 is worth much more than a pawn. I had to suffer to finally make a draw in the ending.

9.

..

ttJd5-f6

.

A slighdy odd-looking move. Black re­ moves his knight from the centre in order to better protect his kingside and also not to allow Black to timely take on dS . A logical continuation - and the main line - is 9 . . . 0-0 and after the logical sequence 1 0 .�c2 !:le8 I I .fVd3 g6 1 2 .f!d l �f8 1 3 .'iVf3 �g 7 1 4.ttJe4 h6, a rather complex battle begins. Black's knight on dS is well placed and his king is also reasonably safe, but White has a little more space and Black has a slight problem to develop his queenside. In the same tournament a couple of rounds later, I had a game with Abramovic, which continued 1 5 .�d2 'iVe7 1 6.�b3 �d7 1 7 .'ifg3 l:led8 1 8 .a3 �e 8 1 9 .'iYh3 ! (preparing a tactical trap) 1 9 .. .'�'f8 2 0 .�ac 1 .

ana lysi s d iagram

The position is objectively approximately equal, but here Black blundered, playing the 'logical' 20 . . Jdac8 ? and missing the pin: 2 l .�xdS ! exdS 2 2 .�b4 and White won after 2 2 . . . �d7 2 3 .�xf8 �xh3 24.�xg7 dxe4 (24 . . . Wxg 7 loses a pawn after 2 S .ttJcS) 2 S .�f6 �d6 2 6.dS ttJaS 27 .�e7 l::Ixc l 2 8 .11xc l . 1 o.

�d3-c2

�b4-e 7

In case of 1 0 . . . �d6 ? Black would likely be a tempo down compared to the

2:

I s o l a t e d Pa wns

game after 1 1 ...tgS ! 0-0 ( 1 1 . .. �xh2 + ? never works due t o 1 2 . �xh2 ttJg4+ 1 3 . �g3 iYxgS 1 4 .f4 'iYhS 1 5 J :th 1 and White wins a piece) 1 2 .a3 , with 1 3 . 'ltV d3 to follow. 0-0

11.

a2-a3

1 2.

'iYd1 -d3

b7-b6

1 3.

�c1 -g5

g 7-g6

1 4.

lla1 -d 1

�c8-b7

1 5.

h 2-h4

Some years later, at the Moscow Olym­ piad 1 9 94, Cebalo had a game against Tisdall which continued I S .'iVh3 (very often the queen is well placed on this natural square) 1 5 . . . llc8 1 6 .l:tfe 1 ttJaS 1 7 .ttJf4! (increasing the pressure and creating tactical possibilities like the d4-dS break or the sacrifice ttJxe6) 1 7 . . . ttJdS (in the event of 1 7 . . . ttJc4 Black has to worry about 1 8 .�b3 ( 1 8 .dS ttJxdS 1 9 . ttJxe6 also looks dangerous) and now 1 8 . . . ttJxb 2 ? loses to 1 9 .l:tb I ! (in case of the more logical 1 9 .�d2 Black has a beautiful tactical solution in 1 9 . . . ttJhS ! 2 0 .ttJxe6 �xg S 2 1 .ttJxd8 �xd2) 1 9 .. .'iVxd4 ( 1 9 . . . �xa3 fails to 2 0 .ttJbS ; 1 9 . . . ttJhS ?? loses immediately to 2 0.ttJxe6) 2 0 .ttJxe 6 ! 'iYxc3 2 1 .'iYxc3 llxc3 2 2 . ttJxf8 and White has a winning endgame) 1 8 .�xe 7 ttJxe7 ( 1 8 . . . ttJxf4? is bad due to 1 9 .'iYh4 'VJIic7 2 0 .�f6) 1 9 . �e4! (by exchanging the light1 09

squared bishops White prepares the d4-d5 break) 1 9 . . . �xe4 2 0 .tDxe4 ttJf5 2 1 .d5 ! e5 2 2 .ttJh5 ! f6 2 3 .d6 and White soon had a winning advantage. 1 5. 1 6.

2:

Chapter

W i n n i n g C h e s s M i d d l eg a m e s

l::t a 8-c8

20.

...

a7-a6

If 20 . . . ttJc4, then 2 1 .�xc4 �xc4 (or 2 1 . .. dxc4 2 2 .d5) 2 2 .tIfe l and White is clearly better, as the black bishop on b 7 is very ive.

I s o l a t e d P awns

39.

�f1 -f2

�f6-c6

40.

l::t d 5-d8!

�g7-f6

41 .

d4-d5

l:rc6-c5

42.

�f2-e3 !

White's centralized king will decide the battle.

ttJe2-f4

42.

...

43.

�b1 -d 3

h6-h5

Getting ready to chase away the black rook. 30.

g 2-g4!

'iVd6-f8

30 . . .fxg4? loses to 3 1 .ttJe4. 31 .

21 . 1 6.

ttJf6-d5?

...

Black is obviously getting nervous about the d4-d5 break and is taking radical measures to prevent it. After the text, however, the pawn structure in the centre is going to transform from an isolated pawn centre to a symmetrical one where White, due to his more ac­ tive bishop, his e-file control and his at­ tacking possibilities on the kingside, will have a clear advantage. It is very important that the dark­ squared bishops are exchanged, so that White can later build on the e 5 -square domination (soon to be seen!) . It was necessary for Black to keep active coun­ terplay with 1 6 . . . ttJa5 ! with 1 7 . . . ttJc4 to follow, which was likely to have been Cebalo's plan when he repeated the line against Tisdall six years later.

f2-f4!

Now, in order to stop the mating attack, Black will have to compromise his pawn structure. 21 .

...

'iVd8-d6

22.

'iVg3-h4

f7-f5

Otherwise f4-f5 , with devastating effect. 23.

I:rf1 -e1

It's time to take possession of the e- file. 23.

...

24.

�d3-b1 !

ttJa5-c6

The bishop will be more effective on the a 2 -g8 diagonal. 24.

...

25.

�e1 -e3

ttJc6-d8

2 5 .ila2 is also strong. 25.

...

26.

ne3-e5

ttJd8-e6

Due to the fact that the dark-squared bishops are gone, the white rook will stay on e5 'forever' .

1 7.

ttJf4xd5

�e 7xg5

26.

...

llf8-e8

1 8.

h4xg5

e6xd5

27.

�d 1 -e1

ttJe6-c7

1 9.

'iVd3-g3

ttJc6 -a5

28.

�b1 -a2

l:te8xe5

20.

�c2-d 3 !

29.

J:te1 xe5

l:i:c8-d8

Preventing Black's counterplay. 1 10

only

way

of

g4xf5

43.

...

h5-h4

Or 43 . . . �e7 44. tIh8.

g6xf5

44.

b2-b4

l::t c 5-c1

�e3-d2

wf6-e7

nd8-h8

nc1 -c7

32.

�a2-b1

�b7-c8

45.

33.

g5-g6!

h 7-h6

46.

34.

�g 1 -f1 ?

A very easy win was 3 4. ilxf5 ! �d6 3 5 .�c2 1:;If6 3 6 .f5 . 34 .

...

I:td8-d6

35.

g6-g7

'iYf8-f6

36.

'iVh4xf6

l:ld6xf6

37.

ttJc3xd5

ttJc7xd5

38.

l::t e 5xd5

wg8xg7

White has not made the best of his attack and we have moved on to the endgame. However, due to his better bishop, strong ed d-pawn and Black's weak­ ness on f5 , White is still on top.

47.

wd2-e3!

Back to the centre. 47.

...

�c8-b7

48.

�e3-d4

l:tc7-d 7

49.

l:th8-h7+

We7-d8

50.

llh7xd7+ wd8xd7

51 .

wd4-e5!

�d7-d8

52.

�d3xf5

�b7-c8

53.

�f5xc8

�d8xc8

54.

�e5-e4

Black resigned.

Now the time has come to open up the black kingside. 111

C hapter

W i n n i n g C h ess M i d d l eg a m e s

11.

Structure 2 . 6

GAME 2 5

Garry Kasparov Anatoly Karpov London/Leningrad Wch

m

1 9 8 6 (22)

1.

d2-d4

tLJgS-f6

2.

c2-c4

e7-e6

tLJg1 -f3

d 7-d5

4.

tLJb1 -c3

itfS-e7

1 0.

0-0

5.

itc1 -g5

h 7-h6

11.

itd3xc4

6.

itg5xf6

6.

...

ite7xf6

7.

e2-e3

0-0

S.

l:Ia 1 -c1

c7-c6

9.

itf1 -d3

tLJ bS-d 7

1 12

e6-e5

1 2.

h2- h 3

e5xd4

1 3.

e3xd4

tLJd7-b6

1 4.

itc4-b3

itcS-f5

1 5.

l::tf 1 -e1

a7-a5

1 6.

a2-a3

l:!fS-eS

White has, due to his well-placed bishop on b3 , certain pressure and also a threat: the tLlf3 -eS jump. Therefore it is advisable for Black to exchange a few pieces. 1 7.

l:re1 xeS+ 'iYdSxeS

1 S.

'iYd 1 -d 2

1 S.

...

Here we have a well-known pawn struc­ ture, which is also often seen in the Moscow Variation of the Slav. The advan­ tage for Black lies in the fact that his dark-squared bishop is already well­ placed and there is no need for him to waste extra tempi on queen moves.

3.

White gives up his bishop pair in order to gain time for quick development. The other main move, 6 .ith4, featured in the game Vyzhmanavin-Beliavsky (Game 2 8) , further on in this chapter.

I s o l a t e d P a wns

This pawn break is a typical way for Black to liberate himself and prepare the development of his bishop on c8 . An isolated pawn position is about to arise, the white trump being a strong bishop, which will be well-placed on b3 and will not be easy for Black to neutralize.

In the central d4/d5 pawn symmetry, often resu ltingfrom a n isolated pawn position, Black's light-squared bishop is (due to the fact that Black's pawn is on a light square ­ d5) almos t per definition inferior to the white light-squared bishop or a kn igh t. Here are two relatively simple examples that highlight Black's problems. The first example is a game that decided the 1986 World Championship match, while the second is an example from my own practice.

QO 1 0 . 1 5 (D 5 5 )

. ..

2:

d5xc4

analysis d i agram

2 1 . . . ite6 ! 22 .itxe6 'iVxe 6 and now: A) 2 3 .'iYb8+ 'iYc8 24.'iYa7 has been tried, but this leads to a forced draw: 24 . . . tLlc4 2 S .�cS 'iVe6 (2 S . . . tLlxb2 should also lead to a draw after 2 6 . tLle4 itd8 ! (2 6 . . . 'iVe6 2 7 .tLlxf6 + gxf6 2 8 .'iVxaS is slightly better for White) 2 7 .'iYc2 'iYfs 2 8 .'iYxb2 'iYxe4 2 9 .�xb7 itf6 , for example 3 0 . 'iYa8 + �h7 3 I . 'iVxaS itxd4 3 2 .tLlxd4 'iYxd4 3 3 .'iYfS + Wg8 34.'iYc8+ Wh7 3 S .'iVxc6 'iVa I + 3 6 . Wh2 'iVxa3) 2 6 .dS cxdS 2 7 .tLlxdS tLlxb2 2 8 .'iYxaS tLlc4 2 9.'iYa8+ Wh7 3 0 .'iYxb7 tLlxa3 3 I .'iYb3 ite7 and a draw was soon agreed in Ftacnik-H. Olafsson, New York Open 1 98 7 . This line illustrates the importance of inserting I S . . . aS I 6 .a3 , as Van der Sterren always stressed - if Black omits this he simply loses the pawn on a7 ; B) 2 3 .'iVc7 tLlc4 24. 'iYxb 7

tLJb6-d7?!

Black indeed wants to cover the eS -square. After this game black players started to look for an improvement and eventu­ ally they found one in 1 8 . . . 'if d7 ! , the idea being that after 1 9 .�e 1 I:te8 20 .11xe 8 + 'iYxe8 2 1 .'iVf4, which seems good for White, Black has an excellent tactical resource found by Paul van der Sterren:

ana lysis d i agram

1 13

Chapter

W i n n i n g C h es s M i d d l eg a m es

24 . . . hd4! - the point of Black's idea. White has nothing better than to force a draw with 2S .ttJxd4 (trying to be 'smart' with 2 S .'iVb8 + ? ! h7 2 6.ttJxd4 doesn't help : 26 .. .'iYe 1 + 2 7 .�h2 'iYxf2 ! (not 2 7 . . . ttJd2 ? 2 8 .h4! 'tixf2 2 9 .'tieS) and White has to fmd the only way to draw with 2 8 .'tic8 ! (Black wins in case of 2 8 .ttJde2 ? ttJe3 2 9 .ttJf4 gS ! 3 0 .ttJce2 gxf4 3 1 .ttJxf4 ttJfl + 3 2 .�h 1 �g3 ! 3 3 .
'iVd 2-f4 !

Securing a powerful position for the white bishop on the a2-g8 diagonal, with long-term pressure on the fl weakness. 1 9.

...

h3-h4

21 .

ttJc3-a4

'iVe8-d8

Black has the bishop pair and a healthy pawn structure. However, White can successfully target the only weakness in Black's camp (fl) and his pieces are coor­ dinating very well, keeping the pressure.

2S . . . �xe4 2 6 .'iVxe4 bxa3 loses a pawn after 2 7 .'ifxc6! na7 (2 7 . . . axb2 ? would blunder a piece due to 2 8 .'tidS ! 'iVf8 29 .'ifxd7) 2 8 .�dS ttJf8 (even worse for Black is 2 8 . . . ttJb6 2 9 .'iYb S �f8 3 0 .bxa3 ) 2 9 .bxa3 and White is a dear pawn up, while still targeting the fl weakness and keeping the initiative. 26.

ttJe4xf6+

27.

b2xa3

ttJ d7xf6

White threatens ttJf3 -eS , further hitting the fl weakness, so Black is forced to shut off the white bishop.

21 .

...

h6-h5

In general not an easy move to play. However, White will have the initiative in the case of 2 1 . . . �c8 2 2 .'iYg4, or 2 1 . . . ttJb6 2 2 .iYg4! �h7 (other moves do not neutralize White 's initiative ei­ ther: 2 2 . . . �h8 2 3 .hS �h7 24.ttJcS , or 22 . . . ttJdS 2 3 .hS �h7 24.ttJeS) 2 3 .ttJcS . 22.

l:tc1 -e1

Control of the e-file is important. 22.

...

b7-b5

27.

...

ttJf6-d5

28.

�b3xd5

c6xd5

In itself this looks like a reasonable deal for Black. A lot of pieces have been ex­ changed, the pawn structure is totally symmetrical and it seems as if Black is about to equalize. However, his troubles are far from over - in fact they are just about to begin. White gets a monster knight on eS , which is dearly superior to Black's virtually useless bishop. It is important to note that the assess­ ment of this position would have been very much different if Black had had the time for . . . fl -f6 , keeping the white knight from its dominant post.

Trying to put up an active defence. 23.

ttJa4-c3

iVd8-b8

24.

�f4-e3

b5-b4

I s o l a t e d P a wns

The rook has done its j ob on the e-file and now the c-file is the one to control - and to penetrate. <;!(g8-h 7

31 .

...

32.

'iWf3-h3!

l:ta6-b6

33.

1:tc1 -c8

'iVd8-d6

34.

'iYh3-g3

White has perfect piece coordination and Black is in terrible trouble. Kasparov's execution is excellent. 34.

...

a 5-a4

Trying to become active with 34 .. J lb 1 + 3 S . �h2 �a6 would result in a probably lost endgame:

a n a lysiS d i agram

3 6 .11f8 ! 'iYfl 3 7 . 'iVf3 'ifh 1 + 3 8 . �g3 a4 (in the case of 38 .. .'iVd 1 3 9 .'ifxdS 'ifb3 + 40.'iVxb3 l::t xb3 + 4 1 .f3 White's d-er is very strong) 3 9 .ttJxfl ! �xfl 40 .'ifxf7 �b3 + 4 1 .f3 'iYe 1 +

�f5-g6

Opposing White's bishop on the a2-g8 diagonal with 1 9 ... �e6 would not solve Black's problems. White would keep the initiative with 2 0 . �e 1 ttJf8 (the tactic 20 . . . �eS does not help, since after 2 1 .dxeS �xb3 2 2 .ttJe4 the white knight will take up a powerful post on d6) 2 1 .dS cxdS 2 2 .ttJxdS . 1 14

20.

2:

25.

ttJc3-e4 !

b4xa3

29.

ttJf3-e5

'iYb8-d8

30.

'iVe3-f3

l::t a 8-a6

31 .

l:te1 -c1 !

analysiS d i agram

42 . \t>f4 ! 'tixh4+ (42 . . .'ifd2 + loses to 4 3 . �fS 'ifc2 + 44 . \t> g S 'iYxg 2 + 1 15

Chapter

W i n n i n g C h es s M i d d l eg am e s

4S .\t>xhS with 'iYg8 and mate next) 43 .�fs 'iYf6 + 44.iYxf6 gxf6 4S J:tf7 + ! � g 8 4 6 . tta 7 ld.xa3 4 7 . � g 6 'It>f8 4 8 .�xf6 �e8 49 .�e6 . 35.

l:tcS-aS

iYdS-eS

3S.

l:taSxa4

iYeS-f5

helpless against the mating threats on the c 1 -h6 diagonal. 43.

l::t b 7-b41

l:td4-c4

Black gets mated in a nice variation af­ ter 43 . . . ld.xb4 44.axb4 d4

QO 1 6 . 1 2 (D3 1 ) Ivan Sokolov

GAME 2 6

Lars Schandorff Reykjavik 2 0 0 1 (5) d 7-d5

1.

d2-d4

2.

c2-c4

e7-eS

3.

ttJb1 -c3

�fS-e7

4.

ttJg1 -f3

ttJgS-fS

5.

'iVd1 -c2 1?

2:

I s o l a t e d P a wns

1 9 8 7 ) 7 .�xf6 �xf6 8 J:td l g6 9 .e3 c6 1 0 . �d3 dxc4 1 l .�xc4 ttJd7 . Now we get what I was talking about. The pawn structure is identical to that in the well-known Tartakower IMakogonov line (see Kasparov-Karpov, Game 2 5 ) , with the difference that the white rook is better placed on d 1 than on the cus­ tomary c 1 , giving White a favourable version of a known line. In Kramnik­ Short, Dortmund 1 9 9 5 , White played a beautiful, instructive game showing all the pros of White's set-up :

analysis d iagram

37.

l:ta4-a 7 1

Preparing a beautiful mating net, as will soon become apparent. 37.

...

l::t b S-b1 +

3S.

�g 1 - h 2

l:tb1 -c1

39.

l:ta7-b7

Preventing 3 9 . . . 'iYb l .

4S .bS ! (the black queen is stuck on fS , having to defend against the mate on f4) 4S . . . d3 46.b6 d2 4 7 . b 7 d 1 'iY 48 .b8'iY iVd2 (only move to prevent the mate on f4) 49 .ttJxg6 'iYxg6 S O .�h8 + iVh7 5 1 . 'iVgxg 7 mate.

39.

...

l:Ic1 -c2

44.

l:ib4xc4

d5xc4

40.

f2-f3

nc2-d2

45.

iYg3-dS!

c4-c3

4S.

'iVdS-d4 1

1 -0

An interesting move, which has been tried many times by a lot of top players. In general, with this move order White tries to get a favourable version of a dif­ ferent type of known lines in the Queen's Gambit Declined. 5.

...

ttJ bS-aS

In the above-mentioned other QGD lines, the text move is a sideline that

41 .

ttJe5-d 7 1

Executing 3 7 . 1:a7 ! .

the

plan

started

41 .

...

�d2xd4

42.

ttJ d7-fS+

�h7-hS

with

Or 42 . . . 'lt>g8 ? 43 .ld.b8. Black's king is now in a mating net and after the d4 rook gets exchanged, Black will be 116

Total domination. After Black's only move to prevent mate on the c 1 -h6 di­ agonal, 46 . . . �h7 , probably the easiest for White is to take 47 . 'iVxc 3 , combin­ ing mate threats with the advance of the a-pawn.

never looked logical to me. Let me give a short introduction to the way play may continue in the main lines : A) S . . . dxc4 6 . e4 ttJc6 7 .eS ttJb4 8 . 'iVb l ttJfdS 9 .�xc4 and a position has been reached that, as regards pawn structure, is similar to the 3 .e4 line of the Queen's Gambit Accepted, with slightly better chances for White; B) 5 . . . 0-0 6 .�gS h6 (6 . . . cS 7 .dxcS dxc4 8 .e4 'iYaS 9 .eS ttJdS 1 0 .�xc4 ttJxc3 1 1 . 0-0 'iYxcs 1 2 .'iYxc3 was sub­ ject to discussion in the World Champi­ onship match Kasparov-Karpov, Sevilla

analysis d i agram

1 2 .h4! 'iYe7 1 3 .a3 ! (two rook-pawn moves on both flanks are part of a well- conceived attacking plan) 1 3 . . . �g7

1 4.�a2

b6

1 5 .�b l

h5 ?!

(giving away the very important g S -square is - in a higher sense - going to cost Black the game. Short was prob­ ably better advised to keep cool and opt for I s . . . 11d8 1 6 .hS gS 1 7 .iYh7 + �f8 1 8 . ttJe4 fs ! (not 1 8 . . . ttJf6 ? 1 9 .tDxf6 'iYxf6 2 0 . ttJ e S and White wins : 2 0 . . . �b 7 2 1 . ttJ g4 'if e 7 2 2 . ttJxh6 ) 1 9 . ttJ g 3 cS ! , creating counterplay) 1 6 . 0-0 �b 7 1 7 .ttJgS (putting a mon­ ster on gS) 1 7 . . J:tfd8 1 8 .�a2 ttJf6 ? ! ( 1 8 . . . ttJf8 , keeping the e 6 / g 6 weak­ nesses under control, was definitely a move to consider) 1 9 .e4 ttJg4 (the 117

C h ap t e r

W i n n i n g C h es s M i d d l eg am e s

knight does very little on this 'active' post) 2 0 .eS �d7 2 1 .ttJe2 ! (setting the stage for a tactical execution) 2 1 . . . �ad8

1 3 .lig 1 iiLfS 1 4.iiLg2 'ifc8 . Black had strong compensation and went on to win in Khalifman-Topalov, FIDE Grand Prix, Dubai 20 0 l . S.

analysis d iagram

2 2 . iiLxe 6 ! fxe 6 2 3 . 'iVx g 6 ttJxeS 24.�h7 + �f8 2 S .4:Jf4, Black resigned. Given the fact that White loses some time making a queen move relatively early in the game, the most principled response for Black here is, in my opin­ ion, to take immediate action in the centre with S . . . cS ! 6 . dxcS d4. Let me give one instructive example of how play may continue in that event: 7 .ttJbS

a2-a3!

Simple and strong, eliminating Black's main idea ( . . . ttJa6-b4 with tempo) . Af­ ter other moves Black would get active play related to ttJa6-b4, j ustifying his previous move, for example : A) 6 .iiLgS ttJb4 7 .'iYb 1 cS ! 8 .dxcS d4 9 .a3 ttJc6 1 0 .ttJe4 ttJxe4 1 1 .iiLxe 7 'iYxe 7 1 2 .'iYxe4 'ifxcs with excellent play for Black; B) 6.g3 ttJb4 7 .'iVb 1 dxc4 8 .iiLg2 0-0 9 . 0-0 lib8 1 0 .a3 ttJc6 1 1 .1:ld 1 ttJaS and in this Catalan-type position Black has a good game. S.

...

c7-c5

7.

e2-e3

c5xd4

S.

e3xd4

d5xc4

9.

jlf1 xc4

0-0

0-0

ttJaS-c7

1 0.

which has also been tried already, is not that dangerous after all, for instance : 1 1. .�d 1 b6 1 2 .ttJeS iiLb 7 1 3 Jid3 bS ! ( creating immediate counterplay) 1 4.iiLa2 as ! l s .iiLgS b4 1 6 J Ih3 ttJce 8 ! (one o f the advantages o f having the knight on c7) 1 7 .ttJbS (the 'aggressive' 1 7 .ttJg4 is easily parried by 1 7 . . . g6) 1 7 . . Jlc8 1 8 .'iVe2 ttJe4 1 9 .iiLxe 7 'iYxe 7 20 .ttJa7 lic7 2 1 .ttJbS lic8 2 2 .ttJa7 lic7 23 .ttJbS draw, Topalov-Karpov, Dos Hermanas 1 9 9 9 . 11.

...

b7-bS

12 .

.l:ra1 -d1

jlcS-b7

1 3.

ttJf3-e5

ttJc7-d5

1 4.

l:If1 -e1

.l:taS-cS

15.

'iVc2-b3

a 7-aS

2:

I s o l a t e d P a wns

2 0 .'iVxa6 1:lc2 2 1 .lde2 , or 1 9 . . . aS ? 2 0 .ttJd7) 2 0 .dxeS as 2 1 .�d4 l:tcs 2 2 .'iVd 1 'iVa8 ! (not 22 . . . 'iVd7 2 3 .b4 axb4 24.axb4 l:tbS and the rook is stranded on b S ) 2 3 .f4 (or 2 3 .b4 axb4 24.axb4 :tc4) 2 3 . . . 'iVc6 ! 2 4.fs !:tc2 with counterplay) 1 9 . . . tIc4 2 0 .ttJe3 lixd4 2 l . ttJxdS ttxd 1 22 J:txd 1 b S . It could well be that 1 7 . . . iiLxdS was a better option for Black.

Black is ready to start kicking back the white pieces with . . . b6-bS , so it is time for White to change the pawn structure. 1 S.

ttJc3-a4!

1 9.

Me1 xe5 !

jlfSxe5

It is important to keep the d4/ dS struc­ ture intact, in which Black's b7 bishop is closed in. 1 9.

...

jlb7-cS

20.

ttJa4-c3

f7-fS

21 .

.J:te5-e3

White controls the e-file, while Black is saddled with a ive light-squared bishop and weak queenside pawns. 1 S.

11. a n a lysis d i agram

7 . . . eS ! 8.ttJxeS a6! 9 .ttJa3 (9 .ttJd6+?? is a terrible blunder due to 9 . . . iiLxd6 1 0 .cxd6 'iVaS + , winning a piece) 9 . . . 0-0 (for the pawns Black has a massive lead in devel­ opment) 1 0 . g 3 (or 1 0 .ttJd3 iiLfS ) 1 O . . . �aS + 1 l .iiLd2 'iVxcs 1 2 .ttJd3 iVc6 1 18

jlc1 -g5

A new plan at the time. I thought that White should have a reasonably good version of the classical type of isolated pawn-position here if he continues with normal piece development, due to the fact that Black has good control of the dS -square, but no real counterplay. The direct attack against the black king,

jlg5xfS!

.ie7xfS

Forced. 1 6 . . . ttJxf6 ? loses to 1 7 .dS ! exdS 1 8. ttJxdS iiLcs ( 1 8 . . . ttJxdS loses after 1 9 .iiLxdS iiLxdS 2 0 .lixdS 'iVc 7 2 l .lid7 'ifcs 2 2 .ttJd3) 1 9 .4:Jf4 'iVc7 2 0 . �d7 ! ttJxd7 2 l .iiLxf7 + �h8 2 2 .ttJfg 6 + hxg 6 2 3 . 'iYh3 mate. 1 7.

jlc4xd5

21 .

...

.l:lfS-f7

22.

lld 1 -e1

.i:lcS-bS

eSxd5

White would also have a slightly better game after 1 7 . . . iiLxdS 1 8 .ttJxdS exdS 19 .ttJg4 (Black would probably equal­ ize after 1 9 .'iYa4 iiLxeS ! (not 1 9 . . . 'iVd6 ? 1 19

Chapter

W i n n i n g C h es s M i d d l eg am e s

23.

�e3-e6!

'fi'dS-d7

24.

h2-h3

g 7-g6

2S.

'ii b 3-b4

White has achieved total domination and the black king is coming under di­ rect attack. The game is over.

Targeting the weak dark squares in Black's camp. 2S .

...

�gS-g 7

32.

...

�g7-f7

33.

l::t e 3-eS!

l::t d 7xdS

34.

'iYa3xdS

ilLcSxa4 'tt>f 7-gS

2S.

a3-a4

l:1bS-b7

3S.

l:ieSxfS+

27.

�eS-d6

as-aS

3S.

'ti'dSxdS+ �gS-g7

2S.

'iYb4-a3

'iVd 7-cS

37.

'ti'dS-eS

29.

�e1 -eS

ilLcS-d7

3 0.

�eS-e3

iLd 7-cS

�g 7-g8

TD 4. 1 6 (D3 4)

GAME 2 7

Alexander Beliavsky Miguel Illescas Cordoba Linares 1 9 9 0 (2) 1.

d 2-d4

2.

c2-c4

e7-e6

3.

tLlb1 -c3

c7-cS

4.

c4xd5

e6xdS

S.

tLlg1 -f3

tLl bS-c6

6.

g2-g3

tLlgS-f6

7.

�f1 -g 2

.ifS-e7

8.

0-0

0-0

d 7-dS

9.

�c1 -gS

cSxd4

1 0.

tLlf3xd4

h 7-h6

11.

�gS-e3

�fS-eS

As usual, Illescas plays the main move. As explained in Petrosian -Sky (Game 2 3 ) , 1 1 . .. �g4 is definitely worth analysing. 3S.

tLlf4xgS

'iYcS-c1 +

39.

�g 1 -h 2

h 7xgS

tLlc3-e2!

40.

rtfSxg6+

�gS-fS

Time to improve the knight.

41 .

'iVeS-hS+ �fS-e7

31 . 31 .

..

32.

tLle2-f4

.

nf7-d7

42.

'ti'hS-h7+ �e7-dS

43.

'iYh7xb7

1 2.

'iYd 1 -c2

2:

I s o l a t e d P a wns

Kasparov- Illescas, Sokolov-T. L.Petro­ sian, etc.) 1 S . . . bxc6 1 6 .�d4 �b4 (hit­ ting the only white weakness, the e2 pawn, Black wants to create counter­ play. However, he also had a standard plan at his disposal : I 6 . . . tLld 7 , with good play) I 7 . l::r d 2 'iYe7 ( 1 7 . . . tLld 7 ? would now b e a blunder due to 1 8 .�xg 7 ! �xg 7 1 9 . 1:Id4) 1 8 .a3 �aS 1 9 .b4 �b6 2 0 .e3 'iVe6 (preparing . . . c6-cS) 2 1 .'ifb2 �xd4 2 2 . l::rxd4 cS 2 3 .bxcS l::r x cs with equal play. Another option that Beliavsky has also experimented with is 1 3 .l::r ad l 'iYd7 (the regular move I 3 . . . �f8 certainly comes into consideration) and: A) If White now tries to immediately target the dS pawn with 1 4.tLlb3 , Black gets good, dynamic play after 1 4 . . . �e6 1 S .tLlcs �xcS 1 6 .jLxcS .

A sideline that Beliavsky has experi­ mented with on a number of occasions. 1 2.

...

iLc8-g4

1 -0

Structure 2 . 7 a n a lysis d i agram

1 3.

Original rook transfer - kingslde attack

One more example emerges from the main line of the Tarrasch Defence. Illescas Cor­ doba shows a very original plan of attackfor the side with the isolated pawn.

1 20

h 2-h3

The first time, Beliavsky opted for 1 2 .iYc2 in his well-known 1 9 8 3 Can­ didates' match against none other than Garry Kasparov. Their sixth game con­ tinued 1 3 J:lfd I �f8 (invariably a useful move here) I 4. �ac l 1:[c8 I S .tLlxc6 (a common plan, already shown in previ­ ous examples like Kramnik-Illescas,

White now has the bishop pair, com­ bined with Black's potential isolated dS pawn weakness. This pawn, however, could also become an asset, creating a lot of tactical possibilities for Black, for example after 1 6 . . . d4! and now: A I ) 1 7 . 'lWa4 nab 8 ! 1 8 . �xc6 ( 1 8 .�xd4? is a blunder due to 1 8 . . . bS 1 9 .�xc6 ( 1 9 .tLlxb S ? tLlxd4) 1 9 ... 'iYxc6 2 0 .'ii'b 4 �h3 ) 1 8 . . . 'iVxc6 1 9 .'iVxc6 bxc6 2 0 .�xd4 nxb2 with about equal chances; 121

Chap ter

W i n n i n g C h es s M i d d l e g a m es

A2) 1 7 .tDb5 �f5 1 8 .iYb3

1 5 .tDxd5 ?! tDxd5 1 6 .�xd5 : 1 6 . . . !:txe3 ! (this blow is worth ing - it works most of the time !) 1 7 .�xc6 (or 1 7 .fxe3 'ifg5) 1 7 . . . bxc6 1 8 .fxe 3 'iYg5 with a strong initiative. 1 4.

...

'iVd8-c8!

1 5.

�g 1 - h2

.ie7-f8

A standard plan, but this time it also prepares an unusual attacking method. 1 6. a n alysis d i agram

1 8 . . . d3 ! 1 9 .exd3 �e6 2 0 .'iVa3 �h3 and Black has good compensation for the sacrificed pawn; B) 1 4.f3 jLh5 1 5 .�f2 �g6 1 6 .'ifa4 a6 1 7 .tDxc6 bxc6 1 8 J :Id2

l1f1 -e1 ? !

In the event of 1 6 .tDxc6 bxc6 1 7 .�d4, Black gets good play with 1 7 . . . tDh7 ! , threatening to start rolling his central pawns. Black has an approximately equal game after 1 6 . tDxd5 ! ? tDxd5 1 7 . �xd5 jLxh3 1 8 .11fe 1 tDe5 ! ( 1 8 . . . �b4? is wrong due to 1 9 .'ifb3 ) 1 9 .�xc8 �xc8 2 0 .\tg l tDg4 2 1 .�c l (if 2 1 .�d2 , 2 1 . . . �d8 ! 2 2 .e4 ttJf6 2 3 .jLa5 11e8) 2 1 . . . �c5 . Now Black can develop his attack nicely.

20.i�J3 �g2 +) 2 0 . . .l::t xh4 ! 2 1 .gxh4 �d6+ and Black soon delivers mate. B) The other defensive option was 1 7 . tDf3 �h5 (here, the classic exchange sacrifice 1 7 . . . 1;Ixe3 again comes into consideration, with strong compensa­ tion after 1 8 .fxe3 'iYb8 ! 1 9 .e4 tDh5 ) 1 8.tDh4 and now: B 1 ) 1 8 . . . g5 ? does not work due to 1 9 . tDxd5 tDxd5 2 0 .�xd5 ; B2) White has some compensation for the exchange after 1 8 . . . d4 1 9 .�xd4 lbxd4 2 0 . Jdxd4 g5 2 1 . 11xd7 ! 'iYxd7 22 .�d l 'iVc7 2 3 .tDf5 ; B3) 1 8 . . . �d6? i s a blunder due to 1 9 .tDxd5 tDxd5 2 0 . 11xd5 llxd5 2 1 .�xd5 �xh3 2 2 .�xf7 + ; B4) 1 8 . . . 4.Jb4 1 9 .'li'b3 �d6 2 0 .f4 as , and in this highly unusual position Black's chances should be preferred. 1 7.

...

lle5-h5!

1 8.

�h2-g1

.id7xh3

1 9.

.ig2-f3

1 9.

�h5-e5

...

19 . . . �g4 is no more than equal after 20J�xh5 �xh5 2 1 .�xh5 tDxh5 22 .tDf3 ; 1 9 ... tDg4 ! ? was interesting. a n alysis d i agram

20.

1 8 . . . 'ife 6 ! (preparing a standard knight transfer) 1 9 .�c l tDd7 and Black's pieces are harmoniously developed, while White is not able to target Black's potential weaknesses effectively. Black is slightly better, Beliavsky-T.L . Petrosian, Moscow Aeroflot 2 0 0 5 . 1 3.

1 4.

.ig4-d7

...

Better and 1 3 . . . �e6 .

more

dynamic

than

lla 1 -d 1

14 .lifd 1 would move the kingside rook away from its king and might allow some classic tactics here after 1 4 . . . �f8 1 22

tbd4xc6

b 7xc6

In the case of the classic exchange sacri­ flce 20 .. .l::t xe3 , White now has . . . 1 6.

...

21 .

.ie3-d4

21 .

...

�h3-f5 !

Other moves, like 2 1 . . . l:tg5 , would al­ low White to damage Black's kingside pawn structure and create counterplay with 2 2 .�xf6 gxf6 2 3 .e4! d4 2 4.ttJe2 c5 2 5 .tDf4. e2-e4

�e5xe4

Also good was 2 2 . . . dxe4 2 3 .�xe5 exf3 24.'ifd2 tDg4 and with the white rook horribly misplaced on h I , Black is clearly better. 23.

tbc3xe4

tbf6xe4

Black has two pawns for the exchange and clear domination. The white rook will remain misplaced on h I for quite some time. 24.

l:te8-e5!

I s o l a t e d P a wns

2 1 .tDxd5 ! (2 1 .fxe3 bxc6 would, as usual, be very good for Black) 2 1 ... tDxd5 2 2 .�xd5 with a messy position, since 22 ... bxc6? is bad due to 2 3 .'iYxc6.

22.

The point of White's defence.

2:

'iVc2-e2

c6-c5?

This rook manoeuvre can also be used in similar positions , which is the main reason that I have selected this game for the book. 1 7.

lle1 -h1 D

White was probably rather surprised by Black's last move and anyhow, it is not easy to find a good defence. For example: A) 1 7 .tDxc6 bxc6 1 8 .�d4 ? loses to 1 8 . . J lh5 1 9 .�xf6 �xh3 2 0 .�h4 (or

a n a lysis d i agram

1 23

Chapter

W i n n i n g C h ess M i d d l eg am e s

I s o l a t e d P a wns

Structure 2 . 8

A tactical blunder. Any regular move, like 24 . . . 'iYe 6 for instance, would have maintained a clear advantage. 2S.

2:

�d4-e3?

White lets the opportunity 2 5 .�xg7 ! ! slip: 2 5 . . . �xg 7 2 6 .�xd5 'iYe6 2 7 J:[xf5 'iYxf5 2 8 .�xe4 'iYe6 2 9 .'iff3 �d8 3 0 .1::t h5 with the better game. 2S.

...

'iYcS-e6

Black is again firmly in control. 26.

�g 1 -g2

an alysi s diagram

dS-d4

26 . . . �e8 is also good. 27.

�e3-c1

1:IaS-eS

2S.

'ife2-bS

a 7-a6

29.

'ifbS-b7

. .. there is 3 6.g4! �g6 3 7 .'ifb7 �d8 3 8 . 'ifc7 , and the rook has no squares left. 36.

�d S-b7

In isolated pawn-positions, it is a common s trategy for the side playing with the isolani to carry through a pawn break (in this case d4-d5), and after exchanging a few pieces obtain a symmetric pawn structure with the better placed pieces in the centre, retaining some initiative. At first sight it often seems that after a few good moves this initiative will evapora te, full equality will be reached and the game will inevitably be drawn. Well, more often than not, for the defending side those few good equalizing moves ' are nowhere to befou nd. Here are two instructive examples.

Trying to get some play. 29.

...

30.

l:ld 1 -e1 1

�e6xa2

30 .�he l ? loses to 30 .. J::te 7 ! 3 1 .'iVa8 'iYb3 !. 30.

...

31 .

�b7-b6

ttJe4-d6

3 7.

31 .

...

.l:teS-cS?

It was correct to return one pawn and coordinate his pieces with 3 1 . . . �e4 ! 3 2 .'ifxc5 'ifa4, and Black is better. 32.

iiLc1 -f4!

Now White has created counterplay.

l:rh 1 -e1 ?

After having created a mess, White does not take his chance. After 3 7 . .t.c 1 ! Black would have had to give back material with 3 7 . . . �c2 3 8 . lIxc2 'iYxc2 3 9 .Vi'xb6 and a draw is the likely outcome. 3 7.

...

3S.

iiLf4-c1

'iYb3-b5

One move too late. The black queen is now defended and White is totally lost. 3S.

...

ttJb6-c4

Black simply has too many pawns.

32.

...

ttJd6-c4

33.

'iVb6-b7

�a2xb2

39.

�b7-dS

iiLfS-d7

34.

'ifb7-d5

�b2-c2

40.

l::t e 2-e7!

iiLfSxe7

3S.

l:le1 -e2

ttJc4-b6

41 .

1:te1 xe7

iiLd7-eS

42.

iiLf3-e4

�gS-fS

White has serious threats. In the case of 3 5 . . . 'ifd3 . . . 1 24

d4-d5 break to get a pawn symmetry with a dominant piece I n the centre

'iVc2-b3

White resigned.

QO 7 . 1 0 (D 5 8)

GAME 2 8

Alexey Vyzhmanavin Alexander Beliavsky Novosibirsk 1 9 9 5 ( 7 ) 1.

d2-d4

ttJgS-f6

2.

c2-c4

e7-e6

3.

ttJg1 -f3

d 7-dS

4.

ttJb1 -c3

�fS-e7

S.

�c1 -gS

h 7-h6

6.

�gS-h4

0-0

7.

e2-e3

b7-b6

S.

iiLf1 -d3

iiLcS-b7

9.

0-0

ttJ bS-d7

Now Vyzhmanavin plays his favourite variation against the Tartakower, which is still considered one of the main lines today.

1 0.

'iYd 1 -e2

White may also start with 1 O .�g3 and after 1 0 . . . c5 opt for 1 1 .cxd5 ( 1 1 .'ti'e2 would bring us back to the game, while after 1 1 . �c 1 l2Je4 1 2 . cxd5 exd5 1 3 .dxc5 l2Jxg3 1 4 .hxg3 bxc5 a strategi1 25

Chapter

W i n n i n g C h es s M i d d l eg am e s

cally different type of position is reached, where Black has the bishop pair and two parallel hanging pawns in the centre) and now: A) 1 1 . . . ttJxd5 1 2 Jk 1 (a long time ago, in Wijk aan Zee 1 9 9 5 against Paul van der Sterren, I seriously mixed up the variations and played 1 2 .'iVe2 here, only to discover that after 1 2 . . . cxd4 1 3 .exd4 ttJ 7f6 1 4. l::t. ac 1 �c8 1 5 .ttJe5 ? (this ' active' move makes matters much worse) 1 5 . . . ttJxc3 1 6 . bxc3 'Ii'd5 ! 1 7 .ttJf3 (sad necessity, as 1 7 .f3 ? ? loses to 1 7 . . J ixc3 1 8 . I;'Ixc3 'iVxd4 + ) 1 7 . . . ttJe4! 1 8 .c4 ttJxg3 1 9 .hxg3 'iVd6 I was left without a bishop pair and with a weak d-pawn to worry about - which I lost later on in the process - and had to fight for another five hours to escape with a draw) 1 2 . . . cxd4 1 3 .exd4 (White is better off not entering the isolated pawn position here and opt for the dy­ namic 1 3 . ttJxd5 �xd5 1 4. e4 . The tacti­ cal tussle is, however, likely to result in an equal position after 1 4 . . . �b 7 1 5 .�c7 'iVe8 1 6 .ttJxd4 �c8 ! 1 7 .�b5 a6 1 8 .�a4 b5 1 9 .ttJxb5 ttJc5 2 0 .ttJd6 'iYxa4 2 1 .�xc5 'iYxd l 2 2 .ldxd l �xe4 and a draw was soon agreed in Atalik-Beliavsky, Ulcinj 1 9 9 7 ) 1 3 . . . l::I c 8 . This version of the isolated pawn -posi­ tion is very comfortable for Black, since White is not able to develop the desired initiative in order to compensate for his pawn weakness. The game Yermolinsky-Beliavsky, Groningen peA 1 9 9 3 , excellently dem­ onstrates the deficiencies of White's strategy: 1 4.�b 1 ttJ 7f6 1 5 .�e5 ttJxc3 ! 1 6 J:lxc3 (or 1 6 .bxc3 'iVd5 1 7 .'iYd3 'iYc4 1 8 .'iVc2 �e4 and again, White has no activity to compensate for his weak pawns) 1 6 . . . �d5 ( 1 6 . . . l:!.xc3 1 7 .bxc3 1 26

'ifd5 1 8 :iYd3 nc8 was also good) 1 7 .Ile3 (moving away the rook so he will only have to worry about one weak pawn - d4 - instead of two - c3 and d4) 1 7 . . J:tfd8 1 8 . �fe l 'iVb5 1 9 .ttJd2

analysis d i agram

analysis d i agram

24 . . . �f6 ! 2 5 .�xe6 (White has already burned his bridges and this sacrifice is desperation caused by a strategic disas­ ter) 2 5 . . . l::I xd4 and Black soon won; B) Another reply is 1 1 . . . exd5 . After 1 2 .�c l ( 1 2 .ttJe5 has also been played)

I s o l a t e d P a wns

1 2 . . . ttJe4 1 3 . dxc5 ttJxg 3 1 4 .hxg 3 bxc5 1 5 .'iVe2 , another position where Black has the bishop pair and parallel hanging pawns in the centre has been reached. A game Zhukova-Skripchenko, Krasnotu­ rinsk 2 0 0 3 , showed how easily things can go wrong here. White scored a quick, instructive victory after 1 5 . . :iVb6

13.

1 9 . . . ttJd7 ! 2 0 .a4 'ifa5 ! (better than 2 0 . . . 'iVb4 ? 2 1 .�b3 'iVxa4 ? 2 2 .�xg 7 ! and White wins) 2 1 .ttJc4 (understan­ ding that the strategic battle has been lost, Yermolinsky is desperately looking for tactics, trying to complicate mat­ ters) 2 1 . . . 'Ii'b4 (2 1 . . .1::t xc4?? would be a beginner's blunder due to 2 2 .'ifd3) 2 2 .iVd3 g6 2 3 .b3 ttJxe5 24J:txe5

2:

analysis diagram

1 6 .e4! d4 1 7 .ttJa4 'iYe6 ? 1 8 .b4! cxb4 1 9 .�c7 �c8 2 0 .e5 ! �b8 2 1 .�c4 'if g4 2 2 .e6. 1 0.

...

c7-c5

11.

.ih4-g3

c5xd4

Black decides to go for an isolated pawn -position. His other option was to keep the tension in the centre. Later on, he will likely play with two parallel hanging pawns in the centre himself, after 1 1 . . . ttJe4 1 2 .cxd5 exd5 1 3 .�ad l 'if c8, keeping the tension and planning - after . . . ttJdf6 - to transfer the queen to e6, is a sound positional plan here and will - along with other possibilities for Black - be examined later in this book in Lautier-Short (Game 40) , un­ der the section 'two parallel hanging pawns in the centre' . 1 1 . . . dxc4 1 2 .�xc4 a6 1 3 .a4 cxd4 14.exd4 would transpose to the game. 1 2.

e3xd4

1 3.

.id3xc4

d5xc4

..

.

a 7-a6

Alternatively, Black may start with 1 3 . . . �b4 in order to take on c3 , creat­ ing a new potential pawn weakness in the white camp while keeping full con­ trol of the d5-square. This could well be the safest and best plan here, for in­ stance : A) The attempt to make use of the fact that the b 5 -square is not protected and to develop an immediate initiative with 1 4.ttJb5 does not seem to bring White anything special: 1 4 . . . ttJe4 1 5 . a3 ttJxg 3 1 6 .hxg3 �e 7 1 7 . l::I fd 1 ttJf6 with ap­ proximate equality in Ftacnik-Beliavsky, Vienna 1 9 8 6 ; B ) 1 4.�ac l i s answered by 1 4 . . . �xc3 1 5 .bxc3 ttJe4 1 6 .�f4 l::I c 8 1 7 .ttJd2 ttJdf6 . The black pawn is well placed on a7 and the position was about equal in Arencibia-Beliavsky, Elista Olympiad 1 998. I t i s indeed better for Black here not to include 1 3 . . . a6 1 4.a4, since this inclu­ sion often forces him to keep his rook on a8 to defend the a6 pawn. 1 3 . . . ttJh5 is another option, eliminat­ ing White's dark-squared bishop but allOWing the d5 break : 1 4 .d5 ttJxg 3 1 5 .hxg3 exd5 1 6 .ttJxd5 ! (the old school probably thought that White should eliminate Black 's bishop pair, so in Gligoric-Petrosian, Zurich 1 9 6 1 , 127

Chapter

W i n n i n g C h es s M i d d l eg a m e s

Black has missed his opportunity to equalize. 1 9 . . .lie 8 ? is no longer good due to 2 0 .iYa2 ! �d6 2 1 .ttJc7 ! �xc 7 22 .�xf7 + cj;>h8 2 3 .�xe8 , winning.

White decided to exchange the light­ squared bishops with 1 6 .�xdS �xdS 1 7 . ttJxdS . However, without the active bishop on c4 White was not able to de­ velop any meaningful initiative and Black was OK after 1 7 .. .lle8 1 8 . blfd 1 �cS 1 9 .'ifc4 �c8) 1 6 . . . �d6 1 7 J :tfd l ttJf6 1 8 . ttJe3 with an initiative for White. 1 4.

a2-a4

l::ra 1 -c1

ttJf6-h5

Black follows a standard plan, which al­ lows a standard reaction. 1 28

l::tf 1 -d1

I s o l a t e d P a wns

22.

ttJf3-h4!

�b7-e4

23.

'ife2-d 2 !

�d6-c7

ttJd 7-f6

�e7- b4

Black may also decide to eliminate White 's dark-squared bishop immedi­ ately with 1 4 . . . ttJhS . Just like in the game, that plan has the drawback that Black loses control of the d S -square, so it allows the typical break I S .dS , for example : l S . . . ttJxg 3 1 6 . hxg3 exdS 1 7 . ttJxdS ! (it is a good idea for White to keep the light-squared bishops on the board here , since it makes it easier for him to develop the initiative. White has only a small plus after 1 7 .�xdS �xdS 1 8 .ttJxdS - it should be noted that should Black choose for this plan , then he is better off to im­ plement it with his pawn on a 7 , as in Gligori c-Petrosian) game the 1 7 . . . �d6 1 8 J ifd 1 . Here it also appears rather difficult for Black to neutralize White 's initiative. In the game Sasikiran -Barsov, Doha 2 0 0 3 , White followed a plan similar to Vyzhmanavin's, and was successful af­ ter 1 8 . . . ttJf6 1 9 .ttJe 3 ! 'iYe7 2 0 .ttJh4 l';Ifd8 2 1 .ttJhfS 'iYf8 2 2 . ttJg4! (the white knights cause a lot of trouble for Black here) 2 2 . . . ttJe4 2 3 . �d4 lIe8 2 S J :1d l Jlad8 �c7 2 4 . 'iYd3 2 6 . �xf7 + ! cj;>xf7 2 7 . 'iV c4 + cj;> g 6 2 8 .'iVxc 7 . 1 5.

20.

2:

1 6.

d4-d5 !

ttJh5xg3

1 7.

h2xg3

e6xd5

1 7 . . . �xc3 ? is a typical blunder that loses to 1 8 .dxe6 �b4 1 9 .11fd l �c6 2 0 .ttJeS . 1 8.

ttJc3xd5

As in several previously-mentioned ex­ amples , White believes that he has better chances for an initiative with the light -squared bishops on the board. Black would gradually equalize in the event of 1 8 .�xdS 'iYb8 ! .

21 .

ttJd5-e3!

'iYd8-e7

It may be surprising, but the white at­ tack is extremely strong and Black's po­ sition can collapse at any moment. 2 1 . . . 'iYb8 does not help either. The fol­ lowing variation is a good illustration of White's attacking potential:

White conducts the attack beautifully, using the potential of his forces to the maximum. 2 3 . . . �xb4 was probably better than the text, but would not solve the problems either: 24. ttJefS ! �xfS (or 24 . . . iYcS 2 S .'iYf4) 2 S .ttJxfS 'iYeS (or 2 S . . . 'iYcs 2 6 . iYf4) 2 6 .ttJxh6 + gxh6 2 7 .'iYxb4 and White has a large advan­ tage. 24.

ttJh4-f5

�e4xf5

Black decides to eliminate one of the white knights. On the other hand, the immediate 24 . . . 'ifeS would allow 2 S .f4 iYe8 2 6 .bS ! (2 6.'iYc3 ? would be a mis­ take, allowing Black counterplay with 2 6 . . . bS ! , opening the a 7 -g 1 diagonal) 2 6 . . . aS ( 2 6 . . J :td8 loses to 2 7 .iYc3 as 2 8 .ttJxh6+) 2 7 .iYc3 and White has a winning advantage. 1 8.

...

�b4-c5

A critical moment in the game. Black could have equalized with the dynamic 1 8 . . .lle 8 ! 1 9 .iYc2 bS ! 2 0 .axbS axbS 2 1 .ttJxb4 bxc4 2 2 .'iYxc4 �xf3 2 3 .gxf3 ttJ e S 2 4 . 'if c3 J:Ib 8 ! 2 S .f4 llxb4! 2 6 .'iVxb4 (or 2 6 .fxeS 'lWb8) 2 6 . . . ttJf3 + 2 7 .Wg2 'ifdS 2 8 . Wh3 ! 'ifhS + 2 9 . Wg2 ttJh4+ 3 0 . gxh4 'iYg4+ with a draw. 1 9.

b2-b4

�c5-d6

a n a lysis d i agram

22.ttJfS ! �xb4 2 3 .iYb2 �xf3 (23 . . . �cS loses to 24.ttJxh6+ gxh6 2 S .'iYxf6 iYxg3 26.�d7 ! �xf2 + 2 7 .Wfl ) 24.gxf3 �cS 2S .�g2 (White's pieces are coordinating perfectly and Black is helpless against the straightforward threat of 2 6 .ttJxh6+) 2S ... bS (or 2S ... ttJe8 2 6 .�dS ! .tIa7 ? 27.!:txcS) 2 6 .�dS bxa4 2 7 .ttJxh6+.

25.

ttJe3xf5

26.

�c4-d3!

'ife7-e5

Due to his poor piece coordination, Black is not able to defend against White's threats. After missing his equal­ izing opportunity with 1 8 . . J le 8 , he re­ mains 'one tempo short' for the rest of the game. Note the attacking power and the great coordination of White's fS knight and light-squared bishop. 26.

...

�c7-d8

1 29

Chapter

W i n n i n g C h ess M i d d l e g a m es

Other moves also lose, for example : 2 6 . . . tDdS 2 7 . f4 'iY e 6 ( 2 7 . . . 'iYf6 2 8 .�e4) 2 8 .'iYb2 or 2 6 . . J�fd8 ? 2 7 .f4. 27.

lld1 -e1

'iYe5-d5

White mates in case of 2 7 . . . 'iVb8 2 8 . tDxh6 + ! gxh6 2 9 . 'iYxh6 'ifd6 3 0 .�h 7 + �h8 3 1 .�fS + <J;>g8 3 2 .�c4 .

4.

�f1 xc4

e7-e6

5.

tDg1 -f3

c7-c5

6.

0-0

a 7-a6

7.

�c4-d3

One of the main moves here, the idea being that White is ready to answer 7 . . . bS with 8 .a4 . 7.

...

tDb8-c6

A) 7 . . . cxd4 is a rather interesting idea that Pre drag obviously had in mind, but he found it a bit risky, so he let his brother try it. Nebojsa Nikolic does not play chess at an international level any­ more, but was a reasonable 1M during his active years. 8 . exd4 and now:

28.

.t1c1 -c4 !

Total domination. Game over. 28.

...

:U8-e8?

Black blunders in a hopeless position. However, other moves also lose, for example : 2 8 . . . 'ifd7 2 9 . �d4 'iYxa4 3 0 .tD xh 6 + ! gxh6 3 1 . 'iYxh 6 ; or 2 8 . . J � a 7 2 9 . �h4! hS (otherwise 3 0 .tDxh6+) 3 0 . �d4 'iVa8 3 1 .tDxg 7 ; or 28 . . . tDe4 2 9 .'iYe2 tDf6 3 0 Jld4 'iYb 7 3 1 . 'iV d 2 and the next move is 3 2 .tDxh6 + . 29.

l:te1 xe8+ tDf6xe8

30.

tDf5-e7+!

1 -0

Black loses his queen after 3 0 . . . �xe 7 3 1 .�h 7 + .

Q G 1 4 . 5 (D2 7 )

GAME 2 9

Ivan Sokolov Predrag Nikolic Elenite 1 9 9 3 (4) 1.

d2-d4

2.

c2-c4

d5xc4

3.

e2-e3

tDg8-f6

1 30

d 7-d5

analysis d i agram

8 . . . g6 ! ? An original plan. Black puts his bishop on g 7 and with his pawn already on g 6 , Black's kingside is considerably less vulnerable to attack. However, this plan takes time and allows White to ex­ ecute the thematic d4-dS thrust: 9.tDc3 �g7 1 0 .�gS h6 and now: A I ) White can opt for 1 1 .�xf6 , part­ ing with his bishop pair, but winning some more time, for instance : 1 1 . . . �xf6 (in the case of 1 1 . . . 'iYxf6 White is clearly better after the standard 1 2 . dS ! 0 - 0 1 3 J :Ie l �d8 1 4 .�c 1 ) 1 2 . lie l ( 1 2 . 'iYd2 �g 7 1 3 .dS 0-0 1 4 .�e4 would not be effective due to 1 4 . . . fS ! l S .�c2 eS 1 6 .fife 1 tDd7 and

2:

I s o l a t e d P a wns

with a bishop pair and his mobile pawns in the centre, Black more than compensates for the advanced white d-pawn ; 1 7 .d6 e4 1 8 .tDd4 tDeS and Black is better) 1 2 . . . 0-0 1 3 .'iVd2 �g 7 1 4.�ad 1 (With all his pieces well devel­ oped, White is ready for the d4-dS thrust) 1 4 . . . 'iYaS (White is better in the case of 1 4 . . . tDc6 I S .�e4 tDe7 1 6 .dS ! exdS 1 7 .tDxdS �e6 1 8 .tDf4 'iVxd2 1 9 .�xd2 ) l S .�c4 tDd7 1 6 .dS ! tDb6 1 7 .�b3 exdS (with all his pieces well developed and coordinated, tactics should work for White. For example : 1 7 . . . �d8 ? 1 8 .'iYf4 ! exdS 1 9 . tDxdS ! tDxdS 2 0 J�xdS �xdS 2 1 .l:Ie8+ <;t>h7 (or 2 1 . .. �f8 2 2 .�xdS 'iYxdS 2 3 .'ifb4 ! '(i'd 1 + 24.tLJe 1 , winning) 2 2 .'iVxf7 �e6 2 3 . �xe6 , winning) 1 8 .tDxdS 'iYxd2 1 9 .tDe7 + <J;>h7 2 0 Jlxd2 �g4 2 1 .tDeS ! �xeS 2 2 .l:IxeS and as usually happens after a well-executed d4-dS break, the pawn structure is symmetri­ cal but White's pieces are better placed, which gives him the upper hand. A2) 1 1 .�h4 tDc6

�g 7 1 7 . tDa4 'iYaS ! 1 8 . �c S 'ifd8 1 9 .�xc6 bxc6 2 0 .�c2 'ltJlie 7 2 1 .tDb6 0-0 2 2 .tDxa8 'ifb 7 and in Petursson-N. Nikolic, Lugano 1 9 8 9 , the game was about even, since White does not have anything tangible in the ending arising after 2 3 . tD c 7 eS 2 4 . tDxeS 'iYxc7 2 S . tDxg6 'iYxf4 2 6 . tDxf4 �xb 2 ; A 2 2 ) 1 2 .�e4 tDe7 1 3 .'iYa4+ �d7 1 4 .'ifa3 gS ! l S .�g3 tDxe4 1 6 .tDxe4 0-0 1 7 .�d6 .tIe8 1 8 .�eS tDfS with an unclear game, Brenninkmeij er-N. Nikolic, Netherlands tt 1 9 9 3 ; A2 3 ) 1 2 .dS ! tD e 7 (taking the pawn does not help : 1 2 . . . exdS 1 3 .1:Ie 1 + �e6 (after 1 3 . . . tDe 7 1 4.'iYa4+ �d 7 1 S .'iYd4 g S 1 6 . tDxdS ! wins ; or 1 3 . . . <;t>f8 1 4. 'iYb3 , with a clear advantage for White) 1 4.�xg6 0-0 I S .�c2 and White is much better) 1 3 .d6! tDc6 1 4.tDe4 gS I s .tDxf6 + 'ifxf6 1 6 .�g3 and Black has terrible problems devel­ oping his pieces. One of the lines that illustrate the problems Black is facing is 1 6 . . . g4 1 7 . tDd2 hS 1 8 . tDe4 'iYd4 1 9 .�c 1 h4 2 0 .�f4 tDb4

analysis d i agram

a n alysis d i agram

White has a temporary lead in develop­ ment and must use this window of op­ portunity to develop the initiative. A2 1 ) 1 2 . l:i.c 1 tDxd4 ! 1 3 .'iYa4+ tDc6 1 4.�fd 1 'iYb6 I S .�xf6 �xf6 1 6 .'iff4

2 1 .�e3 ! ! 'iYxd3 2 2 . 'iVxg4 �xb 2 2 3 .�fd l and White wins : 2 3 . . .'iYbS 24.d7+ �xd 7 2 S .tDd6 + ; B ) 7 . . . tDbd7 i s the other move often played here. Now, apart from the fre131

C h a p t e r 2 : I s o l a t e d P a wns

W i n n i n g C h ess M i d d l eg a m es

quently seen 'regular QGA plan ' 8 . 'iYe2 and 9 .�d 1 , White can also steer the game into Meran-type positions with 8 .�e 1 b 6 9 . e4 cxd4 1 0 . e S ltJdS l l .ltJxd4 ltJcs 1 2 . .ifl �b 7 1 3 .ltJd2 .ie 7 1 4 .'ifg4 0-0 I S .ltJ2f3 �h8 , as in Kramnik -Kasparov, Linares 2 0 0 3 . 8.

a2-a3

One of the standard moves here, S .ltJc3 being the other option. 8.

...

c5xd4

Black decides to determine the situation in the centre, opting for the isolated pawn position. White would have a small advantage in the case of S . . . 'ifc 7 9.dxcS .ixcs 1 O .b4 .ie 7 1 l ..ib2 . e3xd4

�f8-e 7

1 0.

ttJ b 1 -c3

0-0

11.

�d3-c2

9.

I S .ltJa4 bxa3 1 6 .bxa3 ltJaS 1 7 .ltJcS .ic6 1 8 J :tfe 1 and due to the well­ placed knight on cS White was better in Donner-Van Scheltinga, Wijk aan Zee 1 9 69 . 1 2.

11.

. ..

b7-b6

132

�c1 -g5

1 2 . 1:t e 1 immediately would have given Black the chance to carry out a known rook manoeuvre : 1 2 . . J ia 7 ! ? 1 3 . 'iYd3 l:[d7 1 4 .ste3 .ib 7 I s J :iad l g6 as in Vaganian-Bronstein, Tbilisi 1973. 1 2.

..

1 3.

�d 1 -d3

g 7-g6

14.

tta 1 -d 1

l:ta8-c8

1 5.

�c2-b3

.

..tc8-b7

The light-squared bishop has done its duty forcing a weakening of the black kingside, and now has to return to the a2-g8 diagonal. 1 5.

A cautious move that makes sure that the cS-square remains protected. Dur­ ing the game I considered 1 1 . . . bS to be more logical , trying to create counterplay on the queenside as quickly as possible. However, the weak­ ness of the cS-square can play a role here, for instance : 1 2 . 'iYd3 .ib 7 1 3 . .igS ( 1 3 J :Ie l g6 1 4 . .ib3 l::!: c 8 I S . .ih6 .]de8 1 6 . .ia2 b4 was about equal in Naj dorf-Polugaevsky, Mar del Plata 1 9 7 1 ) 1 3 . . . g6 1 4 J �ad l b4

Naturally I decided to prevent Black from shutting in my bishop.

...

b6-b5

I S . . . ltJaS 1 6 . .ia2 ltJdS ? ! would allow White to change the pawn structure from an isolated pawn-position to a fa­ vourable pawn symmetry in the centre with 1 7 . .ixdS exdS I S . .ixe 7 'iYxe 7 1 9 Jj�fe l . As explained earlier in the book (see Sokolov-Cebalo, Game 24) , due to Black's ive b7 bishop and White's control of the eS -square, which he can occupy with his knight or with a rook, White has a long-term advantage here. 1 6.

�b3-a2

ttJc6-a5

Black wants to put his knight on c4 and shut off the white bishop. The other possibility was 1 6 . . . b4, in which case, as previously seen in the comment re­ lated to 1 1 . .. bS , White would be slightly better due to the weakness of the c S -square after 1 7 . ltJa4 bxa3 1 8. bxa3 , with ltJcs to follow. 1 7.

ttJf3-e5

20.

1 7.

...

ttJf6-d 7

1 7 . . . ltJdS is the other logical move, which does not address the problem of the active white knight on eS , but rather secures the blockade on dS . Af­ ter correct play by White this plan does not equalize either: 1 8 . .ixe 7 ! (at first Sight, the ' active ' 1 8 . sth6 looks good, but Black has a strong response in the instructive I S . . . ltJxc3 ! ( I S . . . 1:teS plays into White 's hands after 1 9 . 'iVf3 ! f6 2 0 . 'iVg4 .ifS 2 1 . ltJxg 6 ! .ixh6 2 2 . .ixdS .ixdS 2 3 . ltJ e 7 + W h S 2 4 . ltJ cxdS e x d S 2 S . ltJxc8 'ti'xc8 2 6 . 'iYhS and White is better) 1 9 . bxc3 l::t e 8 2 0 . �fe l .idS . We now have a the­ matic position where White has two weak pawns to worry about - a3 and c3 - and finds it difficult to create enough threats on the other side to keep Black busy) I S . . . ltJxe 7 (in the case of 1 s . . . iVxe 7 White carries out the usual structural transformation from isolated pawn to symmetry in the centre, retaining the upper hand with 1 9 . .ixdS .ixdS 2 0 . ltJxdS exdS 2 1 . �fe 1 and as seen before, White is better here) 1 9 . nfe 1 ltJdS 20 .ltJe4.

d4-d5!

A number of pieces have been ex­ changed and White cannot realistically count on a kingside attack anymore, so the time has come for the thematic d4-dS break, when a symmetrical pawn structure is reached in which White has a dominant piece in the centre - in this case, a knight - and Black, as is often seen in such positions, will find it diffi­ cult to neutralize the white initiative. 20.

...

e6xd5

Ignoring the pawn and shutting in the white bishop with 20 . . . ltJc4 does not solve Black's problems, since White exerts further pressure : 2 1 . 'iYh3 ! (2 l . ltJe4? is not good because of the positional 2 l . . . fS ! ; but 2 l . i.xc4 1:txc4 (or 2 l . .. bxc4 2 2 .'iVh3 ) 2 2 .d6 'iYc6 2 3 .'iV g3 should also promise White some advantage) 2 1 .. .fS 2 2 .dxe6 'iYxe6 2 3 J ife 1 'iYf7 (or 2 3 . . . 'iYf6 24.�d7 �f7 2 S . 1:ixf7 'iYxf7 2 6 . iYh4) 24.'iYh4 1:tfe8 2 S .h3 and White retains an initiative. 21 .

..ia2xd5!

An important move. It is much better for White to have a dominant knight on dS than a bishop. 21 .

...

�g8-g 7

2 1 . . . ltJc6 ? would only make matters worse for Black after 2 2 .ltJe4.

1 8.

..ig5xe 7

'iW d8xe 7

22.

'iYd3-d4+ f7-f6

1 9.

ttJe5xd7

�e7xd 7

23.

.J::!: f 1 -e1

.itb7xd5

1 33

Chapter

W i n n i n g C h ess M i d d l eg a m e s

So far, Black has refrained from captur­ ing the white bishop in order not to ac­ tivate the white knight. Now the pres­ sure on his position has increased and he finally decides to take the bishop. Other moves would not neutralize White's initiative either, for example : A) 2 3 . . . tDc4 ? ? simply blunders a piece after 24.kxc4 'iYxd4 2 5 . l::t x d4 bxc4 2 6 . l::t e 7 + �fl 2 7 JIxfl + �xfl 2 S .l:l.d7 + ; B) 2 3 . . . l::t fdS ? loses to 24 .tDe4; C) After the comparatively best 23 ... !lfeS 24.tDe4 'iYe7 ! (24 . . . kxd5 ? loses to 2 5 . tDxf6 brxe l + 2 6 . Jdxe l tDb3 2 7 . 'iYe 5 ! 'iYe6 2 S . tDh 5 + �h6 2 9 .iff4+ �xh5 3 0 J�xe6 kxe 6 3 1 .h3 with g 2 -g4 to follow) 2 5 .�xb 7 tDxb 7 2 6 .f3 l:IcdS 2 7 . 'iYb6 White executes a plan similar to the game, attacking Black's weak queenside pawns, while at the same time maintaining the threats against the black king. 24.

ttJc3xd5

would be justified after 2 5 . . . 'iYc6? 2 6 .tDxf6 ! �xe 7 (or 2 6 . . . 'iYxf6 2 7 .l:l.xfl + �xfl 2 S .iYd7+, and White wins) 2 7 . tDd5 + \tfl 2 S . tDxe 7 �xe 7 2 9 . 'iYg 7 + �e6 3 0 . I:re l + �d5 3 1 .'iYe5+ �c4 3 2 .'iYc3 + �d5 3 3 J��e5+ \td6 34.'li'd4+ \tc7 3 5 .�e 7 + �bS 3 6 .'iVa7 mate, but it would backfire after 2 5 . . J::I.xe 7 ! 2 6 .'iYxf6+ Wh6 2 7 .'iYh4+ \tg7 2 S .'iYf6 + �h6 and White has no more than a draw : 2 9 .�xe7 'li'xd5 3 0 .'iYe3 + 'iYg5 3 1 . 'iYh3 + �h5 3 2 .'iYe3 + 'li'g5 ( 3 2 . . . \tg7 ? ? would be a road to disaster after 3 3 J id7 +) 33 . �h3 + with perpetual check. But after White simply improves his position with 2 5 .h3 , it is difficult to de­ fend against the multiple threats. For example, if Black tries to become active with 2 5 . . Jlc2 , then 2 6 .�e7 ! quickly decides. 25.

'iYd4-b6 !

With the double threat of 2 6 .tDe7 , or simply 2 6 .'iYxa6, picking up a pawn. 25.

...

tlcS-bS

Black indeed has to give a pawn, since 2 5 . . . 'iYa7 ? ? would lose the queen after 2 6 . l:Ie 7 + , while in the case of 2 5 . . . I!feS , 2 6 .tDe 7 ! wins. 26.

'iVb6xa6

'iVd7-b7

30.

�d 1 -d6

ttJc6-e5

31 .

ttJe6-c5

�b7-c7

32.

ttJc5-e4

32 .

...

24 .

...

ttJa5-c6

134

27.

'iYa6xb7+ I:tbSxb7

2S.

ttJd5-f4

I s o l a t e d P a w ns

3 5 .tDxh 7 tDxb2 3 6 .tDg5 + �f6 3 7 .h4 tDc4 3 S .�e6+ �g 7 3 9 .h5 ! gxh5 (Black gets mated in the case of 39 . . . �h6 40 . l:l.xg6 + �xh5

ttJe5-c4

White's pieces are coordinating very well and apart from being a pawn up, there are also plenty of tactics in the po­ sition working in his favour. Black is looking for an active defence, but the weakened position of his king does not allow him to create any serious counterplay. His problems can also be seen in the following variation : 32 . . . f5 3 3 . ldd5 ! tDc4 (the attempt to bail out in the rook endgame would not work, for example : 3 3 . . . fxe4 3 4 J Ixe5 l:Ic2 35 .�xb5 ld.dS (or 35 ... l:IfS 3 6 . l::t f l �dS 37 .b4 l:Idd2 3 S J le5) 3 6 .h4 ! !ldd2 37 . .a:xe4 ldxf2 3 S . l:l. bS + ! �g 7 39. �b 7 + �h6

Even though White has a material ad­ vantage, Black has to exchange queens and enter an endgame, since otherwise he cannot parry the white threats. It is quite possible that Nikolic missed or underestimated my next move. How­ ever, the position is extremely difficult to defend. Perhaps Black's biggest prob­ lem is that he has no counterplay at all. The most stubborn defence was 24 . . . �fl . Now, the attempt to use ele­ mentary book tactics with 2 5 J:te 7 ?

2:

a n a lysis d i agram

4 1 . �f6 ! �xg 5 4 2 . �dxf5 + �h4 43 J Ig6 nhS 44.�h2 and 45 .g3 mate) 40 .l:Ixf5 tDxa3 4 1 Jlef6 with 42 .tDe6 to follow. Again, with a limited number of pieces left, the white forces are coordi­ nating extremely well, while the black king has no place to hide. 33.

ttJe4xf6+

�gS-f7

Or 3 3 . . . �g 7 34.�de6 . 34.

.tId6-e6

ttJc4xb2

35.

ttJf6xh 7

ttJb2-d3

35 . . . l:l.xa3 ? loses to 3 6 . tDg5 + �g7 3 7 J :t6e2 . 36.

ttJ h 7-g5+

�f7-g 7

37.

l::t e 1 -f1

�g7-h6

White remains two healthy pawns up after 3 7 . . . �xa3 3 S . l:b6 \th6 3 9 . I;'txb5 .

White is a healthy pawn up and has re­ tained the initiative, so the endgame should be won. 2S.

...

l::tf S-aS

29.

ttJf4-e6+

\t>g7-gS

29 . . . �fl ? is a blunder that would lose an exchange after 3 0 .l:l.d6 ! .

analysis d i agram

40.g4! and the black king finds himself in a mating net) 3 4 . tDf6 + �f7 135

Chapter

W i n n i n g C h es s M i d d l eg a m e s

38.

l:te6-e3!

ttJd3-f4

1 0.

Tactics continue to work for White : thus, 3 S . . .l:Ixa3 loses to 3 9 .tLJe6 with 40. ld:h3 mate to follow. 39.

h2-h4

'/J.c7-a7

40.

g2-g3

ttJf4-dS ttJ dS-f6

41 .

l::t e 3-d3

42.

l:tf1 -b1

l::t a 7xa3

43.

l:Id3xa3

:ta8xa3

44.

l::t b 1 xbS

hIf1 -e1

l o .ilgS is the other rather logical move, examined further on in the notes to Kramnik-Anand, Game 34. 1 0.

...

11.

h2-h41?

0-0

The smoke has cleared, and White is two pawns up in an easily won end­ game. 44 .

...

na3-a7

4S.

�g1 -g2

�h6-g 7

46.

l:IbS-cS

1 -0

Sacrificing an isolated pawn to gain the bishop pair and the initiative The following example also falls under the concept of active piece play, though Kramnik sacrificed his isolated pawn in the process.

QG 1 1 . 1 4 (D2 7)

GAME 30

Vladimir Kramnik Robert Hubner Dortmund 2 0 0 0 (8) 1.

d 2-d4

2.

ttJg1 -f3

ttJg8-f6

3.

c2-c4

dSxc4 e7-e6

d 7-dS

4.

e2-e3

S.

�f1 xc4

c7-cS

6.

0-0

a 7-a6

7.

ii.c4-b3

One of the main lines in the Queen's Gambit Accepted. 7.

...

cSxd4

7 . . . bS or 7 . . . tLJc6 are the other main moves here. 8.

e3xd4

ttJ b8-c6

9.

ttJb1 -c3

ii.f8-e7

1 36

Kramnik takes immediate action on the kingside and is ready to sacrifice a pawn in the process. There are many other logical possibilities here : A) I played 1 1 .ilgS myself against Anand in Wijk aan Zee 1 9 9 9 . That game continued 1 1 . . . bS 1 2 .dS (White gets rid of his weak pawn, hoping that after exchanges in the centre, due to his better developed pieces he will have the upper hand in the otherwise symmetri­ cal position) 1 2 . . . tLJxdS 1 3 . .txdS exdS 1 4 . 'iYxdS 'iYxdS I S .tLJ xdS .txgS 1 6 .tLJxgS .tb 7 1 7 .tLJe4 (the position looks very drawish; however, due to his well-placed knights and Black's poten­ tially weak queenside pawn , White has a very small edge) 1 7 .. J�fdS 1 S Jiad 1 tLJd4 (Black wants to exchange a few pieces in order to neutralize White's initiative) 1 9 .tLJb6 (much better was 1 9 .tLJc 7 ! l::I a cS (the tactic 1 9 . . . tLJe2 + does not help after 2 0 . �fl �xd 1 2 1..1d.xd l l::l c s 2 2 .tLJcS ) 2 0 .tLJcS ! (the tactical point behind 1 9 .tLJc7 ! ) 20 . . . h6 (2 0 . . . 11xc7 2 1 .!:txd4) 2 1 .tLJxb7 tLJe2 +

22 .Wfl llxd 1 2 3 .11xd 1 �xc 7 24.Wxe2 I;Ixb7 2 S .�d6 and due to his active rook White keeps an advantage in the endgame) 1 9 . . J :tab S 2 0 .tLJcS .tc6 2 1 .tLJbd7 .txd7 2 2 .11xd4 .te6 and the game was drawn; B) 1 1 .ilf4 (preparing the d4-dS break) 1 1 ... tLJaS ( l l . . . bS 1 2 .dS is better for White) 1 2 .ilc2 (the immediate 1 2 .dS does not bring anything special, since after 1 2 . . . tLJxb3 1 3 .'iYxb3 Black can sacrifice his queen with 1 3 . . . tLJxdS ( 1 3 . . . exdS 1 4. �ad 1 .te6 l S .'iVxb7 .td6 is also OK) 1 4.�ad 1 tLJxf4! l S .�xdS �xdS 1 6 Jtd 1 tLJdS ! 1 7. tLJxdS exdS l S .h3 (or l S .1d.xdS .te6) l S . . . b5 with a dynamic balance) 1 2 . . . bS 1 3 .dS ! (a the­ matic pawn sacrifice, very similar to Kramnik-Anand, Game 3 3 , further on in the book) 1 3 . . . exd5 1 4. 'iYd3 (White creates direct threats against the black king. Another option was to occupy the outpost d4 with the positional 1 4.tLJd4, keeping the as knight out of play) 14 . . . tLJc6 ! l S . .tc7 'iYd7 ( l S . . . 'iYxc 7 ? ? would have been a horrible blunder due to 1 6 .tLJxdS) 1 6 .tLJeS tLJxeS 1 7 .�xeS g6 l S . .txf6 (White decides to take back the sacrificed pawn. The other option was to keep the tension with l S .'ifd4 ifdS 1 9 . .tb 3 ile6 2 0 .l:lad 1 ) l S . . . .txf6 1 9 . tLlxdS ilg 7. White could not capital­ ize on his dominant knight and a draw was soon agreed in Kasparov-Anand, Wijk aan Zee 1 999; 1 1 .a3 is another move here, which will be explored later in the book (see Kramnik-Anand, Game 3 3 ) . 11.

...

2:

I s o l a t e d P a wns

1 3. tLJxdS tLJxd5 14 .hd5 . Even though the pawn structure is symmetrical and it may seem that after a few moves Black will reach equality, White has a strong initiative. The follOwing variation dem0nstrates his attacking potential:

a n alysis d iagram

1 4 . . . �b 7 l S .�gS ! �xgS 1 6 .tLJxgS h6 1 7 . tLJxf7 ! �xf7 1 S . .txf7 + �xf7 1 9 .'iYb3 + �fS 2 0 .11ad 1 . The white forces dominate the board and Black cannot defend against the numerous threats : 2 o . . . ifc7 ( 2 0 . . . 'iYf6 loses to 2 1 .1:;Id7 �bS 2 2 . l:le6 'iYfS 2 3 .'iYe3 ! �gS 2 4.'iVg 3 ) 2 1 .'iVe6 �dS 2 2 .�xdS + 'iYxdS 2 3 J :te4 'iYbS 24.g3 , winning. 1 2.

�b3-c2

b7-bS

1 3.

'ifd1 -d3

.ic8-b7

1 4.

h4-hS!?

An interesting idea involving a pawn sacrifice. The standard 1 4.ilgS g6 l S .tLJeS does not bring anything after l S . . . tLJc6 or l s . . . 11cs .

ttJc6-aS

The white bishop has to be chased away from the a2-gS diagonal. In case of the immediate 1 1 . . . bS ?! White gets a serious initiative after the thematic 1 2 .dS ! exdS 137

Chapter

W i n n i n g C h e s s Mi d d l eg a m e s

1 4.

...

�b7xf3!

Black correctly decides to take the pawn. Other moves would allow an un­ pleasant attack. In the event of 1 4 . . . ttJc4 1 5 .h6, Black has to choose between a damaged kingside pawn structure or a white pawn (after 1 5 . . . g6) remaining on h6, making different kinds of tactics possible. Should Black decide to stop the further advance of the white h-pawn with 1 4 . . . h6, this has some new drawbacks after 1 5 . ttJeS ! (entertaining, but insuf­ ficient for an advantage is 1 5 .�xh6 ! ? �xf3 ! (not l S . . . gxh 6 ? 1 6 .J::r xe 6 I:te8 1 7 .ttJeS ! with 1 8 .!!xf6 and 1 9 .'iYh7 + to follow) 1 6 .�gS �xhS 1 7 .�xf6 �g6 1 8 . �xe 7 'iYxe 7 1 9 . iVe2 �xc 2 2 0 .'iVxc2 as in Stocek-Jirovsky, Karlovy Vary ch-CZE 2 0 04) l S . . . ttJc4 1 6 .ttJg4 �e8 (only move)

a n alysis d i agram

1 7 .I;Ixe6! (a typical tactical motif here) 1 7 . . .fxe6 (should Black decide not to ac­ cept the sacrifice, then White is clearly better after 1 7 . . . �f8 1 8 .ttJxf6 �xf6 1 9 .I;Ixe8 + 'ifxe8 20.�f4 'iYe6 2 1 .dS) 1 8 .�xh6! with a crushing attack. 1 5.

'iVd3xf3

1 6.

�c1 -f4

'iVdSxd4

For the sacrificed pawn White has gained the bishop pair, some attacking possibili138

ties on the kingside and a few tempi re­ lated to the exposed black queen.

Most probably, obj ectively speaking White just has compensation for the sacrificed pawn, but not more. In a practical game, however, such intuitive sacrifices often bring success and Hubner soon loses his way. 1 6.

...

b5-b4?

Most likely it was better for Black to stop the advance of the white h-pawn and eliminate some tactics with 1 6 . . . h6 ! . Also (very important! ) , the pawn on hS would now be a new weak­ ness and something for White to worry about. It is not easy for White to prove adequate compensation here and this may be the reason why Kramnik never repeated this pawn sacrifice. For instance : 1 7 Jhd l iVc5 1 8 .'iYg3 ttJxhS 1 9 . 'iYd3 ttJf6 ! ( 1 9 . . . 'iVfS exfS ttJxf4 2 1 . ttJdS 2 0 . 'iVxf5 2 2 .ttJxe 7 + �h7 2 3 .g3 ttJhS 24.�xfS + g6 2 S .�e4 �ad8 2 6 .ttJdS �g7 2 7 .ttJc7 ttJf6 with equality, was played in Bareev-Rublevsky, Montecatini Terme 2 0 00) 2 0 .ttJe4 'iffS ! repels White's threats , leaving Black with a material advantage. 1 7.

l:!:a1 -d1 !

'iYd4-c4

White has grabbed the initiative. Other queen retreats do not solve Black's

problems either, for example : 1 7 . . . 'tWcS 1 8 J :teS and now: A) In Yermolinsky-Kaidanov, Seattle ch-USA 2 0 0 0 , White beautifully dem­ onstrated his attacking potential after 1 8 . . . 'iVb6 1 9 .'ifg3 ! bxc3 2 0 .�h6 g6 (White is better after 20 . . . ttJe8 2 1 .�xg7 ! ttJxg 7 2 2 .h6 �f6 2 3 .hxg 7 .ixeS 24. gxf8'tW+ �xf8 2 S .'iYxe S ) 2 1 .hxg6 fxg6 2 2 .�xg 6 �h8 2 3 .�xh7 ! Jdf7 24.�g6 l:Ig8 2 S .�e3 ! c2 2 6 . nc 1 'iVxb2 and now White missed the op­ portunity to finish the game in style with 2 7 .'iYh2 + ! �h7 2 8 .�xh7 ttJxh7 (2 8 ... ttJg4 loses to 2 9 .'iYh3 ttJxeS 30.�g 6+ with mate in a few moves) 29.MhS ng 7 3 0 .Ilxc2 ! (3 0 J lxa5 also wins) , but the text is more brutal, since White mates after 30 . . . 'iVxc2 3 1 .iYb8+ llg8 3 2 .�d4+ �f6 3 3 .�xf6 + ; B ) 1 8 . . . 'ifc6 1 9 .'iVg3 ! (this thematic attacking idea works here as well) 1 9 . . . bxc3 2 0 .�h6 ttJe8 2 1 .�xg 7 ttJxg 7 22 .h6 �f6 .

And here White concludes the attack with 2 3 .IlhS ! ! (not 2 3 J :tgS ? �fd8 ! 24 . .tIxg 7 + �f8) 2 3 . . JHd8 24J�xd8 + �xd8 2 S .hxg 7 �xg 7 2 6 .J:txh7 nd 1 + 27 .�h2 . 1 S.

�c2-d3

'ifc4-c6

1 9.

lLlc3-e4

lLlf6-d5

20.

ii!.f4-e5

f7-f5

I s o l a t e d P awns

After just a few moves, the black king finds himself under a terri ble attack with very few defenders around. 20 . . . f6 2 1 .�d4 does not help either. 21 .

'iVf3-g3

22.

lLle4-d2!

llfS-f7

The knight travels to d4. 22.

...

lLla5-b 7

23.

lLld2-f3

lLl b7-c5

24.

lLlf3-d4

'iVc6-b6

The position is ripe for a tactical solution.

25. ana lysis d i agram

2:

lLld4xf5!

e6xf5

Black could have opted for a hopeless endgame with 2S . . . ttJxd3 2 6 .ttJh6+ �f8 2 7 . ttJxf7 'iYxf2 + 2 8 . 'ifxf2 ttJxf2 2 9 .�xf2 �xf7 . 26.

�d3-c4

lLld5-f6

27.

�e5-c7 !

1 -0

A nice final touch. The attack executed by Kramnik is instructive and rather thematic for these positions. 1 39

C h a p t e r 2 : I s o l a t ed P a wn s

W i n n i ng C h es s M i d d l eg am e s

Structure 2 . 9

1 0.

Sacrificing a n isolated pawn by means of the thematic d4-d5 break

One of the important s tra tegic plans for the side playing with an isolated pawn is to temporarily sacrzfice it (mos tly by pushing d4-d5, or d5-d4 for B lack) and ob tain excellen t piece play in return. If this s trategy is executed well, then the initiative developed is worth m u ch more than the small ma terial inves tment. The breaking sacrifice of the isolated pawn is one of the very important typical motifs in this type of position. In the next couple ofgames I will try to give a n um ber of ins tructive ex­ amples. In the next Kasparov game, White always had the possibility to qu ickly regain his pawn with the better game, so even though it never amoun ted to a real sacrifice, the way Kasparov kept his initiative leaving his pawn hanging on d5 for a very long time and never allowing Timman to take it in a convenient way, was true 'power play'. Later we will also delve into some examples of the 'true' pawn sacrifice.

RG 6 . 5 (C42)

GAME 3 1

Garry Kasparov Jan Timman Amsterdam 1 9 94 ( 5 ) 1.

e2-e4

e7-eS

2.

ttJ g 1 -f3

ttJgS-f6 d 7-d6

3.

ttJf3xeS

4.

ttJeS-f3

ttJf6xe4

S.

d2-d4

d6-dS

6.

�f1 -d3

ttJ bS-c6

7.

0-0

iiLfS-e7

S.

.rIf1 -e1

1 40

...

dSxc4

Black releases the pressure in the centre, immediately entering an isolated pawn­ position. This plan has the drawback which is excellently exploited by Kasparov - that the usual blockade on d5 is not maintained, which means that White has tactical possibilities related to the d4-d5 push. After this game, 10 . . . dxc4 has hardly been played any­ more. I myself have had experience with 1 0 . . . 0-0. Black enters a similar isolated pawn-position, with the difference that the blockade on the d5 -square is main­ tained for quite some time, so that White cannot develop his initiative so easily. The game Shirov-Sokolov, Las Vegas 1 9 99, continued 1 1 .cxd5 ttJxd5 1 2 .h3 �e6 1 3 .a3 �f6 1 4.ttJe4

The other main line is 8.c4 ttJb4, while 8 . ttJc3 is a recent attempt by the Topalov / Cheparinov team. S.

...

iiLcS-g4

9.

c2-c4

ttJe4-f6

More or less forced, since the position of this knight was undermined with 9.c4. 1 0.

ttJb1 -c3

Releasing the pressure in the centre with 1 0 .cxd5 did not promise White an opening advantage after 1 O �xf3 (or 1 0 . . . ttJxd 5 ) 1 1 .'iYxf3 'iYxd5 . . . .

an a lysis diagram

14 . . . �f5 ! 1 5 .'iYb3 ttJb6 1 6 .d5 �xe4 1 7 .�xe4 ttJe7 (the knight is heading

for d6, where a perfect blockade would be achieved) 1 8 .�e3 ttJec8 1 9 .1:tac 1 ttJd6 2 0 .�b 1 'iYd7 with a more or less equal game. Black can also decide to accept White's temporary p awn sacrifice with 1 0 . . . �xf3 1 1 .'iVxf3 ttJxd4 and after 1 2 . 'iY d 1 ttJe6 ! , White found it difficult to prove anything tangible in a dynamiC position, which was played at top level in Morozevich-Gelfand, Mexico Wch 2 0 0 7 . That game continued : 1 3 .�f5 ( 1 3 .cxd5 ttJxd5 1 4.�b 5 + c6 1 5 .ttJxd5 cxb5 1 6 .'ifb3 0-0 1 7 . ttJxe 7 + ( 1 7 .�f4 was an interesting try by N aiditsch against Kramnik in Dortmund 2 0 0 8 ) 1 7 . . . 'iVxe 7 1 8 . 'iVxb5 a6 led t o a n equal game in Kasparov-Karpov, 1 5 th World Championship match game, Moscow 1 9 8 5 ) 1 3 . . . d4 ! ? (an ambitious attempt by Gelfand to keep dynamiC play. White had not managed to prove an advantage with the preViously played 1 3 . . . dxc4 1 4.�a4+ c6 1 5 .�xe6 fxe6 1 6 .'ifxc4 0-0 1 7 . 'iVxe 6 + �f7 1 8 .�e3 'iVd6 and a draw was soon agreed in Anand­ Kramnik, Mainz 2 0 0 1 ) 1 4.ttJe2 d3 ! 1 5 .ttJf4 ttJd4 1 6 .�xd3 0-0 1 7 .�e3 �c5 ! ? (a responsible move, relying on tactical resources. White would have a risk -free advantage in the case of 1 7 . . . c5 1 8 . �xd4 cxd4 1 9 . 'ti'f3 ) 1 8 .'iYb 1 ( 1 8 . b4? is a blunder due to 1 8 . . . �xb4 1 9 .�xd4 ? �xe 1 ) 1 8 . . . 'iYd6 ! 1 9 .l:Id 1 'if e5 ! (tactical defence contin­ ues) 2 0 .�xh 7 + ( 2 0 .b4? would be wrong due to 2 0 . . . ttJf3 + ! 2 1 . gxf3 �xe 3 ) 2 0 . . . ttJxh 7 2 1 .ttJd3 iYf5 (Black could have forced a draw with 2 1 . . . ttJe2 + 2 2 .�h 1 ( 2 2 . <;t>fl ?! would be asking for trouble after 2 2 . . . iVxh2 2 3 .ttJxc5 ttJf4) 2 2 . . . ttJg3 + ! 2 3 .Wg 1 ( 2 3 .hxg3 ? is a mistake due to the sim141

Chapter

W i n n i n g C h es s M i d d l e g a m es

pIe 2 3 . . . iYh5 + 24.Wg l �xe3 , and White is left with a wrecked pawn structure) 2 3 . . . ttJe 2 + with perpetual check) 2 2 .tLJxc5 tLJc2 2 3 .b4 and now Black made a tactical oversight with 2 3 . . . 'if g6 ? (correct was 2 3 . . . 'ufeS)

l:tabS 1 9 .1:ta3 .l::I feS 2 0 .�xe S + tLJxeS and a draw was agreed. 1 2.

...

ttJc6-a5!?

The most testing reply. White would have a solid, risk-free advantage in case of 1 2 . . . tLJbS 1 3 .h3 �h5 1 4 .g4 �g6 1 5 .�f4 , or 1 2 . . . tLJb4 1 3 .a3 �xf3 1 4.gxf3 tLJa6 1 5 .�f4 �d6 1 6 .�g3 . 1 3.

�c4-d3

1 4.

h2-h3!

c7-c6

White has to act energetically to seize the initiative here. Black has comfort­ able equality in the event of 1 4.dxc6 tLJxc 6 . 14.

analysis d i agram

. . . which White could have exploited with 24.tLJe6 ! fxe6 2 5 J�d2 and White takes the knight on c2 , remaining a sound pawn up. 11.

1 2.

�d3xc4

0-0

1 5.

...

�h5-g6

Facing a difficult position over the board, Timman chooses the most natu­ ral reply. The other moves would likely have run into Kasparov's home prepara­ tion and they do not solve the prob­ lems, for instance : 1 6 . tLJ xd 5 ! �xf3 A) 1 5 . . . cxd5 1 7 . 'iVxf3 tLJ c 6 I S . tLJxe 7 + tLJxe 7 1 9 .�g5 with a massive advantage. B) Arguably the 'most principled' re­ ply, 1 5 . . . �d6 , allows a strong attack :

I s o l a t e d P a wns

23 . tLJxfS + 'ifxfS 24. b4! - in many lines the bad position of the black knight stranded on as is a key element; B 1 2) 20 . . . �eS 2 1 .g4; B 1 3 ) In the case of 2 0 . . . iYf6 , White executes a nice mating attack with 2 1 .g4! �hS 2 2 .�g5 'iff7

�g4-h5

If Black decides to take the d5 pawn with 1 4 . . . �xf3 ? 1 5 .'ifxf3 cxd5 (cap­ turing the pawn with the knight does not make things better: 1 5 . . . tLJxd5 ? 1 6 .tLJxd5 cxd5 D 1 7 J le5 ! tLJc6 ( 1 7 ... d4? would lose a piece after I S .z:td5 ! 'iYb6 (or I S . . . 'iVc 7 1 9 .�f4) 1 9 . 'iY e4) 1 S .1:Ixd5 and White has a huge advantage) , White gets the pawn back, while keeping the initiative, after 1 6 .�g5 tLJc6 1 7 .�xf6 �xf6 I S .tLJxd5 , as seen in the game Gutman-Hergott, Graz 1 9 S 7 .

a n alysis d i agram

analysis d i agram

1 6 .!Ixh5 ! tLJxh5 B 1 ) Now the thematic 1 7 .�xh7 + Wxh7 I S .tLJg 5 + does not seem to work after I S . . . Wg6 1 9 .'iYd3 + f5 2 0 . tLJe6.

2 3 .tLJe4! ! �c7 24.tLJf6 ! ; B 1 4) 2 0 . . . 'iVd7 ! ! (the only move, but sufficient) 2 1 .g4 tLJf6 2 2 .'iVxf5 + Wf7 2 3 .g5 cxd5 and Black wins. B2) 1 7 .�g5 ! tLJf6 ( 1 7 . . . f6 I S .�d2 with 1 9 .tLJd4 to follow, and Black is left with two horribly stranded knights on as and h5) I S .tLJe4 cxd5 ( I S . . . �e7 ? loses a piece after 1 9 .d6) 1 9 .tLJxf6 + gxf6 2 0 .�h6 tLJc6 2 1 .'iVa4 and White will get his material back, while keep­ ing the attack. C) 1 5 . . . h6 Ioses material :

d4-d5!

There is no doubt that Kasparov had made a thorough analysis of this 1 2th move. In his 1 9 S4 match against Kar­ pov he had continued with 1 2 .�e3 , which led to an equal game after the correct 1 2 . . . �xf3 ! 1 3 . iYxf3 tLJxd4 1 4 . �xd4 �xd4 1 5 . l:txe 7 'iYxc4 1 6 .�xb 7 c6 1 7 .'iVb3 �xb3 I S .axb3 1 42

...

way, combining threats along the fifth rank with those against the black king.

2:

1 5.

l'le1 -e5 !

The key move, probably prepared by Kasparov in his 'home kitchen' . White continues his attack in a very ingenious

analysis d i agram

B 1 1 ) White wins in the case of 20 . . . 'ifbS 2 1 .g4! tLJf4 2 2 .�xf4 �xf4

ana lysis d iagram

1 43

Chapter

W i n n i n g C h e ss M i d d l eg a m e s

1 6 .g4! �g6 1 7 .�xg6 fxg6 1 8 .d6 �xd6 1 9 .�xaS 'iYxaS 2 o .iVxd6 . 1 6.

jLc1 -g5

Kasparov judges that his initiative is al­ ready worth more than just a clear ma­ terial advantage, which could have been easily achieved with 1 6 .�xg6 hxg 6 1 7 .d6 �xd6 1 8 .l:txaS 'iYxaS 1 9 .iVxd6 . 1 6.

...

2 0 . . . 'ifd6 ( 2 0 . . .'ti'fS 2 1 .'ii a4 �xc3 2 2 .bxc3 b6 2 3 .dxc 6 ± ) 2 1 . dxc 6 ± 'iYxd l + 2 2 .�xd l tbxc6 2 3 .tbxc6 bxc6. 1 9.

lIa 1 -c1 !

White brings his last piece into play, underlining, in many lines, the bad po­ sition of Black's as knight.

�e7-d6!

The only move. White wins in the event of 1 6 . . . cxdS ? 1 7 .�xg6 hxg6 1 8 .tbxdS �d6 1 9 .tbxf6 + gxf6 2 0 .l:t dS , or 1 6 . . . tbxdS ? 1 7 .�xe 7 tbxe 7 1 8 .�xg6 tbxg6 1 9 .'ifxd8 J:[axd8 2 0 . I:rxaS .

20.

20.

1 9.

1 7.

l:re5-e2

White also had a rather promising pos­ sibility - which to me looks even better than the text - in 1 7 .�xg6 ! hxg 6 ( 1 7 . . . �xeS loses to 1 8 .iLxf7 + �xf7 1 9 . tbxe5) 1 8 Ji e4! and again the horri­ ble position of the stranded knight on as makes life extremely difficult for Black. 1 7.

...

�d6-b4

1 8.

�g5xf6!

g 7xf6

The other option was to suffer in the ending after 1 8 . . .'ifxf6 1 9 .�xg6 'iYxg6 ( 1 9 . . . hxg 6 ? is bad due to 2 0 .tbe4 'iYd8 2 1 .d6 !) 2 0 .tbeS ! (Black has a reason­ able chance to stay alive in the case of 2 0 . 'iYd4 cS 2 1 .�f4 �xc3 2 2 .bxc3 iVd3 or 2 0 .'�'a4 �xc3 2 1 .bxc3 b6) 1 44

tbc3-e4!

Due to the terrible position of the black knight on as and the clumsily placed black pieces in general, White has plenty of tactics working in his favour.

...

�a8-c8

This allows White to further activate his pieces and underlines the bad position of Black's knight on as . However, other moves would not have solved the prob­ lems either, for example : 1 9 . . . �xc3 2 0 . .ldxc3 cxdS (in the case of 2o . . . iVxd5 ? Black would lose his as knight after 2 1 . �d2 'iVe6 2 2 .b4 �xd3 2 3 . �dxd3 tbc4 24.tbd4 'iYdS 2 S .tbb3 'iYbS 2 6 . iVg4+) 2 1 .iLxg 6 hxg6 2 2 .tbh4 ! (the key move, exerting more pressure. 2 2 J::I d3 tLlc6 2 3 J:l:xd5 'iYb6 would likely result in a draw) 22 .. Jle8 (if 22 . . Jlc8 ?, Black ends up simply a pawn down after 2 3 . nxc8 'iYxc8 24.tbxg6 fxg6 2 S .'iYxdS + , while trying to keep the material would give White a devastating attack after 2 2 . . . d4 2 3 . .t:t:g3 'iYdS 2 4 .'�' d 3 iYhS 2 S J :te4 llfe8 2 6 . neg4) 2 3 .�xe 8 + iVxe8 24 .tbxg6 fxg 6 2 S .'iVxdS + �h8 2 6 .'iYxa5 iVe l + 2 7 . � h 2 'if xf2 2 8 .'�'b4 and Black would have to defend a difficult position.

...

f6-f5

20 . . . cxdS ? loses to 2 1 . tIxc8 'iYxc8 2 2 .tbxf6 + �h8 2 3 .�xg6 fxg 6 24.tbxdS iLd6 2 s .'iYd4+. 21 .

tbe4-g3

22.

a2-a3

ifd8xd5

Kasparov's move underlines once more the clumsy positions of the �b4 and the tbaS . The attempt to go for an im­ mediate win with 2 2 .'iYa4! ? would not work so well :

A) 2 2 . . . iYxd3 2 3 .'iYxb4 'iYbS (or 23 . . . b6 24.h4 hS 2 S .'iYf4) 24.'iif4 and while Black's knight is stranded on as , White develops a deadly attack on the other side of the board;

I s o l a t e d P a wns

B) Black would simply lose a piece in the event of 22 . . . iLd6 ? 2 3 J ld2 �xg3 24.fxg3 ncd8 (or 24 ... b6 2 S .b4 tLlb7 2 6 .�a6) 2 S .�fl 'iYe4 2 6 .'ifxaS 'iYe3 + 2 7 .'lt>h2 ; C) 2 2 . . . cS ! (the only move) 2 3 Jidl (White gets a better ending, but not more, in case of 2 3 .�bS a6 2 4 .�d7 Iicd8 (not 24 . . . bS ? 2 S .'ti' d l 'iYxd l + 2 6 . �xd l �cd8 2 7 .a3 and White is a piece up) 2 S . �d l iVc4 2 6 .tbeS bS ! 2 7 .tbxc4 bxa4) 2 3 . . Jdcd8 24.a3 bS ! 2 S . 'ifxbS iVb3 2 6 .�c2 llxd l + 2 7 .�h2 �xb2 2 8 .�xd l 'iYxa3 2 9 .�c2 and even though White has good com­ pensation due to the misplaced black pieces on the queenside, matters are still far from clear. 22.

...

23.

tbg3xf5!

�b4-d6

Simple and strong. Far less convincing is 2 3 .b4 �xg 3 ! (clearing the path for the queen retreat is the only defence. The immediate 2 3 . . . tbb3 ? loses to 24.�c4 tbxc l 2 s .'iYxc l ! �xg3 2 6 .�xdS cxdS 2 7 .iVh6 �d6 2 8 .h4) 24.fxg3 tbb3 2 S . �eS (2 S .�c4 now leads to an un­ clear position after 2S . . . tbxc l 2 6 .�xdS tbxe2 + 2 7 . 'iYxe2 cxdS ) 2 s . . . tLlxc 1 2 6 .�xdS cxdS 2 7 .�fl 1:lfd8 with a sharp game ahead. 23.

analysis d i agram

2:

...

l:lc8-d8?

This final blunder makes things rela­ tively easy for White. A) Under the circumstances it was probably best to defend a difficult end­ ing which could have been reached af­ ter 2 3 . . . �xfS ! 24.�xfs iVxd l + (not 24 . . . 'iVxfS ? 2 S .'�'xd6 and due to the combination of the misplaced as knight and the weak king's position, this is completely hopeless for Black) 2 s Jlxd l ncd8 26 . .t:1d4; 1 45

Chapter

W i n n i n g C h ess M i d d l eg a m es

B) The ' active ' 2 3 . . . tLJb3 ? allows the game motifs , losing after 24JleS ! 1txeS 2 S .tLJe 7 + �g7 2 6 .tLJxdS tLJxc l 2 7 .1txg6 cxdS 2 8 .1tfS tLJa2 ( 2 8 . . . tLJb3 loses to 29 . 1t c 2 ) 29 . 1txc8 l:lxc8 3 0 .'iYd2 l:rc 1 + 3 l .tLJe l .

Black is once more reminded of his misplaced knight on as in the event of 2 6 .. J:txdS 2 7 .tLJxeS I::txeS 2 8 .1txg6 hxg6 2 9 .b4. 27.

ttJd5-f4

The text is definitely sufficient, but probably even more convincing was 2 7 .'iVe2 ! 1txc l 2 8 .'iVeS + f6 (Black is mated in the case of 2 8 . . . �h6 2 9 .tLJf6 1txd3 3 0 .tLJg4+ Wg6 3 l .tLJh4 mate) 2 9 . 'iVe 7 + �h6 3 0 . 1txg 6 hxg6 3 1 .tLJxf6 . 27.

...

.itg6xd3

Black gets mated after 2 7 . . . .ixc 1 2 8 . 'iVxc l ..txd3 2 9 .tLJhS + �h8 3 0 .'iVh6 . 2S. 24.

ne2-e5 !

.itd6xe5

Other moves lose even more qUickly: 24 .. .'iVb3 2 S .�xaS 'iVxd l + 2 6Jixd l 1tc 7 2 7 . tLJe 7 + W g 7 2 8 .tLJxg 6 , or 24 . . .'it'xd3 2 s .'iVxd3 1txeS 2 6 .'tWc2 . 25.

ttJf5-e7+

�gS-g7

26.

ttJe7xd5

.ite5xb2

ttJf4xd3

.itb2xc1

28 . . . I:lxd3 loses to 2 9 .'iVxd3 �xc l 3 0 .'iVc3 + . 29.

'ii'd 1 xc1

30.

'ii'c1 -g5+

�dSxd 3

Black resigned. He loses his as knight and remains un­ der attack.

Structure 2 . 1 0

As Kramn ik-Illescas Cordoba (Game 21), next is another game with the Tarrasch Variation that excellently illus trates some s trategic issues around the isolated pawn.

1 46

QG 1 3 . 1 5 (D2 7)

2:

I so l a t e d P a wns

GAME 3 2

Tigran V Petrosian Boris Sky Moscow Alekhine-mem 1 9 7 1 (6) 1.

d2-d4

2.

c2-c4

d5xc4

3.

ttJg1 -f3

ttJgS-f6

4.

e2-e3

e7-e6

5.

.itf1 xc4

c7-c5

6.

0-0

a 7-a6

7.

a2-a4

d 7-d5

For many years this was considered the main line of the Queen's Gambit Ac­ cepted. 7.

...

ttJ bS-c6

In the isolated pawn-pOSItIOn that is about to arise, Black usually prefers White to have his fl rook developed to d 1 rather than to e 1 . Apart from the text, Black sometimes also takes on d4 immediately, which indeed leads to the position where the white fl rook will be developed to e 1 . In this line I can show a few examples of my own: 7 . . . cxd4 8 . exd4 tLJc6 9 . tLJ c3 �e 7 1 0 . I:t.e l 0-0 1 1 ..igS (a few years later, in the KO World Championship in Tri­ poli against Simutowe, I tried 1 1 . .if4 and obtained a small advantage after 1 1 . .. tLJdS 1 2 . .ig3 tLJcb4 1 3 .tLJeS b 6 1 4. tLJxdS exdS ( a typical transforma­ tion , often seen in this book, has taken place : from an isolated pawn-position into a symmetrical structure in the centre where White has the upper hand) I S .1tb3 1tfS 1 6 .'iVf3 1te 6 1 7 J :tac l l::t c 8 1 8 . nc3 ! ) 1 l . . . h 6 1 2 .1th4 'iYb6 1 3 .'iVd2 �d8 1 4.1:ad l 'tWaS . With all White's pieces well developed , the time is ripe for the thematic break­ through:

an alysis d i agram

I S .dS ! . A ) Not surprisingly, the tactics work for White in the case of 1 5 . . . tLJxdS 1 6 . tLJxdS �xh4 1 7 :iVxaS tLJxaS 1 8 . tLJ b 6 ! tLJxc4 1 9 . tLJxa8 tLJxb 2 2 0 J;'I xd 8 + .ixd8 2 l . � c l ! .id7 2 2 .tLJeS 1tc6 2 3 .tLJxc6 bxc6 2 4 J lxc6 , winning ; B) O r 1 5 . . . exdS 1 6 . tLJxdS l::t x dS 1 7 .1txdS 1tb4 1 8 .'iYd3 1txe 1 1 9 . .ixf6 gxf6 2 0 .�g 6 + and Black gets mated; C) I S . . . 'iYcs 1 6 .'iVe2 exdS 1 7 . .ixf6 1txf6 1 8 . llxdS ! (better than 1 8 . tLJxdS 1tfS 1 9 .tLJxf6+ gxf6 2 0 . l:k 1 �aS with only a small advantage for White) 1 8 . . . 'iVf8 1 9 . �xd8 1txd8 2 0 .'iVe8 gS ! (only move) 2 1 . .ixf7 + and White was clearly better in Sokolov-Anand, Prague (Eurotel) 2 0 0 2 . S.

'ifd1 -e2

c5xd4

Black may also decide to keep the pawn structure in the centre intact and continue to develop his pieces with 8 . . . 'iVc 7 , which is the other main line here. I have played several games in this line. One of them, against Rublevsky, POikovsky 2 0 0 2 , continued 9 .tLJc3 1td6 1 0 .1td2 0-0. A) 1 l . �ac 1 b6 and now: A I ) In the case of 1 2 .dxcS Black is well-advised not to be afraid of a poten­ tial pawn weakness and take 1 2 . . . bxcS ! . 1 47

Chapter

W i n n i n g C h ess M i d d l eg a m e s

a n alysis d iagram

His well-placed pieces and good con­ trol of the central squares offer ample compensation for the visually damaged pawn structure on the queenside. Ob­ jectively, the position is approximately equal , for example : 1 3 .h3 �b 7 1 4J:tfd 1 Itfd8 1 5 .�e 1 h6 1 6 .�d3 lLJeS 1 7 .lLJxe S �xeS 1 8 . f3 lLJdS with even chances, Sokolov-Shirov, Sarajevo 2 0 0 2 ; A 2 ) 1 2 .�d3 �b7 1 3 .lLJe4 (this ex­ change is a typical plan here. After the swapping of knights, the white bishop will be well placed on e4, while the other bishop could be developed to c3 ) 1 3 . . . lLJxe4 1 4.�xe4 'ife7 I s .Sfd l �ac8 1 6 .dxcS (otherwise Black takes . . . cxd4) 1 6 . . . �xcS (taking with the pawn 1 6 . . . bxcS ? - is not a very good advice here, since White is better after 1 7 .�c3 . In comparison with my game against Shirov, White is the one controlling the central squares here, while the potential problems related to Black's damaged pawn structure remain) 1 7 .h3 (in the ab­ sence of direct play, this useful move is normally a good idea) 1 7 .. J�fd8 1 8 .�c3 h6 1 9 .1LJd2 lLJb4 (White's well-placed bishop on e4 has to be exchanged) 2 0 .�xb 7 'iVxb7 2 1 .'ifg4 (forCing a weakness) 2 1 .. .fS 2 2 . 'iVc4 lLJdS 2 3 . �b3 bS 24 .lLJf3 and due to the weakness of Black's e6 pawn I had a small advantage. 1 48

B) The other strategic option is the immediate push I I .dS ! ? exdS 1 2 .lLJxdS lLJxdS 1 3 .�xdS nd8 (it is important for Black not to allow White to capital­ ize on his well-placed bishop) 1 4.h3 (White has to spend time on this move, since the immediate 1 4 .e4? is wrong due to 1 4 . . . �g4) 1 4 . . . �h2 + I S .lLJxh2 tIxdS 1 6 .�c3 �e6 1 7 .lLJf3 .l::t a d8 1 8 .aS and due to his better pawn structure (Black's 3 vs 2 pawn majority is fixed, whereas White's 4 vs 3 pawn majority is mobile) White had slight pressure in Kramnik-Anand, Monaco rapid 1 9 9 9 . 9.

1::rf 1 -d1

�f8-e7

1 0.

e3xd4

0-0

11.

ttJb1 -c3

ttJf6-d5

This is one of the main positions of the Queen's Gambit Accepted with 7 . a4. Compared to the many other regular isolated pawn-positions, White has a pawn on a4 here (in most other cases this pawn is on a3 ) , which has its pros and cons. With the pawn on a4 , the white bishop on c4 is rather stable and it will also take Black more time to put his c8 bishop on the a8-h l diagonal, since . . . b7 -bS is never possible. On the other hand, now the b4-square is, and will remain, chronically weak and is a nice place for a black knight.

The main move here. A) White can also decide to first de­ velop his c 1 bishop with 1 2 .�d2 , which is perhaps a little tame. Black gets an equal game after 1 2 . . . �f6 1 3 .'iVe4 lLJcb4 1 4. h4 �d7 ! (a correct decision; Black is j ust in time to create adequate counterplay) I S .�gS (it is rather obvious that Black would have at least enough compensation for the pawn in the case of 1 S .�xdS exdS 1 6 . lLJxdS lLJxdS 1 7 .'iVxdS �c6) I S ... �c6 1 6 . 'iHg4 �xg S 1 7 . hxg S lLlxc3 ! (not giving White the time to jump with his f3 knight to e S ) 1 8 .bxc3 �xf3 1 9 .'iYxf3 lLJ d S with an equal game in Sokolov-Van den Doel, Dutch Championship , Leeuwarden 2002. B ) The other plan i s t o transfer the c4 bishop to the b I -h7 diagonal, where it will be an essential piece for the attack on the black king : 1 2 .�d3 lLJcb4 1 3 .�b 1 and now: B 1 ) A logical plan for Black is to place his bishop on b7 with 1 3 . . . b6 1 4.lLJeS �b7 , and then the best option for White is to follow a well-known attack­ ing plan:

analysis d i agram

1 2.

'ife2-e4

1 S Jh3 ! . Transferring the rook to the kingside, with, characteristically, a

2:

I s o l a t e d P a wns

sharp struggle ahead. It is important to note that this rook transfer is one of the specifics related to the isolated pawn-positions with the white pawn on a4. It may seem like a clever idea to first jump to the centre with the c3 knight with 1 S .lLJe4? and only then execute the na3 -g3 transfer. However, Black then has a typical reaction which is worth ing, since it mostly works in these positions : I S . . . fS ! 1 6 .lLJc3 �c8 1 7 . lla3 ? ? N ow, two moves too late, this i s a terri­ ble blunder.

a n a lysis d i agram

1 7 . . . lLJc2 ! 1 8 . �a2 ( 1 8 .�xc2 simply loses an exchange after 1 8 . . . �xa3 1 9 .1LJxdS 'iVxdS ) . With the text White hopes that Black has nothing better than a move repetition with 1 8 . . . lLJcb4 1 9 . �a3 . Well, sadly for White there is a nice tactical solution in this position : 1 8 . . . lLJxd 4! 1 9 . 'if e 1 (Black remains a healthy pawn up in the ending after 1 9 . .sxd4 lLJxc3 2 0 .bxc3 'iVxd4 2 1 .cxd4 �xc 1 + 2 2 . 'iV fl I;Ixfl + 2 3 . Wxfl ) 1 9 . . . lLJb3 and with an extra pawn and a much better position, Black soon won in Malaniuk-Yakovich, Moscow 1 99 2 . B 2 ) 1 3 . . . �d 7 1 4.lLJeS �c6 l S .lLJe4 and here Black has to be careful. 1 49

C h a p t e r 2 : I s o l a t e d P a wns

W i n n i n g C h es s M i d d l eg am e s

White has attacking potential on the kingside, Sky-Pachman , Manila Interzonal 1 9 7 6 . Once more I would advise the reader to note and the importance of the rook transfer a l -a3 -g3 , since this is a standard attack­ ing plan for White here. 1 2.

an alysi s d iagram

B2 1 ) A logical-looking developing move like 1 S . . . �c8 ? could lead to disas­ ter after 1 6 .�a3 ! (this thematic rook transfer is extremely strong here. White is attacking with literally all his pieces now and Black is in serious trouble) 1 6 .. .f6 ( 1 6 . . . <1:Jf6 does not help any longer due to 1 7 . <1:Jxf6 + �xf6 1 8 .�xh7 + ! (a typical motif here) 1 8 . . . \t>xh7 1 9 .'iVhS + \t>g8 2 0 . 11h3 and White delivers mate) 1 7 . �hS ! and Black is mated after 1 7 . . . fxeS ( 1 7 . . . g6 loses 1 8. <1:Jxg 6)

...

ttJc6-b4

1 2 . . . <1:Jf6 , harassing the white queen, has been played frequently here. How­ ever, White seems to be able to run away with his queen and get some ad­ vantage. Perhaps a good example is the game Marin-Ibragimov, Odorheiu Secuiesc 1 9 9 3 , which continued 1 3 .'iVh4 <1:JdS 1 4 .'iVg3 \t>h8 I S .�b3 <1:JaS ? ! (a risky concept. Better was 1 S . . . <1:Jc b4 in order to control the dS -square and not to allow the transfer of the b3 bishop to the b I -h7 diagonal) 1 6 .�a2 <1:Jxc3 1 7 .bxc3 �d7 1 8 .�f4 nc8 1 9 .�b I ! (as mentioned earlier, White's light-squared bishop placed on the b 1 -h7 diagonal is an essential piece for the attack) 1 9 .. J:txc3 2 0 .dS ! exdS 2 1 .�eS �f6 2 2 .�xc3 �xc3

'iVxf6 2 6 .�a3 and White is a sound ex­ change up, keeping the initiative; B) Another good example of how play may continue is the game Polugaevsky-Hort, Manila Interzonal 1 9 7 6 : 1 4 .�g4 <1:Jf6 l S .'li'g3 <1:JhS 1 6 .'iYh3 <1:Jf6 (finally the white queen is not under attack, so it's time to develop pieces) 1 7 .�gS <1:Jb4 1 8 .'li'g3 �e8 1 9 .c1JeS <1:JfdS 2 0 .�h6 �f8 2 1 .�ac l b 6 (this weakens the c6-square and allows White to change the pawn structure, achieving (as we saw in several previous examples) the upper hand with a pawn symmetry in the centre. I think that it was better for Black not to weaken the c6 -square in his camp and play 2 1 . . . �d 7 instead) 2 2 . <1:JxdS ! exdS 23 .�b3 �e6 24.h3 �c8

knight with 2 S . . . f6 , but White has an unpleasant initiative after 2 6 . <1:Jg4) 2S . . . 'iVxc8 2 6 .�f4 'ifb 7 2 7 .<1:Jg4 and White had some advantage. 1 3.

ttJf3-e5

1 3.

...

liaS-a 7

Black is getting ready to play . . . b7 -b6 and then . . . �c8 -b 7 , completing his de­ velopment. A logical question is : why waste time on 1 3 . . . �a7 and not play 1 3 . . . b6 immediately? Black's problem is that after 1 3 . . . b 6 White has a nice tacti­ cal solution in 1 4. <1:Jc6 ! , and after 1 4 . . . <1:Jxc6 (White is clearly better in case of 1 4 . . . <1:Jxc3 ? ! 1 S .<1:Jxe 7 + 'iYxe7 1 6 .bxc3 �b7 1 7 .'iYg4 J:lfc8 1 8 .�gS ) 1 S . <1:JxdS

analysis d i agram

analysis d iagram

1 8 . 'iVxh 7 + ! ! \t>f7 (or 1 8 . . . \t>xh 7 1 9 . <1:Jf6 + \t>h8 2 0 . �h3 mate) 1 9 .<1:Jd6+ ! (the bl bishop must be involved in the action) 1 9 . . . �xd6 2 0 .�g 6 + \t>f6 2 1 .�g S + ! �xg S 2 2 .!ig3 + \t>f6 2 3 .�h4 mate ; B2 2 ) I S . . . <1:Jf6 ! 1 6 . <1:Jxf6 + �xf6 1 7 .�a3 g6 1 8 .�h6 l::r e 8 1 9 . .a.g3 and 1 50

analysis d i agram

2 3 .<1:JgS ! (energetically played, not giving Black a break to consolidate) 23 . . . �f6 ( 2 3 . . . �xa l ?? loses to the well-known tactical motif 24.�h4 h6 2 S .<1:Jxf7 + ) 24.<1:Jxh7 �e8 2 S . <1:Jxf6

2 S .�xc8 (here White could have de­ cided on a different (and, for Black, more unpleasant) strategic concept, as mentioned in the comments on the 7 th move in my own game against Simutowe, with 2 s .l:k3 ! . Black has a difficult life here, since the rook on c3 cannot be taken, while at the same time White threatens to build on his posi­ tional pressure. Under the circum­ stances it looks sensible for Black to at least kick out White's dominant eS

an alysi s d i agram

I S . . . 11a7 ! (the best defence ; I S . . . exdS simply loses a pawn after 1 6 .�xdS �g4 1 7 .f3 ! (not 1 7 .r!e I ? �d 7 ! 1 8 .�xc6 151

Chapter

W i n n i n g C h es s M i d d l eg a m e s

iii.x c6 1 9 .'iYxc6 'iYxd4 with ' equality) 1 7 . . . �d7 I S .iii.x c6 �xc6 1 9 .'iVxc6 �cS 2 0 .'iYe4 �f6 2 1 .iii.e 3 �eS 2 2 .'iYd3 , while White has a clear endgame advantage in the case of I S . . . iii.b 7 1 6 .tLlxe7 + 'iYxe7 1 7 . d S ! (direct and strong) 1 7 . . . exdS ( 1 7 . . . tLlaS ? is a blun­ der which simply loses a pawn after I S .iii.d 3 g6 1 9 .�h6 �feS 2 0 .'ifd4 eS 2 1 . 'iYxb 6 , while after 1 7 . . . tLlb4 I S . 'iVe2 l::f. ad 8 1 9 . dxe6 �xd l + 2 0 .'iYxd l fxe6 2 1 .iii.e 3 , the problems related to Black's damaged pawn struc­ ture would soon begin to tell) I S .iii.x dS 'iYxe4 1 9 .�xe4) 1 6 .iii.d 3 ! (an impor­ tant intermezzo. Black would gradually equalize after 1 6 .tLlxe 7 + tLlxe 7 1 7 .�gS 'ii' d 7 ! I S .�xe7 'iYxe 7 1 9 .dS exdS 2 0 . iii.x dS 'iff6 ) 1 6 . . . fS 1 7 .tLlxe 7 + tLlxe 7 I S .'iVe 2 , White keeps a small, lasting advantage. The move played by Spas sky eliminates the tLlc6 tactic, but also loses precious time. 1 4.

�c4-b3

White is preparing to take tLlc3xdS at a suitable moment. Another rather logical continuation is to aim directly at the black king with 1 4 . 'iY g4. After 1 4 . . . �hS (a preventive move. Wrong is 1 4 . . . fS ? , since after I S . 'iYf3 b 6 1 6 .tLlxdS ! exdS 1 7 .tLlc6 ( 1 7 .�b3 is also good) 1 7 . . . 'iYd7 l S .tLlxe7 + 'iYxe 7

1 9 .iii.b 3 ! White has a massive advan­ tage. It is important to note that keeping the bishop pair and exploiting the weaknesses in Black's position is much better than taking the pawn on dS) I S .'iVhS 'iYeS 1 6 .tLlxdS (otherwise 1 6 . . . f6 ) 1 6 . . . exd S 1 7 .�d3 tLlxd3 I S .tLlxd3 b6

a n a lysis d i agram

1 7 .iii.e 3 ! �b 7 I S .dS ! tLlbxdS 1 9 .aS .

analysis d i agram

1 9 .�e 1 . A pawn structure transforma­ tion - already seen many times in this book - has taken place and White has a pleasant advantage. Thanks to his bishop pair Black should have reasonable com­ pensation should White decide to take a pawn - and open the light -square diago­ nals - with 1 9 .'iYxdS . 1 4.

l S2

...

'iYe4-h4

b 7-b6

Black is going to finish his development by placing the cS bishop on the long di­ agonal and it seems that he has a good game. Petrosian now develops and keeps the initiative in a masterful way, playing with great power. 1 6.

'iYh4-g3 !

1 7.

�c1 -h6

1 8.

l::!: a 1 -c1

ttJf6-e8

Preparing the central break. It should be noted that the immediate I S .dS ! ? was possible, but would lead to an unclear game:

I s o l a t e d P a wns

White has a strong and energetic solu­ tion:

1 9.

d4-d5 !

Beautiful and highly instructive. With all his pieces harmoniously developed, White is ready for this central break. This is a typical intuitive sacrifice. Petrosian rightly judges that in the en­ suing complications the tactics will work for him. 1 9.

...

e6xd5

1 9 . . . tLlxdS loses to the spectacular 2 0 . tLlxdS exdS

ttJd5-f6

In the event of 1 4 . . . b 6 , White trans­ forms the pawn structure in the centre with I S .tLlxdS and he will enjoy all the advantages of the 'central pawn symmetry/ upper hand-position' after I S . . . exdS 1 6 .'iYf3 . 1 5.

an alysis d iagram

In case of the preventive 1 6 . . . �hS , White would get a clear advantage after the nice and strong :

2:

�c8-b7

a n a lysis d i agram

1 8 . . . tLlxdS 1 9 . tLlxdS (or 1 9 J:td4 �f6 ; 1 9 . . . iii.d 6 is less good, since White is clearly better after 2 0 . tLlxdS i1LxdS 2 1 .i1LxdS �xe S 2 2 . 'iVxeS exd S 23 .l:1xdS Iid7 24.�xd7 iYxd 7 2 S . iii.e 3) 19 ... exdS 2 0 . .tt d4 �f6 2 1. . J:te l tLld6 22 .i1LxdS 'iYe 7 . 1 8.

...

�g8-h8

It seems that now after 1 9 .iii.e 3 tLlf6 , Black will have a good game. However,

a n a lysis d iagram

2 1 .l:1c7 ! ! gxh6 (only move, since 2 1 . .. 'iVxc 7 loses to 2 2 .i1Lxg 7 + tLlxg7 2 3 .tLlxf7 + llxf7 24.'lWxc 7 and Black loses more material) 2 2 ..tId 7 �h4 2 3 . �xdS �xg 3 2 4.hxg3 (due to his terrible piece coordination, removing the queens from the board does not l S3

Chapter

W i n n i n g C h ess M i d d 1 e g a m es

help Black) 2 4 . . . �g7 2 S .!id7 ! tLJf6 2 6 J:re7 - despite being a pawn up, Black is completely tied up and has a hopeless position here. 20.

..th6-e31

l:Ia7-a8

In the case of 2 0 . . . �cS White remains a pawn up :

22.

..te3xb6

23.

ttJc4-aS

'ifd8-b8

Also good was 2 3 . tLJe3 , and White has a winning advantage after 23 . . . 1;lc8 24.�d4 �f8 2 S .tLJcxdS . 23.

...

ttJd6-fS

By exchanging queens , Black hopes to make his troubles more bearable. 24.

'iVg 3xb8

�a8xb8

2S.

ttJaSxb7

1.lb8xb7

26.

a4-aS

..te7-gS

27.

.l:lc1 -b1

2 1 . tLJxdS ! �xd S 2 2 J ixc S ! 2 3 .�xcS tLJf6 24.�xa 7 . 21 .

ttJeS-c41

31 .

...

.:tc8-b8

32.

f4xgS

ttJe7xdS

33 .

..ta4-c6

l:tb7xb6

34.

aSxb6

ttJdS-e3

3S.

b6-b7

3 5 .�d7 also wins easily, for instance : 3 5 . . . tLJxd 1 3 6 .�xfS tLJe3 3 7 .b7 and the pawn queens. ttJe3xd 1

36.

Mc1 xd 1

g 7-g6

37.

g2-g41

ttJfS-g7

38 .

.:td1 xd4

ttJ g 7-e6

39.

.r!d4-d 7

1 -0

QG 1 l . 1 4 (D2 7 )

ttJe8-d6

In the event of 2 1 . . . tLJf6 White is much better after the simple 2 2 .tLJxb6 �b8 23 .as , while in reply to 2 1 . . . �cS White has a nice tactic again:

Reaching the ending has not solved any of Black's troubles. Petrosian's execution is flawless. The difference in coordina­ tion between the white and black forces remains as striking as ever. 27.

...

dS-d4

28.

ttJc3-dS

ttJb4-c6

29.

..tb3-a4 1

�f8-c8

30.

f2-f41

ttJc6-e 7

a n a l ysis d i agram

2 2 .tLJaS ! bxaS 2 3 .�xcS �g8 24.�xb4 axb4 2 S .tLJxdS , with a winning advan­ tage. In all these lines , there is a striking dif­ ference in coordination between the white and the black pieces.

31 .

Rb1 -c1 1

GAME 3 3

Vladimir Kramnik Viswanathan Anand Dos Hermanas 1 9 9 9 (3) 1.

d2-d4

d 7-dS

2.

c2-c4

dSxc4

3.

ttJg1 -f3

e7-e6

4.

e2-e3

ttJg8-f6

S.

..tf1 xc4

c7-cS

6.

0-0

a 7-a6

7.

..tc4-b3

cSxd4

8.

e3xd4

ttJb8-c6

9.

ttJb1 -c3

..tf8-e7

l:tf1 -e1

0-0

1 0.

1 54

...

A great game by Petrosian. bxcS

I s o l a ted P a wns

Not giving Black the slightest chance of survival. A highly instructive combina­ tion of tactical and technical motifs in the execution.

3S .

an a lysis d iagram

2:

Kramnik has vast experience in this line, his game against Karpov (Frank­ furt rapid 1 9 99) also being very in­ structive. Anatoly Karpov tried to find an improved move order for Black and played 1 O . . . tLJaS 1 1 .�c2 bS . Now Kramnik correctly felt that the position required immediate, energetic action and proceeded with:

a n a lysis d i agram

1 2 .dS ! (in this sharp struggle White's ini­ tiative is very strong) 1 2 . . . b4 (searching for active counterplay. Taking the pawn on dS is not good either: 1 2 . . . exdS 1 3 .�gS �e6 ( 1 3 . . . 0-0n is a terrible blunder due to 1 4.iYd3 g6 l S .nxe7 'ifxe7 1 6.4.Jxd5) 1 4.4.Jd4 and Black's position will soon collapse; or 1 2 ... tLJxdS 1 3 .tLJxdS 'iVxdS 1 4.�d2 ! tLJc6 ( 1 4 . . . tLJb7 loses to I S .�e4 iVd6 1 6 .'iYc2 ! 'iYcs 1 7 .iVxc5 �xcS 1 8 .I;Iac l and White wins material) l S .�e4 'iYd6 1 6.'iVc2 ( 1 6.�c3 gives White an advantage in the endgame after 1 6 . . . 'iVxd l 1 7 .1;laxd 1 �b7 1 8 .�xg7 �g8 1 9 .�c3 fS 2 0.�c2 tLJaS 2 1 .�f1 b4 2 2 .�eS) 1 6 . . . �b7 1 7 .�ad l and White's initiative is worth much more than a pawn) 1 3 .�a4+ Wf8

analysis d i agram

1 4.�f4 ! bxc3 I S .d6 tLJdS (Black has no good options left , since I S . . . cxb 2 1 6 .dxe7 + Wxe7 . . . 155

Chapter

W i nn i n g C h e s s M i d d l eg a m e s

ana lysis d i agram

. . . meets with 1 7 .'iYc2 ! �d7 1 8 .'iVxb 2 ! �xa4 1 9 .'iYb4+ \t e 8 2 0 .'iVxa4+ �f8 2 l .I;Iad 1 and Black loses his knight on as , for example : 2 l . . . lbdS 2 2 . .,tgS 'if c7 23 .1lc 1 'tli'b 6 2 4 . lib 1 'if cS 2 S .l:t.ec l ) 1 6 .dxe 7 + 'ifxe7 1 7 .�eS �b 7 1 8 .bxc3 nd8 1 9 .1bd4 (in a diffi­ cult position the 1 2th World Champion suffers a tactical oversight) 1 9 . . . lbc4 ? 2 0 .�xg 7 + �xg 7 2 1 . lb fS + exfS 2 2 J:Ixe 7 CiJxe 7 2 3 .'ife2 and White soon won, since Black is not able to protect his knights. 11.

a2-a3

ttJc6-a5

1 2.

�b3-c2

b 7-b5

isolated pawn position with 1 3 . 'tli'd3 �b 7 1 4.�gS g6 l S .11ad l :lc8 1 6 .CiJeS . My opponent, Danish grandmaster Curt Hansen, exploited the deficiencies of my set-up well with 1 6 . . . lbc4! (first ex­ changing the attacking knight on eS) 1 7 .lbxc4 bxc4 ! (and then recapturing correctly with a pawn, closing the a2-g8 diagonal) 1 8 .'tli'h3 lbdS 1 9 .�h6 �e8 2 0 .lbe4 'ifb6 (hitting White's newly-created weakness) 2 1 .�c l

13.

d4-d5!

White sacrifices a pawn in order to ob­ tain an outpost on d4 for his knight and disturb Black's natural piece develop­ ment. In one of my own games , I opted for a traditional attacking set-up in an 1 56

ana lysis diagram

1 8 .lbh4 ! (threatening with immediate annihilation by 1 9 .lbxg6) 1 8 . . . �f7 1 9 .'if g3 :lg8 2 0 .lbf3 ! (the knight has done its j ob on h4 and now is the time to create new tactical motifs) 2 0 . . . CiJc4 (Black tries to get his stranded as knight back into play, but it's too late) 2 1 . 'iff4 ne8 (2 l . .. 'tli'd6 2 2 .iYh4, with 2 3 .�f4 to follow, is also unpleasant for Black) 2 2 .CiJgS + �g8

2 l . . . fS ! (a timely, well-executed . . . f7 -fS often removes the last glimmer of hope of a white attack in such positions) 2 2 .lbc3 �f6 - Black was better and I was fortunate to draw this game, Sokolov-C. Hansen , Malmo 1 9 9 8 . 1 3.

.. .

I s o l a t e d P awns

B) The other way for Black, similar to Anand's reaction, is to leave the white pawn on dS as it is and bring the as knight back into play with 1 3 . . . lbc4 - a logical reaction in itself which, how­ ever, does not solve Black's problems af­ ter 1 4. 'if d3 lle8

analysis d i agram

lifS-eS!

A good, cool reaction. Black takes care of the development and coordination of his pieces, making it difficult for White to develop an initiative. Taking the d5 pawn gives White the opportunity to capitalize on the horribly misplaced black knight on as and set up a swift at­ tack on the black king. A) One nice, instructive example I found went 1 3 . . . lbxdS 1 4.lbxdS exdS I S .'tli'd3 g6 1 6 .�h6 lie8 1 7 .'iYc3 ! (forcing Black to further weaken his kingside) 1 7 . . . f6

2:

ana lysis d i agram

I S .a4 ! (a strong move, highlighting the drawbacks of 1 3 . . . lbc4) I S . . . exdS 1 6 .axbS as 1 7 .b3 lbb6 1 8 .CiJd4 (the strong knight on d4 ensures White an advantage here ; take note of Black's weak pawn on dS , which is clumsily standing in the way of any active black counterplay) 1 8 . . . �b 7 1 9 .f3 �c8 2 0 .lba4 �f8 2 1 . .,tg S and White was clearly better in Kasparov-Ivanchuk, Linares 1 9 9 9 . 1 4.

analysis d i agram

23 .CiJxh 7 ! (time to storm the barri­ cades) 23 . . . \txh7 24.'ifh4 �fS (per­ haps the best for Black was to suffer in a difficult endgame after 2 4 . . . :lh8 2S .�f8 + <;t;g8 2 6 .�xe7 'ifc7 ! 2 7 .'iYg3 \i'xg3 2 8 .fxg3 <;t;f7 2 9 .�cS) 2 S .�xfS gxf5 2 6 .�e6 ! (White coolly brings more pieces on, increasing the pres­ sure) 26 . . . \tg6 2 7 .�ae l Jda7 2 8 .'ifgS + 1 -0, P.H. Nielsen-Bentzen, Denmark tt 1 9 99.

�c1 -g5

1 4.iYd3 , forcing 1 4 . . . g 6 , has also been tried, but it brought White no more than equality. Black is still not disturbed by the white pawn hanging on dS and simply improves his position. Staying cool, making constructive moves in the face of White's seemingly dangerous pawn on dS , is often important for Black in these positions : 1 s . b4 lbc4 1 6 .dxe6 .,txe6 1 7 .lbd4 �dS , Illescas Cordoba-Anand, Dos Hermanas 1 9 9 9 . 1 4.

...

h7-h6

157

Chapter

W i n n i n g C h ess M i d d l e g a m es

Somewhere around here Black probably made a mistake. The text is definitely logical , but, as seen further on, it gives White some extra attacking possibili­ ties. The other critical move for the assessment of this position is the immediate 1 4 . . . ttJxd5 .

White has a dangerous attack after 1 6 .ttJg5 g6 1 7 .'iYd4 .ib 7 I S .'iYh4 h5 1 9 J !ad 1 ) 1 6. 'iY e 2 (Black should be at least equal after 1 6 . ttJ g 5 'iVxd l ! 1 7 . .ixh 7 + �fS I S .liaxd l ttJc4) 1 6 . . . .ib 7 ! (Black comes under attack after 1 6 . . . ttJg6 1 7 .�ad 1 'iVe7 I S.�xg6! hxg6 1 9 .ttJd5 'iVa7 2 0 .ttJg5) 1 7 .ttJg5

analysis d i agram

A) 1 5 . ttJxd5 exd5 ! ( 1 5 . . . �xg 5 1 6 .ttJxg5 'iVxg5 looks better for White after 1 7 . 'iYd3 g6 I S . 'iYc3 ! 'iYxd5 1 9 .�e4) 1 6 .l:rxe 7 �xe 7 1 7 .'iVd3 ! (the principled way to continue; 1 7 .�xe7 'iYxe 7 I S .'iVxd 5 .ib 7 1 9 .'iVd3 g6 2 0 J :re 1 'iYf6 2 1 .ttJe5 IleS is slightly better for Black) 1 7 . . . g6 I S .'iYc3 �aa7 .

analysis d i agram

1 7 . . . h6! (don't be afraid of ghosts!) 1 S . .ih 7 + �fS 1 9 . ttJxf7 �xf7 2 0 .'iYxe6 + �fS 2 1. .]dad l 'iVcS ! 22 . .t'!d7 'iYc6 2 3 .�g4 h 5 2 4 . 'iYf4+ (only move) 24 . . . 'iVf6 2 5 .'iYb4 ,gacS and Black repels the attack, 2 6 .'iYxa5 ? ? be­ ing a horrible blunder due to 26 . . . ttJc6. Finally, instead of the text move, 1 4 . . . ttJc4?! would allow White to con­ tinue his initiative with 1 5 .ttJd4! . 1 5.

Probably a crucial mistake that will ulti­ mately cost Black the game. A) 1 5 . . . ttJc4? is again wrong due to 1 6 . ttJd4, for example : 1 6 . . . ttJxd5 ( 1 6 ... ttJxb2 ? loses to 1 7 .'iYc l ttJc4 l S .ttJc6 'iVc 7 1 9 . ttJ xe 7 + 'iYxe 7 20.ttJe4) 1 7 .'iYd3 g6 I S .ttJc6 'iVd7 1 9 .ttJxe 7 + ttJxe 7 20 .'iYxd7 (20 .'iVh3 also looks good for White) 2 0 . . . �xd7 2 1 ..ixe 7 IIxe 7 2 2 .ttJd5 and White wins an exchange; B) 1 5 . . . exd5 ! looks very risky, but could in fact be critical here, since it is far from easy to find anything clear for White. Let me give a few lines : B l ) 1 6 . .ixf6 �xf6 1 7 .�xe S + 'iYxeS l S .ttJxd5 'iVdS 1 9 .iVd3 g6 2 0 . �e l �g 7 with an approximately equal game; even 20 . . . �b 7 ! ? 2 1 .�e S + �g 7 22.�xdS �xdS comes into consider­ ation; B2) 1 6 J!xe7 �xe 7 ! (not 1 6 . . . 'iYxe 7 ? 1 7 .ttJxd5 'iY d S l S . ttJxf6 + ! gxf6 1 9 .'iYc l ! �g 7 2 0 .'iVf4 !Ia7 2 1 .'iVg3 + �hS (even worse is 2 1 . . . �fS 2 2 .�h7) 2 2 .ttJe5 ! I:rxe5 2 3 .'iYxe5 fxe5 24 . .ixdS and White should win this endgame) 1 7 . .ixf6 gxf6 I S .�d2 �g7 1 9 .'iVd3 [5 .

�g5-h4

ana lysis d i agram

This messy position could be critical for this variation and needs to be analysed. B) 1 5 .�xe 7 , and now the best way for Black to proceed seems to me to be 1 5 . . . ttJxe 7 ! (in the case of 1 5 . . . trxe 7 I SS

ana lysis d i agram

1 5.

...

ttJf6xd5?

White's attack looks promising, but it is difficult to find anything concrete :

2:

I s o l a t ed P a wns

2 0 .ttJxd5 J:te6 2 1 .'iVc3 + �gS 2 2 . l::t d l l:ld6 ! 2 3 .'iVd4 ttJc4 24 . .ib 3 �e6 ! 2 5 . .ixc4 bxc4 2 6 .ttJe 7 + ( 2 6 .ttJf6 + ? 'li'xf6 ) 2 6 . . . 'iVxe 7 2 7 .'iVxd6 �xd6 2 S . blxd6 I;IbS and Black is better, since 2 9 .�d2 ? ? is a terrible blunder which loses immediately to 29 . . . c3 ; B 3 ) Another interesting line is 1 6 .'iYd3 g 5 ! (courageous and strong) 1 7 . ttJxg 5 ! ( a principled reaction, in the spirit of the position) 1 7 . . . hxg 5 l S . .ixg5 ttJc4! 1 9 .'iYg3 ( 1 9 .b3 looks like one of those 'little moves' that can throw Black off balance, but it does not achieve more than an approximately equal ending :

an a lysis d i agram

1 9 . . . ttJxa3 ! 2 0 .ttJxd5 ! ( 2 0 .'iYg3 ttJh5 ) 2 0 . . . 'li'xd5 2 1 . .ixf6 'iYxd3 2 2 . .ixd3 bldS ! 2 3 .�xe 7 Ilxd3 24.blxa3 �e6) 1 9 . . . ttJh5 2 0 .'iVh4 (continuing in style. Black has reasonable counterplay in the ending after 20 . .ixe 7 + ttJxg3 2 1 . .ixdS !IxdS 2 2 . hxg 3 d4) 2 0 . . . .ixg 5 2 1 . �xe S + 'iVxe S 2 2 . 'iYxg 5 + ttJ g 7 2 3 .ttJxd5 'iVe5 24.ttJf6 + �fS 2 5 .ttJh7 + �gS (wrong is 2 5 . . . �eS ? 2 6 .'iVxe5 + ttJxe 5 2 7 . 1de l .ie 6 2 S . �xe5 IIcs 2 9 .�e2) 2 6 .ttJf6 + (White may go one step too far by playing 2 6 .'iVh6 f5 ) 2 6 . . . �fS and a draw by perpetual check concludes this beautiful variation. 1 59

Chapter

W i n n i n g C h es s M i d d l eg a m es

1 6.

ttJc3xd5

e6xd5

More or less forced. 1 6 . . . ibch4? is not good, since after 1 7 . 'if d3 ! fs (Black gets mated in case of 1 7 . . . �f8 ?? 1 8. ttJxh4 �b 7 1 9 .�h7 'iVxdS 2 0 J:lad l �c6 (or 2 0 . . . 'ifcs 2 1 . �d 7 ) 2 1 .'iVh 8 + � e 7 2 2 .ttJfS + �f6 2 3 .'iVxg 7 mate, while 1 7 . . . g 6 ? loses to 1 8 .ttJxh4 exd S 1 9 . ttJxg 6 or 1 8 . . . 'iVxh4 1 9 .�c3 ! ) 1 8 .ttJxh4 �b7 ! (best under the circum­ stances. 1 8 . . . �xh4? loses more material after 1 9 .ttJc7 �d8 2 0 .'iYc3 ) 1 9 .ttJxfS ! �xdS ( 1 9 . . . exfS ? loses to the prosaic 2 0 . ttJ e 7 + .l:ixe 7 2 1 .�xd 8 + I:txd8 2 2 .�xe7) 2 0 .ttJxh6 + ! \tf8 (20 . . . gxh6 ? loses the queen after 2 1 .�g6+ �f8 2 2 .'iVxh 6 + � e 7 2 3 . 'ii' g 7 + �d6 24 . .t:tad l �g8 2 S .'ii' e S +) 2 1 .'iVxdS �xdS 2 2 .ttJg4 and White is a sound pawn up in the endgame. 1 7.

'iYd1 -d3

1 S.

'iYd3-e3

g7-g6

White wants to capitalize on the fact that there is a pawn hanging on h6. Also very promising was 1 8 .!:Ixe 7 brxe 7 1 9 .'iVd2 ! (making use of the fact that the black pawn is on h6) 1 9 . . . ttJc4 ( 1 9 . . . gS ? 2 0 . ttJxgS ) 2 0 .'iYxh6 'iff8 2 1 .'iYf4 and White is clearly better. 1 S.

...

20.

�e3xh6

ii.c2xg6 !

ii.e7xh4

'iVdS-f6 !

Only move. 2 0 . . . fxg 6 ? loses to 2 1 .'iVxg 6 +
ii.g6-h7+ �gS-hS

22.

ii.h 7-g6+

23.

�h6-h7+ �gS-fS

24.

ttJf3xh4

�hS-gS

ii.cS-e6

The best defence. 1 8 . . . gS ? ? loses quickly to 1 9 .�d3 ! fS 2 0 .�g3 ttJc4 2 1 .ttJd4 (the black king is too exposed, without any pieces to protect him) 2 1 . .. ttJxb2 2 2 .'iVe2 �d7 2 3 .'iVhS 'iYb6 24.ttJe6 and Black gets mated. 1 8 . . . ttJc6 ? does not help either, since af­ ter the sim pIe 1 9 . 'if xh 6 �f8 ( 1 9 . . . �xh4? loses to 2 0 .�xg 6 ! fxg6 2 1 .'iVxg 6 + �f8 2 2 . 'iYxc6 IIxe l + 2 3 J :Ixe 1 �b8 2 4 . 'iVh6 + ) 2 0 .'iVd 2 White has regained his sacrificed pawn , while his attack continues. 1 60

1 9.

bly to take the pawn with 24 . . . 'iVxb2 ! . White has the upper hand, but things are far from clear: A) Black is very much in the game after 2 s .iVh6 + 'iVg7 2 6 .iff4 �g8 2 7 .�fS (or 2 7 .�d3 �g4) 27 . . J::rac8 ! (not 2 7 . . . ttJc4 ? 2 8 . �ad 1 ) 2 8 .ttJf3 lIc4 ; B) The straightforward 2 S . ttJfS ? does not work after 2 S . . . fxg6 2 6 .'iVh6+ �f7 2 7 . ttJ d 6 + rJil e 7 2 8 . ttJxe 8 nxe 8 29 .'iVxg6 'iff6 and Black is better; C) 2 S .�fS ! 'iYf6 2 6 . 1:t.ad l �xfs (2 6 . . . ttJc4 ? ? is a terrible blunder due to 2 7 .�xe6 l:Ixe6 2 8 .1dxe6 fxe6 2 9 .ttJg6+ �e8 3 0 . 'if g 8 + ) 2 7 . l:Ixe 8 + !;Ixe 8 2 8 .ttJxfS ttJc4 2 9 .g4 ! lleS 3 0 J ld3 . 25.

�h 7xg 7+ wfSxg7

26.

ii.g6-d3

White is a sound pawn up and in the hands of Kramnik this means that the game is more or less decided. It is nice to enjoy Vladimir's technique, but as far as understanding isolated pawn-posi­ tions is concerned, the game is over. I give the rest with very limited comments.

32.

2:

I s o l a t e d P a wns

lld1 -d3!

White is not in the mood to allow drawing chances , which could have been obtained after 3 2 .�xd4 �xf3 3 3 . .l:!xd8 nxd8 3 4.gxf3 ttJd2 3 S .�g2 (or 3 S .�e2 �e8) 3 S ... .tId3 3 6 .f4 ttJc4. 32.

...

ttJc4-b2

33.

l::I d 3xd4

ii.g4xf3

34.

nd4xdS

l::t eSxdS

35.

g2xf3

Now White has taken the black d-pawn under much better circumstances

35.

...

l:tdS-cS!

36.

Ya1 -a2!

ttJ b2-a4

In the case of 3 6 . . . nc2 the simplest road to victory is 3 7 .a4 ! bxa4 3 8 .�xa6 , creating a ed pawn.

26.

...

ttJa5-c4

27.

b2-b4

ttJc4-b2

2S.

ii.d3-f1

d5-d4

3 7.

na2-d2

�cS-c6

29.

ttJh4-f3

l:t.aS-dS

3S.

f3-f4

ttJa4-b6

30.

l:le1 - b1 !

ttJ b2-c4

39.

�g 1 -g2

ttJ b6-c4

31 .

llb1 -d1

ii.e6-g4

40.

J:t.d2-d3

ttJc4-b2

41 .

l::t d 3-g3+

�g7-hS

Or 41 . . . �f6 42 .h4.

24.

...

'iVf6-g7?!

Having struggled for quite some time in a difficult position, Anand misses the best practical defence - and the chance to stay in the game -, which was proba-

42.

ii.f1 -e2

43.

ii.e2-h5

l:Ic2-c7

44.

f4-f5

�hS-h7

45.

ii.h5-e2

.tlc6-c2

4s .f6 or 4S . �f3 should be a simple win. � h 7-h6

45 .

...

46.

h 2-h4

l:tc7-c2

47.

ii.e2-f3

ttJ b2-c4

161

C h a p t e r 2 : I s o l a t e d P a wns

Winn i n g C h es s M i d d l e g a m es

48.

�f3-dS

A blunder would have been 48 .�b 7 ? ttJd6 49 .�xa6? �c6. 48.

...

tbc4-d6

Now 48 . . . ttJeS loses to 49 .1Lb 7 . 49.

.J:Ig3-d3!

�h 6-g 7

White wins after 49 . . . ttJxfS S O .�b 7 ttJxh4+ S 1 . �g3 (S l . �g l should also be enough, but S 1 .�h3 ? ttJg6 probably is not) S I . . . ttJ fS + S 2 . �f3 ttJh4+ S 3 .�e3 ttJfS + S 4.�e4 �xf2 S S .i.xa6 .

ficulty to cope with: the one on the rook file. S7.

...

tbd4xc6

S8.

�a6xbS

tbc6-aS

S9.

�g 1 -g2

wg 7-f6

60.

�g2-f3

�f6-eS

61 .

�f3-e3

tbaS-b7

62.

.tbS-c4

f7-f6

Even more hopeless for Black would be 6 2 . . . fs 6 3 .f4+ �d6 64.i.d3 �e6 6 S . �d4. 63.

a3-a4

tbb7-aS

64.

�c4-f7

tbaS-c6

6S.

�e3-d3

�eS-d6

66.

�d 3-e4

tbc6-e7

67.

a4-aS

\t>d6-cS

68.

as-a6

tbe7-c8

68 . . . �b6 loses to 6 9 . i.c4 c:t;a7 7 0 .�f4 �b6 7 1 . � g4 � a 7 7 2 . �h S �b6 7 3 . �h6 and White eventually collects Black's f-pawn. SO.

�dS-f3 !

tbd6xfS

S1 .

�f3-b7

tbfSxh4+

S2.

Wg2-g 1

.1::t c 2-e2

S3.

�b7xa6

�e2-eS

S4.

l::r d 3-c3

tbh4-fS

SS.

l:Ic3-cS

1:[eSxcS

S6.

b4xcS

tbfS-d4

69.

69.

cS-c6!

White forces Black to fight against the er that the knight has the most dif1 62

1.

d2-d4

d7-dS

2.

c2-c4

dSxc4

3.

tbg 1 -f3

e7-e6

4.

e2-e3

tbg8-f6

S.

�f1 xc4

c7-cS

6.

0-0

a7-a6

7.

�c4-b3

cSxd4

8.

e3xd4

tbb8-c6

9.

tbb1 -c3

�f8-e7

1 0.

�c1 -gS

In two previous examples (Kramnik­ Hubner, Game 3 0 and Kramnik-Anand, Game 3 3 ) we have analysed 1 0 . �e 1 . The text move is the other main line here. Note that Kramnik has taken his time to thoroughly analyse this position and a considerable part of his success came as a result of his superb under­ standing of isolated pawn -positions. ...

...

�cS-d6

.thS-f3

1 -0

Black's situation is hopeless, for exam­ ple: 7 0 . . . ttJa7 7 1 . �d4 ttJc8 7 2 .i.b 7 ttJa7 7 3 . �c4 o r 7 0 . . . �c6 7 1 .�fS + �b6 7 2 .i.b 7 .

A) 1 1 . . . h6 and now: A I ) The idea to eliminate the f6 knight and try to create threats on the b I -h7 diagonal, does not work here : 1 2 .�xf6 ? ! �xf6 1 3 Jiad l ttJaS 1 4.i.c2 bS I S .b3 1Lb 7 1 6 .ttJe4 �c8 1 7 .i.b l 1Lxe4 1 8 .�xe4 ttJc6 1 9 .I;He 1 ttJe7 2 0 . i.b l ttJdS 2 1 .'iVd3 g6 and Black was slightly better in Gelfand-Shirov, New Delhi rapid 2 0 0 0 ; A2) 1 2 .�h4, o n the other hand, is a rather logical move and I am surprised that it has not been tried on grandmas­ ter level. Should Black decide to follow our main game idea with 1 2 . . . ttJaS 1 3 . �c2 bS , the vulnerability of the g6-square (resulting from 1 1 . .. h7 -h6) would be immediately felt after 1 4 .'iYd3 g6 I S . ttJeS ; A 3 ) 1 2 . �f4 � e 8 1 3 . I:rad l �f8 1 4 .ttJeS ttJe7 I S .'iVd3 (White is now setting up a classical isolated pawn at­ tack) 1 5 . . . ttJedS

0-0

�f7-hS!

Or 69 . . . �b6 7 0 .1Le2 . 70.

GAME 34

Vladimir Kramnik Viswanathan Anand Dortmund 2 00 1 (9)

1 0.

White's bishop, combined with the ex­ tra a-pawn, is easily dominating the black knight.

S7.

Q G 1 1 . 1 4 ( D2 7 )

an alysis d iagram

11.

'iYd 1 -d2!

As we see, White is not going to follow a traditional type of isolated pawn at­ tack (lae l , a3 , i.c2 , �d3 ) , but has a rather fresh idea in mind. 11.

...

tbc6-aS

Black has tried two other natural con­ tinuations, but both fail to equalize:

1 6 .�c l ( 1 6.ttJxdS is another way to proceed here, giving White advantage after 1 6 . . . ttJxdS (in the event of 1 6 . . . exdS White will have a pleasant and risk-free advantage after 1 7 .i.a4 ! bS 1 8 .�c2 - the knight on eS is very dominant, while the attacking possibilities along the b I -h 7 diagonal 1 63

C h ap t e r 2 : I s o l a t ed P a wns

W i n n i n g C h ess M i d d l eg a m es

remain) 1 7 .ii,c2 fS 1 8 .'iVg 3 ttJxf4 (or 1 8 . . . �f6 1 9 .�b 3 ) 1 9 .'iVxf4 ii,d6 2 0 Jife l as 2 1 . h4 and White was defi­ nitely better in Barsov-Lesiege, Mon­ treal 2 0 0 2 ) 1 6 . . . bS 1 7 .'iYg3 ii,b 7 (Black hopes to have enough tactical re­ sources to withstand the attack) 1 8 .ii,xh6 ttJxc3 1 9 .bxc3 ttJe4 2 0 .'iVg4 ttJf6 2 1 .'iVgS (White considers he has more than a draw) 2 1 . . . 'iVe 7 2 2 Jid3 ( 2 2 .ii,c2 , preventing Black's . . . ttJe4 drawing mechanism, deserved atten­ tion) 2 2 . . . ttJe4

B) 1 1 . . . ttJdS allows White to steer the game into a ' central pawn symmetry / upper hand position' after 1 2 .ttJxdS exdS 1 3 .ii,xe 7 ttJxe 7 1 4 Jlfe 1 .

2 3 .'iYg4 (because of his weak c3 pawn, White decides to avoid a pawn-up end­ ing. It was, however, definitely interesting to consider 2 3 .'iVxe7 l::lxe7 24.ii,f4 and it is not easy for Black to get his pawn back, for instance: 24 .. J:tc8 2 s J�k l llec7 2 6 .c4! gS 2 7 .ii,e3 f6 ? 2 8 . cS fxeS 29 .ii,xe6+ �g7 3 0 .�xc8 ii,xc8 3 1 .dxeS ii,fS 3 2 .c6 ! ttJxf2 3 3 J:td7 + ii,xd7 34.cxd7 �xd7 3 S .�xf2 and White has a won endgame) 2 3 . . . ttJf6 24.'iYgS ttJe4 2 5 . 'iY g4 ttJf6 with a draw by move repe­ tition, Kramnik-Anand, Mainz 20 0 1 . As you have probably noticed, Kramnik and Anand have carried on an extensive de­ bate in the Queen's Gambit Accepted, which went favourable for Kramnik in most cases; 1 64

1 3.

As explained before, with the d4/ dS pawn symmetry in the centre and the dark-squared bishops exchanged, the light-squared bishops almost invariably work in White's favour. Here, White also has good control of the e- and c-files and he will put a dominant knight on eS . White has a long-term, risk-free advantage. The 7 th match game Leko-Karpov, Miskolc 2 0 0 6 , is a good example : 1 4 . . . 'ifd6 l S J!ac 1 �g4 1 6 .ttJeS ii,fS 1 7 . l:Ic3 l:rac8 1 8 .�ce3 ttJg6 1 9 .ttJxg6 hxg6 2 0 . lieS ii,e6 2 1 .h4. White has e-file control, with first his knight and now his rook well-placed on e S . White's light-squared bishop heavily dominates its black counterpart here and this plays a crucial role. This typical position , with the d4/ dS pawn symmetry, the white bishop on b3 or d3 , and the black bishop on e6 or b 7 , occurs very often and almost invari­ ably. White has long-term, risk-free pressure, while Black has very limited to zero counterplay. 1 2.

�b3-c2

b7-b5

�d2-f4 !

The queen goes to h4 - this is the idea behind 1 1 .'iYd2 -, where it will be ex­ cellently placed for all sorts of attacking motifs and, above all, to exercise pres­ sure along the h4-d8 diagonal, making the potential d4-dS thrust rather pow­ erful. This idea in itself does not belong to Kramnik, but, if I am well informed, to Vladislav Tkachiev. 1 3.

a n a lysis diagram

ana lysis d i agram

The immediate 1 2 . . . ttJc4? is a horrible move, since after 1 3 .'iVe2 b S ? ? blunders a rook to 1 4 .�xf6 �xf6 I S .�e4.

...

l::t a 8-a7

This rook transfer does not work well here. It takes Black quite some time and meanwhile his pieces remain clumsy. A) In the stem game, the natural 1 3 . . . �b7 was tried. However, White has strong attacking potential and this move does not solve Black's problems : 1 4.�ad 1 . A I ) Black was probably not in a mood to weaken his kingside immedi­ ately, but he is having a difficult life. If he starts with the logical 1 4 .. J:lc8 , the white attack will break through after l S .'iYh4 h6 ( l S . . . g6 runs into the the­ matic 1 6 .dS ! �xdS ( 1 6 . . . exdS is bad due to the simple 1 7 J�fe 1 ) 1 7 .ttJe4! and Black gets mated)

a n alysis d i agram

And now the standard sacrifice 1 6 .�xh6 ! gxh6 ( 1 6 . . . �xf3 1 7 .gxf3 changes very little) 1 7 .�xh6 (the black king is caught in a mating net and once his only defender, the knight on f6 , gets eliminated, that will be the end of the game) 1 7 . . . b4 (the only shot at counterplay) 1 8 . ttJ e S ! nxc3 (on 1 8 . . . bxc3 ? 1 9 J �d3 mates) 1 9 .bxc3 'ifdS 2 0 .f3 with 2 1 .ttJg4 to follow, and White mates; A2) Keeping the white queen from the attacking position on h4 with 1 4 . . . ii,xf3 , and after I s .'iYxf3 trying to get counterplay on the c-file with I S . . J dc8 , would not work after 1 6 .dS ! exdS ( 1 6 . . . ttJxdS ? simple loses a piece after 1 7 . 'iVe4 g6 1 8 . �xe 7 �xe 7 1 9 .ttJxdS ) 1 7 .�xf6 �xf6 1 8 JlxdS (the white rook is to be transferred to the h-file and then, with all the white pieces attacking, Black is defenceless) 1 8 . . . iVb6 1 9 . �hS h6 2 0 .ttJ dS ( 2 0 . llxh6 should also win after 20 . . . gxh6 2 1 .'iffs l:lfd8 2 2 .'ifh7 + f8 2 3 . 'iYxh6 + �e 7 2 4 . �e 1 + � d 7 2 S .ttJdS 'iY d 6 2 6 .�fs + \t> c 6 2 7 .ttJxf6 l:lh8 2 8 .�e4+) 2 0 . . . 'iYd6 (20 . . . 'iYd4 2 1 .nd 1 'iYxb2 2 2 .l:Ixh6) 2 1 . l:Ixh6 ! gxh6 (2 1 . . Jlxc2 ? loses immediately to 2 2 .ttJxf6 +) 2 2 .ttJxf6 + \t>g 7 2 3 .'iffs llxc2 24 .ttJhS + Wh8 2 S .'ifxc2 and White has won a healthy pawn, while his attack continues ; A3) 1 4 . . . g6 l S .�h6 ttJhS ! ( 1 S . . . l:Ie8 1 6 . ttJeS is bad for Black - White has terrible pressure on f7 ) 1 6 .'iYg4 fS ? ! (Black wants t o solve his kingside prob­ lems in a radical way and also to drive the white queen to h3 , where it would be vulnerable to tactics related to Black's . . .ttJf4 jump. Carrying on chaSing the queen with 1 6 . . . ttJf6 ! 1 7 .�g 3 ttJhS is a 1 65

Chapter

W i nn i n g C h e s s M i d d l eg am e s

A) 14 ... <1Jc4 1 5 .<1Je5 ! �c7 (taking a pawn with 1 5 . . . <1Jxb2 ? would lose after the principled 1 6 .'iVh4 ! (if White for­ gets about developing his attack and gets greedy with 1 6 .<1Jc6 ?, then Black will be OK after 1 6 . . . <1Jxd I ! 1 7 .<1JxdS <1Jxc3 I S .'iVf3 nxdS 1 9 .�xc3 �ad 7 ) 1 6 . . .h 6

standard QGA motif and could well be good advice here, since it seems that Black holds his own after I S . 'iVh3 .a:eS 1 9 . <1Je5 jdcS 20. nfe 1 ! (the threat is 2 1 .<1JfI ) 20 . . . �g5 ! 2 1 .�xg5 �xg 5 2 2 .�e4! (after exchanging the light-squared bishops, the white c3 knight threatens to jump powerfully to e4) 2 2 ... <1Jf6 ! 2 3 . 'iVf3 �c7) 1 7 .'iih 3 �cS

an alysis d i agram

I S .d5 ! (consistent and strong. Naturally White is not in a hurry to take the ex­ change with I S .�xfS ?, since Black gets a strong attack after I S . . . 'iVxfS 1 9 .�d3 <1Jf4 2 0 .�g3 �d6) I S ... b4 (desper­ ately hoping for counterplay, but White is one step ahead. If I S . . . exdS , then simply 1 9 .<1Jd4 and White has total po­ sitional domination and plenty of dev­ astating threats) 1 9 .dxe 6 �d6 2 0 .�fe I ! �eS 2 1 . �a4 ! ld e 7 2 2 . �d 7 ! bxc3 23 . .ldxd6 cxb2 24.�g5 and White soon won in Tkachiev-Lesiege, New Delhi! Teheran 2 0 0 0 . B ) Realizing that the . . . g 7 - g 6 weaken­ ing seems to be almost inevitable, black players have recently resorted to the im­ mediate 1 3 . . . g 6 . One instructive exam­ ple is 1 4.�h4 ReS ! , defending the e7 bishop in order to prevent d4-d5 breaks. 1 66

2:

I s o l a t e d P awns

2 1 .d6 ! (2 1 ..a:xd l should also win) 2 1 . .. 'iVxd6 2 2 .<1Jxf6 + �xf6 2 3 .�h 7 + ! � h S 24.�g6+ � g S 2 5 . <1Je4 ! 'iVe5 2 6 . <1Jxf6 + 'iYxf6 2 7 . �h 7 + �hS 2 S .'iVxf6+) 1 6 .<1Jxc4 ! (having a good feeling about the position, Kramnik correctly forces a favourable change of pawn structure) 1 6 . . . bxc4 ( 1 6 . . J�xc4 allows 1 7 .�b 3 )

ana lysis d i agram

1 5 .d5 ! (this principal thrust seems to be working in most of the lines here) 1 5 . . . <1Jxd5 1 6 .<1Jxd5 exd5 1 7 .1';Iae 1 <1Jc6 (not advisable for Black is 1 7 . . . �xg5? I S .<1Jxg5 h5 1 9 .'ifg3 with <1Jxfl being a terrible threat, while 1 7 . . . �a7 ! ? could be an improvement) I S J ixe 7 ! llxe 7 (hanging on to the exchange would be a bad decision, since after the simple I S . . . <1Jxe7 ? 1 9 .�f6 'iVd6 2 0 Jle 1 'iic6 2 1 .�c3 ! White has a mating attack, for example : 2 1 . . . f6 2 2 .<1Jg5 ! fxg5 2 3 .'iVd4) 1 9 .<1Je5 ! <1Jxe5 20 .�xe 7 'iYc7 2 1 .�b3 �e6 2 2 .�g5 - Black is certainly missing his dark-squared bishop. Due to the weaknesses around the black king, White has a lasting initiative, probably worth more than a pawn, Bruzon-L. Dominguez , Cuban Championship, Santa Clara 2 0 0 6 . 1 4.

l:ta 1 -d 1

�c8-b7

Black's rook on a7 looks clumsy after this move and he is never going to get time for . . . �b7 -as to bring the rook into play. Black is one tempo away from reaching harmonious development. En­ ergetic action is required by White and Kramnik seizes the moment. As already mentioned, Kramnik and Anand have a long-lasting debate in the Queen's Gambit Accepted and one of their games went:

ana lysis d i agram

1 7 .�xh6 ! (thiS standard sacrifice works also here) 1 7 . . . gxh6 l S .'iixh6 �c7 1 9 .<1Jg4! (again, eliminating the only black defender, the knight on f6 , is the key) 1 9 .. J�eS 2 0 .d5 ! (adding pressure, after which Black's position collapses) 20 . . . <1Jxd l ( 2 0 . . J :ixc3 ? ? would volun­ tarily remove a defender of the fI -square and run into a quick mate : 2 1 . <1Jxf6 + �xf6 2 2 .�h 7 + �hS 2 3 .�g 6 + � g S 2 4 . 'iVh 7 + �fS 2 5 .�xfl mate)

analysis d i agram

an alysis d i agram

1 7 .�xf6 ! �xf6 I S .d5 ! (creating a strong ed pawn) I S . . . e5 1 9 .'iVf3 �b 7 2 0 . 'ii e 4 g 6 2 1 . �xc4 J:.xb2 2 2 .�b3 �g5 2 3 .d6. Now White was clearly better and he won easily in Kramnik-Anand, 3rd match game, Leon man + computer 2 0 0 2 ; B ) A logical continuation of Black's plan started with 1 3 . . J la7 would have been 1 4 . . J�c7 . It is quite possible that the immediate rook transfer to c 7 was Anand's original idea, but he simply did not like the position arising after 1 5 . l:Ife 1 (White's d4-d5 thrust needs some preparation. The immediate 1 5 .d5 would not be that effective, since after 1 5 . . . exd5 1 6 .�xf6 �xf6 1 7 J �hd5 �d 7 ! I S . l::t fd l �xc3 1 9 .1:t xd 7 ( 1 9 .�xh 7 + ? �xh 7 2 0 .<1Jg5 + �gS 2 1 . 'iVh4 JleS) 1 9 . . . �xd 7 2 0 .bxc3 'iVcS ! Black would gradually equalize) 1 5 . . . �b7 . 1 67

C h a p t e r 2 : I s o l a t ed P a wn s

W i n n i n g C h e s s M i d d l e g a m es

a n a lysis d i agram

analysis d i agram

1 6 .dS ! (now this typical temporary pawn sac seems to be rather unpleasant for Black again. Please note that the d4-dS thrust worked very well in a huge number of lines shown ! ) 1 6 . . . �xd5 ( 1 6 . . . exd5 ? ? i s very bad; Black immediately gets mated after 1 7 . nxe 7 ! 'ifxe 7 I S .�xf6 gxf6 1 9 .'iYg4+ �hS 2 0 .'iff5 ; or 1 6 . . . tLlxd5 1 7 . tLlxd5 .ixd5 I S . .ii!. x h 7 + (or 1 S .ldxd5 , transposing to the main line) I S . . . �xh 7 1 9 . 'ifh4+ �gS 2 0 .�xe 7 'iYxe 7 2 1 . tLl g 5 'iYxg 5 2 2 . 'iYxg 5 ) 1 7 . tLlxd5 tLlxd5 ( 1 7 . . . exd5 ? i s again bad : I S .�xe 7 ! 'iYxe7 1 9 .�xf6 gxf6 2 0 .�f5 ! and White has a mating attack)

2 3 J ie l ! (2 3 J :ta7 100ks active, but Black escapes with 2 3 . . . 'iYb6 24.�d4 'ifxd4 2 5 .tLlxd4 ldcS ! 2 6 .g3 tLlc6 2 7 .tLlxc6 Ilxc6 into a drawn endgame) 23 . . . Me8 24.ldd l - Black has a weak d-pawn and an exposed king and he is facing a diffi­ cult defence.

1 5.

d4-d5 !

This thematic pawn thrust works excel­ lently here. It soon becomes clear that White will get his pawn back while his attack continues. 1 5.

analysis d iagram

I s .ldxd5 ! exdS 1 9 .�xh 7 + �xh 7 2 0 . �xe 7 ldxe 7 2 1 . fih4+ � g S 2 2 JIxe 7 f6 (only move; 2 2 . . . tLlc6 ? ? blunders into a mate after 2 3 .tLlg5) 1 68

...

.ib7xd5

Forced. Other captures lose : A) 1 5 . . . tLlxd5 ? 1 6 . tLlxd5 �xdS 1 7 .ldxd5 ! (White has a winning attack, but still he has to be careful with the ex­ ecution , for example : 1 7 .�xh 7 + ?
I S . . . I:rxe 7 , 1 9 .'iYh4 and White is a piece up) 1 9 . .ii!.x h7 + �xh7 2 0 .tLlgS + wins) I S .jlxh 7 + �xh7 1 9 .iYh4+ �gS (or 1 9 . . . �g6 2 0 .tLle5 + �f5 2 1 .'iVg4+ ! �xe5 2 2 .�f4+ �f6 2 3 .lde l with mate next move) 2 0 .�xe 7 iYxe7 2 1 .tLlgS and Black must give his queen in order to prevent mate; B) I S . . . exdS ? does not help either: 1 6 .�h4 h6 ( 1 6 . . . g6 loses to 1 7 .tLle4 ! , while 1 7 J:[fe 1 should also win) and now the standard piece sacrifice decides again: 1 7 .�xh6 ! gxh6 I S .'iYxh6 and White soon delivers mate. 1 6.

ttJc3xd5

e6xd5

1 6 . . . tLlxd5 1 7 .ldxdS ! transposes to the variation mentioned in the previous comment ( 1 S . . . tLld5 ) . 1 7.

'iYf4-h4

1 7.

...

1 8.

l';1f1 -e1 !

White is calmly increasing the pressure. 1 8.

...

ttJa5-c6

It may seem that Black, by bringing his knight back into play, is going to escape, but Kramnik finds a way to break the fortress. Anyhow, it is difficult to give Black good advice as other moves would not suffice either, for instance : A) I S . . . 'iYcS 1 9 .tLld4 �g4 (trying to find refuge in the endgame) 2 0 .'iYxg4 hxg4 2 1 .�d2 ! �dS (2 1 . . . b4? loses to the simple 2 2 J ixe 7 �xe 7 2 3 .�xb4) 2 2 .�b4 .hIeS 2 3 J :txeS + ttJxeS 24.�fS - White's bishops are dominating the board and Black will soon start losing material; B) Or I S . . Jdc7 1 9 .ttJd4 ttJc4 2 0 .ttJfS ttJxb2 2 1. .I1c l �eS 2 2 .'iYg3 and White should win in the attack.

h 7-h5!

In a difficult position, Anand is putting up the best defence. Other moves would make White's task considerably easier; after 1 7 . . . h6 , for example, the standard sacrifice works again : 1 8 .�xh6 ! gxh6 1 9 .�xh6 l:Ic 7 (or 1 9 . . . J::t e S 2 0 . l::I d4 �fS 2 1 . ldg4+ , mat­ ing) 2 0 .tLld4 ! (as usual, the involve­ ment of extra forces in the attack de­ cides the battle) 20 . . . l:Ixc2 2 1 .�d3 ! tLlg4 2 2 Jig3 'iVd7 2 3 .�hS wins. After 1 7 . . . g 6 1 S .l::l fe 1 , Black loses mate­ rial.

1 9.

92-94!

�d8-d6

If 1 9 .. Jlc7 , 2 0 .gxhS simply continues the attack. 1 9 . . . �eS ? loses by force to 2 0 .�xf6 .txf6 2 1 . 'ifxhS Ile6 (or 2 1 . . . g 6 2 2 . .txg 6 ! ldxe l + 2 3 . ldxe l fxg6 24.'ifxg 6 + �fS 2 S J le6) 2 2 J Ixe6 fxe6 2 3 . l:Ie l 'iYd7 24 . .ii!.f5 ! ttJdS 2 5 . gS .txb2 2 6 .ttJh4 and either Black gets mated or he loses tons of material. In the event of 1 9 .. J :td7 , White will be a healthy pawn up after 2 0 .gxhS ttJe4 1 69

Chapter

Wi n n i n g C h ess M i d d l eg a m e s

2 1 .�xe4 dxe4 22 J lxd7 'iYxd 7 2 3 .'iYxe4 . 1 9 . . . ttJxg4 ? is bad, losing to 2 0 . 'ifxhS fS 2 1. .�xdS 'iVxdS 2 2 .�b3 . 20.

g4xh5

'iYd6-b4

Black is trying to exchange queens , hoping to find it easier to neutralize the attack. 20 . . . ttJh7 does not solve his problems, since the attack continues after 2 1 .�f4! 'iVcs (2 1 ... 'iVb4 10ses to 2 2 .'iYg3 'iYxb 2 2 3 .�xh 7 + �xh 7 2 4 . ttJ e S ttJd8 2 S .ttJ d 7 ! with 2 6 .�eS to follow) 2 2 .�xh 7 + �xh 7 2 3 .'iVg4. 21 .

hS-h6!

27.

27.

...

J::l cS-c5

The attempt at counterplay 2 7 . . . .!:tc2 loses to 2 8 .�g4+ �h7 2 9 . �f6 ttJg6 (or 29 . . . ttJg8 3 0 J :tf3 ttJh6 3 l .�gS l::rxb2 ? 3 2 .l:IhS) 3 0 .ttJfS �ac7 3 l .h4! . 2S.

l:te4-g4+

�g7- h 7

29.

ttJ h4-f3 !

ttJe7-g6

30.

ttJf3-g5+

�h7-g7

31 .

Kramnik correctly judges that his attack also continues strongly after the queen swap.

J::l d 1 -d6!

White correctly judges that there is much more in the position than just the pawn-up rook endgame resulting from 2 7 .1'dxe7 !:rxe7 2 8 .tLJfS + �f6 2 9 .ttJxe 7 Wxe 7 .

ttJgSxf7

A rook ending with two extra pawns is something Kramnik is prepared to con­ sider.

21 .

...

�b4xh4

31 .

l:la7xf7

22.

ttJf3xh4

ttJf6-e4

32.

�d 6xg6+ �g7-h7

23.

h6xg 7

nfS-cS

33.

l:lg6-gS

lic5xg5

34.

.ug4xg5

'fJ.f7-c7

24.

.igSxe7

ttJc6xe7

25.

�c2xe4

dSxe4

26.

J:re1 xe4

c;t>gSxg7

It seems as if Black has managed to keep the damage limited and, in the ensuing endgame, has reasonable chances of survival . Black's main problem here, however, is that the white attack is far from over. It is nice to see how Kramnik, with relatively little material left, manages to keep a firm initiative until the end of the game.

1 70

The only thing White has to take care of is to keep at least one queenside pawn on the board in order to avoid a theo­ retical rook + h- and f-pawn versus rook ending.

3S.

a2-a3

b5-b4

36.

a3xb4

l:tc7-c1 +

3 7.

�g1 -g2

�c1 -b1

3S.

ng5-aS

l:tb1 xb2

39.

�a5-a4 !

1 -0

Even though the next game in itself does not really belong in a chapter on isolated pawn-positions, it is an excel­ lent example of the dynamiC power of central pawn breaks and also a good in­ terlude between this and the next chap­ ter, on hanging pawns.

QO 8 . 1 2 (DS 8)

GAME 3 5

Miguel Illescas Cordoba Nigel Short Pamplona 1 9 9 9 / 0 0 (5) d 7-dS

1.

ttJg1 -f3

2.

d2-d4

ttJgS-f6

3.

c2-c4

e7-e6 �fS-e7

4.

ttJb1 -c3

5.

�c1 -g5

h 7-h6

6.

�gS-h4

0-0

7.

e2-e3

b 7- b6

The Tartakower IMakogonov Variation is one of the oldest and strategically soundest variations in the classical openings. S.

�f1 -e2

.icS-b 7

9.

�h4xf6

�e 7xf6

c4xd5

e6xdS

1 0.

One may wonder why White first 'loses ' a tempo with 6 .�h4 and then takes with 9 .�xf6 . Well, White wanted to wait for Black to place his bishop on b 7 , then take on f6 and capture with the pawn on dS , forcing the . . . exdS recap­ ture and obtaining a pawn structure where his light-squared bishop should be better placed than its black counter­ part. This position resembles those with the d4 I dS pawn symmetry in the centre where (as we have already established) White's light-squared bishop is almost per definition better placed than its black colleague.

2:

I s o l a t e d P a wns

Here, however, there is a rather impor­ tant difference, since White has a mo­ bile pawn on e3 and Black has one on c7 . Most of the time it is advantageous for White if Black decides to push . . . c7 -cS and then captures . . . cxd4 , al­ lowing White to recapture e3 xd4 and obtain the advantageous central sym­ metry with a better bishop. The problem for White, however, is that after playing . . . c7 -cS , Black usually keeps the tension in the centre, relying on his bishop pair (the bishop being well placed on f6) . Should White try to take advantage of Black 's hanging pawns and at some stage take d4xcS , then Black gets the opportunity (as we will see later in a number of examples in Chapter 3 on hanging pawns) to play the dynamiC . . . dS -d4 pawn thrust, of­ ten developing a serious initiative. This is one of the main positions in the Tartakower IMakogonov, which is also important for a strategic understanding of pawn structures in the middlegame in general. So I shall give a number of lines that are not of direct interest for this game, but serve to create a better understanding of the pros and cons of the position.

11.

0-0

Another rather logical and frequently played move here is 1 1 . b4 (White 171

Chapter

W i n n i n g C h ess M i d d l e g a m es

seizes space on the queenside, trying to discourage Black from making the lib­ erating move . . . c7 -cS) and now: A) The immediate 1 l . .. cS is possible, but plays rather into White's hands. Black normally lands in a slightly infe­ rior, ive position after 1 2 .bxcS bxcS 1 3 . �b l 'iVaS 1 4 . 'ifd2 cxd4 I S .ttJxd4 �xd4 1 6 .exd4 and again, the central d4/dS pawn symmetry with White having the upper hand due to his superior light-squared bishop, has been reached; 1 6 . . . �a6

B 1 ) Black can first develop his pieces with 1 2 . . . Iie8 1 3 .'iYb3 ttJd7 , but then he will remain ive after 1 4J!ad 1 ! (putting pressure on the dS -square, which means that Black's . . . a7 -as thrust will never work because it would be met by b4-bS , hitting on the potential weakness of his dS pawn) 1 4 . . . a6 I S .a4 ttJf8 1 6 .aS ! bxaS 1 7 .bxaS (Black's b 7 bishop i s merely a pawn here) 1 7 . . J:tb8 1 8 .ttJa4 ! ttJe6 (taking a pawn with 1 8 . . . 'iYxaS ? leads to total white domi­ nation after 1 9 .ttJcS) 1 9 .'ifc3 and White had a clear advantage in Kaspa­ rov-Ehlvest, Belgrade 1 9 8 9 ; B2) 1 2 . . . 'iVd6 1 3 .'iVb3 ttJ d 7 1 4. 1:Ife l �e 7 and after 1 S . :tab 1 Black can search for counterplay with I S . . . aS 1 6 .bxaS �xaS .

a n a lysi s d i agram

1 7 .ttJbS ! 'ifd8 1 8 . 0-0 ttJd7 1 9 J ifc l ttJf6 2 0 .f3 and White was better in Vaganian-Geller, New York 1 99 0 ; B ) 1 1 . . . c6 (for the time being, Black agrees to play a relatively ive positi­ on, planning to develop counterplay as­ sociated with . . . a7 -as) 1 2 .0 -0

a n a lysis d i agram

1 72

ana lysis d i agram

White now fixes Black's weakness on b6 and prepares the e3 -e4 break. The black bishop on b7 remains ive, while after White executes e3 -e4, Black's pawns on b6 and c6 will be much weaker than the white ones on a4 and d4. The following game gives an excellent illustration of White's pluses : 1 7 .a4 I:le8 1 8 .�fl �f8 1 9 .'ifc2 ! (pre­ paring e3 -e4. The immediate 1 9 .e4? would have been a mistake due to 1 9 . . . dxe4 20 . ttJxe4 'iY a3 ! ) 1 9 . . . g6 (now the time has come for White to

execute the thematic central break, favourable under circumstances. 1 9 . . . �a6 ! , exchanging the light­ squared bishops and eliminating plenty of tactical possibilities for White, was advisable) 2 0 .e4! dxe4 2 1 .ttJxe4 'iYf4.

analysis d i agram

This is one of the thematic positions in this type of Tartakower /Makogonov Variation. Black has weak pawns on b6 and c6, while White easily defends his potential weaknesses on a4 and d4 . Black's weakness on f7 and White's im­ minent control of the e-file provides him with plenty of tactical possibilities. It is important for the reader to know this thematic position, because it can arise frequently in this variation and it almost invariably favours White. Karpov himself has won many games as White, proving over and over again that the pawn struc­ ture White: a4, d4, Black: b6, c6 . . .

2:

I s o l a t e d P a wns

. . . favours White (find his games in this line in a database and study them on your own ! ) . 2 2 .�c4! (by now Black definitely regret­ ted not exchanging his bishop with 1 9 . . . �a6) 2 2 . . . �g7 2 3 .l:t:e2 (control of the e-file is essential here) 2 3 . . . cS (Black's position is already difficult and this at­ tempt at counterplay does not help) 24.dS l:taa8 2 s Jibe i �ad8 26.iVb3 ! (an excellent, multifunctional move. White targets b6 and f7 and prepares to chase away the black queen with g2-g3) 26 ... �a8 2 7 .g3 'iYb8 2 8 .d6 and now: B2 1 ) Black cannot solve his problems tactically with 2 8 . . Jlxe4 2 9 . �xf7 + �h7 3 0 J :txe4 �xe4 3 l .bf.xe4 ttJf8 (if 3 1 . . . ttJf6 , then 3 2 .ttJeS ! wins immedi­ ately) because of 3 2 . ttJ e S ! �xeS 3 3 . 1:txeS 'iYxd6 3 4 J �dS 'iYe 7 3 S .�xg 6 + ! ttJxg 6 3 6 . �xd8 'iYxd8 3 7 . 'iYf7 + �h8 3 8 . 'ifxg6 'ifd l + 3 9 . � g 2 'iYxa4 4 0 . 'iYxh6 + � g 8 4 1 .'iYg 6 + � h 8 42 .h4! 'ifa8 + 43 .f3 'iVa2 + 44.�h3 iVdS 4S .'iVf6 + �g8 46 .g4 c4 4 7 .hS and White wins; B2 2 ) 2 8 .. J�J8

analysis d i agram

Now the situation is ripe for a tactical

Structure a4/d4 v s b6/c6

solution, and Karpov's execution is bril­ liant: 2 9 .�xf7 + ! b(xf7 3 0 .ttJegS hxgS 3 l . ttJxg S 1:(f8 1 73

C h a p t e r 2 : I s o l a t e d P a wns

Winn i n g C h e s s M i d d l eg a m e s

analysis d i agram

3 2 . 1:le8 ! (precision is important; 3 2 .ttJxf7 ? is wrong due to 32 . . .'ifb7 ! 3 3 .ttJd8 + 'ifdS 3 4 . 'ifxdS + �xdS 3 5 . ttJe6 brf7 and Black is slightly better) 3 2 . . . 'ifxd6 3 3 .'ii' xf7 + �h8 34.ttJe6! 1 -0 , Karpov-Kir. Georgiev, Tilburg 1 9 94. B3) 1 2 . . . aS ! , looking for immediate counterplay. Now White has three pos­ sibilities : B3 1 ) White's standard plan, 1 3 .bxaS �xaS 1 4.'�'b3 , trying to hit on Black's potential b6 and c6 weakness, does not work that well here after 1 4 . . . �c8 ! (im­ proving the bishop) 1 5 .�d3 (if I S .1dab l ? , then I S . . . �fS 1 6 .�b2 ttJd7) 15 ... �e6 1 6 .a4 cS ! (the time is right for concrete counterplay) 1 7 .�bS ttJa6 1 8. brad 1 (trying to force . . . cS -c4 ) 1 8 . . . c4 1 9 .'ifb l ttJc7 20.ttJeS ttJxbS 2 1 .axbS �xeS ! 2 2 .dxeS 'iVa8 and thanks to his ed c-pawn and imminent con­ trol of the a-file ( . . . �a3 , . . .'ti'aS ) , Black had good counterplay in Sokolov-Lutz, Garmisch Partenkirchen 1 994; B3 2) Should White decide to main­ tain his pawn pressure with 1 3 .a3 , then Black gets good play with 1 3 . . J::re8 1 4.'iVb3 bS ! with . . .ttJd7 -b6-c4 to fol­ low. This is one of the standard plans in these positions, worth ing ; B 3 3 ) 1 3 .bS cS 1 4.ttJeS (this knight jump is the essence of the plan started 1 74

with 1 3 . bS. White wants to disturb the harmonious development of Black's pieces) 1 4 .. .'iYc7 ( 1 4 . . . ttJd7 ? is a blun­ der due to l S .ttJc6 , while 1 4 . . . �xeS l S .dxeS d4 1 6 .exd4 cxd4 is a logical plan. Black has opened up his b7 bishop, while White has a grip on the weak b6 pawn, for instance : 1 7 .ttJa4 'if g S ! 1 8 .�g4 'iYxeS 1 9 .ttJxb6 �a7 and Black has dynamic play, holding the balance) 1 S .ttJg4 (the point behind White's plan)

terrible endgame) 2 0 . ttJ e 2 ! �cS 2 1 .ttJf4 'ife 7 2 2 .'iYeS and Black's weak dS pawn will fall soon; B3 3 2 ) I S . . . ttJd7 ! 1 6 .�f3 ldad8 ! (dy­ namic defence. 1 6 . . . cxd4 is not good due to 1 7 .ttJxf6 + ttJxf6 1 8.'iVxd4 with a typi­ cal isolated pawn advantage) 1 7. ttJxdS �xdS 1 8 .�xdS cxd4 1 9 . 1dc l (or 1 9 .exd4 'iYf4 2 0 J�c l 'iYxd4 2 1 .'ifxd4 �xd4 2 2 .1dfd l �cS 2 3 .�c6 ttJf6 with a drawn endgame) 1 9 . . . 'ifd6 2 0.e4 ttJcS and Black has strong counterplay, for ex­ ample: 2 1 .1dc4 d3 2 2 .f4 �fe8 . 11.

a n a lYSis d i agram

B3 3 1 ) l S . . . �e7 ?! is no good due to 1 6 .�f3 bId8 1 7 .dxcS ! 'ti'xcs 1 8 .�c 1 'ifd6

...

'iYd8-e7

This move prevents the standard b2 -b4 advance and is a logical option here. The drawback of the move is that, should White successfully execute �e 1 and e3 -e4, the black queen can prove to be unfortunately placed on the e-file. The immediate 1 1 . . . cS has also been tried and is fully playable. Play may then continue 1 2 . dxcS bxcS ! (this direct re­ capture offers active counterplay and is better than the relatively ive 1 2 . . . �xc3 1 3 .bxc3 bxcS 1 4 .�b l 'ifc7 l S .ttJeS ! ttJd7 1 6 .ttJxd7 'iYxd7 1 7 .'iYb3 �c6 1 8 .'iYa3 'iYe 7 1 9 .�f3 and due to the weakness of the cS and dS pawns Black had to defend an inferior position in Kramnik-Khalifman, Linares 2 0 0 0 ) 1 3 .'iYb3 �c6 1 4. �fd l

1 4 . . . ttJd7 ! (to obtain active counterplay it is essential for Black to keep his f6 bishop. A classic mistake would have been 1 4 . . . �xc3 ? l S . 'ifxc3 ttJd7 due to 1 6 .b4! c4 1 7 .bS �b 7 1 8 .a4 and White has a typical positional advantage) 1 5 .�b S ! ? (an inventive attempt. l S . ttJxdS brings nothing due to I S . . J:lb8) l s . . . ldb8 1 6 .�xc6 (the point of White's previous move) 1 6 .. J 'lxb3 1 7 .axb3 �xc3 1 8 .bxc3 ttJf6 1 9 .c4 d4 2 0 .exd4 cxd4 2 1 .l:Ixd4 'ifb6 2 2 .�bS 'if cS with approximate equality in Piket-Yegiazarian, Ohrid Ech 2 0 0 1 . 1 1 . . . �e8 is in itself a logical developing move, but it also has its drawbacks, as is aptly shown in the follOWing game : 1 2 . b4 (as we have seen, this is a stan­ dard plan for White. Black will be late with his counterplay now) 1 2 . . . c6 1 3 . 'ifb3 'iYd6 1 4. !:Ife I ! ttJd7 1 5 .�f1 (a timely executed e3 -e4 is the key to White's advantage)

ana lysis d iagram

analysis d i agram

1 9 .'iYd4 ! (a positional move, relying on tactical motifs) 1 9 . . . 'if e6 ( 1 9 . . . hS 2 0 .ttJeS �f6 fails to 2 1 .ttJc4! dxc4 (2 1 . . .'iYe 7 ? 2 2 .'iYxb6) 2 2 .'ti'xd6 �xd6 2 3 .�xb 7 �a7 24. ttJe4! �d8 2 S .�c6 ttJxc6 2 6 .bxc6 bS 2 7 .a4 and Black has a

a n a lysis d i agram

l S . . . �e7 1 6 . 11ab l as 1 7 .bxaS IlxaS 1 8 .a4 �a6 1 9 .�xa6 �xa6 2 0 .e4! dxe4 2 1 . ttJxe4 'if g 6 2 2 . �e3 - this is a type of position that is important to remem­ ber. Here, Black's b6 and c6 pawns are per definition weaker than White's a4 and d4 pawns, so Black is sentenced to a long, unpleasant defence, as was seen in Karpov-Bonsch , Baden-Baden 1 9 9 2 . 1 75

Chapter

W i n n i n g C h es s M i d d l e g a m es

1 2.

'ifd 1 -b3

13.

l:.f1 -d1

�f8-d8

Another plan here is 1 3 J �ad 1 , waiting for Black to play . . . c7 -c6 , and then putt­ ing another rook on e l and trying to execute e3 -e4. Now: A) In the case of 1 3 . . . c6 (White, as said, focuses on the e3 -e4 break) 1 4.l:Ife l <1:Ja6 ? ( 1 4 . . . <1:Jd7 is a better move) 1 5 .e4! dxe4 1 6 .<1:Jxe4 <1:Jc7 1 7 .�c4 <1:Jd5 1 8 .'iYc2 White's initiative is very unpleasant : 1 8 . . . g6 1 9 .h4 h5 2 0 .a3 �g 7 2 1 .<1:Jeg5 'ifd6

active counterplay related to the dynam­ ics of the . . . d5 -d4 pawn thrust) 1 8 . . . 'iVe6 1 9 .Iid l ifb6 2 0 .<1:Je l ! (plan­ ning �g4 in order to attack the black knight on d7 that defends the c5 pawn)

also possible. In general it is very im­ portant for Black here to force b 2 -b4, since Black's strategic idea is to transfer his knight to c4, and White would not be able to prevent this with b2-b3) 2 1 . b4 <1:Jc7 (it is time for the knight to head for c4) 2 2 .<1:Jc3 g6 (Black has to be careful with the execution, since 2 2 . . . <1:Je8 ? would run into 2 3 .e4) 2 3 . �e l �g 7 2 4 . �d3 (preparing e3 -e4) 24 . . . �g4! 2 5 .<1:Jd2 <1:Je8 2 6.h3 (if 2 6 .e4? ! , then 26 . . . 'iVg5) 26 . . . �e6 2 7 .<1:Je2 <1:Jd6 2 8 .<1:Jb3 <1:Jc4 - Black has achieved his strategic objectives and has an excellent game. 1 4.

a n a lysis d i agram

analysis d i agram

2 2 . <1:Jxf7 ! (a standard tactical motif) 2 2 . . stt xf7 2 3 .<1:Jg5 + and with 24Jle6 to follow, Black resigned in Podgaets­ Gusev, Odessa 1 99 1 . B) 1 3 . . . c5 is probably the best move. Black should not be afraid of the cre­ ation of hanging pawns in the centre and should look for active play: 1 4.dxc5 �xc3 1 5 .'iVxc3 bxc5 - a standard struc­ ture with parallel hanging pawns in the centre has been created. There is a dy­ narnic balance; however, things can go wrong quickly, which the follOwing game illustrates very well : 1 6 Jk 1 <1:Jd7 1 7 Jic2 ld:ab8 1 8 . b3 (a strategically tense position has been created. White is try­ ing to target Black's hanging pawns in the centre, while Black has to organize 1 76

2 0 . . J ibc8 ( 2 0 . . .' ti'g 6 , preventing �g4, would have given Black good play) 2 1 .�g4 'iYg6 ? (now this move is wrong. Better was 2 1 . . .1::1 c7) 2 2 .�h3 l:Ic7 2 3 .<1:Jd3 ! <1:Jf6 24 .�a5 ! (this possi­ bility was obviously overlooked by Black when he played 2 1 . . . 'if g6?) 24 . . . <1:Je8 2 S .l:Ixc5 ld:xc5 2 6 .'iYxc5 and in this hopeless position Black reSigned in the 5th Candidates' match game Kortchnoi-Geller, Moscow 1 97 1 . 1 3.

...

1 4.

�e2 -f1

...

ttJ b8-a6

Black correctly first develops his knight, keeping his bishop on b 7 for the time being. He may also decide to improve his light-squared bishop with 1 4 . . . �c8 1 5 . g3 �g4 . The drawback of this plan is that his c6 pawn will be easier to tar­ get. The following game is a good illus­ tration: 1 6 .�g2 <1:Jd7 1 7 Jhc 1 �ac8

2:

I s o l a t e d P a wn s

3 0 .l:Ic l (Black is lost here) 30 . . . g5 3 1 .<1:Jh5 <1:Je6 3 2 .�xe4 dxe4 3 3 .�g2 'iYa4 34J:tc8+ <1:Jf8 3 5 .ld:xf8 + ! �xf8 3 6 .<1:Jf6+ �g 7 3 7 .'ifxa4 bxa4 3 8 .<1:Jxd7 axb3 3 9 .axb3 �d6 40 .g4 �g6 4 1 .f3 exf3 + 42.�xf3 f5 43 .e4 fxe4+ 44.�xe4 �f7 45 .�d5 �g3 46.'litc6 rbe7 47 .d5 1 -0, Piket-Beliavsky, Bugojno 1 99 9 . 1 5.

l:!:d1 -d2

ttJa6-c7

1 6.

a2-a4

ttJc7-e6

1 7.

a4-a5

b6-b5!

1 8.

'iVb3-a2

1 8 .a6 �c8 is risky for White, given the fact that he will have to defend the a6 pawn later on, but probably it had to be played. Black now quickly develops a strong initiative. 1 8.

...

1 9.

l:ta1 -c1 ?

a 7-a61

White does not sense the danger. 1 9 J �ad I , strengthening the control of the d4-square, had to be played.

c7-c6

Targeting the black queenside with 1 4.a4 <1:Ja6 1 5 .a5 does not bring any­ thing after the correct 1 5 . . . b5 ! . The game P. Nikolic-Beliavsky, Reykj avik World Cup 1 9 9 1 , instead continued (better was 1 6 . 'if c 2 1 5 . . . 'iYb4 1 6 .'ifxb4 ! <1:Jxb4 1 7 .<1:Ja4, with advan­ tage) 1 6 . . . 'ife 7 1 7 .<1:Ja4 ? (forcing Black to make a good decision) 1 7 . . . b5 1 8 .<1:Jc3 �c8 ! (improving the bishop. 1 8 . . . <1:Jc 7 ? ! is wrong due to 1 9 .e4! b4 2 0 .e5 and White is better) 1 9 .<1:Ja2 �d7 2 0 .lIab l �dc8 (2 0 . . . .tIac8 was

a n a lysis d i agram

1 8 .<1:Je2 ! (starting to put pressure on c6) 1 8 . . . <1:Jf8 1 9 .h3 �f5 2 0 .<1:Jf4 g6 2 1 .'ifa4! itg 7 2 2 .b4 (ready for 2 3 .llxc6 �d7 24.b5) 22 . . . b5 2 3 .'iYa6 'iVxb4 24.�xc6 (now the d5 pawn is very weak) 24 .. J�xc6 2 5 .'ifxc6 'ifa4 2 6 Jh l ifc4 27 .'iVa6 lid7 28.<1:Jd2 'iYb4 29 .<1:Jb3 �e4

1 9.

...

c6-c5!

20.

d4xc5

d5-d4!

Now the bishop on b 7 becomes a monster. 21 .

ttJf3xd4

�f6xd4

22.

e3xd4

ttJe6xd4

23.

\!(g 1 -h 1

N ow Black finishes off the attack very energetically, not giving White any time to breathe. 177

W i n n i n g C h ess M i d d l eg a m e s

Chapter 3 Paral lel H a ngi ng Pawns in the Centre

23.

...

ttJd4-f3!

Excellent execution by the English grandmaster. 24.

I:!d 2xd8+

24.�dd l loses to 24 . . .'ifgS ! (threaten­ ing . . . ttJd2 or . . . 'iYf4) 2 S .ttJe2 'iYh4 ! 2 6 .h3 llxd l 2 7 .�xd l 'iYxf2 2 8 .'iVa3 ttJe 1 ! . 24 .

...

�a8xd8

25.

c5-c6

iLb7xc6

26.

ttJc3-e2

26.

...

'iYe7-h4!!

27.

g2xf3

'ii' h4xf2

28.

ttJe2-f4

iLc6xf3+

Also winning was 2 8 . . . 'iYe3 2 9 . l:hc6 'iYxf3 + 3 0 . Wg l IId l 3 1. .�c8 + Wh7 . 29.

iLf1 -g2

30.

l:[c1 -g 1

ldd8-d2

White also loses after 3 0 .l:Ic8 + �h7 3 1 .'iYb l + g6 3 2 .'iYg l l:ld l . 30.

...

iLf3-e4!

A nice final touch. White resigned.

I ntrod uction In this chapter I have kept the concept similar to the chapters on doubled pawns and isolated pawns, selecting diagrams that show pawn structures (and games) related to the most principled plans for both sides : with parallel hanging pawns and fighting against them. Hanging pawns in the centre are by definition connected; for two hanging pawns on the same rank we will use the term 'parallel' in this book. They normally arise from various 1 .d4 openings, like the Tarrasch Defence, the Tartakower Variation of the Queen's Gambit Declined, the Queen's Indian Defence, the Meran Variation, etc. In four of our five commented games , the Tartakower Variation of the Queen's Gam­ bit is played. For a club player willing to learn the strategic pros and cons of these po­ sitions and, subsequently, to either employ this variation with black or to learn better how to tackle it with white in practical tournament games , it will certainly pay off to spend time on a thorough analysis of these games and comments. Structure 3. 1 & Structure 3 . 2 (Game 3 6 Fischer-Sky) . This World Championship classic re­ mains an excellent learning example. After the opening, one of the most typical positions with hanging pawns in the centre arises. Fischer then executes a standard plan, executed by White on a number of other occasions in dif­ ferent openings - one of the main lines of the Tarrasch Defence in particular: first he takes the black bishop on e6 with ttJd4xe 6 , forcing . . . f7xe6 , and then attacking the black centre with a timely e3 -e4 pawn thrust, forcing Black to make a positional concession. Spas sky qUickly collapses under the pressure. The comments to this game also provide a good opening gUideline for a club player in this Tartakower Variation of the Queen's Gambit.

3.1

In Game 3 7 (Sokolov-Khalifman) we see White building up pressure against Black's hanging pawns in the centre, to the point where Black is forced to look for tactical solu­ tions.

1 78

1 79

Chapter

W i n n i n g C h es s M i d d l e g a m es

Structure 3 . 3 (Game 3 8 - Sokolov-Short) . In this game White executes the standard e3 -e4 plan, hitting Black's hanging pawns in the centre, with the difference com­ pared to the Fischer-Sky game that there has been no piece exchange on e 6 , so the black pawn is still on f7 . An interesting strategic battle ensues, showing how easily things can go wrong in such complicated positions. This game and its comments are a good learning example of this Tartakower line, as well as of the strategic pros and cons of the position.

P a ra l l e l H a n g i n g P a wns i n t h e C e n t re

Structure 3 . 1

3.3

Game 3 9 - Kramnik-Yusupov is an interesting tussle, where both sides demonstrate an abundance of tactical possibilities such as are often present in these dynamic positions. Again , besides improving his understanding of these positions in general, the club player will find some useful opening information about this Tartakower line. Structure 3.4 (Game 40 - Lautier-Short) . Excellent play by Short with the parallel hanging pawns shows the dy­ namic possibilities related to a well-executed pawn break - in this case, . . . d5 -d4 . A standard plan for developing an initiative for the side with the hanging pawns. Again, apart from strategic objectives , useful opening explana­ tions have been provided.

3:

Structure 3 . 2

3.4

We will start this chapter o n parallel hanging pawns i n the centre with one of the classic examples. Perhaps I have gone too far giving and explaining the many differ­ ent lines and ideas in this famous game, but I have done this with a reason. This par­ ticular opening position, the plans related to it and the middlegame arising from it, are classic and occur in many of the positions with hanging pawns in the centre. Therefore, I would advise the reader not to be lazy, but to delve deeply into the com­ ments and try to u nders tand them.

GAME 36 QO 6 . 6 (D5 9) Robert James Fischer Boris Sky Reykjavik World Championship 1 9 7 2 (6)

1 80

3.

d2-d4

tb g8-f6

4.

tbb1 -c3

�f8-e7

S.

�c1 -gS

0-0

1.

c2-c4

e7-e6

6.

e2-e3

h 7-h6

2.

tb g 1 -f3

d 7-dS

7.

�gS-h4

b 7-b6

181

Chapter

W i n n i n g C h ess M i d d l eg a m es

8.

c4xd5

11.

One of the main 'old' lines of the Tartakower/Makogonov Variation. Mod­ ern theory focuses more on 8 .jLd3 or 8.jLxf6 . Both these moves have been in­ vestigated in the Isolated Pawn chapter see Vyzhmanavin-Beliavsky (Game 2 8 ) and Illescas Cordoba-Short (Game 3 5 ) . 8.

...

ttJf6xd5

It is considered to be a good idea for Black to exchange a few pieces here in order to have more space to manoeuvre. White has better chances in the posi­ tions arising after 8 . . . exd5 9 . jLd3 jLb 7 1 0 . 0-0 CLJbd7 1 1 .'iVc2 c5 1 2 .l:tad l , with CLJf3 -e5 and jLd3 -f5 to follow. 9.

�h4xe7

'iYd8xe7

1 0.

ttJc3xd5

e6xd5

11.

1da1 -c1

The strategic situation is rather clear. White has agreed to exchange a few pieces in order to create a potential weakness : Black's c-pawn. Black, on the other hand, will play . . . c7 -c5 and then , after a pawn swap on c5 (dxc5 bxc5) , a position would arise with two parallel hanging pawns in the centre. Black is harmoniously developed and should have a good game in general. White, on his part, will use the queen manoeuvre iVa4-a3 in order to pin the black queen along the a3 -f8 diagonal - on e7 -, try­ ing to create some problems for Black. 1 82

. ..

�c8-e6

As often after a pawn swap on d5 , Black keeps his bishop on the h3 -c8 diagonal, since from b7 it would hit its own pawn on d5 . This general logic is questionable here, since after . . . c7 -c5 White is likely to take on c5 and then, if Black has two parallel hanging pawns in the centre (c5 and d5) , it is quite useful to have the bishop placed on the long diagonal, sup­ porting various tactics related to the . . . d5 -d4 break. Anyhow, Black has dy­ namic play after 1 1 . . . jLb7 and White has not shown a way to an advantage after this continuation. It is fully playable and I will give a few examples: A) With 1 2 .iVa4 White follows a stan­ dard plan, like in our main game; rely­ ing on the pin along the a3 -f8 diagonal. 1 2 . . . c5 1 3 .'iVa3 ldc8 . Now, 1 4.jLd3 , with the idea to pin a black knight on d7 with jLf5 , is well answered with:

ana lysis d i agram

1 4 . . . CLJc6 ! , and if 1 5 .dxc5 ( 1 5 . 0-0 c4! 1 6 .iVxe 7 CLJxe 7 1 7 .jLb 1 b5 is better for Black) , then 1 5 . . . d4 ! - the reader is ad­ vised to these classic motifs. After 1 4.jLe 2 : A I ) A very interesting plan is 14 . . . a5 ! ? 1 5 . 0 -0 CLJa6 1 6 .jLxa6 l:txa6 and Black has enough dynamic possi­ bilities to compensate for his weak c5 pawn, for example : A l l ) 1 7 .a.fd 1 iV e4! and Black has an excellent game, since if White should decide to take a pawn with 1 8 .dxc5 ? , then Black's b 7 bishop becomes a mon­ ster : 1 8 . . . bxc5 1 9 .�xc5 l:txc5 2 0 .'iVxc5 a.g 6 ! 2 1 .CLJe l d4 ; A 1 2) 1 7 .h3

3:

P a r a l l e l H a n g i ng P a wns in t h e C e n t re

A2 1 ) Another move, played by Sky six years earlier, is 1 5 . . . iV f8 (a standard way to solve the queen pin) 1 6 . dxc5 bxc5 1 7 . :ac2 'iVd6 ! (transferring the queen to b 6 , where it will be well placed to the . . . d 5 -d4 break) 1 8 . :a fc l 'ifb6 1 9 . CLJd2 ! (transferring the knight to b3 to increase the pressure on the c5 pawn) 1 9 . . . l:te8 (Black understands that the c5 pawn cannot be adequately defended , so he prepares a tactical so­ lution) 2 0 .CLJb3 . Black now has to go for a pawn break he has planned for the past few moves, which involves a pawn sacrifice :

analysis d iagram analysis d i agram

1 7 . . . iV e4! 1 8 .dxc5 bxc5 1 9 .:axc5 l:txc5 2 0 .'lWxc5 l:tg6 2 1 .�h l d4 2 2 .'ifxd4 �xf3 2 3 . gxf3 jLxf3 + 24.�h2 l:tg2 + with a draw by perpetual check in Yermolinsky-Lputian , Telavi 1 9 8 2 . A2) 1 4 . . . CLJd7 1 5 . 0-0 CLJf8 (the typi­ cal knight transfer to e 6 , already seen in a few earlier examples , works here as well) 1 6 . l:tfd l l:tc7 (threatening c5 -c4) 1 7 .dxc5 bxc5 1 8 .jLa6 jLxa6 1 9 .'iVxa6 :ad8 2 0 . iVd3 CLJe6 2 1 .iVf5 d4 and Black has achieved his strategic objec­ tives , and he had a good game in Flohr-Vidmar, Groningen 1 946 .

2 0 . . . d4 ! 2 1 . CLJxc5 CLJxc5 2 2 . ldxc5 dxe 3 2 3 . fxe3 l:te 7 . Black seems to have enough compensation to hold the balance. Therefore, the idea with 1 5 . . . CLJf8 seems more sound to me. For example : 2 4 . :ab 5 (White decides to simplify into a drawn endgame. Should he decide to hang on to his pawn with 2 4 . b4 , then , due to his strong bishop on b7 and White 's weak e3 pawn , Black gets good compensa­ tion after 2 4 . . . l:t d 8 ) 24 . . . 'iVxe 3 + 2 5 . iVxe 3 l:txe 3 2 6 . ldxb 7 l:txe2 2 7 . l:tcc7 l:td8 2 8 J �d7 (a terrible mis­ take would have been 2 8 . �xf7 ? ? due 1 83

Chapter

W i n n i n g C h es s M i d d l eg am e s

1 5 . . . 'iVa7 ! (the point behind 1 2 ... a5) 1 6 .�b5 tLla6 1 7 .0-0. Now Black dis­ plays an excellent feeling for the position and plays a beautiful series of moves: 1 7 . . . tLlc7 ! 1 8 .�c6 c4 ! 1 9 .b4 'iVa6 ! 2 0 .bxa5 tLlb5 2 1 .�xb5 �xb5 2 2 .tLle5 �ha5 2 3 .'iVd6 l:xa2 and Black was win­ ning in Furman-Zaitsev, Tallinn 1 99 1 .

to 2 8 . . J ld l + 2 9 J ifl I:tdd 2 ) 28 .. J�c8 2 9 . � b c 7 draw, Petro sian-Spas sky, Santa Monica 1 9 6 6 . Other moves do not pose any problems for Black, for example : B) 1 2 .�e2 c5 1 3 .dxc5 bxc5 1 4.0 -0 �c8 ! (a mistake is 1 4 . . . tLld7 ? , because White has the thematic 1 5 .b4! c4 1 6 .tLld4 with a clear advantage) 1 5 .'iVa4 tLlc6 1 6 .�d3

an alysis d i agram

1 7 . . . �a6 ! (now the white king is for­ ever confined to the middle of the board, and this for the price of one lousy pawn) 1 8 .tLld4 tLle4 1 9 .c6 'iVb4+ 2 0 . �d 1 �d3 ! (a simple motif, surely missed by White) 2 1 .'iYxd3 tLlxf2 + and Black soon won in Timman-Rukavina, Sombor Parcetic Memorial 1 9 7 2 . 1 2. analysis d i agram

1 6 . . . tLld8 ! (a nice plan, worth remem­ bering. On e6 the knight will be well placed) 1 7 .'ufe l tLle6 1 8 .e4 (trying to provoke d5 -d4 , but this comes one tempo too late) 1 8 . . . c4! 1 9 .�b l tLlc5 2 0 . 'iVa3 dxe4 Szabo-Pirc, Saltsjobaden Interzonal 1 94 8 ; C ) 1 2 .�d3 c5 and now: C l ) In case of 1 3 .dxc5 bxc5 1 4. 0- 0 llc8 1 5 . :te l tLld7 1 6 .'iVa4 tLlf8 ! , Black again transfers his knight to e 6 , with approximate equality; C2) In case of 1 3 . 0-0 Black gets a good game by taking space on the queenside with 1 3 . . . c4; C 3 ) The 'ambitious' 1 3 .�b l ? ! tLld7 1 4.'iVc2 tLlf6 1 5 .dxc5 ? ! �fc8 1 6 .b4? (snatching a pawn and defending it with the king still in the middle of the board is not a good idea) 1 6 . . . bxc5 1 7 .bxc5 1 84

'iYd 1 -a4

c7-c5

A rather viable alternative is 1 2 . . . a5 ! ? The idea i s to vacate the a7 -square for the black queen, so that Black can effi­ ciently solve the problem of his pawn being pinned on c5 , and at the same time defend this pawn. I like this plan very much. One of the stem games con­ tinued 1 3 .rIc3 (or 1 3 .�b5 �c8 1 4.0 -0 'iYb4 1 5 . a3 (if 1 5 . tLl e 5 , 1 5 . . . c 5 ) 1 5 . . . 'iYxa4 1 6 .�xa4 c 5 1 7 Jic3 c4 1 8 .b3 �d7 ! with a good game for Black) 1 3 . . . �c8 1 4.�d3 c5 1 5 .�a3

analysis d i agram

1 3.

'iYa4-a3

1 4.

�f1 -b5

3:

P a ra l l e l H a n g i ng P a wns in t h e C e n t r e

1 9 . . . c4 ! (good judgement. Black sur­ renders control of the d4-square in or­ der to enter with his rook on the second rank) 2 0 . 11xb8 + �xb8 2 1 .'iVxa5 1:txb2 and Black was slightly better in Winants-Kasparov, Brussels 1 9 8 7 .

l::tfS-cS

The idea is to hit the black knight once it is developed to d7 , and so to under­ mine the defence of the black c5 pawn. Another frequently played move here is 1 4 .�e 2 . Now Black gets a comfortable game with 1 4 . . . a5 . He chooses the same plan as in Furman-Zaitsev (see above after 1 2 . . . a5 ! ?) : 1 5 . 0-0 'iVa7 ! . The queen is no longer pinned, so . . . c5-c4 is a positional threat White has to reckon with. 1 6 .b3 tLld7 1 7 .�b5 �c7 1 8 .1:Ifd l l::t a c8 and Black had noth­ ing to complain about in Karpov­ Kavalek, Linares 1 9 8 1 . 1 4 . . . �f8 is also a good option to solve the problem of the pin on the queen, which was played by Kasparov on one occasion: 1 5 .dxc5 bxc5 1 6 . 0-0 as ! (this move is almost invariably useful for Black in these positions) 1 7 . �c3 tLld7 1 8 J :Ifc l .J:Icb8 1 9 .�b3

1 4.

...

a 7-a6

Here Black has a number of possibilities to achieve a satisfactory game: A) 14 . . . �f8 1 5 . dxc5

a n alysis d i agram

analysis d i agram

1 5 .. ,llxc5 ! (a correct decision. White has the advantage in the case of 1 5 . . . bxc5 1 6 . 0- 0 and now, if 1 6 . . . �g4?, then the simple 1 7 . l:lfd l will do) 1 6 . 1:txc5 'iYxc5 ! (exchanging the queens is OK for Black here, since his king is close to the fray and will find a natural post on the d6-square) 1 7 .'iVxc5 + ( 1 7 . �d2 is equal after 1 7 . . . 'iVxa3 1 8 . bxa3 tLld7 1 9 .�xd 7 1 85

Chapter

W i nn i n g C h es s M i d d l eg a m es

�xd7) 1 7 . . . bxcS 1 8 .<;t>d2 <;t>e7 1 9 .blc l <;t>d6 with equality; B) The most ambitious, and probably the best, for Black is 1 4 . . . �b 7 ! , since after I S .dxcS bxcS 1 6 . �xcS blxcs 1 7 .'ifxcS . . .

a n a lysis d i agram

a n alysis d i agram

. . . Black has a strong response in 1 7 . . . ttJa6 ! (not good is 1 7 . . . a6 , because the position arising after 1 8 .�d3 'iYxb2 1 9 .0 -0 ttJd7 ( 1 9 . . . 'iYxa2 ? is no good, since Black cannot finish his queenside development after 2 0 . ttJd4) 2 0 .'iYc6 �b8 2 1 .ttJd4 'iVb6 2 2 .ldc l is clearly better for White) 1 8 .�xa6 (in the case of 1 8 .'ifc6 'iYxc6 1 9 .�xc6, Black is better after 1 9 . . . blb8 ! (the immediate 1 9 . . . blc8 allows White to defend with 2 0 .�a4 ! ) 2 0 .0 -0 (2 0. b3 ? ? is now a terrible blunder, losing to 2 0 . . . blc8) 2 0 . . . blxb2 2 1 .a4 ttJb4 2 2 .ttJd4 bla2) 1 8 . . . 'iYxa6 . The white king stays in the middle and Black's compensation is very strong. The following game illustrates the trou­ ble White has gotten himself into : 1 9 .iVa3 'iYc4 2 0 .<;t>d2 iVg4 2 1 .blg l . Although the white king is in the mid­ dle, it seems that, with the relatively limited number of forces left on the board, he will be able to survive the as­ sault. But. . . 186

2 1 . .. d4 ! (opening the position further) 2 2 .ttJxd4 (if 2 2 .exd4, then 2 2 . . . �dS ) 2 2 . . . 'lWh4 2 3 .ble l 'iYxf2 + 24.l:Ie2 'lWfl 2 S .ttJxe 6 fxe6 2 6 .'lWd6 <;t>h8 (even though there are only heavy pieces left, the black attack is still very strong) 2 7 .e4 IIc8 2 8 .<;t>e3 �f8 2 9 .�d2 ? (a crucial mistake. 2 9 . �d2 had to be conSidered)

ana lysis d i agram

29 . . . eS ! (after this, White finds himself in a mating net) 3 0 .'iYxe S 'ife l + 3 1. .tIe2 'iVg l + 3 2 .<;t>d3 �d8 + 3 3 .<;t>c3 'iVd l 34.'lWbS 'lWd4+ 3 S .<;t>c2 a6 ! . This little move throws White off balance. 3 6 :iVxa6 'iVcS + 0- 1 , Timman-Geller, Hilversum 1 9 7 3 . Note that this game was played shortly after the 1 9 7 2 World Championship match, where Geller was one of Sky's seconds , so most likely,

1 4 . . :iVb 7 ! had been found in Reykjavik. It would be interesting to know whether Fischer had analysed 1 4 . . . 'li'b 7 and if so, what he had planned to do about it. C) The standard 1 4 . . . aS ! ? plan might even work with one tempo less : 1 5 . 0-0 and now: C 1 ) I S . . . ttJa6 does not equalize after 1 6 .ttJeS ttJb4 1 7 .'lWa4 cxd4 (or 1 7 . . . f6 1 8 .ttJg6 'iYf7 1 9 .ttJf4) 1 8 .exd4 blxc l (if 1 8 .. .f6, then White is better after 1 9 .ttJg6 'lWd6 2 0 .a3 �fS 2 1 .blxc8 + blxc8 2 2 .axb4 �xg6 2 3 .bxaS bxaS 24.'iYxaS blc2 2 S .'iYa3 ) 1 9 . ,gxc l 'iVgS 2 0 . �c 7 'iYd2 2 1 .a3 ;!; ttJa6 2 2 .�xa6 �xa6 2 3 . 'iVe 8 + <;t>h7 24.'iVbS �a8 2 S .'iVd3 + 'iYxd3 2 6 .ttJxd3 with some advantage for White. As often seen in the Isolated Pawn chapter, White has the favourable upper hand-position with symmetry in the centre; C2) I S . . . ttJd7 ? ! is not in the spirit of the position, since White gets a nice blockading post on d4 for his knight af­ ter 1 6 .dxcS ttJxcS ( 1 6 . . . bxcS ? simply loses a pawn after 1 7 .ttJd4 ! a4 1 8 .ttJxe6 fxe 6 1 9 . �xd7 'iVxd7 2 0 . nxc S ) 1 7 .ttJd4 ; C3) I S . . . 'iYa7 ! 1 6 .�c3 �d7 ! 1 7 .�e2 c4 with complicated play. 1 5.

d4xc5

b6xc5

Black could also have taken with the rook, though with correct play White seems to have some edge: 1 S . . . �XCS and now: A) 1 6 .�xcS bxcS (for the sake of achieving dynamic play it is in principle always advisable for Black to keep queens on the board. However, Black holds the endgame after 1 6 . . . 'ifxcs 1 7 .'iVxcS bxcS 1 8 . �a4 �d7 ! . 1 8 . . . ttJd7 , on the other hand, is slightly better for White after 1 9 .�xd7 ! �xd7

3:

P a r a l l el H ang i n g P a wns in t h e C e n t re

2 0 .�d2) 1 7 .0-0 �g4 1 8 .�e2 ttJd7 with an approximately equal game; B) 1 6 .0 -0 'lWc7 (the point behind 1 5 . . . �xcS . White is forced to take on cS and the black b6 pawn will end up on cS anyway. 1 6 . . . �g4? is not good, since after 1 7. ttJd4 'if c7 White has 1 8 . ttJ b3 ! ) 1 7 .�xcS 'iYxcs (wrong is 1 7 . . . bxcS ? ! , since after 1 8 .�c 1 'lWb6 1 9 .�a4 c4 2 0 .ttJd4 White, by forCing . . . cS -c4 , has achieved his strategic obj ective. He has a strong, dominant knight on d4, while Black no longer has any tactics associ­ ated with the . . . dS -d4 break) and now: B 1 ) 1 8 .'iVxcS bxcS 1 9 .�a4 does not promise anything, since Black has an excellent game after 1 9 . . . �d7 ! . Note how many times this 'little' move solves Black's queenside development prob­ lems. Hence, it is handy for Black to keep the knight on the board: 2 0 .�b3 c4 (here this move is justified. Black surrenders the d4-square, but gets strong pressure on White's b2 pawn) 2 1 .�c2 ttJc6 2 2 .�a4 ttJeS ! 2 3 .�xd 7 ttJxd 7 2 4 . b 3 (or 2 4 . �d l ? �b8 ! ) 2 4 . . . 1db8 ; B 2 ) 1 8 .�e2 'iYxa3 1 9 .bxa3 ttJd7 2 0 .ttJd4 ttJcS 2 1 . �b l ! bS 2 2 . �c l nc8 2 3 .ttJxe 6 fxe6 24.a4! bxa4 2 S .�xa6 1da8 2 6 .�bS ttJe4

a n a lysis d i agram

1 87

Chapter

W i n n i n g C h ess M i d d l e g a m es

2 7 .�d3 ! cJtfl 2 8 .�xe4 dxe4 2 9 .g4! a3 3 0 J::t c 3 and White is better in the end­ game, since Black's rook is doomed to ivity. 1 6.

0-0

But there is a tactical justification, since after 1 7 .�xc6 l:txc6 1 8 . ttJeS �c7 1 9 .ttJd3

na8-a7

This is one of the logical moves here; however, moving the queen to a better square looks to me to be perhaps more appropriate : A) 1 6 . . . 'iYb7 promised Black a good game after 1 7 .�a4 'ifb6 1 8 .ttJeS

ttJc6 ; while in the case of 2 0 .ttJd4 Black uses the aforementioned equalizing mechanism with 20 . . . ttJc6 2 1 .ttJxe6 fxe6 2 2 .bxc4 ttJeS) 20 . . . dxc4 2 1 .!:lc3 l:tb7 . Black tries to get active in order to com­ pensate for his c4 weakness.

an a lysis d i agram

analysis d i agram

1 8 . . . aS ! . As already mentioned, this move is always useful for Black in those positions. Black has a comfortable game and violent attempts like 1 9 .f4? would only get White in trouble after 1 9 . . J�a7 ! 2 0 .[S �d7 2 1 .f6 'ifb4 ! . I f 1 9 . ttJd3 , then - thanks t o 1 8 . . . as ! 1 9 . . . ttJa6 . B) Another good standard plan here (as already shown in a number of previ­ ous comments) is 1 6 . . :iYa7 , and Black is doing OK after 1 7 .�a4 as ! . C ) Another interesting option was 1 6 . . . ttJc6 ! ? According to general princi­ ples this move should be wrong, since: 1 . Black should try to keep the knight on the board, and : 2 . Black should also avoid the . . . cS -c4 push, which would surrender the d4-square. 1 88

1 9 . . . c4! the weakness of White's b2 pawn, which will be hit by Black's rooks, more than compensates for the general inferiority of Black's bishop versus White's knight here. It is impor­ tant to note that with rooks on the board, the white b2 pawn weakness here prOVides Black with an excellent game, while without rooks on the board Black would have been positionally lost 0 ) . 1 7.

�b5-e2

tLlb8-d7

A) Black could also opt for the earlier explained strategy of surrendering the d4-square for the sake of pressure on the b2 pawn with 1 7 . . . c4 ! ? 1 8 :ifxe 7 J::Ix e7 , although Black's dynamic play does not seem sufficient for equality this time, and White keeps the advan­ tage after 1 9 .b3 ! (immediate action is needed; Black has nothing to worry about in the case of 1 9 .ttJd4 ttJc6 2 0 .ttJxe6 fxe6 2 1 .b3 .tIec7 ! 2 2 .bxc4 <1:JeS) 1 9 . . . l:f.ec7 2 0 .bxc4! (eliminating the b-pawn weakness once and for all and creating a weak c-pawn for Black to worry about. Drawish is 2 0 .l:tc3 cxb3 ! 2 1 . �xc7 l:txc7 2 2 .axb3 l:Ic3 2 3 .ttJd4

3:

P a ra l l e l H a n g i n g P a wns in t h e C e n t r e

Still using the motifs based on the pin on the black queen. It is quite possible that Sky reckoned chiefly with 1 8 .b4, when after 1 8 . . . c4 Black has active play. 1 8.

...

'ife7-f8?

Black's first real mistake in the game. Defending the queen with 1 8 . . . �f8 is not good due to 1 9. b4 and, compared to the previous 1 8 .b4 comment, White has made a useful move (ttJd4) while the black king can now only become exposed on f8 . But Black had a rather satisfactory re­ sponse in 1 8 . . . ttJf6 1 9 .ttJb3 .

analysis d i agram

2 2 .nfc l ! nb2 2 3 .ttJd4 ttJd7 24.�f1 ! 11ldS 2 S .f3 ! with e4 to come, and Black is in trouble ; B ) The standard 1 7 . . . aS does not work so well now, because of the vul­ nerability of the cS pawn, for instance: 1 8 .l::t c 3 ttJd7 1 9 .1:tfc l �e8 ? ( 1 9 . . . �g4 was better) 2 0 .�b S and White was clearly better in Furman-Geller, Mos­ cow 1 9 7 0 , since 20 . . . �g4 . . .

analysis di agram

. . . fails to 2 1 . l::lx cS ! �xf3 2 2 .�c8 ! ' 1 8.

tLlf3-d41

a n a lysis d i agram

White would be better in the endgame after 1 9 . . . ttJe4 2 0 .f3 c4 2 1 .'ifxe 7 �xe 7 2 2 .ttJd4; or 1 9 . . . c4 2 0 .'iVxe 7 1::t. xe 7 2 1 .ttJd4 ; retreating the knight with 1 9 . . . ttJd7 is ive and does not solve anything after 2 0 J dc3 with �fc 1 to come. But the tactical defence 1 9 . . . l:tac7 ! of­ fers Black good prospects : A) 2 0 . �xa6 ? is a blunder which loses material after 20 . . . l;Ia8 2 l . l:txcs (or 2 1 .ttJxcS �c8 2 2 .b4 'ifd6 2 3 .e4 (23 .bS ? ttJe4) 2 3 . . . ttJd7 24.l::t fd 1 �xa6 2 s .IlxdS 'ife7 ) 2 1 . .. l:tca7 2 2 .'ifb4 ttJe4 2 3 .a3 ttJxcS 24.ttJxcS lic7 2S .ttJxe6 'ifxb4 2 6 .axb4 fxe6 2 7 .bS l:tc2 ; 1 89

C h a p t e r 3 : P a r a l l e l H a n g i n g P a wns i n t h e C en t r e

W i nn i n g C h e s s M i d d l eg a m e s

B) In the case of 2 0 .'iYxa6 ? ! Black has the simple 2 0 .. J la7 2 1 . 'ifb6 c4 (or 2 1 . . Jib 7 2 2 .'iVaS c4 2 3 .4Jd4 �xb2) 2 2 .4Jd4 l::r b 7 2 3 .'iVa6 l:Ixb 2 , in all cases with better prospects for Black; C) 20.l:lfd 1 , and now Black continues his dynamic play with 20 . . . d4! (the side with parallel hanging pawns in the centre should play energetically !) 2 1 .4JaS ! (White has to be careful not to end up worse; 2 1 .�xa6? is not good, losing ma­ terial due to 2 1 . .. �aS , and if 22.4JxcS �cS 2 3 .exd4 ha6) 2 1 . . . �dS ! 2 2 .�c4 'iYe6! 2 3 J:te 1 'iVb6 24.�xdS 4JxdS 2 S .4Jc4 'iVg6 with a dynamic position, where Black is not worse. 1 9.

tLld4xeS

20.

e3-e4 !

f7xeS

Thematic play. Black can now either al­ low the position to be opened, when White's light-squared bishop would be powerful and Black's central pawns would be targeted, or - a much worse option, chosen in the game - try to keep the position closed by responding with . . . dS -d4.

does not have any counterplay and is doomed to wait in agony for the inevi­ table to happen. Black's position is already difficult, but it could have been better defended. One way is to try and provoke White to play e4-eS with the black pawn on dS , so that Black can set up a pawn chain with e6, dS , c4, and try to constrain the white bishop : A) 2 0 . . . 4Jf6 2 1 .'iYh3 ! (forcing Black to defend the e6 pawn with a rook. If 2 1 .eS , 2 1 . . . lDd7 2 2 .f4 c4 2 3 .'iVh3 4JcS ) 2 1 . . Jlc6 2 2 .eS 4Je4 ( 2 2 . . . 4Jd7 2 3 .f4 with an attack) 2 3 .f3 ( 2 3 .'iVe3 , intending f2 -f3 , is also possible) 23 . . . 4JgS 24.'iYhS c4 2 S .h4 'ifcS + 2 6 . Wh2 4Jf7 2 7 .f4 �e3 2 s . ldc3 'ifd2 2 9 J:i�g3 with an immediate attack; B) 2 0 . . . c4? would ultimately lose ma­ terial after 2 1 . 'iYh3 �f7 (or 2 1 . . . 4JcS 2 2 .b4 cxb3 2 3 .axb3 ) 2 2 .�hS 'iYe7 (or 2 2 . . . g6? 2 3 .�g4) 2 3 .exdS exdS 24.�f3 �cS 2 S .b4! cxb3 2 6 . .t'rxcS 'iVxcs 2 7 .�e6+ �hS 2 S .�xdS 4Jf6 2 9 .�xb3 ; C) Probably the best for Black is to try and keep his stuff together with 2 0 . . . 'iVd6 ! . Now the black pieces look clumsy, but it is difficult to find any­ thing concrete for White, for example: 2 1 .exdS (or 2 1 .1dfd 1 �eS ) 2 1 . . . exdS 2 2 .l:Ifd l (or 2 2 .�g4 �c6 ! and, clumsy or not , Black holds : 2 3 J�fe 1 4JeS) 22 . . . lDf6 2 3 .b4 'iVe6 ! 24.�e 1 cxb4. 21 .

20 .

...

d5-d4?

An inexplicable mistake for a player of Sky's calibre. Did he really fear Fischer so much? White now gets a free hand to attack on the kingside, and his bishop gets open diagonals, while Black 1 90

f2-f4

In a higher sense, Black is lost now. White has achieved the dream scenario in the 'parallel hanging pawns in the cen­ tre' -structure: the timely e3 -e4 has pro­ voked the desired effect, White's light­ squared bishop has full control of the weakened a2-gS and b 1 -h7 diagonals, his pawn majority on the kingside is advanc-

ing with devastating effect, and Black has no counterplay whatsoever. Having Bobby Fischer sitting on the other side of the board does not help ei­ ther. 21 .

...

'iVfS-e7

22.

e4-e5

!IcS-bS

If 2 2 . . . 4Jb6 : A) Wrong is 2 3 .fS ? because of 2 3 . . . c4! (23 . . . exfS ? ? is a terrible blunder due to 24.'iVb3 +) 24.iVaS 'iVcS ! and Black is fine, since the seemingly strong forced line 2 S .b4? 'iVbS 2 6 .a4 4Jxa4 2 7 .�xc4 �xc4 2 S .'iVdS + h7 2 9 .fxe6 does not work as Black has 29 . . J:IcS ! ; B ) 2 3 .'iYd3 4JdS 24.�e4 and White gets a deadly �d3 I'i¥ e4 battery. 23.

�e2-c4

�gS- hS

On 23 . . . 4Jb6 White has 24.'iVb3 ! .

31 .

e5-eS !

32.

'iVg3-e5

.tt b 7-c7

A sad situation for Black. Without any counterplay he is confined to a few pas­ sive moves, waiting for annihilation.

32 .

...

33.

a2-a4

'iYdS-eS

White has plenty of time for everything.

24.

'iYa3-h3

tLld 7-fS

33.

...

'iYeS-dS

25.

b2-b3

as-a5

34.

l:tf1 -f2

'iYdS-eS

35.

!If2-f3

'iYeS-dS

3S.

�c4-d3

'iVdS-eS

37.

'iVe5 -e4

tLl h 7-fS

3 7 . . . l:Ixe 6 runs into the forced 3 S JUs + 4JxfS 3 9 . 11xfS + 'iVxfS 40 .�h7 mate.

Now the time is right for White to start direct action on the kingside. 2S.

f4-f5

eSxf5

27.

�f1 xf5

tLlfS-h7

2S.

J:tc1 -f1

It is never too late to do something stu­ pid and lose a winning position with a blunder like 2 S .1lf7 ? ? 4JgS .

3S.

l::tf5xfS

g 7xfS

39.

l::rf3xfS

�hS-gS

40.

�d3-c4 1

Threatening 4 1 JH7 .

2S.

...

'iYe7-dS

29.

'iYh 3-g3

'fda7-e7

40.

...

3 0.

h2-h4

ttbS-b7

41 .

'iVe4-f4

�gS-hS 1 -0

191

Chapter

W i nn i n g C h ess M i d d l eg am e s

Positions with parallel hanging pawns in the centre are difficult to play for both sides, since there are lots of dy­ namic possibilities involved and a keen eye for detail is needed. There are a number of well-known plans, but it has to be thoroughly checked whether they can be applied in the exact situation on the board. Being a l .d2 -d4 player I have had plenty of such positions in my own games (mostly playing against the hanging pawns) and I have seen the pros and cons. The following two ex­ amples from my own practice show the sensitivity of these positions and how easily things can go wrong.

SL 8 . 3 (D4S )

GAME 3 7

Ivan Sokolov Alexander Khalifman Pardubice 1 9 94 (3) 1.

d2-d4

d 7-d5

2.

c2-c4

c7-cS

3.

ttJ b 1 -c3

ttJ gS-fS

4.

e2-e3

e7-eS

5.

ttJ g 1 -f3

ttJ bS-d 7

S.

�d1 -c2

�fS-e7

cussion. In almost all lines, the bishop is better placed on d6, where, among other things, it s the liberating move . . . e6-eS . 7.

I guess that the main idea behind the choice for this move (as opposed to 6 . . . iLd6) is to avoid a theoretical dis1 92

...

b7-bS

In the absence of the . . . e6-eS break, Black has to develop his light-squared bishop to b 7 . We also have this type of lines with the black bishop on d6 in­ stead of e 7 , and again this looks (to me at least) like the more natural develop­ ment. You may have understood by now that I am not fond of the 6 . . . iLe 7 line. S.

�f1 -d3

�cS-b7

9.

0-0

0-0

�c1 -b2

1:raS-cS

1 0.

11.

l:ta 1 -d 1

11.

. ..

P a ra l l e l H a n g i n g P awns in t h e C e n t re

sacrifiCing material) 1 8 . . . iLgS 1 9 .fxe 6 fxe6 (or 1 9 . . . iLxe 3 + 2 0 .�h l fxe6 2 1 .'iVg4, transposing to the game) 2 0 .'iVg4

b2-b3

The most common move here, since Black does not have . . . e6-e S . White can also opt for 7 .iLd3 , but then Black has the opportunity to try and prove that his bishop is well-placed on e7 by play­ ing 7 . . . cS , and after 8 . 0-0 cxd4 9 .exd4 dxc4 1 0 . iLxc4 0 - 0 1 1 . �d l ttJb6 1 2 .iLb3 iLd 7 1 3 .ttJeS iLc6 1 4J:td3 (applying a rook transfer we have al­ ready seen in the Isolated Pawn chapter) 1 4 . . J::tc8 I S .'iYe2 ttJbdS 1 6 Jig3 ttJxc3 1 7 .bxc3 iLe4! 1 8 .iLh6 iLg6 1 9 .ttJxg6 hxg6 and the game was more or less equal in Tkachiev-Vaisser, French Championship, Aix-Ies-Bains 2 0 0 7 . 7.

Kramnik -Van Wely, shown further on in the comments.

3:

In the case of 1 0 . . . cS , 1 1 .cxdS exdS 1 2 . l::t ad l �c8 would transpose to the game. Black can also postpone all decisions and first play a useful move with 1 0 . . . h6. Then, after 1 1 . �ad l 'iVc7 , White can either take action in the cen­ tre with 1 2 .e4 dxe4 1 3 .ttJxe4 ttJxe4 1 4.iLxe4 ttJf6 I S .ttJeS , or play 1 2 .ttJeS and follow plans similar to

cS-c5

Instead of taking action in the centre, Black can also opt for the logical devel­ oping move 1 1 . . .'1jy c7 . In the following game, White's play was instructive and the plan he used can be employed in similar types of position: 1 2 .ttJeS ! ? (another standard plan here is 1 2 .e4) 1 2 . . . h6 1 3 .'iVe2 �fd8 1 4.f4.

an alysi s d i agram

2 0 . . . �e7 ? A losing mistake. The pawn on e 3 had to be taken after all : 2 0 . . . iLxe3 + ! 2 1 . �h l ttJf4 ! 2 2 . ld.xf4 'iYxf4 2 3 .'iYxe6+ btfl 24. l::l f l iLf2 2 S .dS ( 2 s .iLg 6 ? ? allows mate in two 2 s . . . iLxg 2 + 2 6 .�xg2 'iYf3 mate ; while in the case of 2 S .'ifg6? ld.d8 there is no follow-up for White) 2 S . . . l::t d 8 2 6 .iLc4 ( 2 6 . iLg 6 ? iLxdS 2 7 .iLxf7 + 'iYxf7 2 8 . 'li'xfl + ? iLxfl ) 2 6 . . . iLc8 ! 2 7 . VJIi e2 iLg4 and Black fights off the attack. 2 1 .'li'e4 ttJf6 2 2 .'iYg6

analysi s d i agram

The point behind 1 2 . ttJeS . White's wing attack is now correctly answered with central counterplay : 1 4 . . . cS ! I S .cxdS ttJxdS 1 6 .ttJbS 'iYb8 1 7 .ttJxd7 l::lx d7 1 8 .fS ! (White has al­ ready taken a substantial positional risk, so there is no way back; he has to pur­ sue his original plan, even if it involves

a n a lysis d i agram

2 2 . . . �h8 2 3 .dxcS (with the involve­ ment of the b 2 bishop, the white attack 1 93

C h a p t e r 3 : P a ra l l e l H a n g i n g P awns i n t h e C en t r e

W i n n i n g C h es s M i d d l eg a m e s

cannot be stopped) 2 3 . . J ixc5 24J lxf6 gxf6 2 5 .�xf6 + �xf6 2 6 .iVxf6 + 1:tg7 2 7 . 'iYxh 6 + �g 8 2 8 . �xe 6 + �h8 2 9 .�h6+ �g8 3 0 .e4 and White soon won in Kramnik-Van Wely, Wijk aan Zee 2 0 0 7 . 1 2.

c4xd5

e6xd5

1 3.

d4xc5

b6xc5

So we have reached one of the typical positions with parallel hanging pawns in the centre. White is going to put pressure on the c5 and d5 pawns (mostly d5) , trying to force Black to make a decision in the centre - pushing c5 -c4 or d5 -d4 - while Black will try to coordinate his pieces and look for dy­ namic possibilities related, mostly, to the . . . d5 -d4 pawn break. Capturing 1 3 . . . .uxc5 ? and opting for an isolated pawn structure instead of hanging pawns in the centre, would have been a grave positional mistake, since Black cannot develop any active piece play here to compensate for the isolated pawn. After such positional res­ ignation, things can go very wrong very quickly: 1 4 .'ii' b l lIe8 1 5 .ttJe2 ! �c8 1 6 .ttJf4 h6? (a small move with big consequences) 1 7 .�f5 ! .a:c7 1 8 .ttJd4 �a6 1 9 .ttJde6 1 -0 Polugaevsky-Comas Fabrego, Palma de Mallorca GMA 1 9 8 9 . 1 4.

�c7 ! , Black seems to be doing OK, since after 1 6 .ttJf4 he can choose be­ tween 1 6 . . . �d6 and 1 6 . . . d4. 1 5.

�f5-h3

a7-a6

Black would like to put his rook on c 7 (in order t o unpin his d7 knight) , but could not do this immediately due to 1 6 .ttJb 5 . 1 6.

tbc3-e2!

14.

...

g 7-g6?

Not a bad move in itself, but in this case it does weaken the a I -h8 diagonal, which will be one of the main reasons for Black's later problems. It was better to postpone this decision. A logical option was to play a useful move like 1 4 . . J �e8 and after 1 5 .ttJe2

A) If 2 0 .�fe l ? ! , then 20 . . . iiLxf4 (not 20 . . . d4 because of 2 1 .ttJd3 !) 2 1 .exf4 d4 2 2 .ttJe5 ttJxe5 2 3 . fxe5 ttJd5 and Black has an excellent game, since 24.g3 ? runs into 24 . . . ttJf4! ; B) 2 0 .iiLxd 7 ! �xf4 2 1 .iiLh3 �d6 2 2 . 'if c3 ! (back to the long diagonal, re­ storing the battery) 22 . . . d4! . Black has to execute this thematic break.

White is trying to take advantage of the weakened a 1 -h8 diagonal. 1 6.

...

l:Ic8-c7

1 7.

'ifc2-c3

tbd7-b6

Black was worried about the d-file pres­ sure and a different kind of tactics based on a battery with �c3 / �b2 . Another option was 1 7 . . J �e 8 . 1 8.

'ifc3-a5

tb b6-d7

White has pressure along the d -file, and his bishops are coordinating well, so Black has problems. 1 8 . . . ttJc8 , hoping for a transfer to e4 via d6 , would run into 1 9 .ttJf4. 1 9.

tbe2-f4

�d3-f5!

Pinning the d7 knight and increasing the pressure on the d5 pawn (the c5 pawn is well defended and cannot be hit for the time being) is a standard plan here and it is the correct way to proceed. One of the standard strategic plans in these positions is 1 4 .e4. The idea is to force . . . d5 -d4 (in the case of a massive swap on e4, Black's c5 pawn would re­ main rather weak and easy to target) 1 94

and then transfer the c3 knight via d2 to c4 and reach the constellation �d3 / ttJc4; or to leave that knight on d2 , put the bishop on c4 and transfer the other knight via e 1 to d3 with the constellation ttJd2 /ttJd3 / �c4 , taking advantage of the blockade on d3 and c4 and, after that blockade is established, start advanCing the e- and f-pawns (f2-f4 and e4-e 5 ) . The problem for White is that Black gets space and possibilities for a kingside attack. Moreover, the idea out­ lined above is difficult to execute, for example : 1 4 . . . d4 (Black also gets good counterplay in the case of 1 4 . . . c4 1 5 .bxc4 dxc4 1 6 .�e2 �c 7) 1 5 .ttJb l ttJg4! (Black has to be qUick and must not give White time to achieve ttJa3 -c4, h 2 -h3 , �b 2 -c l ) 1 6 . ttJbd2 �c 7 ! 1 7 . ttJc4 f5 ! and Black has taken the ini­ tiative; his pieces coordinate very well. It is clear that White's strategy has failed.

1 9.

...

d5-d4?

Black is under pressure, so he sacrifices a pawn in order to change the character of the game. For the sacrificed pawn he will not get any meaningful compensa­ tion. A better defence was 1 9 . . . iiLd6 ! , and now :

White is better, but he has to be accu­ rate : B 1 ) 2 3 . exd4? is wrong due to 23 . . . ttJd5 24.�d2 ttJf4 2 5 .dxc5 �xf3 2 6 .cxd6 ttJxh3 + 2 7 .gxh3 (2 7 . �h l ? loses to 2 7 . . . 'ii'a 8 ! 2 8 J lg l �d7 ! 2 9 . gxf3 'ii'xf3 + 3 0 .Rg2 ttJxf2 +) 2 7 . . Jic5 ; B2) In the case of 2 3 .�c4 Black has at least equality with 2 3 . . . ttJd7 ! ; B 3 ) 2 3 . �d3 ! �xf3 (if 2 3 . . . ttJd7 ?, then 24. exd4 ; or 2 3 ... �e4 24. 'ii'xa6) 24. gxf3 ttJd5 2 5 . f4 and White should be better, but the position is rather complicated and Black has his chances. 20.

e3xd4

�b 7xf3

21 .

g2xf3

c5-c4

22.

b3xc4

�c7xc4

23.

'iYa5xa6

White has a damaged kingside pawn structure, but he has reasonable control over the position and it is difficult for Black to prove the correctness of his sacrifice. 1 95

Chapter

W i n n i n g C h ess M i d d l eg a m es

23.

...

l::Ic 4-c2

If 23 .. .tt:Jb6 , then 24.�fe l l:ta4 2 5 .'iYb5 and now: A) White will have a massive advan­ tage in the case of 25 . . . �d6 2 6 .�c I ! , see : A I ) 2 6 . . Jlxa2 ?? is a blunder due to 2 7 .ttJe6; A2) With the white rook on e 1 2 6 . . . �xf4? 2 7 .�xf4 ttJfd5 does not work because of the simple 2 8 .�h6 ! ttJc3 2 9 .'iYe5 . B) 2 5 . . . �b4 ! 2 6 . �e2 �d6 and now: B 1 ) 2 7 . ttJ d3 ! .l:[xa 2 2 8 . 'ifb3 and White remains a sound pawn up ; B 2 ) 2 7 .�c 1 can be played, but it is less good after 2 7 . . . �xf4 2 8 .�xf4 tDfd5 2 9 .�h6 ttJc3 3 0 .'iYe5 tDxe 2 + 3 1 .�h I f6 3 2 .�e 6 + W h 8 3 3 . 'iYxe2 ( 3 4 . . . ttJxd5 ? 'iYd6 3 4 .d5 �e8 3 5 .'iYb 5 ) 3 5 .�e3 and White has com­ pensation, but it is difficult to say if he has more.

If 2 6 . . . ttJb6 , then 2 7 . i:'1a I ! (2 7 .d5 ?! is not good due to 2 7 . . . ttJc4 and Black has compensation) 2 7 . . J �icxa 2 2 8 .�xa2 �xa2 2 9 .l:lb I ttJbd5 3 0 .�fl and White is a sound pawn up. 27.

nd 1 -b 1

�a2-b2

28.

ttJf4-d3

llb2xb1

29.

::tf1 xb1

White is a clear pawn up ; however, due to his damaged kingside pawn structure Black keeps some drawing chances. 29.

...

30.

�b1 -b5!

l::r c 2-c7

Not allowing a blockade of d5 . 30.

...

31 .

�g1 -g2?!

33.

::tb5-bS

34.

n bS-aS

ke 7-d6

34.ttJe4? blunders a pawn to 34 . . . ttJxe4 3 5 .fxe4 l:ie7 3 6. ldb6 11xe4. 34.

...

�c7-e7

35.

.te3-g5

::te7-e1 ?

The final mistake. Black would have had reasonable surviving chances after 3 5 . . . h6 ! 3 6 .�h4 g5 3 7 .�g3 h5 .

3:

P a r a l l el H a n g i ng P a wns in t h e C e n t re

Time pressure does its job. Much better was 3 9 . .id2 �d I 4o J lxe8 f5 4 1 .ttJc3 , winning a piece. 39.

...

40.

ke5-g3

f7-f6

Black still cannot prevent further mate­ rial loss. 40.

...

�g 7-f7

41 .

�h3-d7

tLJeS-c7

36.

tLJc5-e4!

�d6-b4

42.

J:tbS- b 7

ne1 -c1

37.

�aS-b8!

�b4-a3?

43.

�d 7-a4

g6-g5

3 7 . . . �a5 ! was a better response, in or­ der to answer 3 8 .�f4? (3 8 .�e3 should win for White) with 3 8 . . . �c 7 .

Or 43 . . . f5 44 . .ie5 �e7 45 .ttJf6 ttJxf6 46 .�xc7 �f8 47 .�f4 1Ic8 48 .�h6+ �g8 49 .�b3 + . 44.

�gS-g 7

�g3xc7 !

Black was in time pressure and I wanted to keep pieces on the board, but objec­ tively better was 3 I . ttJ e 5 ! ttJxe5 3 2 .dxe5 ttJd7 3 3 .f4 �c5 3 4 .�d2 , and White should win. 31 .

...

tLJf6-eS!

3S.

kg5-f4 !

Now Black loses more material. 3S.

...

39 .

..if4-e5+

tLJf6-d5

And Black resigned in view of 44 . . . ttJxc 7 45 . ttJ c 5 ; or 44 .. J �xc7 45 . .ib3 .

Structure 3 . 3

32. 24.

�b2-c1

I wanted to return one of my two extra pawns and keep things simple. It was, however, better to be greedy and try to keep both pawns, by playing 24.ttJd3 ! ttJd5 2 5 .�c I �d2 2 6 .�fd 1 . 'iW dS-aS

24.

...

25.

'iVa6xaS

l::IfSxaS

26.

kc1 -e3

�aSxa2

1 96

tLJd3-c5?

This is j ust wrong. Better was 3 2 .ttJe5 or 3 2 .i�J4. 32.

...

tLJd 7-f6?

He does not take his chance with 3 2 . . . tDxc5 ! 3 3 .dxc5 f5 ! (shutting off the h3 bishop) 3 4.�g I ttJf6 3 5 . .ifl ttJd5 (not 3 5 . . . ttJd7 ? ? 3 6 .c6) 3 6 .�d4+ �f6 3 7 . c 6 �xd4 3 8 . �xd5 �b6 3 9 .�b5 ide7 and Black saves the game. 197

C h a p t e r 3 : P a ra l l el H a n g i n g P a wns in t h e C en t re

W i n n i n g C h ess M i d d l eg a m e s

QO 8.5 (DS 8)

GAME 3 8

Ivan Sokolov Nigel Short Saraj evo 1 9 9 9 (6) 1.

d2-d4

ttJ gS-f6

2.

c2-c4

e7-e6

3.

ttJg1 -f3

d 7-dS

4.

ttJb1 -c3

�fS-e7

S.

�c1 -gS

h7-h6

6.

�gS-h4

0-0

7.

e2-e3

b7-b6

S.

�a1 -c1

One of the main lines of the Tartakower IMakogonov Variation , which was rather popular in the late 1 98 Os and through the 1 9 9 0s. S.

...

9.

�h4xf6

�cS-b7

After Black has placed his bishop on b 7 , White parts with his bishop pair and then takes on dS , forcing Black to re­ capture with the pawn, an idea we have seen in several games in this book. 9.

�e7xf6

1 0.

c4xdS

e6xdS

11.

�f1 -d3

c7-cS

1 2.

0-0

This is considered to be the starting po­ sition of the variation. One may wonder what White's idea could be here ? After all, Black has the bishop pair, the dark-squared bishop in particular being 1 98

well placed on f6 , and with his last move . . . c7 -cS Black even gets some pressure on White's central pawn d4. Well, White's main idea in this varia­ tion is to force a swap on d4 and end up playing with a favourable d41 dS pawn symmetry in the centre, where his light-squared bishop on d3 is superior to its colleague on b 7 , an idea we have already seen in a number of games in Chapter 2 on the Isolated Pawn. If White's idea is well executed, he will have a small, lasting advantage. 1 2.

...

ttJbS-a6

With the text Black wants to avoid the aforementioned exchange on d4 and keep the tension in the centre - which is a reasonable plan in general. How­ ever, the black knight is a bit clumsy on a6, and the text move is considered to be one of the sidelines. The position is interesting from a stra­ tegic prospective. It requires both sides to be on the alert for central exchanges, because three different pawn structures may easily arise here : with an isolated pawn, with central pawn symmetry, or with hanging pawns in the centre. For a better understanding of the pros and cons of the position I shall give a few lines. A) 1 2 . . . tLJd7 is considered to be the main line. Quite some time ago, in Biel 1 9 8 9 , I had a game against Hellers which continued 1 3 .�fS , increasing the pressure on Black's central pawns and trying to force Black to make a de­ cision in the centre. 1 3 .�b 1 has also been tried, but this allows Black to continue normal development with 1 3 . . . 1:te8 and after 1 4 .dxcS tLJxcs I S . ttJd4 g6 Black has a comfortable iso­ lated pawn-position .

After 1 3 .�fS there followed: A I ) 1 3 . . .cxd4 (definitely not a mis­ take; however, it allows White to fulfil his strategic obj ectives) 1 4.exd4 g6 I S .�d3 �g 7 1 6 .'iVb3 (White has slight pressure on dS , his d3 bishop is considerably better placed than its black counterpart, and White's f3 knight is likely to enter on eS . Black, however, has a strong bishop on g 7 and his pres­ sure along the a I -h8 diagonal can be an important source of counterplay) 1 6 . . . a6 1 7 .nfe l !:lc8 1 8 . a4 tLJf6 1 9 .!rcd 1 (eliminating the tactics related to Black's . . . tLJf6-e4) 1 9 . . . 'iVd6 2 0 .tLJeS (the f3 knight has found its dominant post) 20 . . . �c7 2 1 J le2 !re8 2 2 . f4 nce7 2 3 .�b l (time to move the bishop to the a2-g8 diagonal)

A2) He can stay calm and play the useful 1 3 . . . Iie8 . Surprisingly enough, this continuation has rarely occurred in practice ; nevertheless, Black's position is solid and healthy and White cannot prove an advantage, for example : A2 1 ) After 1 4 .�xd 7 'iYxd7 I S .dxcS bxcS 1 6 .tLJa4 c4 1 7 .tLJd4, we reach one of the classic positions with hanging pawns in the centre.

ana lysis d i agram

analysi s di agram

2 3 . . . ttJe4? (getting short of time, Black goes for a desperate counterplay at­ tempt, which simply loses a pawn. However, the position was far from clear and Black had to undermine the dominant eS knight with 2 3 . . . gS ! 24.fxgs hxgS , threatening . . . tLJf6-g4) 24. tLJxe4 dxe4 2 S . �xe4 �xe4 2 6 .I1.xe4 and I was a sound pawn up and soon won. Black had two other acceptable continu­ ations.

White has a perfect blockade with his knights on c3 and d4, Black has an ex­ cellent bishop on f6 and a ive one on b 7 ; and also, Black has a little more space. There is a dynamic balance and the position is about equal. A similar pawn structure with the same blockade (White : tLJc3 , tLJd4 - Black: pawns c4, dS , bishop on b 7 and bishop on eS) , but more favourable for White, was seen in Game 1 7 of the Capa­ blanca-Alekhine World Championship match 1 9 2 7 . Despite these better cir­ cumstances , the legendary Cuban was not able to take advantage of his perfect blockade and the game ended in a draw. A2 2 ) 1 4.'iYa4 tLJf8 I S .Rfd l g 6 ! (an important move. White has some pres­ sure in the event of I S . . . tLJe6 1 6 .dxcS bxcS 1 7 .'iVg4) 1 6 .�h3 ttJe6 1 7 .dxcS bxcS ! (the correct decision. White has 1 99

Chapter

W i n n i n g C h es s M i d d l eg a m es

some advantage in the isolated pawn­ position ensuing after 1 7 . . . tLlxcs 1 8 .'iYc2) 1 8 .'iVb3 �b8 1 9 .tLlxdS �xdS 2 0 .�xdS �xb3 2 1. .�xd8 �xd8 2 2 .axb3 �xb 2 2 3 .l:rb l �c3 24.l:Ic l (otherwise 24 . . J l b 8 , threatening 2 S . . . c4) 24 . . . �b2 with equality. A3) Also possible for Black is 1 3 . . . g 6 and after 1 4.�xd7 ,

D) Finally, 1 2 . . . cxd4 has also been played many times. It immediately hands White his strategic obj ectives (a d4/ dS central pawn symmetry with the superior light-squared bishop) , but now Black keeps the possibility to de­ velop his knight to the natural c6 -square : 1 3 .exd4 ( 1 3 .tLlxd4 tLlc6 1 4.tLlxc6 �xc6 I S .tLle2 'iVd6 1 6 .'ifd2 , playing against the isolated pawn, gives White a very small plus) 1 3 . . . tLlc6 1 4.�b l l:Ie8 I S .'iVd3 g6 1 6 .I;Ife l iYd6 1 7 .a3 I;Ixe 1 + 1 8 J ixe 1 �d8 1 9 .�a2 tLle7 2 0 .tLleS and White had a standard, well-known , small advantage in Karpov-Portisch, Skelleftea World Cup 1 989.

200

1 7.

1 3.

�d3-b1

1 4.

a2-a3

1 4.

... l:lf1 -e1

a n a lysis d i agram

1 7.

...

d 5-d4

l'lf8-e8

9 7-96

Preparing to hit Black's hanging central pawns with e3 -e4 (after the pawn swap d4xcS b 6xcS ) . 1 5.

...

1 6.

d4xc5

'iVd8-d7

Putting the question to Black which type of pawn structure he wants to play. 1 6.

�

b6xc5

A I ) 24 . . . 'iVa2 does not solve Black's problems after 2 S .'iVe 8 + �g7 2 6 . �b 7 'iVe6 2 7 .'iVxe6 �xe6 2 8 .�d3 ! .

e3-e4

With his last two moves White is get­ ting ready to transfer his bishop to the a2-g8 diagonal, at the right time. 1 5.

P a r a l l e l H a n g i n g P a wns in t h e C e n t r e

Executing the plan prepared with a3 , �b l , �e l .

Black is not worried about the weaken­ ing of his light squares on the fl -a6 and a2-g8 diagonals. I think that better, and more principled, was 1 7 . . . Yad8 ! when the bishop on f6 is strong and Black has a good game : 1 8 .exdS l:.xe l + 1 9 .'iVxe 1 �xdS 2 0 . tLlxdS (or 2 0 J��d l 'ife6 2 l . �xe6 �xe 6 2 2 . 1:!xd8 + �xd 8 2 3 .tLleS tLlb8) 2 0 . . . 'iVxdS 2 1 .'iYe2 �d6 2 2 .h4.

analysis d i agram

. . . the zWischenzug 1 4 . . . cxd4 ! (a mis­ take would have been 1 4 . . . 'iVxd7 ? due to I S . dxcS bxcS 1 6 . tLle4 �xb 2 1 7 .tLlxcS 'iVe 7 1 8 .�b l 'iVxcs 1 9 .�xb2 and now White achieves a perfect blockade with a strong knight against a bad bishop) 1 S .�c6 (winning a tempo compared to I S . tLlxd4 'iYxd 7 ) I S . . . �xc6 1 6 . tLlxd4 �b 7 and the strong bishop on f6 fully compensates for the isolated dS pawn, Kasim­ dzhanov-Kramnik, Wijk aan Zee 1 9 9 9 . B ) 1 2 . . J�re8 has also been tried in practice, with the idea to keep the ten­ sion in the centre, while making a use­ ful move; C) 1 2 . . . 'iV e 7 was played once by Helgi Olafsson against me in Wijk aan Zee 1 9 9 1 . Black tries to achieve the same objectives as with 1 2 . . J :t:e8 , but this has the drawback that after 1 3 . lde 1 White threatens the unpleasant e3 -e4;

Playing an isolated pawn-type position after 1 6 . . . tLlxcs was a reasonable choice.

3:

Black 's knight on a6 is doomed. 28 . . . �d4 2 9 .�c4 I;If6 3 0. �fl ! (with the other pieces already on perfect posts, it's time for White to improve his king) 3 0 . . . gS (a counterplay attempt that does not work) 3 1 .hxg S hxgS 3 2 . tLlxgS l:.xf2 + 3 3 . �e l lif6 (3 3 . . J dxg 2 ? loses a piece to the simple 3 4.tLle6+ cJ;th6 3 S .tLlxd4 cxd4 3 6 .�xa6 �a2 3 7 .�b3 ) 3 4 .tLlxf7 �g6 3 S .g4 and White wins; A2) 24 . . . 'iix a3 ? ? is a terrible blunder due to 2 S .'iVe8 + �g7 2 6 . �b3 ; A 3 ) 2 4 . . . l:l e 6 2 s . 'it'c 4 'iYxa3 (if 2S . . . 'iic 3 , then 2 6 .�b3 'iYxc4 2 7 .�xc4 �c6 2 8 .l:t:b7 tLlc7 2 9 .I;Ixa7) 2 6 .�xg6 tLlb4 2 7 .�fS �e7 (2 7 .. J lc 6 ? loses to 2 8 . l:!e 1 ) 2 8 .'iVxcS and due to his weak­ ened kingside Black is in trouble. B) 2 2 . . . cJ;tg7 ! 2 3 .�e4 'iVa2 24.lic2 ne6 2 S .hS gS with a sharp game where Black is definitely not worse. 1 8.

ana lysis d i agram

actIVIty A) Immediate with 22 . . . 'iib 3 ? is wrong due to 2 3 .�c2 ! iYxb2 24.�b 1 and now:

�b1 -a2

This is one of the typical positions with the hanging pawns in the centre-struc­ ture. With e3 -e4, White has provoked . . . dS -d4, which has given White strong control of the fl -a6 and a2-g8 diago20 1

C hapter

W i n n i n g C h ess M i d d l e g a m es

nals and an excellent blockade on d3 and c4 in many cases. Black's cS pawn is also rather weak. The position is obj ec­ tively better for White, but exact play is needed. Black has the bishop pair (both of them are working now) and a poten­ tially strong ed d-pawn, so - as we are about to see - the tables can turn quickly. Please note that this pawn structure is far better for White than the similar one discussed in the previous game Soko­ lov-Khalifman , where I suggest 1 4 .e4 in the comments, instead of the game move 1 4.i�JS . Here White does not have the dark-squared bishop on b 2 ; he has exchanged that bishop for Black's knight on f6 , which is rather useful. The text looks logical in itself, but White had a strong and straightforward way to achieve an advantage: 1 8 .tLla4 �ac8 1 9 .i.d3 (focusing on the weak cS pawn) 1 9 .. .'ifd6

analysis d i agram

2o .'ifd2 ! i.g7 (going for ' counterplay' with 20 . . . c4? 2 1 .i.xc4 i.xe4 ? ? is a di­ sastrous blunder after 2 2 .i.xa6 l:rxc 1 2 3 .iYxc l 'iYxa6 24.tLlcS 'iYc6 2 S .tLlxe4 'iYxc 1 2 6 .tLlxf6+) and now: A) Playing for a light-square blockade with 2 l .b3 tLlb8 2 2 .tLlb2 tLld7 2 3 . tLlc4 leads to a very sharp game, since Black 202

is harmoniously developed and ready for counterplay should the position be opened, for example: 2 3 . . . 'iYb8 24.h4 tLlf6 ! 2 S . e S tLlg4 2 6 . 'iYf4 i.xf3 2 7 .tLld6 !

i.xa6 24.lLJxcS i.bS 2 S .a4 i.c6 2 6 .b4 d3 2 7 .tLld2 ; or 2 7 .bred 1 ) 23 . . . i.f8 (or 23 . . . i.f6 2 4.f4) 24.eS and White is better. In most of these lines, Black is suffering because he is not able to cre­ ate enough counterplay to compensate for the weakness of his cS pawn. 1 8.

...

na8-b8!

Hitting one of the weaknesses in White's position: the b2 pawn. l 8 . . . tLlc7 ? is bad due to 1 9 .lLJa4±.

P a r a l l el H a ng i n g P a wns in t h e C e n t re

A 2 ) 1 9 . . . �g7 2 0 . tLl dS i.c6 (if 2 0 . . . i.xdS 2 l . i.xdS , there is a striking difference in activity between the white and the black bishop) 2 l .h4. The posi­ tion is sharp, but White's chances seem to be preferable. B) Interesting for White was 1 9 .�e2 ! ? White has a weak b2 pawn and that problem is now being solved. Now, 1 9 . . . lLJc7 ? is not good due to 2 0 . lLJa4 lLJa6 2 1 .i.b 1 with �d3 to follow (2 1 .i.c4 i.c6 offers Black counterplay) . However, the position looks rather un­ clear after 1 9 . . . i.c6 ! 2 0 .lLJdS i.g 7 . 1 9.

analysis di agram

2 7 . . .'iYxb3 ! (not afraid of ghosts ; 2 7 . . . tLlxeS is not good due to 2 8 .�xeS ! !:lf8 2 9 . �excS !:lxcs 3 0 .�xcS with a massive advantage for White) 2 8 .i.c4 �b 2 ! (the threat to take on f2 forces Whi te to lose a crucial tempo) 2 9 .i.xfl + ( 2 9 . �xf3 lLJxe S 3 0 .'iYf4 ebh7 ! 3 I .lLJxc8 �xc8 is better for Black' or 29 .'iYxfl + 'it>h7 3 0 .iYxf3 tLlxeS ) 29 . . . ebh8 3 0 .'ifxf3 lLJxeS 3 l .'iYh3 �f8 3 2 .lLJxc8 nxfl 3 3 .'if g 3 ; B ) 2 1 . 'li' as ! , continuing t o hit the cS weakness : B 1 ) Very likely, Black is forced to ex­ change queens; see 2 l . .. hS 2 2 .i.bS ! (not 2 2 .i.xa6? �xa6 2 3 .�xa6 i.xa6 24.l:rxcS (2 4.lLJxcS ? loses the exchange after 24 . . . i.h6) 24 . . . i.d3 ! 2 S .b4 i.c2 2 6 .�xc8 :rxc8 2 7 .tLlcS d3 and Black's initiative is worth more than a pawn) 2 2 . . . t(e 7 2 3 .eS ! iYd8 24.'iYxd8+ idxd8 2 S .i.xa6 �xa6 2 6 .lLJxcS and White is a sound pawn up ; B2) So, 2 1 . . .'iVd8 2 2 .'iYxd8 �exd8 2 3 .lLJd2 (taking a pawn is now very much worth considering: 2 3 .i.xa6

3:

...

�b7-c6!

For some reason I had completely lnissed this strong reply. 1 9 . . . tLlc7 ? is again not good because of 2 0 .lLJa4. The tables have already turned and White is on the defensive. 20. 1 9.

�a2-c4?

Completely missing Black's response. White had two reasonable ways to con­ tinue : A) 1 9 .'�Vd2 and now: A I ) 1 9 . . . tLlc7 and now: A l l ) 2 o .'iVxh6 ? is a bad mistake, since White only gets a few checks for the piece after 2 0 . . . dxc3 ! 2 1 .'iYxg 6+ i.g 7 2 2 .tLlgS tLldS ! 2 3 .'iVh 7 + �f8 ; A 1 2 ) 2 0 .�f4 'ife7 (if 2 0 . . . �g 7 , White nevertheless plays 2 1 . tLla4 ! 'li'xa4 (the ' tactic' 2 l . . . lLJdS ? 2 2 .i.xdS 'iYxa4 loses to 2 3 .i.xfl �xfl 24.eS) 2 2 .'�xc 7 �e7 2 3 .tLleS ; or 2 0 ... i.g 7 ? 2 l .lLJa4 ! ) 2 l .tLla4 gS ( 2 1 . . . �c6 2 2 .lLJxcS llxb2 2 3 .i.c4 i.g 7 24.eS ; 2 1 . . . lLJa6 2 S .eS ; l:[b6 2 4 . tLl d3 2 2 .iYxh6) 2 2 .'ifg3 d3 2 3 .tLlxcS ! i.xb2 24 .h4 i.xc i 2 S J:txc i and White has a strong attack.

tDc3-d5

Black is better after 2 0 .i.xa6 dxc3 2 l .'iYxd7 i.xd 7 2 2 .bxc3

a n a lysis di agram

2 2 . . . 1:Ib3 ! (better than 2 2 . . . i.c6 2 3 .eS i.g 7 24.i.c4 i.xf3 2 S .e6 fxe6 2 6 . I:!xe6 l'dxe6 2 7 .i.xe6+ �f8 2 8 . gxf3 and White escapes with a draw) 2 3 .eS Idxa3 . 20.

...

21 .

e4xd5

�c6xd5

203

Chapter

W i n n i n g C h ess M i d d l eg am e s

White does not have enough compen­ sation for the sacrificed pawn after 2 1 .�xdS nxb2 2 2 .'iYd3 r!b6 2 3 .�c4 CiJc7 24.eS �g 7 . 21 .

...

22.

'iVd1 -d2

tDa6-c7

Better was 2 2 J :rxe S + CiJxe S (or 22 . . . �xeS 2 3 . 'iYd2 ) 2 3 .'ifd2 �g 7 , reaching the game position. 22.

...

�gS-g7

Black now had an extra opportunity in 22 . . J:i�xe l + 23 ..the l Wg 7 and White has problems defending his dS pawn, since Black is better after 24.CiJeS �xe S 2 S .1dxeS 'iYd6 . 23.

24.

ne1 xeS

tDc7xeS

tDf3-e1

Black has a perhaps very small, but last­ ing and unpleasant advantage. A) If White tries to simplify with 24.b4, he is still worse after 24 . . . CiJd6 2 S .�d3 ( 2 S .bxcS ? is wrong due to 2 S . . . CiJxc4 2 6 . !:txc4 'iYxdS) 2 S . . . cxb4 2 6 .axb4 CiJb5 2 7 . t. k6 �b6 ! (2 7 . . .tLlC3 ? ? blunders a piece after the simple 2 S .l:txf6 �xf6 2 9 .'iYf4+) 2 S .�xbS �xbS 2 9 .�f4 ( 2 9 J :lxf6 does not work after 2 9 . . . �xf6 3 0 . 'iVxd4+ � e 7 3 1 .'iYeS + � d S ) 2 9 . . .'iVdS ! ; B) Or 24.�a2 CiJd6 ! ; C) Solving the problem o f the b2 weakness with 24.b3 does not help after 24 . . . CiJc7 2 5 .'iYaS CiJxdS 2 6.'iVxcS CiJf4. 204

24.

...

tDeS-d6

2S.

tDe1 -d3

'li'd 7-fS !

Keeping the initiative. 26.

'iVd2-c2

If 2 6 . tLlxcS , unpleasant for White is 2 6 . . . tLlxc4 2 7 .�xc4 'iVb l + (2 7 . . . �xdS 2 S J lc l �gS ! 2 9 .f4 �f6 is also better for Black) 2 S J :tc l �xb2 ! 2 9 .d6 'ifa2 . 26.

...

.l:IbS-eS

27.

h2-h3

�f6-gS

2S.

l::t c1 -d1

h6-hS

29.

b2-b3

hS-h4

Seizing space on the kingside. 3 0.

'iic 2-b2

31 .

'iVb2-c2

�gS-f61

Like it or not , White is forced into a ive defence. I was definitely not in the mood to calculate the consequences of 3 1 . tLlxc5 d3 (the simple 3 1 . . . 'if cS 3 2 .tLld3 CiJxc4 3 3 .bxc4 'iVxc4 34.tLlf4 (or 34 .'ifb7 l:te 7) 34 . . . 'iVc3 ! 3 s .'ifxc3 dxc3 3 6 .d6 l:tdS also looks good enough for Black) 3 2 . 'if c 1 �d 4 3 3 .CiJxd3 CiJe4 or 3 3 . . J Ie 2 . I f 3 1 . �d2 (planning 3 2 . �e2 ) , 3 1 . . . �dS ! 3 2 .CiJxcS (3 2 .'iVc2 CiJe4) 3 2 . . . 'iYcS 3 3 .tLld3 CiJxc4 34.bxc4 'ifxc4 and Black is a sound pawn up. The black pieces are already very well placed and it is difficult to improve them. Needless to say, Black's next move came as a surprise to me. 31 .

...

32.

f2 -f3??

A truly horrible move, surrendering all the squares and bringing the game to an end. Black's . . . gS -g4 was not such a terrible threat at all, and even if it were, it is better to get mated than to allow such positional capitulation. Looking at this horrible, ugly move I managed to produce, I cannot help but recall Yasser Seirawan's quote : 'Thank God the grandmaster title is for life ' . I still had a n acceptable defence in 3 2 .�f1 ! (planning 'iY d 1 , to cover the g4-square) 3 2 . . . g4 3 3 .hxg4 fVxg4 34.'�d l . With his last few active moves, Black has also created some weaknesses of his own and the game has become rather unclear.

3:

P a r a l l e l H a n g i n g P a wns i n the C e n t r e

39.

�a6-d3

tDfS-g3

40.

'iVe1 -d1

'li'f6-eS

41 .

'iYd 1 -c2

'li'eSxdS

42.

b3-b4

cSxb4

43.

a 3xb4

'ifdS-eS

And White reSigned.

QO 7 . 1 0 (D S S)

GAME 3 9

Vladimir Kramnik Artur Yusupov Dortmund 1 9 9 8 (6) 1.

tDg 1 -f3

d 7-dS

2.

d 2-d4

tDgS-f6

3.

c2-c4

e7-e6

4.

tDb1 -c3

�fS-e7

S.

�c1 -gS

h 7-h6

6.

�gS-h4

0-0

32 .

...

treS-e3

7.

e2-e3

b 7- b6

33.

'iVc2-d2

�f6-eS

S.

�f1 -d3

�cS-b7

9.

All these vital squares White has surren­ dered with one stupid move. 34.

�c4-a6

'li'fS-f6

3S.

l::t d 1 -e 1

�eS-f4

36.

'iVd2-d1

.tt e3xe1 +

37.

'iVd1 xe1

�f4-e3+

3S.

tDd 3-f2

Sad reality: 3 s sii) h l tLlfs 3 9 .�c4 tLlg3 + 40 .Wh2 'ifd6 forces White into a mat­ ing net. White is completely paralysed and many moves win here.

0-0

tDbS-d7

1 0.

'iYd 1 -e2

c7-cS

11.

�h4-g 3

tDf6-e4

For 1 1 . . . cxd4, and also for comments on White's alternative 1 0 .'iYe2 , we refer you to Vyzhmanavin -Beliavsky - Game 2 S in the Isolated Pawns chapter. 1 2.

c4xdS

1 3.

l:ta 1 -d1

e6xdS

g6-gS!

3 S.

...

tDd6-fS

This is one of the critical positions in this popular variation. In grandmaster practice, Black has tried all the logical possibilities here. 205

C h a p t e r 3 : P a r a l l el H a n g i n g P a wns in t h e C e n t re

W i n n i n g C h e ss M i d d l eg a m e s

1 3.

...

ttJe4xg3

1 8 .f4 ( 1 8 .jt,c2 deserves attention, since if Black tries to become active with 1 8 . . . tLle4? 1 9 .f3 jt,d6 he loses material after 2 0 .g3 tLlcs 2 l .f4 'iVe8 2 2 .tLlfS 'iYe6 2 3 .c4) 1 8 . . . 'iVd6 1 9 .tLlb3 l:ife8 2 0 . tLlxcS bxcS 2 l .c4 'iVe6 2 2 .jt,xf6 'iVxf6 2 3 .cxdS jt,xdS 24.e4 c4 2 S .jt,b l and White started rolling his pawn ma­ j ority on the kingside, obtaining the initiative, Vyzhmanavin-A. Petrosian, Palma de Mallorca 1 9 8 9 ; C) For 1 3 . . . 'iVc8 , according t o current theory the best reply, see the next game; D) Putting the dark-squared bishop on a natural square, on the long diago­ nal, with 1 3 . . . jt,f6 ? ! does not work well after 1 4.jt,xe4 ! dxe4 I S .tLleS jt,xeS 1 6 .dxeS 'iVe 7 1 7 .tLldS jt,xdS 1 8 .ldxdS .

Black decides to eliminate White's dark-squared bishop. However, given the fact that the white knight on c3 ex­ erts pressure on Black's dS pawn and White's g3 bishop is not causing any particular trouble at the moment, Black can keep the knight on e4 (keeping the option of exchanging White's c3 knight at any moment) and maintain tension in the centre by playing, for example, 1 3 . . . 'iY c8 , as in our next featured game. Let me give a short survey of the op­ tions in this position. A) 1 3 . . . cxd4?! (the wrong plan. Black voluntarily releases the tension in the centre, giving White a free hand to take advantage of his well-placed pieces) 1 4.jt,xe4! (instead of playing an isolated pawn-position, White prefers a favour­ able symmetry) 1 4 . . . dxe4 I S .tLlxd4. White controls the d-file, his knight is excellently placed on d4, his bishop is working well on g3 and Black's b7 bishop is ive, hitting its own pawn on e4. All in all, White is clearly better; B) 1 3 . . . tLldf6 1 4 .dxcS ! (changing the pawn structure and creating an isolated pawn on dS for Black, while the white weakness on c3 will hardly be felt) 1 4 . . . tLlxc3 I S .bxc3 jt,xcS 1 6 .tLld4 'ife7 1 7 .jt,h4 'iV eS

Preparing . . . cS -c4, to be followed by . . . b6-bS . However, White's reaction will prove this to be almost a loss of tempo. Black has few useful moves here: A) A logical developing move like 1 4 . . . tLlf6 has the shortcoming that at a critical moment it gives White an im­ portant tempo, after I S .dxcS bxcS (with again a typical position with hanging pawns in the centre) and now: A I ) 1 6 .e4! (like in my game against Short, this typical break is an essential part of White's strategy) 1 6 . . . d4

ana lysis d i agram

analysis d i agram

206

1 4.

h 2xg3

1 7 .eS ! (seizing the initiative. Regular light- square blockade play with 1 7. tLl b 1 results in an unclear game after 1 7 . . . jt,d6 1 8 .tLlbd2 l:ie8) 1 7 . . . tLle8 1 8 . tLle4 jt,dS 1 9 . tLlfd2 ! ? (choosing an ambitious plan involving a pawn sacri­ fice. White could have obtained a risk­ free advantage with 1 9 .jt,c4) 1 9 . . . 'iVb6 2 0 .f4 'ifxb 2 2 l . l:ib l 'iVa3 (2 l . .. 'iVxa2 ?? loses the queen after 2 2 .na l 'iVb2 23 .l:ifb 1 ) 2 2 .fS jt,gS 2 3 .tLlxgS hxg S 24. tLle4 c4 2 S . tLlxgS ! (threatening 2 6 .'ifhS ; 2 S .jt,xc4? is a blunder due to 2S . . . jt,xc4 2 6 . 'iVxc4 'lWe3 + 2 7 .tLlfl 'lWxeS)

a 7-a6

a n a lysis d i agram

2 S . . . f6 ? (the losing mistake. The correct defence was 2S . . . g 6 ! 2 6 .'iVg4 tLlg7) 2 6 . jt,xc4 and White soon won in Moi­ seenko-Sargissian, Istanbul Ech 2 0 0 3 ; A 2 ) Another strategic idea is 1 6 .jt,a6 ! ? The swap of the light-squared bishops is in general a good idea here. Black's b7 bishop is de­ fending the potential weakness on dS and is also an instrumental piece in any counterplay associated with the . . . dS -d4 break : 1 6 . . . 'lWb6 1 7 .jt,xb 7 'iVxb 7 1 8 . l:id2 l:ifd8 1 9 .l:ifd 1 l:id6 2 0 . tLle l �ad8 2 l .tLld3 tLle4 2 2 .�c2 jt,f6 2 3 .tLlf4 jt,xc3 24.bxc3 and White now kicked the black knight out of the

centre with �h2 and f3 , obtaining a small advantage in Vyzhmanavin-Kolev, Elenite 1 9 9 3 ; B ) Placing the dark-squared bishop on its natural post with 1 4 . . . jt,f6 is a principled and rather logical decision (this position resembles the lines dis­ cussed in Sokolov-Short , Game 3 8 ) , and I do not see any advantage for White, for example : B 1 ) Exerting pressure on dS with I S .jt,b l l:ie8 1 6 . l:id2 a6 1 7 . b1c l l:ic8 1 8 . 'iY d 1 , in order to provoke black ac­ tion in the centre, is a fruitless exer­ cise for White, since after 1 8 . . . c4 ! Black starts rolling his queenside pawns and takes over the initiative , as in the game Vyzhmanavin-Pigusov, Moscow 1 9 8 7 ; B2) Protecting the knight on c3 be­ fore taking on cS with I S .'iVc2 prom­ ises White nothing, since Black will achieve a comfortable position with parallel hanging pawns in the centre af­ ter I S . . . g6 1 6 .dxcS bxcS ; B3) I S .dxcS jt,xc3 ! (an important in­ termezzo) 1 6 .c6 (trying to take advan­ tage of the d-file pin with 1 6.bxc3 tLlxcs 1 7 .c4? would have bad conse­ quences for White after 1 7 . . . tLlxd3 1 8 . l:ixd 3 jt,a6) 1 6 . . . jt,xc6 1 7 .bxc3 'iY f6 ; both sides have a weakness to worry about and the position is roughly equal; C) The set-up with 1 4 . . . 'iVc7 I S .jt,c2 l:iad8 gives White somewhat better chances in an isolated pawn-position after 1 6 .dxcS tLlxcs 1 7 .tLld4 ; D) Another logical and ambitious op­ tion for Black is the immediate release of White's potential d-file pressure with 1 4 . . . c4 ! and after I S .jt,b l a6, Black is ready to roll his queenside pawns , 207

Chapter

W i n n i n g C h ess M i d d l eg a m e s

when White i s forced t o take immedi­ ate central action: 1 6 .e4

a n a lysis d i agram

a n a lysis d i agram

1 6 . . . dxe4 ( 1 6 . . JieB l ? deserves atten­ tion, and if 1 7 .eS then 1 7 . ..tbfB - a standard plan in these positions which works most of the time) 1 7 . tLlxe4 bS . Black has a good game and White should be careful not to end up worse. The following game is a good example : I B .tLleS (or I B .dS �eB) I B . . . WVc7 1 9 .tLlg4 fS (an option was 1 9 . . . braeB and if 2 0 .tLle 3 , 20 . . . �b4) 2 0 . tLlc3 fxg4 2 1 .'iVxe 7 1daeB 2 2 .'iYb4 tLlf6 .

3 1 .I;IhS ! and White wins) 2 7 . . . �xg2 2 B .'iYc2 1:txe l + 2 9 . l::Ixe l l:if6 ! and Black was better in Babula-Bansch, Ger­ many Bundesliga 1 9 9 9 / 0 0 . 15.

d4xc5!

Not giving Black another chance to close the centre with . . . cS -c4. 1 5.

...

b6xc5

Black is forced to enter a typical struc­ ture with parallel hanging pawns in the centre, and to acknowledge that his 1 4 . . . a6? has been a loss of a tempo. 1 5 . . . tLlxcS ? simply loses a pawn after 1 6 .�c4 ; 1 5 . . . �xcS ? loses a pawn to the tactic 1 6 .tLlxdS ! �xd5 1 7 .�h7 + Wxh7 I B .IIxdS l:ta7 1 9 .1dfd l 'iYe 7 (a blunder is 1 9 . . . WgB ? 2 0 .tLleS and Black is likely to lose more material) 2 0 . 'iY d3 + and White regains his piece, remaining a healthy pawn up.

Kramnik correctly increases the pres­ sure on Black's only weak spot, the dS pawn. The standard 1 6 .e4 d4 1 7 .tLlb 1 (since the black knight is still on d7 , there are no extra possibilities like e4-eS for White here, as in Moiseen­ ko-Sargissian) , with tLlbd2 to follow, promises White a small advantage at best. It should be noted that a beautiful blockade often promises White much less than it seems to at first sight. 1 6.

...

tLJd7-b6

1 6 . . . tLlf6 ? ! simply loses a pawn after 1 7 .tLlxdS ! tLlxdS (or 1 7 . . . �xdS I B .e4) I B .e4. 1 7.

a2-a4!

Continuing to build the pressure on dS . With the black knight on b 6 , 1 7 .tLlxdS ? ? i s a horrible blunder due to 1 7 . . . �xdS 1 B .e4 �c4 and White has lost a piece for nothing. 1 7.

...

�e7-f6

1 7 . . . as is an ugly move to play and also, White's pressure remains unpleasant af­ ter I B .'ifc2 g6 (or I B . . . fS ? ! 1 9 .94) 1 9 .�a2 . 1 8.

'iYe2 -c2

1 9.

a4-a5

97-96

analysis d i agram

Now White needs to create counterplay immediately: 2 3 .a4 ! tLldS ! 24.tLlxdS �xdS 2 S .axbS 'ifb 7 2 6 . nfe l ! axbS 2 7 .'ifd2 ? ! (correct was 2 7 .�eS ! with an unclear game, since 2 7 . . . �xg 2 ? ? is a bad blunder due to 2 B .l:txbS 'iff3 2 9 . 'iYxc4+ �hB 3 0 .WVc2 g 6 208

1 6.

�d3-b1 !

White wants more than just a safe ad­ vantage , which could have been achieved with 1 9 .�a2 ! (by the way, to me this looks like obj ectively the best

3:

P a r a l l el H a n g i n g P a wns in t h e C en t r e

move here) 1 9 . . . �xc3 2 0 . 'iVxc3 ! tLlxa4 2 1 . �c2 �d7 2 2 .�b3 tLlb6 2 3 .'iVxcS and Black's dS pawn remains a prob­ lem that is impossible to solve. Note the huge difference that the presence or absence of Black's dark-squared bishop (placed on f6 , for example) makes in the assessment of these posi­ tions. 1 9.

...

ttJb6-c4

Another defence was 1 9 . . . �xc 3 2 0 .'ifxc3 (if 2 0 . axb 6 , then 2 0 . . . 1tb4 2 1 .�a2 �eB) 2 0 . . . tLlc4! (20 . . . tLld7 ? is not good due to 2 1 .�a2) 2 1 .e4! (Black would be doing well in the endgame in case of 2 1 .�a2 'iYxaS 2 2 .�xaS tLlxaS 2 3 .�xdS �fbB 24.tLleS �xdS 2 S .�xdS �xb2 2 6 . lixcs 1! b S ) 2 1 . . . 'iYxaS 2 2 .'ifc l \t>g 7 2 3 .exdS \i'b4 24.1ta2 tLld6 2 S .tLleS and White has the better chances. 20.

tLJc3xd5

The path that White chooses looks logi­ cal, but 2 0 .e4 ! may well have been better, for example : A) In the event of 20 . . . d4?, White is much better after 2 1 . tLla4 tLlxaS 2 2 . tLlxcS ; B) In the event of 2 0 . . . 'iYxa S ? 2 1 .tLlxdS ! (wrong is 2 1 .exdS ? because of 2 1 . . . �g 7 ! and Black has an excel­ lent , Grunfeld-type initiative) 2 1 . . . �xdS 2 2 J lxdS 'iYb4 2 3 .eS �g 7 2 4.�a2 , due to the open a2-gB and b I -h7 diagonals White obtains a win­ ning advantage; C) The prophylactic 2 0 . . . �g 7? is also not good, due to 2 1 . tLlxdS tLlxb2 2 2 . IId2 'iYxaS 2 3 .tLle7 + �h7 24.eS tLlc4 2 5 .tLlxg 6 ; D) 2 0 . . . �xc3 ! i s best for Black : 2 1 .'ifxc3 'iYxaS 2 2 . 'iYc l ! �g 7 2 3 .exdS 'iVb4 24.�a2 tLld6 209

Chapter

W i n n i n g C h e s s M i d d l e g a m es

analysis diagram

2 5 .ttJe5 and by transposition, we have reached a position discussed in a previ0us comment ( 1 9 . . . i1Lxc3 instead of 1 9 . . . ttJc4) . 20.

...

the consequences of his stranded knight on b2 : 25 . . J�b5 ! 2 6 .�e4 (obviously not 2 6 .!Ixb 2 ? ? �h5 and White is mated) 2 6 . . .'iV g4 2 7 .f3 ! 'li'xg 3 + 2 S .�g2 'tlVbS 2 9 .�xg6 � g 5 3 0 .�e4 �dS 3 1 .i1Ld5 ! and Black is lost due to the terrible position of his b2 knight; B) Black's best response is 24 .. J;Ib5 ! 2 5 .'li'c2 ! ? (White has a very slight ad­ vantage, far from enough for any realis­ tic winning chances, in the case of 2 5 . iVd4 'iYxd4 2 6 . �xd4 �xa5 2 7 .�e4) 25 . . . 'iVxf3 2 6 .�d6 ! ldfbS (not 2 6 . . J ih5 ? ? 2 7 .�xg 6 +) 2 7 .�a2 !:rh5 !

ttJc4xb2!

Black is not afraid of having his knight almost encircled and lost. 2 0 . . . �xd5 2 1 .e4 would be good for White. 21 .

ttJd5xf6+

'if d8xf6

22.

1:td 1 -d2

�b7xf3

23.

g2xf3

na8-b8

a n alysis d i agram

2 S .'iYxg 6 + ! �fS ( 2 S . . . WhS ?? 2 9 .'iYf6+ 'tlixf6 3 0 .z:txf6 �xa5 3 1 .�xf7 Wg7 3 2 J if4 ttJd3 3 3 .ldd4) 2 9 .�xf7 + �xf7 3 0 .�xf7 Wxf7 3 1 .�xa6 and with three pawns for a knight and an exposed black king, White should be better. 24.

f3-f4

It is quite possible that Kramnik overes­ timated his chances related to Black's misplaced knight on b 2. Very soon this knight will become a monster. White could have taken a pawn with 24.'iVxc5 , and now: A) 24 . . . �xf3 ? is bad because of 25 .'iVd4 and now Black is going to feel 210

e4-e5

J:td8xd2

27.

'ifc2xd 2

�f6-d8

28.

"iVd2-e2!

White has to keep the queens on the board, in order to have attacking poten­ tial against the black king to compensate for Black's domination on the queenside. 2 S .'iYc3 ? would be a blunder in view of 2 S . . . l:f.b3 ! . 28.

2 3 . . . c4 ?? blunders a knight due to 24.'tlVxb2 c3 2 5 .'iVa l ! .

24.

26.

...

c5-c4

Now the black knight will be well sup­ ported on d3 . Also, the pawn on c4 is very useful as it closes the a2-gS diagonal. 25.

e3-e4

�f8-d8!

It is useful for Black to exchange a pair of rooks. 25 . . . c3 ? would be a losing mistake because the pawn on c3 would be difficult to defend and the a2-gS di­ agonal would be re-opened for the white bishop: 2 6 .e5 �c6 2 7 .ldd6 'iYc7 2 S . l:ic l l:IfcS 2 9 .�a2 Wg 7 3 0 JH6 .

...

'ifd8-d4!

Faced with a choice between greed and centralization, Black correctly opts for the latter. 2 S . . . 'iVxa5 would give White a strong initiative after 2 9 . e6 (or 2 9 .f5 ) 2 9 . . . 'it' d 5 3 0 . e 7 'li'b5 3 l . f5 gxf5 3 2 .�xf5 . 29.

e5-e6

Also possible was 29 .f5 . 29.

...

ttJ b2-d3

An interesting possibility was 29 . . . f5 ! ? in order to control the light squares and limit the white bishop.

3:

P a r a l l el H a n g i n g P a wn s i n t h e C e n t r e

30.

�b1 -c2

White wants to bring his bishop back into play and to create motifs with �a4. 3 0 .�d l , pinning the knight, was proba­ bly better, since 3 o . . . lldS (or 3 o . . . llb2 3 1 .�c2) 3 1 . exf7 + ( 3 1 . e 7 �eS 3 2 .�xd3 cxd3 3 3 .lixd3 I:rxe7 34Jlxd4 �xe2 35 . .a:d6 is a draw in view of 3 5 . . J�re6) 3 1 . . .
...

f7xe6?!

There was no need at all for Black to al­ low the white queen to enter. Natural and good was 3 0 . . . neS 3 1 .exf7 + ( 3 1 .f5
'iVe2xe6+ �g8-g 7

32.

�c2-a4

Again, attacking with the queen only is wrong: 3 2 .f5 ? gxf5 3 3 .'tlixf5 ? ? ldfS , while 3 2 .'iYxa6 ldb2 is likely to lead to a draw: 3 3 .hd3 cxd3 34.'li'e6 .tIxf2 3 5 .�xf2 d2 3 6 .'iYe7 + WgS 3 7 .'iYeS+ �g 7 . 32.

...

c4-c3??

Probably in time pressure, Black wants to make sure that he is 'on time' with his counterplay Unfortunately, White now regains light-square control and his bishop becomes a monster again. A) 3 2 . . . ttJxf2 ? ? would be a blunder after 3 3 . .a.xf2 'iY a 1 + (or 3 3 . . J� b 2 3 4 .'li'e5+) 3 4.<;t>g2 'iYxa4 3 5 .'iYe5 + ; 211

Chapter

W i n n i n g C h es s M i d d l eg am e s

B ) Keeping the white bishop under control with 3 2 . . . ttJcS ! was the way to proceed. After 3 3 .'iVe7 + �g8 34.�c2 (34.�e8 ?! ldb 7 ! 3 S .'iVh4 'ife4 3 6 .'iYd8 �g7 brings White in trouble) 34 . . . ttJd3 , it seems that White has nothing better than to repeat moves or enter the line with 3 S .'iYe6+ �g7 3 6 .'iYxa6 l:ib2 , which is also likely to lead to a draw. Now this interesting and well-played game comes to an abrupt end.

33.

QO 7 . 1 0 (DS 8)

�a4-c2!!

This 'small' retreat must have been completely missed by Black. Now, the knight on d3 , which was a well-de­ fended, dominant piece only one move earlier, becomes a terrible liability. Black loses in one move from a bal­ anced position by pushing his ed pawn one square closer to promotion, according to regular chess principles. Logic sometimes follows strange rules in chess. 33 .

...

RbS-b2

A desperate attempt at counterplay. Or 3 3 . . J:rf8 3 4.'ifxa6 . 34.

'iYe6-e7+ �g7-gS

3S.

�e7-e2!

Now Black loses material. 3S.

ttJd3xf4

36.

g3xf4

wgS-f7

37.

llf1 -d 1

1 -0

Structure 3 .4

GAME 4 0

Joel Lautier Nigel Short Pamplona 1 9 9 9 / 0 0 (8) 1.

d 2-d4

e7-e6

2.

ttJ g 1 -f3

ttJgS-f6

3.

c2-c4

d 7-dS

4.

ttJ b 1 -c3

�fS-e7

S.

�c1 -gS

h 7-h6

6.

�gS-h4

0-0

7.

e2-e3

b7-b6

S.

�f1 -d3

�cS-b7

9.

0-0

ttJbS-d7

1 0.

�h4-g3

c7-cS

11.

'iVd1 -e2

ttJf6-e4

1 2.

c4xdS

e6xdS

1 3.

l::t a 1 -d1

For a number of different lines and ex­ planations about this popular variation, see Vyzhmanavin -Beliavsky (Game 2 8 ) and Kramnik-Yusupov (Game 3 9) . 1 3.

...

'iVdS-cS!?

If I am not mistaken, this interesting idea was invented by Short and introduced into practice in this game. Black wants to put his d7 knight on f6 and be able to re­ capture on cS with his b6 pawn (which is why he moves his queen away from the d-file pin) . Also, after . . . ttJdf6 the black queen can be transferred to e6, which is one of the best squares for the queen in these positions.

3:

P a ra l l e l H a n g i n g P a wns i n t h e C e n t re

1 4.

�d3-b1 ? !

White wants to exert pressure on the dS pawn and set up a battery on the b 1 -h7 diagonal, but both these motifs will not have any practical significance in this game. White cannot try for the upper hand/ symmetrical pOSItIOn with 1 4 .�xe4? dxe4 1 5 .ttJeS , since after l S . . . ttJxeS 1 6 . �xe S , Black has 1 6 . . . �a6 ! (thiS is also one of the pluses of 1 3 . . . 'iYc8) 1 7 .ttJbS 'iYe6 ! 1 8 .a4 'iYb3 and White is in trouble. Since the game move ( 1 4 . �b 1 ) brought White nowhere, two other moves have since become the main continuations : A) 1 4 .a4. The idea is to slow down Black's potential queenside advance (should he ever play . . . cS -c4) and also wait for Black's . . . ttJdf6 in order to jump to eS with the f3 knight. I doubt that, obj ectively, this plan can yield White any advantage, for example : A I ) 1 4 . . . ttJdf6 l S .ttJeS ! 'iYe6 1 6 .f4 (in case of 1 6 .f3 ttJxg3 1 7 .hxg3 �d6 1 8 .f4, White would be half a tempo down compared to the game) 1 6 . . . cxd4 1 7 .exd4 ttJxg 3 1 8 .hxg3 .

analysis diagram

White's strategy has been successful, since the g4-gS push will indeed be un­ pleasant for Black: 1 8 . . . �b4 1 9 .94! 212

213

Chapter

W i n n i n g C h es s M i d d l eg a m e s

�ae8 2 0 .gS hxgS 2 1 .fxgS tLle4 2 2 .�xe4 (White wants more than just an extra pawn in the endgame with heavy pieces after 2 2 .tLlxe4 dxe4 2 3 .�xe4 �xe4 24.iVxe4 �d6 2 S J:rde l �xeS 2 6 .dxeS ) 2 2 . . . dxe4 2 3 .'iYhS �xc3 (if 2 3 . . . e3 ?, then 2 4. �f4! e 2 2 S .ldh4 exd l 'iV+ 2 6 .tLlxd l and Black is mated) 24.bxc3 e3 2 S .�de l and White soon got a winning advantage in Sasikiran­ Kasimdzhanov, Skanderborg 2 0 0 3 . Despite White's success in this game, 1 4.a4 has rarely been repeated. It was perhaps better for Black to also make a useful move, waiting with his knight transfer and keeping it on d 7 , where it controls the e S -square. A2) In reply to 1 4 . . . a6 ! ? White still cannot play for the 'upper-hand sym­ metry' with I S .�xe4? ! dxe4 1 6 .tLleS , since Black will have a good game after 1 6 . . . tLlxeS 1 7 .�xeS 'iYe6 . B) The most frequently tried move so far has been 1 4 .�c l tLldf6 I S .dxcS ?! (the creation of hanging black pawns in the centre is not a good idea for White here. As can be seen from the present instructive example and also from our main game, Black obtains excellent piece play - the bishop pair, also a little more space, and White is very far from able to seriously target Black's pawns on cS and dS . It was better for White to maintain the status quo in the centre and play I S .tLleS ! , following up with plans similar to the above-mentioned game Sasikiran-Kasimdzhanov, with good chances to get an opening advan­ tage) I S . . . bxcS 1 6 .�fd l 'iVe6 1 7 .�h4 . Probably the white player thought that he was being smart and that the bishop on h4 would be annoying for Black. I have chosen to show this game because 2 14

Black now demonstrates an attacking plan which can often be used in posi­ tions like this and his execution is fine, which makes this game well worth ing.

while Black executes his plan of a mat­ ing attack along the h-file. It was high time for White to organize active counterplay with 2 3 . �e 1 ! , threatening e3 -e4, and if 2 3 . . . tLle4, then 24.tLldxe4 dxe4 2 S .�c4 'iYeS 2 6 .lded l with ap­ proximate equality) 2 3 . . . 'iYeS 24.�bS hS 2 S . tLlf4 Ith8 ! . Now Black's plan has been executed. The coming . . . h S -h4 cannot be stopped. White was now definitely fully aware of his problems and he shuffles his knight in order to try and prevent immediate annihilation - without suc­ cess : 2 6 . 'iVe 2 h4 2 7 .tLl£1 �d6

analysis d i agram

1 7 . . . gS ! 1 8 .�g3 �fd8 1 9 .tLld2 ! (cor­ rectly forcing Black to make a decision concerning his e4 knight) 1 9 . . . tLlxg3 2 0 .hxg3 g4! (fixing the white kingside and getting ready for a direct attack) 2 1 .b3 �g7 ! .

an alysis d i agram

ana lysis d i agram

It becomes clear that Black will try to open the h -file and create mate threats. White needs to find counterplay quickly: 2 2 .iV£1 (2 2 .�e l , preparing e3 -e4, was an option) 2 2 . . . aS 2 3 . tLle2 ? (White is obViously not aware of the trouble he is in and comes up with a se­ ries of useless , time-wasting moves,

2 8 .�d3 ? (in a difficult position, this new tempo loss definitely doesn't help. The counterplay attempt with 2 8 . �d7 ! 'iYgS 29 .'iYbS had to be considered) 2 8 . . . 'iYgS 2 9 .'iYd2 'iYh6 3 0 .e4 (White's first active move is far too late) 30 . . . hxg3 3 1 .fxg3 c4 ! 3 2 .�b l �ae8 3 3 .tLle3 �cS and White resigned in Sargissian-Halkias, Antalya Ech 2 0 04. The way Sargissian suffered this terrible defeat in this game is not that uncom­ mon for white players in such positions. Rather often, the side playing against hanging pawns in the centre, instead of concentrating on the e3 -e4 pawn break or straight pressure on Black's cS or dS

3:

P a r a l l el H a n g i n g P a wns i n the C en t re

pawns (mostly cS) makes the mistake of believing that a senseless shuffling of pieces will for some reason make one of Black's hanging pawns weak by itself, and will force Black onto the defensive. In the meantime, Black will naturally use his space advantage to either prepare a successful . . . dS -d4 break - as in our main game Lautier-Short - or - as in Halkias's case - a kingside attack. Black can also, for the time being, keep his knight on d7 in order to prevent White 's planned tLlf3 - e S j ump : 1 4 . . . tLlxg3 I S .hxg3 a.e8 1 6 .�fd 1 c4 (another option is to keep the tension a little longer with 1 6 . . . tLlf6 1 7 .tLleS �d6 , since after 1 8 .f4, compared to the game Sasikiran-Kasimdzhanov, the white rook is misplaced on d 1 and it will cost White time to bring it back to £1 ) 1 7 .�b l tLlf6 1 8 .tLleS �d6 1 9 .f4 a6 2 0 .g4 bS with a complicated game in Riazantsev-Onischuk, Sochi 2 0 0 8 . Playing a useful developing move and keeping all options open with 1 4 . . J�d8 is the most flexible for Black: 1 5 .lIfd 1 tLldf6 1 6 . tLleS 'iY e6.

a n a lysis d i agram

Since White's £1 rook has moved to d 1 and is misplaced for ing f2 -f4, White now opts for a different plan : 1 7 .�h4 �ac8 1 8 .f3 cxd4 1 9 .exd4 215

Chapter

W i nn i n g C h es s M i d d l eg a m e s

ttJxc3 2 0 .l::r x c3 l';Ixc3 2 1 .bxc3 ttJd 7 ! (Black keeps exchanging pieces and with limited material left, it will be dif­ ficult for White to take advantage of his better-placed light-squared bishop and create activity on the kingside) 2 2 .ilxe 7 'ti'xe 7

analysis d i agram

2 3 .ttJxd 7 (this leads to an immediate draw. In order to keep some life in the position and to show aspirations, White had to continue with 2 3 .f4) 2 3 . . Jlxd 7 24J:Ie l 'iYxe2 2 S .l;Ixe2 wfS 2 6 .J:I.e3 l::l e 7 27 .�f2 �xe3 2 S . �xe3 cJ;)e 7 draw, Bacrot-Ivanchuk, Odessa A Grand Prix 2 0 0 7 . 14.

...

tb d7-f6

1 5.

�e2-c2

l::US -eS

Black has a number of useful moves and he plays them correctly, keeping the tension in the centre.

This move does not fit in any plan and the bishop will only be clumsy on e S . I am not fond of the plan with ilb 1 1'iY c2 and I think that White has to be careful not to land in a worse position here. Probably some central exchanges were needed : 1 6 .dxcS bxcS 1 7 . ttJxe4 ! ttJxe4? ! (the attempt to keep his central pawns connected may bring Black into trouble. The game is about equal after 1 7 . . . dxe4 1 8 .ttJeS 'ife6) l S .ttJd2 ! fS 1 9 .ttJxe4 fxe4 2 0 .f3 ! ilgS 2 1 .fxe4 ilxe3 + 2 2 .ilf2 ilxf2 + 2 3 .'iYxf2 d4 ( 2 3 . . . dxe4 ? ? is a huge blunder due to 24 .iVf7 + �hS 2 S .�d7) 24.b4! ila6 (24 . . . cxb4? ? is a terrible blunder due to 2 S .'iYf7 + \t>hS 2 6 .nc l 'ifbS 2 7 .l::r c 7) 2 S .nfe l and White has some initiative, for example : 2 S . . . cxb4 2 6 .'iYxd4 'iVc3 2 7 .'iVdS + \t>hS 2 s .iVd6 ilb7 2 9 .e S . 1 6.

216

�g3-e5?

'ifcS-e6

1�

d4xc5

b6xc5

A position with parallel hanging pawns in the centre has been reached, where it is difficult for White to exert any pres­ sure on Black's pawns on dS and cS , while it is rather easy for Black to keep improving his pieces. It is probably fair to say that Black is slightly better here. 1 S.

1 6.

...

The queen arrives at its natural position and underlines the clumsiness of the bishop on eS .

1 S.

...

tbe4-g5!

Conquering the bishop pair. 1 9.

tbf3xg5

20.

tbg5-f3

'iY e6xe5 'iV e5-e6

21 .

�f1 -e1

iLe7-fS

Black's cS and dS pawns are well-de­ fended and his pieces are harmoniously placed, while White has to worry about the . . . dS -d4 pawn break. Also, Black has enough time to prepare this break and execute it at a favourable moment, since White has no meaningful counterplay. 22.

�a4-f4

'iYe6-b6

3:

P a ra l l e l H a n g i ng P a w n s in t h e C e n t re

due to 2 7 . . . ilxe4 2 S .ilxe4 d3 - the black d-pawn is strong, while White's pieces are very badly placed; the ile4 and 'iYf4 in particular - 29 .l::r d 1 (even worse is 2 9 .'iYfS ? after the simple 29 . . . g6 3 0 .'tWf4 �g7 and all Black's problems remain) 2 9 .. :iYb4 3 0 .ttJd2 nd4 3 1 .a3 (3 l .f3 ? loses to 3 1 . .. ttJdS 3 2 .'ifh4 ttJe3) 3 1 .. :iYc3 and Black can­ not prevent material loss , for instance: 3 2 .'ife3 ttJxe4 3 3 .ttJxe4 'ifxb3 etc. 27.

...

2S.

Ite1 -c1

2S .

...

'iVb6-c6

Also interesting was to try and lock out the white queen on f4 with 22 . . . ttJe4 ! ? , and if 2 3 . ttJxe4 (or 2 3 .'iVc7 �abS) 23 . . . dxe4 2 4 . ttJ d 2 ild6 ( 2 4 . . . c4 2 S .'ifc7 ildS is also better for Black) 25 .'ifh4 �adS 2 6 .b3 ilc7 2 7 .ttJc4 ildS 2 S J�c 1 fS and Black is better. The white bishop on b 1 is rather ive, while the queen is definitely misplaced on h4. 23.

b2-b3

naS-dS

Black focuses on preparing the . . . dS -d4 break. A standard plan like 23 . . . ild6 24.'ifh4 as is also good for Black. 24.

:td 1 -c1

�c2-a4

d5-d4!

24 .

...

25.

e3xd4

lleSxe 1 +

26.

�c1 xe1

c5xd4

27.

tbc3-a4!

Sadly, the knight has to j ump to the edge of the board. 2 7 . ttJe4? is not good

'iYc6-b5 1?

Setting a trap that White falls into. Better was the intermediary 2 S . . . ttJdS ! , allowing Black to improve his queen : 2 9 :�'d2 ( 2 9 .nxc6 ? is a blunder which loses quickly after 29 . . . ttJxf4 3 0 . l::r c 7 (or 3 0 J lc4 ..tdS ! ) 3 0 ... ilxf3 3 1 .gxf3 d3 3 2 .ttJb2 d2 3 3 .ilc2 neS 3 4.h3 IIe l + 3 S . �h2 ttJd3 ! 3 6 .ttJxd3 ild6 +) 29 .. .'iYf6 3 0 .�e4 ! ( 3 0 . 'ifxd4 ? ? blun­ ders a piece after 30 . . . 'ifxd4 3 1 .ttJxd4 ttJf4) 3 0 . . . ila6 3 1 . l::t d l (probably a better, or at least a more practical de­ fence is to eliminate the black knight with 3 l .ilxdS ! and after 3 l . . . l::r x dS play 3 2 .b4, with ttJcS to follow) 3 1 . . . ttJf4 and, with his knight alive, Black has plenty of annoying tactics. There follows : 3 2 . g 3 (3 2 . l::r e I ? is wrong due to 3 2 . . . d3 , with ttJe2 to fol217

Chapter

W i n n i n g C h e s s M i d d l eg a m e s

low) 32 ... ttJe2+ 3 3 .�g2 l':reS 3 4.'iVc2 ! (White's bishop has to stay on the h I -aS diagonal. A blunder is 3 4 .�d3 ? due to 34 . . . ttJc3 3 5 .ttJxc3 (or 3 5 .l':re 1 ? �b 7 ; or 3 5 .�xa6 ttJxd l 3 6 .�b5 l:lbS) 35 . . . dxc3 3 6 .'iYc2 �b 7 ) 3 4 .. .'ife 6 .

a n a lysis d i agram

A) Now White 's best defence is 3 5 .�h 7 + ! �hS 3 6 .�f5 ! 'iYd5 3 7 .�d3 and it seems as if he survives, for exam­ ple : 3 7 . . . ttJf4+ 3 S .gxf4 �b 7 3 9 .�e2 'iYe6 40.f5 ! 'iYxe2 4 1 . 'iYxe2 l':rxe2 42 . .ldxd4 Ilxa2 43 .l':rdS �gS 44.ttJc5 �c6 45 . l:lcS ! Ilc2 4 6 . l':rxc6 l':rxc5 4 7 .Ilxc5 �xc5 4S .ttJe5 . B) 3 5 .ttJd2 ? is worse : 3 5 . . . d3 (simple and strong is also 3 5 . . . �b5 ! , which prevents White's 'iYc6 and leaves Black with a big advantage) 3 6 .'iVc6 (3 6 .�xd3 ?? is wrong due to the simple 36 . . . ttJf4+ 3 7 .gxf4 'iVg4+ 3 S . �h l �xd3) and now:

an a lysis d i agram

218

Strongest is 3 6 . . . ttJf4+ ! 3 7 .�g l �b4! 3 S .'ifxe6 ttJxe 6 . Now the queens have been swapped, but the endgame is an improved ver­ sion for Black due to the knight check. His d-pawn is strong, his pieces are per­ fectly coordinated and the position is very difficult for White, for example: 3 9 .ttJc4 �xc4 40.bxc4 d2 4 1 .a3 �a5 , with . . . ttJd4 to follow, and Black should win. If Black plays 3 6 . . . �b4 ? ! immediately, after 3 7 . 'iYxe 6 .ldxe 6 White has good saving chances : 3 S .�d5 (another op­ tion is 3 s .f3 �b5 3 9 .ttJb l ! and if 39 . . . �xa4 4 0 . bxa4 d2 , White has 4 1 . ttJ xd 2 ! Il d 6 4 2 . l':r b I Ilxd2 4 3 . l':rxb4 f5 44 . Ilb S + �f7 45 .�b 7 , with good drawing chances) 3 S . . . l':rd6 and now:

3:

P a ra l l e l H a n g i n g P a wns in t h e C e n t r e

B2 1 ) Perhaps 4 1 .ttJb 1 can be consid­ ered, though White's position is rather ive and probably lost, since after 4 1 . . . �xa4 4 2 .bxa4 d2 4 3 . �f2 ttJc l (43 . . . ttJc3 44.ttJxc3 �xc3 45 .�e2) 44.a3 �a5

43 . . . ttJxe4+ ! 44.ttJxe4 f5 ! 45 .�e3 l:ld5 and Black wins. An unexpected denouement. The white knight is caught in the middle of the board: 46 .ttJd2 (or 46.ttJf2) 46 . . . �c5 + 4 7 .�f4 g5 mate. ana lysis d i agram

and now the only attempt at counterplay, 45 .�e3 , loses to 45 . . . �b6 46 .�c2 (46 . Ilxd2 ? ? is a blunder on ac­ count of 46 . . . Ild3 + 4 7 .�f4 �e3 +) 46 . . . l':rdS + 47 .�e4 �a5 4S . �e3 g 6 ! 49 .ttJxd2 ? �b6 + 5 0 . �e4 (or 5 0 . �f4 ttJe2 + 5 1 . �g4 ttJc3 ) 5 0 . . . f5 + 5 1 . �e5 �f7 5 2 .ttJc4 �c 7 + 5 3 .ttJd6+ Ilxd6 54.l':rxd6 �e 7 ; B 2 2 ) 4 1 . �f2 �xa4 42 .bxa4 ttJc3 43 .Ilc 1

29.

ttJf3xd4?

30.

�f4xd4

30.

...

l:id8xd4

iVb5-g5 !

And White reSigned.

analysis d i agram

B l ) Best is 3 9 .�c4 ! �b 7 + 40.f3 ttJd4 4 1 .ttJe4! (4 1 .�xd3 ?! does not solve the problems after 4 1 . . . g 6 ! (to prevent the discovered check �h7 +) 42 . �f2 �xd 2 43 .Ilxd2 ttJxf3 ) 4 1 . . . d2 4 2 .�f2 �xe4 43 .fxe4 and White has good chances to hold; B2) 3 9 .�e4? proves to be a decisive mistake : after 3 9 . . . l':rd4 40.f3 �b5 ! the knight on a4 is an important defender, which controls the c3 -square : 219

Chapte r 4 Pawn M ajority i n the Centre I ntroduction In this chapter I have systematized the structures and games slightly differently com­ pared to the previous three chapters, since the various thematic ideas for both sides often must be considered and finally chosen, depending on the peculiarities of the position. Therefore, 7 structures are given on the following pages, related to the thematic ideas. Five commented games then follow, and often, in one game or its comments, we encounter different thematic ideas. The reader may find that I am perhaps a little biased, favouring the side playing with the pawn majority. This opinion could be correct, given the fact that I am a l .d4 player, and throughout my career I have mostly played on the side with the pawn ma­ jority in the centre. For a club player, apart from the fact that he will enhance his understanding of thematic ideas , the five selected games also offer an excellent guide to the opening variations featured, and if you take time to study the material seriously, the knowl­ edge you acquire can be very valuable in tournament play. Game 4 1 (Sky-Petrosian) is a guide to one of the main lines of the Semi-Tarrasch, featuring classic and contemporary games. Game 44 (Sokolov-Howell) explains an interesting sideline of the Griinfeld In­ dian and includes some previously unpublished analysis of this game by the author. The games 42 (Kasparov-Najdorf) , 43 (Galliamova-Tiviakov) and 45 (Sokolov-­ Polgar) all feature the Petrosian Variation of the Queen's Indian, a variation that the young Kasparov built a considerable part of his success on almost 30 years ago. This particular variation has also served me very well. In the comments to these games I give the relevant strategic plans , important games, personal opinions and analyses.

Structure 4. 1 is a regular starting position with a pawn majority in the centre.

4.1

22 1

Chapter

W i nn i n g C h ess M i d d l eg am e s

Structure 4.2 relates to White's plan to build a kingside attack with e4-e5 , explained in the comments to Game 4 1 (Sky-Petrosian) , and in the comments to 1 1 . . . ttJd7 (instead of the game move 1 1 . . . ttJc6) - the games Lobron-C.Hansen and Radjabov-Van Wely in par­ ticular.

Structure 4. 3 relates to the regular central pawn push d4-d5 that White often executes in these positions (and Black mostly takes . . . e6xd5 ) .

4.2

4.3

4:

P a wn M aj o r i t y in the C e n t re

dark-squared bishop, opening up Black's position and creating a strong attack along the a 1 -h8 diagonal. Game 42 (Kasparov-Naj dorf) is a more classic example. Kasparov's victory was swift, but the analyses given show that Black had plenty of defensive resources. Structure 4. 7 shows a nice attacking example from my own practice - in fact, this was one of the best games I have ever played. White makes the standard d4-d5 pawn sacrifice, and after . . . e6xd5 proceeds with e4-e5 , with a very strong knight on d4, making maximal use of the cre­ ated outpost, Black's f-pawn being on f5 here.

Structure 4.4 relates to the white plan to push d4-d5 and then create a strong ed d-pawn - see Game 4 1 (Sky -Petrosian) .

Structure 4.5 relates to the white plan of sacrificing his d-pawn, and, after d4-d5 e6xd5 , proceeding with e4-e5 to build a kingside attack on the open diagonals a 1 -h8 , b 1 -h7 and c 1 -h 6 , a plan which is nicely illustrated in Game 43 (Galliamova-Tiviakov) . The other aspect of this pawn sacrifice is that White gets a beautiful attacking outpost for his knight on d4 , as featured in Games 42 (Kasparov-Naj dorf) , 44 (Sokolov-Howell) and 45 (Sokolov-Polgar) . Structure 4.6 relates to a double pawn sacrifice. After al­ ready sacrificing his d-pawn (as mentioned under Struc­ ture 4. 5 ; d4-d5 e6xd5 and White proceeds with e4-e5) White proceeds with ttJf3 -d4, sacrificing his e5 pawn in order to win some extra time and launch an attack against the black king. In Game 44 (Sokolov-Howell, a Griinfeld Indian Defence) , Black's pawns were on g6 and f5 and White exchanged his knight for Black's fianchettoed 222

223

C h a p t e r 4: P a wn M aj o ri t y i n t h e C e n t r e

W i n n i n g C h es s M i d d l eg a m e s

TD 2.9 (D4 1 )

GAME 4 1

Boris Sky Tigran V Petrosian Moscow World Championship 1 9 69 (5) 1.

c2-c4

ttJg8-f6

2.

ttJ b 1 -c3

e 7-e6

3.

ttJg1 -f3

d 7-d5

4.

d2-d4

c7-c5

5.

c4xd5

ttJf6xd5

6.

e2-e4

ttJ d5xc3

7.

b2xc3

c5xd4

8.

c3xd4

within the limited space available to him. The other option for Black is to slightly delay the bishops swap and make use of a tactical element to aim for quicker de­ velopment, since after 8 . . . ttJc6 9 .�c4 Black has the tempo move 9 . . . b S . Prac­ tice has shown that, with correct play by White, it is questionable whether Black, who is slightly underdeveloped, is ready for such an active undertaking. The following game is a good example.

a n alysis d i agram

One of the main positions of the Semi-Tarrasch, and a perfect one for the study of pawn maj orities in the centre. 8.

...

iil.f8-b4+

The pawn structure we have here is similar to the Petrosian Variation of the Queen's Indian, which we will see in the games Kasparov-Najdorf, Gallia­ mova-Tiviakov and Sokolov-Polgar fur­ ther on in this chapter. Compared to that variation, Black has a slightly more convenient situation here, since he can immediately force the ex­ change of the dark-squared bishops. Given the fact that White has consider­ ably more space, it is generally quite useful for Black to exchange his dark-squared bishop for its white coun­ terpart, in order to make it easier for himself to manoeuvre his pieces , 2 24

A) 1 0 .�d3 may look active, but promises nothing after 1 0 . . . �b4+ 1 1 .�d2 �xd2 + 1 2 .'iYxd2 a6 1 3 .a4 O - O ! 1 4 . 'if c3 (or 1 4 .axbS ttJxd4) 1 4 . . . jLb 7 I S .axbS axbS 1 6 .0 -0 (the bS pawn is taboo, since after both 1 6 .�xbS ? �xa l + 1 7 . 'iYxa l iYb6 and 1 6 J ixa8 'ifxa8 1 7 .�xbS ? ttJa7 , Black regains his pawn, with a better game) 1 6 . . . 'iYb6 1 7 J lab 1 b4 and now: A I ) 1 8 . 'iV d2 leads to a forced draw: 1 8 . . . ttJxd4 1 9 .ttJxd4 'iYxd4 2 0 J 'rxb4 'iVd7 and the draw was soon agreed in the 9th match game Sky-Fischer, Reykj avik 1 9 7 2 . Fischer used the Semi­ Tarrasch in a number of important games. He believed in Black's counter­ play; A2) The active attempt 1 8 .dS also leads to a peaceful outcome after

1 8 . . . exdS 1 9 .exdS bxc3 (getting 'ambi­ tious' with 1 9 . . . 'if as ? is a blunder due to 2 0 .�xh7 + \t>xh7 2 1 .�c 2 + with dxc6 to follow, and Black will be a pawn down - in the best -case scenario for him) 2 0 J :1xb6 ttJaS 2 1 . d 6 jLxf3 2 2 . gxf3 �ac8 2 3 .Itc l ttJc4. B) 1 0 .�e2 ! (the bishop goes back to e2 in order for White to be ready for the d4-dS push) 1 0 . . . �b4+ 1 1 .�d2 'iYas (one of the problems of the early . . . b7 -bS is that, after the initial tempo gain, Black has to lose time defending his bS pawn. Following in Fischer's footsteps with 1 1 . . . �xd2 + 1 2 .'ii'x d2 a6 would not work here, since White now has 1 3 .dS ) 1 2 .dS ! (this typical pawn thrust is almost by definition un­ pleasant for Black in these structures) 1 2 . . . exdS 1 3 .exdS ttJe7 1 4. 0 -0 �xd2 1 S . ttJxd2 0-0 1 6 .ttJb3 'iYd8 . B 1 ) Now White can aim for more than a small advantage with 1 7 .ii.xbS �b 7 1 8 .d6 ttJfS ( 1 8 . . . ttJg6 ! ? is an op­ tion) 1 9 .ttJcS ( 1 9 .d7 ? 'iVgS) 1 9 . . . ttJxd6 2 0 .ttJxb7 ttJxb 7 ; B 2 ) 1 7 . �f3 ttJfS (Since Nimzo­ witsch, the knight is known to be the best ed pawn blocker, so the black knight goes to d6) 1 8 . lic l ttJd6 .

ana lysis d i agram

Sadly for Black, pawns cannot go back. If he had his b-pawn on b7 instead of

bS , he would be equal at least. Now, with the c6-square weak, it is a differ­ ent story. White is better due to the fact that Black cannot prevent his knight from entering on c6. 1 9 .'ifd4 'ifb6 20.'iff4 �d7 2 1 .ttJd4 �fe8 2 2 .ttJc6 ttJc4 2 3 Jife l and due to the excellently-placed knight on c6, White had a clear advantage in Yusupov­ Ribli, Montpellier Candidates' 1 9 8 5 . iil.c1 -d2

�b4xd 2+

1 0.

9.

'iYd1 xd2

0-0

11.

iil.f1 -c4

Since this is a one of the typical posi­ tions with a pawn majority in the cen­ tre, this is a good moment to take stock and try to explain some general plans. Black is , per definition, going to de­ velop his bishop to b 7 , where it will be actively placed, hitting the white centre. Black's a8 rook will be developed to c8 , where normally White leaves it unop­ posed , since rook exchanges are gener­ ally good for Black. The black knight is sometimes - as in this game - developed to c6 in order to hit on White's d4 pawn and to have ma­ noeuvres like . . . ttJc6-aS or . . . ttJc6-e 7 ; or t o d7 i n order not t o step into a d4-dS tempo. Queen swaps and other types of ex­ changes are generally good for Black here. White is supposed to organize his forces behind a mighty pawn centre and either execute the d4-dS thrust in order to get a ed pawn, or sacrifice the d4 pawn in order, after d4-dS e6xdS , to push e4-eS , obtain an outpost on d4 for his f3 knight and get attack­ ing prospects on the black king. If the black knight is developed to d 7 , White also has a possibility to leave his 225

Chapter

W i n n i n g C h es s M i d d l e g a m es

pawn on d4 and to push e4-eS in order to limit this d7 knight and seize some extra space to organize an attack on the kingside. The fact that White has sur­ rendered the central d S -square is mostly less relevant than it may seem. White should generally try to avoid ex­ changes and keep as many pieces on the board as possible. These are the general plans in the posi­ tions with a pawn majority in the cen­ tre. They also apply, as we shall see, to the Petrosian Variation of the Queen's Indian, where both sides have dark­ squared bishops on the board. I shall give a sufficient number of top-class practical examples to explore the above-mentioned plans.

11.

..

.

black bishop is strong. Moreover, Black has two pawns versus one on the queenside, which offers him winning chances if the white attack is stopped. The position is very unclear) 1 9 . 1ih3 ! 'tWe7

2 0 .ttJh7 ! l:Ic7 (if 2 0 . . J lfd8 , 2 1 .�xh6 ! gxh6 2 2 .'iYxh6 wins) 2 1 .c1Jxf8 - White was a clear exchange up and went on to win in Lobron-Cu. Hansen, Wijk aan Zee 1 9 9 3 . As we have seen, Black's 'beautiful' knight on dS was rather ir­ relevant in the whole story. A2) 1 4 . . . :rc8 .

a n a lysis d i agram

1 7 .eS ! ttJdS 1 8 .c1JgS h6 ? (a crucial mis­ take. Black needed strong nerves and had to go for 1 8 . . . ttJxe3 ! 1 9 .�xh 7 + �h8 2 0 . 'iVxe3 g6 (20 . . . 'tWc3 ? is bad due to the simple 2 1 .std3 ) 2 1 .stxg6 (it could be that White has something better here) 2 1 . . . fxg6 2 2 .ttJxe6. White would have three pawns for a piece, but his attack runs out of steam and the

P a wn Ma j o r i t y in t h e C e n t r e

Lobron-Cu.Hansen, continuing with 1 S .std3 . As we are about to see, the sur­ render of the dS -square (which is part of the plan) did not play any Significant role in this game either : 1 S . . . h6 1 6 .j!"b l �e8 1 7 J !e3 (as in Lobron­ Hansen, the rook transfer via the third rank is an essential part of the plan) 1 7 . . . 'iYc7

analysiS d i agram

tLlb8-c6

Black can also first develop his bishop to b 7 , and later decide where to put his knight, by opting for 1 1 . . . b6 ? ! . This move, however, has the drawback that it allows the immediate 1 2 .dS ! (one of the reasons behind 1 1 .�c4) . White seizes a lot of space and has the better chances after 1 2 . . . �a6 ! (the c4 bishop has to be exchanged) 1 3 .stxa6 c1Jxa6 1 4.d6! c1JcS I S .eS , as in Kortchnoi­ Mecking, Hastings 1 9 7 1 / 7 2 . So i t seems that due to the possibility of d4-dS , it is better for Black to first de226

velop his knight, and modern theory shares this view. I l . . . c1Jd7 is, as mentioned, another main alternative for Black here. Now the vast majority of games have contin­ ued with the almost compulsory 1 2 .0-0 b6 1 3 J�Je l stb 7 1 4Jlad l . A) White has a rather interesting plan in 1 4.�e3 ! ? As far as I know, this move has been played only once at top level, but it probably deserves more attention. White wants to execute the aforemen­ tioned e4-eS plan, surrendering the dS-square and obtaining a kingside at­ tack in return. Let's see how this worked out in practice : A I ) 1 4 . . . lIc8 I S .std3 (in the plans with e4-eS , this bishop is needed on the b l -h 7 diagonal) I S . . . 'iYc7 1 6 .I:rae l ttJf6 ? ! . Playing into White's hands. Better was 1 6 . . J lfd8 . Now White can nicely execute his plan :

4:

a n a lysis di agram

Now White has to decide on an attack­ ing plan to proceed with. A2 1 ) White can retreat his bishop to the b 1 - h7 diagonal and play for e4-eS , attacking on the kingside, similar to

a n a lysis d i agram

1 8 . e S ! . At first Sight it may look a little surprising that White develops a strong kingside attack here. After all, the diagonal of the black bishop is now open, the dS -square is firmly under black control - one might even argue that White is effectively a pawn down -, any kind of ending would be more or less winning for Black, and White still does not have any threats for the moment. Black's problems, on the other hand, are that his b 7 bishop indeed looks wonderful, but does not create any threats at the moment ; his knight on d7 and, later, f8 is ive ; it is difficult to create counterplay, Black 's forces are cut off from the defence of the kingside, and therefore Black will be forced to play . . . f7 -fS at some stage, which will make his e6 pawn weak af­ ter exf6 . Also, for Black it is difficult 227

C hapter

W i n n i n g C h ess M i d d l eg a m e s

practically almost impossible - to ex­ change any pieces. 1 8 . . . ttJf8 1 9 . ttJe l ! (the white knight goes via e 1 , d3 and f4 to its perfect at­ tacking post on h S ) 1 9 .. .'iVe7 2 0 J lg 3 'iVh4 2 l .h 3 ! (White threatens with ng4, followed by the execution of the knight transfer ttJd3 -f4-hS . Then, the only black piece defending the king will be his f8 knight, so Black will be forced to involve other pieces in the defence of his king by playing . . . f7 -fS . As said, this creates a weakness on e6 and still doesn't solve the problems on his kingside) 2 1 . . . fS ! 2 2 .exf6 �xf6 2 3 .ttJ d3 �ed8 2 4 .ttJf4 (the knight continues its j ourney to the desired square) 24 .. J ics (this defence will not work, but the position was already difficult for Blac k . If 2 4 . . . �c 7 , 2 S . �g4! ( 2 S .ttJhS ? is a mistake due to 2 S . . . 'iVh4 and now 2 6 . 'iVxh6 ? ? is a of bec ause blunder horrible 26 . . . 'iVxg 3 2 7 .ttJf6 + �f7 ) 2S . . . eS 2 6 .ttJhS �e6

a n a lysis d i agram

2 7 .iLc2 ! (time to move the bishop to the a 2 - g 8 diagonal) 2 7 . . . �h8 (2 7 .. J::txd4? ? is a blunder in view of 2 8 . l:rxd4 exd4 29 . .ii!. b 3 iLdS 3 0 .ttJf4) 2 8 .iLb3 �c6 2 9 .dS and White is clearly better) 228

tioned plans with his central pawns) 2 1 . . . exdS 2 2 .eS (this type of d-pawn sacrifice is standard here and will be seen in many games in this chapter)

analysis di agram

2 S .dS ! (cutting off Black's defence of the hS-square) 2 S . . . gS (the kind of move that Black was definitely not happy to play) 2 6 .h4! J:tc4 2 7 .ttJhS �eS 2 8 .hxgS ld:xdS 2 9 .ttJf6+ and Black reSigned in Radjabov-Van Wely, Khanty-Mansiysk 2 0 0 S . It is good to notice that Van Wely suffered a rela­ tively straightforward loss without making any obvious mistake. The at­ tacking plan demonstrated by Radjabov (and also seen in Lo bron -Hansen) is per definition more dangerous than it may seem, while the importance of the control of the d S -square is less signifi­ cant than it looks. A2 2) I S .iLb3 (aiming for a d4-dS push) 1 S . . . ttJf6 (the featured game was played in 1 9 3 8 . The text is definitely logical and OK, but modern theory likes the f6-square to be kept for the black queen; thus, in one of his games Ivanchuk played I S . . . �f6 1 6 .h3 �fd8) 1 6 .'iYf4 'iVc7 1 7 .�h4 l:tfd8 1 8 J�e3 bS 1 9 .�de l as 2 0 .a4 b4 (20 . . . bxa4 was also possible. White now uses one of the well-known standard plans : ) 2 1 .dS ! ? (it should b e noted that Black generally continues to improve his po­ sition, whereas White will sooner or later be forced to take a decision in the centre, choosing one of the above-men-

analysis d i agram

2 2 . . . ttJd7 ? (better was 2 2 . . . ttJe4! , when the white attack may look dangerous, but Black can successfully fight it off: 2 3 J�rxe4 (or 2 3 .e6 �f8 ! (not 2 3 . . .fxe6 ? due t o 2 4 . �xe4 dxe4 2 S . ttJ g S ) 24. exfl + \th8 !) 2 3 . . . dxe4 24.ttJgS h6 ! (wrong is 24 . . . �c3 ? due to 2 s .iLxfl + \tf8 2 6 . �f1 ) 2 s .iLxfl + (2 S .ttJxfl ? loses to 2 s . . . iLdS ) 2 S . . . �h8 and Black is better, for example : 26 . .ii!.e 6 I::f. f8 2 7 .iLxc8 �xc8 2 8 . e 6 b3 ! ) 2 3 .ttJgS ( 2 3 .ttJd4 would have been a typical fol­ low-up, according to the positional rules of the standard plan with the d-pawn sacrifice. After Black takes exdS and White proceeds with e4-eS , he will get an excellent outpost for his knight on d4) 23 . . . ttJf8 ? (a crucial mistake. Good nerves were needed, and Black had to opt for 23 . . . h 6 ! 24.e6 hxgS 2 S . exf7 + xf7 2 6 . ne 7 + �g 6 ! 2 7 . nxg 7 + ( 2 7 . � l e6 + ? ? loses to 2 7 . . . ttJf6) 2 7 . . . �xg 7 2 8 .�xgS + �h8 2 9 .�h6+ and White has to give per­ petual check) 24.ttJxh7 ! ttJxh7 2S . l::lh 3 and White soon won in Keres-Fine, Os tend 1 9 3 8 .

4:

P a wn M aj o r i t y i n t h e C e n t r e

So I think, also for the sake of getting to know some theoretical aspects of this variation, that it is fair to say that I s .iLd3 , as played in Radj abov-Van Wely, is probably more dangerous for Black. 1 2.

0-0

1 3.

l:la1 -d 1

b7-b6

It took white players some time to reach the conclusion that the rooks should be put on d 1 and e 1 here. The 1 8th World Championship match game Alekhine­ Euwe ( 1 9 3 7 ) continued 1 3 .nfd l iLb 7 1 4 .�f4 ldc8 (nowadays, 1 4 . . . 'iYf6 ! would have been considered a standard response)

analysis d i agram

I S . dS ! exdS 1 6 .iLxdS 'iVe7 1 7 .ttJgS

analysis di agram

1 7 . . . ttJ e S ! 1 8 .iLxb 7 ttJ g 6 1 9 .iYfS 'iVxb 7 2 0 .�d7 'iVa6 (Black should be 229

C h a p t e r 4: P a wn M a j o ri t y in t h e Cen t r e

Wi n n i n g C h e s s M i d d l eg a m e s

able to hold after 2 0 . . Jdc7 2 l .�ad l l:Ixd 7 2 2 .�xd 7 'ii a 6 2 3 .h4 'iYxa 2 24.�h2 (or 24.h5 'iVa I + 2 5 . �h2 �e5 +) 24 . . . h6 2 5 .h5 hxg 5 2 6 .hxg6 fxg 6 2 7 .'iVe 5 l'lf7 2 8 .�e 8 + �f8 2 9 .'ifxg6 �fl) 2 l .h4 (if 2 l .�xfl , Black defends with 2 1 . . . �xa2 ! - with a white rook on f1 instead of a I , such tricks are never possible - 2 2 . bI.xf8 + ttJxf8) 2 l . . Jlc5 with a dynamic balance. 1 3.

...

14.

:f1 -e1

�c8-b7

White develops his last piece before taking action with his central pawns. In some lines , leaving out 1 4.l:i:fe 1 .a:c8 gives White some advantages ; therefore, it is also possible to start with the im­ mediate 1 4.d5 ! ? and now: A) In the case of 1 4 . . . ttJa5 , the ab­ sence of the moves 1 4 .l'lfe 1 �c8 would indeed give White an extra possibility:

a n a lysis d i agram

White's ed pawns are very strong. 1 9 . . Jlc8 ! ! (the only move to stay in the game. The idea is to move the rook away from the e 6 - e 7 tempo. If 1 9 . . . il.xf3 ? ?, 2 0 .e6 and White pro­ motes; or 1 9 . . . il.d5 ?? 2 0 .�d l ; while 1 9 . . . il.c8 loses to 2 0 . �e l ! g6 2 l .ttJg5 ttJxe5 2 2 .f4) 2 0 .e6 ( 2 0 .nd l is a draw after 2 0 . . . il.xf3 ! 2 l .e6 ttJe5 2 2 . gxf3 ttJxfl 2 3 .exfl g6) 2 0 . . . il.d5 2 l .ttJg5 ttJd6 22 .l'ld 1 il.xe6 (22 . . . il.xa2 ? loses to 2 3 .h4) 2 3 .ttJxe 6 ttJxfl 24.h3 with lld7 to follow, and Black has a difficult life ahead of him. B) Black should continue with 1 4 . . . exd5 ! 1 5 .il.xd5 'iVe 7 and in this position , the inclusion of 1 4.l'lfe 1 �c8 would definitely have favoured White, since a rook on e 1 would have been very useful here.

a n alysis d i agram

1 5 .dxe 6 ! (the standard idea 1 5 .il.d3 exd5 1 6 . e 5 was another option) 1 5 . . . ttJxc4 (White remains a pawn up in case of 1 5 . . . 'ifxd2 ? 1 6 .exfl + \tih8 1 7 .ttJxd2 ) 1 6 .exfl + �h8 (White re­ gains the sacrificed piece, remaining a pawn up after 1 6 . . . �xfl ? 1 7 .�f4+ 'iVf6 1 8 . 'ifc7 +) 1 7 . 'ifxd8 �axd8 1 8 .l'lxd8 �xd8 1 9 .e5 . 230

14 .

...

J:i.a8-c8

The idea behind the text is obvious. Black plays a useful developing move, postponing the decision as to where to jump with his knight : e7 or as . Let's see what happens if Black jumps with his knight immediately. A) The deficiencies of 1 4 . . . ttJe7 were excellently shown in Petrosian­ Kortchnoi, 6th Candidates' match game, II Ciocco 1 9 7 7 : 1 5 .d5 ! (White has de­ veloped all his pieces and it is time for action) 1 5 . . . exd5 1 6 .exd5 ttJf5 (the knight heads for d6 to block the advance of White's ed d-pawn, as seen in the above-mentioned comments on 8 . . . il.b4 in Yusupov-Ribli) 1 7 .ttJe5 ! (Black's pro­ blem is the weakness of the c6-square, as in Yusupov-Ribli) 1 7 . . . ttJd6 (otherwise 1 8 .d6)

highly instructive and a pleasure to watch:

a n a lysis d i agram

2 2 .g3 h6 2 3 .'ife5 ! liac8 24.'iVd5 (the queen is placed in a dominant position) 24 . . . �h7 2 5 .1de4 \tg8 (Black is obvi­ ously doomed to ivity, but how does White improve ?) 2 6 . \tg2 a6 2 7 .h4 b5 2 8 .g4 ! h7 2 9 .l'le2 �h8 3 0 .g5 h5 3 1 .l'ld2 �fe8 3 2 .'iVf3 ! g6 3 3 . I! 2 d 5 ! (the threat is 3 4 . 1d d 7 ) 3 3 . . J lf8 34JH6 ! (another rook enters on d7) 34 . . . 'iYe7 3 5 .ldd7 'iVe8

an alysis d i agram

1 8 .ttJc6 ! il.xc6 ( 1 8 . . . ttJxc4? is bad due to 1 9 .ttJxd8 ttJxd2 2 0 .ttJxb 7 1::t ab8 2 l .�e7 ttJe4 2 2 .d6 and White has a winning endgame) 1 9 . dxc6 ttJxc4 2 o .'iVf4 (the point behind 1 8 .ttJc6 ; White regains material and ends up with a well-defended ed pawn on c6) 20 . . . ttJd6 2 1 .l'lxd6 'iVc 7 . It i s obvious that White i s better. How­ ever, it is also very obvious that it is far from easy to convert this advantage into a full point. Petrosian's execution is

a n a lysi s d i agram

3 6 . ldxg 6 ! (the relation between Petrosian and Kortchnoi was known to be extremely bad; still, sometimes it is better to resign and, among other things, not to give your hated opponent the pleasure of mating you) 3 6 . . . 'if e5 3 7 .'iVxh5 mate. Probably the only game 23 1

C ha p t e r

W i n n i n g C h e s s M i d d l eg am e s

in the history of the Candidates' matches that ended with an actual mate! B) 1 4 . . . <1JaS is the other knight jump, which most of the time simply trans­ poses to 1 4 .. .l:: k 8 I S .dS <1Ja5 lines : 1 5 .�d3 h6 (this move, always useful in general, is the independent idea behind 1 4 . . . <1JaS ; I s .. J�k8 1 6 .dS , which is played in most cases , is indeed a simple transposition)

B 2 ) 1 6 .dS ! (one o f the main strategic ideas is executed again) 1 6 . . . exd5 1 7 .eS <1Jc4 1 8 .'ti'e2 'iYc7 1 9 .<1Jd4.

analysis d i agram

B 1 ) The attempt to improve the posi­ tion further and then aim for the d4-dS thrust does not bring success, because Black organizes his forces as well : 1 6 .'iYf4 'ti'f6 1 7 .'iYg3 'iYe 7 1 8 .h4 �fd8 1 9 .dS �d6 ! 2 0 .eS

a n alysis d i agram

20 . . . 'ti'xdS 2 1 .�h 7 + �xh7 2 2 .l:txdS jdxdS and Black had a good game in PNikolic-Ruban, Groningen PCA 1 9 9 3 ; 232

A nice diagram position to understand the point behind a standard sacrifice like 1 6 .dS . White has given a pawn in order to have an excellent knight on d4, which can easily become a very strong attacking piece by jumping to f5 . White's d3 bishop is a strong attacking piece, which may easily form a battery with the queen and is perfectly placed on the open diagonal. White's eS pawn has attacking power of its own (e5 -e6) and can soon be ed by the f-pawn (f2 -f4) rolling on the kingside. Black's pieces are doomed to ivity: the b7 bishop is merely a pawn, the fS -square is weak due to the e6 pawn being moved to dS , his kingside pawn structure is vulnerable to attack in gen­ eral, the knight on c4 is doing little and does not the kingside defence, and finally, Black's counterplay is very limited as well. All this for the price of a single pawn is a good foundation for a positional pawn sacrifice and in gen­ eral, not (too many) direct variations are needed to justify its soundness. I know that I am being a little biased here - on the other hand, isn 't life itself

P a w n M aj o r i t y i n t h e C e n t re

biased? - when I say that in such posi­ tions I would almost per definition pre­ fer White. After 1 9 . . . �ae8 2 0 . f4 bS 2 1 . �h l White had good compensation for the pawn in Gulko-Yusupov, Tal Memorial, Riga 1 9 9 5 . All White's pieces are well-placed, so there is no more time to waste. 1 5.

analysis d i agram

4:

d4-d5!

e6xd5

This leads to a position which is better for White. Another main move is I S . . . <1Ja5 (as mentioned under 1 4 . . . <1JaS I S .�d3 �c8 1 6 .dS) and now: A) With rooks on e 1 and c8 , 1 6 . dxe6 is, compared to the comment after 1 4.d5 <1JaS 1 5 .dxe 6 ! , much less dan­ gerous for Black. Complications lead to a more or less forced drawn ending, see:

analysis d i agram

The white pawns are rolling and Black is forced to qUickly return the piece : 2 0 . . . �c8 ! 2 1 . e6 �xe 6 2 2 J :txe 6 g 6 ! (2 2 . . . I;lc 7 ? i s not accurate due to 2 3 .�e8 �xf7 24.�xf8 + �xf8 2 S .l:td7 �a8 2 6 . �b7 and White is better due to his active rook) 2 3 Jie7 lla4. B) So, play now continues with 1 6 .�d3 exdS 1 7 .e5 (again, a standard position for the d4-dS pawn sacrifice has been reached) 1 7 . . . <1Jc4 (it may seem that this situation is very different due to the fact that the white knight is still on e 2 and compared to Gulko­ Yusupov, there is no 'ti'e2 , but : ) 1 8 .'iVf4 ! (White trusts his attacking re­ sources! ) 1 8 . . . <1Jb2

analysis di agram

A I ) A bad mistake is 1 6 . . . <1Jxc4? 1 7 .exf7 + �h8 (or 1 7 . . . �xf7 1 8 .'iVf4+ 'iYf6 1 9 . �d 7 + ) 1 8 .'iVxd8 �cxd 8 1 9 . !Ixd8 llxd8 2 0 .eS and White wins. Compared to the previously analysed position after 1 4.dS <1JaS 1 5 .dxe 6 <1Jc4, there is an extra tempo with the white rook on e 1 instead of fl ; A 2 ) 1 6 . . . 'ifxd 2 ! 1 7 .exf7 + �h8 1 8 .<1Jxd2 <1Jxc4 1 9 .<1Jxc4 !Ixc4 2 0 .eS .

a n a lysis d i agram

1 9 .�xh7 + ! ! (the idea behind 1 8.'iYf4) 1 9 . . . �xh7 2 0 .<1JgS + �g6 2 1 .h4! (not only defending the knight, but also threatening a deadly 2 2 .hS + !) and now: 233

W i n n i n g C h ess M i d d l eg a m e s

B 1) After 2 1 . . . tLJxd I ? , White's idea becomes clear: 22 .h5 +! �xh5 (or 2 2 . . . �h 6 2 3 .tLJe6+ <;t>h 7 24. tLJxd8 ncxd 8 2 5 J :lxd 1 ) 2 3 . g4+ ! \t> g 6 2 4 . 'iVf5 + �h6 2 5 . 'iVh 7 + \t>xg 5 2 6 . 'li'h5 + \t>f4 2 7 . 'iff5 mate; B 2 ) 3 4 years ( ! ) after the original 1 9 .jLxh7 + ! ! sacrifice saw the light of day, another top-level game was played which continued 2 1 . . . 'iVe7 2 2 .l::t d 2 �c4 (if 22 ... tLJc4, White continues the attack with 2 3 .�d3 )

Chapter

from the above-mentioned line : 2 6 .h6 ! l:Ih4 2 7 .�d4 !) 2 6 .'ifxf6 gxf6

better square) 2 9 . . . �h7 3 0 .tLJf5 �exe6 3 1 .1:txe6 �xe6 3 2 .J:tc2 11c6 3 3 .lde2

4:

P a w n M aj o r i t y in t h e C e n t r e

2 1 . . . tLJb7 ! (cool and instructive de­ fence. In such positions, the tables can turn easily, and all of a sudden White realizes that he is in serious trouble) 2 2 . ,ge7 (if 2 2 .�d3 ? ! , 2 2 . . . f6 ! ) 2 2 . . . f6 2 3 .tLJe6 'ifxf4 24.tLJxf4 btfl and Black had a good game in Khalifman-Ruban, Russian Championship, Elista 1 9 9 5 , since 2 5 . d 6 ? i s not good due to 2 5 . . . �f8 ! and White loses his d-pawn.

an alysis d i agram a n a lysis d i agram

analysis d i agram

2 3 . 'if g 3 (White misses a nice forced win with 2 3 .h5 + ! \t>h6 (if 2 3 . . . �xh5 ?, Black still gets mated by 24.g4+ \t>h 6 2 5 .�h2+) 24. tLJxf7 + �h7 2 5 .'iff5 + �g8 2 6 . e6 tLJa4 (otherwise White sim­ ply collects the knight) 2 7 .h6 ! (threat­ ening mate in one) 2 7 . . . l:!:h4 2 8 .lJ:d4! and Black is helpless. Note that in all of these lines, apart from defending a pawn on d5 , the bishop on b7 is ive and totally out of play. This is one of the big strategic points behind the d4-d5 pawn sacrifice) 23 . . . �h6 24.�xb2 and White went on to win in Dreev­ Yusupov, Mainz 2 0 0 3 ; B 3 ) 2 1 . . J �c4 2 2 . h5 + �h6 (2 2 . . . �xh5 ?? 2 3 .g4+ i s the same old story) 2 3 .tLJxfl + \t>h 7 24.'iVf5 + �g8 2 5 .e6 'iff6 ! ( 2 5 . . .'ife 7 ? loses to a motif 234

2 7 .�d2 ! (a forced draw is 2 7 .tLJd6 tLJxd 1 2 8 .e7 .J::l c 1 29.exf81::I. + �xf8 3 0.tLJxb7 d4 3 1 .tLJd6 d3 3 2 Jle8 + \t>g7 3 3 .tLJf5 + \t>f7 34.tLJd6+ \t>g 7 3 5 .tLJf5 + �h 7 3 6 .,ge7+ �g8 3 7 Jde8+) and now: B3 1 ) 2 7 . . Jdb4, defending the knight, looks logical, but loses after 2 8 .a3 ! , throwing the b4 rook off-balance: 2 8 . . Jdb3 29 .e7 �xf7 3 0 .exf8'if + \t>xf8

3 3 . . . jLc8 ? (this allows the white rook to enter the seventh rank, which proves decisive in this spectacular game. Better was 33 . . . �c7 , though it could well be that Black is lost anyway after 3 4J �e6) 3 4.1:te 7 + �h8 3 5 .tLJh4 ! f5 3 6 .tLJg6+ �g8 3 7 .�xa7 and Black resigned in Polugaevsky-Tal , Moscow Soviet Championship 1 9 6 9 . 1 6.

�c4xd5 !

On d5 the bishop will be very domi­ nant and Black will have to exchange it, creating a ed d-pawn for White af­ ter all. The immediate 1 6 .exd5 gives Black a tempo with 1 6 . . . tLJa5 and is not that convincing : 1 7 . jLfl iY d6 ( 1 7 .. J:k5 1 8 .d6 jLd5 1 9 .tLJg5 ! is dan­ gerous for Black) 1 8 . tLJ g 5 llcd8 1 9 .'tliPd3 'li'h6 ! 2 0 .'iVf5 jLc8 2 1 .'iff4

analysis d i agram

3 1 .h6 ! �g8 (if 3 1 . .. jLc6 , then simply 3 2 . �e6 ; or 3 1 . . . tLJc4 3 2 . h 7 � g 7 3 3 .lie 7 + �h8 3 4 . .a:d4 and White Wins) 3 2 . 1:! e 7 �c6 3 3 . l:!:de2 tLJc4 3 4. 1:t g 7 + �h8 3 5 Jiee 7 and White de­ livers mate in a few moves; B 3 2 ) 2 7 .. . !:I c6 2 8 . �xb 2 .J::l e8 2 9 . tLJh6 + ! (transferring the knight to a

analysis d iagram

1 6.

...

ttJc6-a5

Very likely, Black's queen will have to be moved to e7 or c 7 , and Black might well have decided to do this at once : A) 1 6 . . . 'iVe7 1 7 .e5 ! (note that com­ pared to the line 1 4.d5 ed5 1 5 .�d5 'if e 7, White's rook is very useful on e 1 here, like in most of the other lines, so it is clever for White to include 1 4.l:te 1 1:tc8 after all ! ) 1 7 . . . tLJd8 1 8 .jLxb 7 tLJxb 7 1 9 .tLJd4 g6 2 0 . f4 and White is better with his pawn majority danger­ ously rolling towards the black king; B) 1 6 . . . 'iVc 7 , on the other hand, has some other deficiencies after 1 7 . 'if g 5 ! h 6 1 8 . 'Ii' g4 �fd8 1 9 .h3 (a useful move here) 1 9 . . . tLJe7 ( 1 9 . . . tLJb4? is bad due to the simple 2 0 .�xb 7 �xb7 2 1 .tLJd4, and the white knight enters via f5 ) , and now both 2 0 .�b3 and 2 0 .jLxb 7 'ifxb 7 2 1 .e5 look promising for White. 1 7.

'iVd2-f4

'iYd8-c7

235

Chapter

W i n n i n g C h es s M i d d l eg am e s

2 7 . . . g6 loses to 2 S . :axdS gxf5 2 9 . nxfS + �xfS 3 0 . lIeS + . Or 2 7 . . . 'iVxd4 2 S . lIxdS lIxdS 2 9 .lIe S + +- .

After 1 7 . . . VJlie 7 , White again transfers his knight to f5 with l S .t2J d4. 1 S.

'iVf4-f5!

Keeping the pressure and forcing Black to eliminate the dominant bishop on d5 . 1 S.

...

1 9.

e4xd5

2S.

tLJd4-eS !

29.

tLJeSxdS!

tLJ b 7-dS

�b7xd5

Now White has a fast ed pawn .

22.

4:

P a wn M a j o r i t y in t h e C e n t r e

scored many decisive victories with it, taking advantage of his excellent feeling for White's attacking potential, which is stored behind his central pawn majority. The positional idea behind 4.a3 is a rather simple, straightforward fight for the central squares d5 and e4. Con­ cretely, the move prevents the �b4 pin.

'iVa2-e4

23.

'iVf4-f5

24.

l:.d1 -e1 !

h 7-hS

Preparing to enter on the seventh rank. 24.

1 9.

...

'iYe7-e2

Black wants to chase the white queen away from the b 1 -h7 diagonal in order to prevent the creation of dark-square weaknesses after White's ttJf3 -g 5 . Black does not have time for the knight transfer to d 6 , because 1 9 . . . ttJc4 loses to 2 0 .ttJg5 g6 2 1 . 'iYh 3 h5 2 2 .ttJe4 ttJd6 2 3 .ttJf6 + �g 7 24.VJlig3 ! . A forced ending ensues after 1 9 . . . 'iY d6 2 0 .ttJg5 ! VJlig6 2 1 .'ifxg6 hxg 6 2 2 .d6 ttJb7 2 3 .d7 llcdS 24Jle7 - this is likely lost for Black. 20.

'iYe4-aS

25.

:te1 -e7

2S.

tLJf3-d4

bS-b5

2 6 .l:IeS should also win (if 26 . . . 'iVd6 , 2 7 Jk s 'iVd 1 + 2 S . ttJe 1 ) , but it is un­ derstandable that White wants to keep one rook on the first rank. 2S.

.i::t e 7-eS!

Black's fate is sealed. 27.

...

tLJa5-b7

A nice line is 2 7 . . . b4 2 S .Iie S ! 'iYxd4 2 9 . �xfS + lIxfS 3 0 . lIxfS + �xfS

'iYf5-f4!

The ending after 2 0 . VJlixc 2 �xc2 2 1 . I:Ie 7 is clearly better for White, but Spas sky wants more. 20.

...

'iYe2xa2

Black has a point. His position is diffi­ cult anyway; now at least he has a pawn. 21 .

d5-dS!

22.

dS-d7

::teS-dS

It is obvious that White's strong pawn on d7 and his complete domination are worth more than the sacrificed pawn. 236

29.

...

30.

tLJdS-eS

tLJdSxf5

Black resigned.

a n a lysis d i agram

3 1 .'iYc5 + ! 'iYxc5 3 2 .dS'iY mate.

�eS-b7

4.

'iYaS-bS

...

26 . . . 'iVd6 is perhaps better than the text, but does not save Black after 2 7 .ttJxb5 'iYd2 2 S . l:Ib 1 a6 2 9 .ttJc3 ttJb3 3 0 .lId l . 27.

Nice execution until the very end.

QI 7 .4 (E 1 2 )

GAME 4 2

Garry Kasparov Miguel Najdorf Bugojno 1 9 8 2 (3) 1.

d2-d4

tLJgS-fS

2.

e2-e4

e7-eS

3.

tLJg1 -f3

b7-bS

4.

a2-a3

This variation was brought on the big scene by the 1 0th World Champion Tigran Petrosian, who regularly em­ ployed it since 1 9 5 S , and it was named after him. Nowadays it is one of the main lines of the Queen's Indian. Most of the time, the middlegame arising from this variation features a typical structure with a pawn majority in the centre. The young Kasparov picked up this dy­ namic line, studied its ins and outs and

5.

tLJb1 -e3

d 7-d5

S.

e4xd5

tLJfSxd5

Another main line is 6 . . . exd5 . 7.

e2-e3

Another popular move is 7. 'iV c2 ; see Game 45 (Sokolov-Polgar) further on in this chapter. 7.

...

�fS-e7

All the normal moves for Black in this position have been tried hundreds of times in grandmaster practice : 7 . . . g 6 ; 7 . . . ttJf6 ; 7 . . . ttJ d 7 (see Galliamova­ Tiviakov, Game 43 , further on in this chapter) and 7 . . . ttJxc3 . S.

�f1 -b5+

This check is an idea of Petrosian him­ self. The idea is to slightly disturb Black's development. e7-eS

S. 9.

�b5-d3

tLJd5xe3

1 0.

b2xe3

eS-e5

11.

0-0

tLJ bS-eS

237

Chapter

W i n n i n g C h es s M i d d l eg a m e s

A) 1 6 . . . tLJe 7 ? does not solve the prob­ lems of the black king stuck in the mid­ dle : 1 7 .c4 'iVd6 (not 1 7 .. .'iVc6? because of the simple 1 8 .�e4) 1 8 .tLJeS ! �d8 (if 1 8 . . .'iVc7 , White has a strong attack af­ ter 1 9 .'�' h S ; or 1 9 . 'iVa4+ �c6 2 0 .'iYc2) 1 9 .�a4+ �f8 2 0 .llad l �c 7 2 1 . �e4 ! �xd l 2 2 . �xd 1 �xe4 2 3 .tLJd 7 + �g8 24.tLJxf6 + gxf6

The main dilemma for Black here is where to put his knight. The other main line is, naturally, 1 1 . . . tLJd 7 .

an alysi s d i agram

1 2.

e3-e4

Very logical; however, White can also decide to place his dark-squared bishop on the long diagonal and then (instead of e3 -e4) push c3 -c4, aiming for a posi­ tion with hanging pawns in the centre. One of Kasparov's spectacular wins went as follows : 1 2 .�b2 l::r. c 8 1 3 .'iVe2 0-0 1 4.�ad l 'iVc7 I S . c4 cxd4 1 6 .exd4 (a typical 'hanging pawns in the centre' position) 1 6 . . . tLJaS .

White is harmoniously developed, but how to continue ? A few pieces have al­ ready been exchanged and Black is about to consolidate. 2 1 .�xg 7 ! ! (this nice, thematic sacrifice is based on good calculation and intu­ itive tactical feeling) 2 1 . . . �xg 7 2 2 .tLJeS ! I1 fd8 2 3 .'ifg4+ Wf8 2 4 .'�'fS ! f6 2 5 .tLJd7 + l:txd7 2 6 . ldxd7 �cS 2 7 .�h7 and the black king remains under attack, Kasparov-Portisch, Niksic 1 9 8 3 . 1 2.

...

0-0

Black may also delay castling and first develop his rook: 1 2 . . . �c8 1 3 .�b2 (I assume that in the case of 1 3 .�e3 we would eventually get a transposition to the main game) 1 3 . . . �f6 .

analysi s d i agram

P a wn M a j o r i t y in t h e C en t re

2 2 .' iYa4 ? ! (thiS move is probably a mistake , since now, the white bishop on b2 will be totally out of play for a while. It was better to exchange this ive bishop for Black's good one with 2 2 .c4! and Black has terrible problems to connect his forces after 2 2 . . . �xb2 2 3 .'li'xb2 hS 24.Md6 ! h4 2 S . 'iYd2 'iY b 8 2 6 . ld d 7 ) 2 2 . . . h S ? ! (better was 2 2 . . . b S ! 2 3 .'iYf4 h S 2 4 . bI d6 �h6 ; 2 4 . . . �c 7 2 S . �fS ) 2 3 . 11 d 7 (Kasparov-Ivkov, Bugojno 1 9 8 2 ) 2 3 . . .'iYc6 24.'iYxa7 �g8 and White is better, although Black has good draWing chances based on White 's weak queenside pawns. 1 3.

Now, 2 S .'iYe 8 + is not good; it will get White into trouble after 2 S . . . \t>g 7 2 6 .�xf6 + �xf6 2 7 .'iYxh8+ �g6; But after 2 S .�xf6 ! , Black is paralysed : 2 S . . . tLJc6 (or 2 S . . . �c8 2 6 .'iYd7 �xd7 2 7 .�xd7 tLJc6 2 8 J::r c 7 and White wins) 2 6.f3 �fS 2 7 .'ii' xc6 and Black is mated. B) So Black is forced to part with his right to castle: 1 6 . . . Wf8 1 7 .'iYc2 tLJeS (understandably, Black would like to ex­ change pieces) 1 8 .�e4 tLJxf3 + 1 9 .�xf3 �d7 20.�ad l �c7 2 1 .�xb7 'iYxb7

4:

�c1 -e3

A logical question one may ask is : why not simply push the d4 pawn , seizing more space and sending the black knight to the edge of the board? Well, things are not that simple, let 's see : 1 3 .dS exdS 1 4.exdS tLJaS 1 5 .c4.

a n a lysis d i agram

a n a lysis d i agram

Black hits White's c4 weakness. 1 7 .dS ! (With his pieces well developed, White goes for the central pawn thrust, as we have often seen in the previous chapter) 1 7 . . . exdS 1 8 . cxdS �xdS 1 9 .�xh 7 + �xh7 2 0 . l:ixdS �g8 . 238

ana lysis d i agram

With Black's king still on e 8 , Kasparov does not hesitate to sacrifice a pawn in order to open the centre : 1 4.dS ! exdS I S .exdS 'iVxd S 1 6 Jle 1 and now:

analysis d i agram

The black knight has indeed been pushed to the edge of the board and White seems to have strong central con­ trol, while Black's b7 bishop is also temporarily out of play. White's prob­ lem , however, is that his central pawns (c4 and dS) can easily been under­ mined with I S . . . bS ! and after 1 6.cxbS : 239

C h a p t e r 4: P a wn M a j o r i t y in t h e C e n t r e

W i n n i n 8 C h e s s M i d d l e 8 a m es

A) Surprisingly enough, it is now wrong for Black to ' gain a tempo' and play 1 6 . . . j2,f6 ? ! 1 7 .l:rb l 'ifxd5 , since Black's as knight now indeed becomes stranded and becomes an easy target af­ ter 1 8 .'iYc2 .tIfe8 1 9 .�d2 ! and Black's otherwise well-coordinated , active pieces cannot compensate for the hope­ less knight, for example : 1 9 . . . 'ifh5

a n alysis d i agram

2 0 . �fe I ! (not 2 0 .j2,xa5 ? j2,xf3 2 1 .gxf3 �e5 2 2 .f4 iYg4+ 2 3 . �h l 'iVf3 + 24.�g l 'iVg4+ and Black has at least a draw) 2 0 . . . �xf3 2 1 . gxf3 �e5 (if 2 1 . .. ttJb7 , White is clearly better after 2 2 .�e4 l:le 7 2 3 .b6 ! ) 2 2 J::l xe5 ! 1::l x e5

analysis d i agram

2 3 .�e4 ! ( 2 3 .j2,xa5 ?? is a terrible mis­ take due to 23 .. Jdg 5 + 24.�fl ge8 and Black Wins) 2 3 . . . �d8 24 .j2,xa5 . The smoke has cleared - White is clearly 240

better and went on to win in Aru­ tinian-Melkumyan, Martuni 2 0 0 7 . B ) 1 6 . . . 'iVxd5 ! 1 7 .'iYc2 ttJb3 ! . That's the difference. Black solves the prob­ lems of his stranded knight and he is at least equal. 1 3. 1 4.

c5xd4 c3xd4

A good moment to take stock and com­ pare this position to the previous Semi­ Tarrasch game. The pawn structure is the same, but the difference is that here the dark-squared bishops are still on the board. In my opinion, in general this should be favourable for White, since : - Black has conSiderably less space to manoeuvre his pieces; - Black queen sorties (to f6) are not pOSSible; - In the case of a central pawn blockade (after d4-d5 and e6-e5 ) , it is, in general, positive for White to have the dark­ squared bishops on the board. Should a black pawn appear on h6, White may get sacrificial motifs with j2,xh6 and, in gen­ eral, in many lines White simply has one more piece for his kingside attack. The plus for Black - perhaps the only one - is the fact that the . . . £7 -f5 idea, to under­ mine White's pawn centre and take con­ trol of the d5 -square, is much easier to execute with the dark -squared bishops still present.

1 4.

...

l:la8-c8

If we consider this only black plus of having the dark-squared bishops on the board, then a rather interesting option - surprisingly enough, very little seen in practice - is : A) 1 4 . . . �h8 ! ? (the immediate 1 4 . . . f5 allows 1 5 . 'ifb3 ) 1 5 . 'iV e 2 ? ! (White does not take Black's plan seriously, but he should have done so and considered moves like 1 5 . ifb 1 ) 1 5 . . . f5 . Black correctly undermines the white centre and is about to take over control of the light squares (namely, d5) . White does not have the possibility to solve the problem with a d4-d5 type sacrifice here - as we see further on in the chap­ ter, in the game Sokolov-Howell while after 1 6 .exf5 exf5 , Black has a very comfortable game. White does not want to make any posi­ tional concessions and decides to tem­ porarily sacrifice a pawn : 1 6 .11ad 1 f4 ! 1 7 .j2,c l ttJxd4 1 8 .ttJxd4 'iYxd4 1 9 .j2,b2 if c5 (Black had a strong intermezzo in 1 9 . . . f3 ! 2 0 .gxf3 'iYc5 , when White would have had a difficult time of it with his damaged kings ide pawn struc­ ture) 2 0 .'iYg4 'ifg5 (20 . . . e5 2 1 . �c 1 'ifd6 2 2 .�fd l ) 2 1 .'iYxe6 Iiae8 2 2 .iYd7 f3 ! 2 3 .g3 j2,c8 24.'iYxa7 �c5 25 .l:lfe 1

2 5 ... 'iYh6 ! 2 6 .�fl ? (White blunders in a very difficult position; on the other hand, 2 6 .h4 weakens the g 3 -square and loses to 2 6 . . . 'iY g 6 ! ) 2 6 . . . �xf2 + 2 7 .�xf2 'iYxh2 + 2 8 .rJ;te3 �xb2 and Black won in Vaganian-Razuvaev, Yaro­ slav1 1 9 8 2 . B) To use the presence of his dark­ squared bishop in order to attack the white centre with 1 4 . . . j2,f6?! does not work nearly that well: 1 5 .j2,b 1 l:rc8 1 6 .'iYd3 g6 1 7 .�a2 ! 'iVd 7 1 8 .ldad 1 ttJa5 1 9 .j2,h6 l1fe8 2 0 .ttJg5 ! and White had a promising kingside attack and soon won in Furman -Panno, Madrid 1 9 7 3 . 1 5.

'tWd1 -e2

In one of his earlier games , Kasparov tried to develop his a 1 rook, using a second-rank transfer to d2 . This plan, though logical in itself, did not work well here : 1 5 Jla2 ! ? �f6 1 6 .j2,b l

an alysi s d i agram

analysis d i agram

1 6 . . . g 6 ! (in many lines , this protects the black king better against the kingside attack) 1 7 .'ifd3 (I guess that in the event of an immediate 1 7 . �d2 , White did not like 1 7 . . . ttJ a 5 ) 1 7 . . . 'ifd 7 1 8 .l:td2 �fd8 1 9 .'iYe2 ttJa5 . White's central pawns cannot move, while the ttJa5 -c4 transfer is threaten­ ing. So, White's next move is more or less forced: 2 0 .ttJe5 24 1

Chapter

Wi n n i n g C h e s s M i d d l eg am e s

pawn is always worth sacrificing for a gain of three tempi. Well , here White is gaining only two : 1 7 . . . �xh4 1 8 Jh d l �e 7 1 9 .dS �cS 2 0 .�f4

analysis di agram

20 . . . �xeS ! (Black is not afraid of ghosts. The white bishop on f6 alone cannot deliver mate, while his queen cannot be transferred to h6 to help) 2 l .dxeS 'liVc 7 2 2 .lixd8 + �xd8 2 3 .�gS nd4 24.'liVe3 (sacrificing a pawn, since after 2 4 . �f6 'ifc 3 ! 2 5 . 'ifg4 �a6 2 6 .'liVgS 'iVd2 ! , Black is better) 24 . . . 'iVxeS 2 S .f4 'iVd6 2 6 .�e 7 'ifd7 2 7 . �f6 Md l 2 8 .�c2 . Respect can sometimes help a lot. Black was clearly better here, but still accepted a draw in Kasparov-Karolyi, Graz 1 9 8 1 . 1 5.

...

<1Jc6-a5

I S . . . rJ;th8 ! ? , preparing . . . fl -fS , is again positionally sound and was in a practi­ cal sense a very interesting alternative. After 1 6 Jifd l ttJaS 1 7 .a4, the posi­ tional idea 1 7 . . . fS ! , surprisingly not of­ ten used in practice, is very logical, en­ gaging in a straightforward fight for domination of the central squares: 1 8 .exfS exfS 1 9 .�f4 ttJb3 ! 2 0 . I:rab 1 �xf3 2 1 .'liVxf3 ttJxd4 2 2 .'ifb 7 and White was just in time with his draw offer in Browne-Smyslov, Los Polvo­ rines 1 9 8 0 . 1 6.

l:tf1 -e1

'ii d 8-d6

1 6 . . . rJ;th8 ! ? again came into consider­ ation, and now: A) 1 7 .h4 is probably a dubious pawn sacrifice. Capablanca once said that a 242

a n a lysis d i agram

20 . . . f6 ! 2 1 .dxe 6 �e 7 (it is difficult for White to justify his pawn sacrifice) 2 2 .�b l 'liVxe6 2 3 .ttJh4 ? (White had definitely miscalculated. Somewhat better was 2 3 . ttJd4) 23 . . . gS ! 24.�a2 'iYxe4 2 S .'iVxe4 �xe4 2 6 .I:.xe4 gxf4 2 7 .Mxf4 Mcd8 . Again, respect is a won­ derful thing, since in this clearly better ending Black soon agreed to a draw in Kasparov-Groszpeter, Graz 1 9 8 1 . By the way, 2 7 . . . �xa3 ?? would have been a terrible blunder due to 2 8 .ttJg6 + ! hxg6 2 9 .lih4+ �g 7 3 0 .l;Id 7 + and wins. B) An interesting option is 1 7 . Mad 1 ! ?, since after 1 7 . . . �xa3 1 8 .ttJgS the white attack seems to compensate for the sacri­ ficed pawn : 1 8 . . . h6 1 9 .�hS �g8 .

analysis d i agram

It seems that in this pOSition, White has gotten closer to Capablanca's theo­ retical three tempi for the sacrificed pawn. 2 0 .eS ! (thiS thematic way of building the attack is dangerous here as well, but the thematic sacrifice 2 0 . dS exdS 2 1 .eS was definitely to be consid­ ered) 20 . . . 'ifdS (Black tries to make use of his only trumps, the open a 8 -h 1 diagonal and the dS -square) 2 1 .ttJf3 (threatening 2 2 .�h6) 2 l . .. fS (or 2 1 . . . 'iYd7 2 2 .�xh 6 ! ) 2 2 . �xh6 gxh6 2 3 . 'liVg 6 + rJ;th8 2 4 .'ifxh6 + � g 8 2 S .'iYg6+ �h8 .

analysis d i agram

2 6 . . .' �d 7 No! 2 6 .�e4! Draw? ( 2 6 . . . fxe4 loses t o 2 7 . ttJ g S 'iYd7 2 8 . �e 3 ) 2 7 . ttJ g S M c 3 (otherwise 2 8 .�e3) 2 8 .�xb 7 ttJxb 7 2 9 .ttJxe6 Mfl 3 0 .dS and White's rolling pawns, com­ bined with the exposed black king, will decide the battle; C) The game develops in a highly complicated way after 1 7 .a4 fs (Black executes his plan) 1 8 . exfS exfS 1 9 .�f4 �b4 ! 2 0 . �ed l ttJb3 2 1 . 1:Iab l ttJxd4 2 2 .ttJxd4 'iYxd4 2 3 .�eS 'iVdS 24.f4. It is obvious that in this pOSition, White has difficulty to prove his compensa­ tion.

4:

P a wn M a j o r i t y in t h e C e n t r e

analysis d i agram

24 . . . �cS + ? ! (24 . . . aS ! was a possible improvement. Now White's strong eS bishop, combined with Black's exposed queen, should offer White enough dy­ namic possibilities) 2 S . �h l �cd8 ?! (pro bably not feeling comfortable about his position, the former World Champion chooses a forced sequence of moves which lands him in a lost end­ game) 2 6 .�xfS �xg 2 + ? ! 2 7 . 'iYxg2 Mxd l + 2 8 . !Ixd l �xg 2 + 2 9 . �xg2 lixfS 3 0 J id7 ( 1 - 0 , 46) , Schmidt­ Smyslov, Moscow 1 9 8 0 . A n interesting counterplay idea, also seen in similar lines, is 1 6 . . Jic3 ! ? 1 7 .a4 ttJb3 1 8 . �ad l 'liVa8 and Black's well-co­ ordinated pieces exert pressure on White's pawn centre.

Here Kasparov opts for a thematic pawn sacrifice, which we have already ex­ plored in this chapter. The problem 243

C hapter

W i n n i n g C h es s M i d d l eg a m e s

with this sacrifice is that, even if it is sound from a positional/principled standpoint, there are always specifics of the position that are important to deter­ mine the correctness of the sacrifice. Those specifics are also present here : 1 7.

d4-d51

e6xd5!

Correctly accepting the sacrifice. De­ clining with 1 7 . . . e5 gives White an easy, risk-free advantage after 1 8 . nab 1 �a8 (otherwise 1 9 .�d2 comes any­ way) 1 9 .�d2 tbb7 2 0 .a4 tbc5 2 1 .a5 . 1 8.

e4-e5!

White continues executing his thematic and principled plan. 1 8.

...

1 9.

ttJf3-d4

'iVd6-e6

As already explained, one of the impor­ tant aspects of this thematic pawn sacri­ fice is the beautiful outpost on d4 for the white knight. Conquering the bishop pair with 1 9 .tbg5 does not give White enough compensation after 1 9 . . . �xg5 2 0 .�xg5 h6 ! (exact play is needed; 20 . . . tbb3 ? is wrong due to 2 1 .�f6 ! ) 2 1 .�f4 tbb3 ! 2 2 J Iad I tbc5 and Black is better. 1 9.

...

20.

ttJd4-f5

'fi e6xe5

Another important aspect of the d4-d5 pawn sacrifice in these positions is the transfer of the white knight via d4 to 244

the excellent attacking post on f5 . Kasparov is playing a classical, book ex­ ample attack. There are, however, spe­ cifics of the position and Black has a de­ fence which can turn the tables. 20.

...

iLe7-f6?!

It is obvious that the e7 bishop has to move. However, there was a tempo move : 2 0 . . . �d6 ! . I assume that Black did not like this move for the principled reason that his bishop would remain under attack by the f5 knight. However, this small tempo gain gives Black a crucial break and it is difficult for White to prove that his initiative is indeed worth the sacrificed material. Interestingly, 20 . . . �d6 ! is not mentioned at all by Kasparov in his comments (in Chess Infonnant and some other publica­ tions) and he praises the game move 20 . . . �f6 as the best defence, giving it an exclamation mark. It seems to me that af­ ter 20 . . . �d6 ! Black is better: 2 1 . 'ifh5 (trying to keep matters com­ plicated. Other moves simply do not of­ fer enough compensation: 2 1 .f4 'iYf6 2 2 .'iVh5 (or 2 2 .�d4 �c5 ) 2 2 . . . h6 ; or 2 1 .tbxd6 'iVxd6 2 2 .'ifg4 life8) 2 1 . . . g6 2 2 .tbh6+ Wh8 ! (it is important to leave the g 7 -square vacant for the queen's re­ treat; 2 2 . . . �g 7 ? 2 3 .'lWh4 is good for White) 2 3 .'iYh4 and now:

a n a lysis d i agram

A) Active defence with 2 3 . . . tbb3 al­ lows White to quickly develop an attack which is at least worth the sacrificed material: A I ) 24.tbg4? ! is not good due to 24 . . . �e7 ! 2 5 .'iYh6 'iYh5 ; A2) 24.�g5 is definitely possible; the follOWing line is entertaining : 2 4 . . . 'iYxh2 + 2 5 . 'iYxh 2 �xh 2 + 2 6 . �xh2 f6

4:

P a wn M a j o r i t y in t h e C e n t r e

one pawn down and has strong com­ pensation. B) 23 . . . 'iVg 7 ! (this queen retreat may not be logical, but it is quite strong) 24.�d4 f6 (Black is on the defensive, but he is two pawns up and White seems to be one tempo short with his attack) 2 5 .l:Ie6 (if 2 5 .tbg4?, 25 . . . tbb3 ! wins)

analysiS d iagram analysis d iagram

2 7 .�h4! (better than 2 7 . �ad 1 fxg 5 2 8 .�e7 tbc5 2 9 .tbfl + klxfl 3 0 . �xfl c:;t>g8 3 1 .�e7 �f8 3 2 .f3 lifl and with three pawns for an exchange, Black is better) 2 7 . . . tbxa 1 2 8 . �e 7 �c6 2 9 .tbfl + and now: A2 1 ) 29 .. J�:txfl ? ! is a risky decision, since from the three missing pawns White will collect two back and with his bishop pair and Black's exposed king , he will have plenty of compensa­ tion after 3 0 . ld.xfl g5 3 I .�g3 tbb3 (3 1 . . . �g8 3 2 .�xf6 �d 7 (otherwise �d3-f5 -e 6 and �g 3 -e 5 ) 3 3 J :td6) 3 2 J�xh7 + �g8 3 3 .�xa7 etc. ; A2 2 ) 2 9 . . . �g8 3 0 .tbh6+ c:;t>h8 with a draw. A 3 ) 2 4 . 11ab I ! 'iVc3 (only move) 2 5 .�ed 1 f6 (again, this seems like the only move ; 2 5 . . . �e5 ? fails to 2 6 .'iYe 7 ! ) 2 6 .�xg 6 . White is now only

B I ) 25 . . . tbb3 is not that convincing after 2 6 J :txd6 ! g 5 ? (this is a losing move. The white monster on d4 defi­ nitely had to be destroyed with 26 . . . tbxd4 ! , though after 2 7 . 'iYxd4 'iYxh6 2 8 .11d7 ! �c I + 2 9 . �fl llxa l 3 0 .�xa 1 �c8 3 1 .lixa7 , due to his ac­ tive rook, White should be able to hold) 2 7 .'iYh3 ! (2 7 .'iYh5 ? allows 2 7 . . . tbxd4) 2 7 . . . tbxa 1 ( 2 7 . . . tbxd4 ? loses to 2 8 . 1::l d 7) 2 8 .f4 ! !;Ic I + 2 9 .�f2 lld 1 (trying to bring the bishop back into play does not help : 2 9 . . . �c8 3 0 .'iYe3 lid 1 3 1 .I:txf6 ! nxf6 3 2 .'iVe 8 + and White wins) and now: B I I ) 3 o . 11d7 ? tIxd3 ; B I 2) 3 0.tbfl + 'it>g8 3 1 .tbxg5 �d2 + 3 2 . �e 1 Ilxd3 3 3 .'iYxd3 'iYc7 34.�xf6 ; B I 3 ) 3 0 .tbf5 ! and now: B I 3 1 ) 3 0 . . . 'iYg 6 . Now Black gets caught i n a nice mating net. 245

Chapter

W i n n i n g C h e s s M i d d l eg a m e s

analysis d i agram

ana lysis d i agram

3 1 .ldxf6 ! ldxf6 3 2 .ttJe7 'iVxd3 (or 3 2 .. :iYf7 3 3 .�xf6+) 3 3 .�xf6 mate; B I 3 2 ) 3 0 . . . �c8 (pinning the white knight, the bishop now at least partici­ pates in the defence; however, White storms the barricades) 3 1 .ldxf6 ! ldxf6

And Black wins. As we have seen, the position is very complicated, it is easy to go wrong and, after all, 2 0 . . . �f6 looks at first sight more logical than 20 . . . �d6 . However, had Naj dorf chosen for 2 0 . . . �d6 , Kasparov would have had a hard time seeing all those 'promising lines' and finding out that none of them worked. 21 .

Let's see what happens if Black puts an­ other rook on e8, which is much better than the game move, since Black will then have plenty of tactics related to the activity of both his rooks and White's weak back rank: A) 2 1 . .. ldfe 8 . Now, as in the game, energ etic action is required from White : 2 2 .�d2 ! (an attempt to make some normal moves and play for com­ pensation would not work, for exam­ ple : 2 2 . ldad l g6 2 3 . ttJ h 6 + �g 7 24.lde2 'iVc7 )

iVe2-g4

analysis d i agram

2 2 . . . 'iVxa I ! (forced , since 2 2 . . . 'iVb2 fails to the nice mating line 2 3 . ttJd6 ! ! 'li'xd2 (or 2 3 . . . ldxe 1 + 24. ldxe 1 'iYxd2 2 S .'iVxc8 + !) 24.ttJxe8 �xa l

analysis d i agram

3 2 :iVhS ! ! (Black is helpless against White's double threat - 3 3 :iVe 8 + or 3 3 .ttJxg 7 ) 3 2 . . . �xfS (or 3 2 . . . �d 7 3 3 .ttJxg 7 ldxf4+ 34.�e3) 3 3 .'iVe 8 + 'iV g 8 34.�xf6 mate. B I 3 3 ) 3 o . . . ldd2 + 3 1 . � e I 'iVc7 3 2 .�xd2 ttJb3 + 3 3 .�e2 ttJxd4+ (or 3 3 . . . ne 8 + 3 4 . �f2 ttJxd4 3 S .ttJxd4 , winning) 3 4 .ttJxd4 and White wins be­ cause the rook on d6 is taboo due to the mate threat on h7 ; B2) 2 S . . . �cS ! is the winning move : 2 6 .�xf6 (or 2 6 .�b2 d4) 2 6 . . J:rxf6 2 7 .ldxf6 �e7 246

21 .

...

4:

P a wn M aj o r i t y i n t h e C e n t r e

analysis d i agram

A I ) Now the game idea 24.ttJxg 7 ? �xg 7 2 S . �h6 does not work for White after 2 s . . . ldc I + ! 2 6 .�xc I (2 6 .�fl ? blunders into mate after 2 6 . . . ldxfl + 2 7 . �xfl �a6 +) 2 6 . . . .t're I + 2 7 .�fl �a6 and Black wins; A2) In his comments on the game Kasparov opines that White is clearly better after 24.h4. This is perhaps opti­ mistic , but White definitely has strong attacking potential , for ins tance : 24 . . . lde 6 ! 2 S .hS ! (if 2 S .ttJh6 + , then 2 S . . . �f8 2 6 .�b4+ ldcs and it is difficult to find a follow-up for White) 2 S . . . ttJb3 (if 2 S . . . ttJc4, 2 6 . ttJxg 7 ! �xg 7 2 7 .�c3 )

ldc8-e8

Arguably the most logical defence. Black pins the white bishop along the e-file and vacates the c8-square for his b7 bishop, in order to control the fS -square. This sensible move, however, puts Black on the edge and it will re­ quire very exact defence to save the game. White's pieces are now very har­ monious , with strong attacking poten­ tial, and White definitely has at least compensation for his sacrificed pawns.

analysi s d i agram

analysis d i agram

2 S .ttJf6 + ! �xf6 2 6 .'iVxc 8 + �xc 8 2 7 .lde8 mate) 2 3 . ldxa l �xa l

A2 1 ) 2 6 . ttJxg 7? is now a blunder, los­ ing to 26 . . . �xg 7 2 7 .h6 ldg6 2 8 .�xg6 hxg 6 2 9 .hxg 7 ttJxd2 3 0 .'iYh4 ldc I + 3 1 .�h2 ttJfl + 3 2 . �h3 �xg 7 ; 247

Chapter

W i n n i n g C h e s s M i d d l eg am e s

A 2 2 ) 2 6 .'iYd l ! �c3 (or 2 6 . . . ttJxd2 2 7 .iVxa l f6 2 8 .'ifd l ttJc4 2 9 .�a4) 2 7 .�xc3 �xc3 2 8 .�c2 and White's ini­ tiative continues. Note that the strong knight on f5 is a crucial attacking piece in all these lines ; A3) 24.ttJd6 �c l + ! (the ive de­ fence 24 .. J�rf8 loses after 2 5 .ttJxc8 �xc8 2 6 .'li'h4 ! h6 (or 26 . . . g6 2 7 .�b4 l:ie8 2 8 .�b 5 ) 2 7 .'iVe 7 ! ttJc6 2 8 .�h 7 +) 2 5 .�xc l (2 5 .�f1 does not make too much sense, since if he has nothing better, Black can still transpose with 2 5 . . . �a6 2 6 .�xc l l:Ie I ) 2 5 . . .lle l + 2 6 .�f1 �a6 .

B) Probably the best defence for Black was 2 1 . . . 'li'c3 ! , when White can : B I ) Force a draw by perpetual with 2 2 .ttJh6 +\th8 2 3 .nac l ! (not 2 3 .�f5 ? ttJ b 3 ) 2 3 . . . iYxd3 2 4 . ttJxf7 + �g8 2 5 .ttJh6+ \th8 2 6 .ttJf7 + ; B 2 ) O r choose 2 2 . ttJe 7 + �xe 7 ( 2 2 . . . \ith8 2 3 .�xh 7 ) 2 3 .�d4 �xd4 24.'ifxd4 �f6 2 5 .'ifg4 �xa l 2 6 J lxa l .

The only way to stay in the game was 2 2 . . . 'ifb2 ! , which is now considerably better than after 2 1 . . . Ilfe 8 , since White does not have any visible mating tactics. This should, as far as I can see, still lead to a draw after 2 3 . ttJh6 + ! \th8 24.ttJxf7 + \itg8

analysis d iagram

Here, Kasparov himself was optimistic about White's chances, but I am not that convinced. 22.

�e3-d2!

a n a lysis d i agram

248

P a wn M aj o r i t y in t h e C en t r e

A I ) 3 2 . . . \th8 ? fails to 3 3 .ttJf5 �e4 34.�c l ! �xf5 3 5 .'iYxf5 ; A2) 3 2 . . . gxh6 3 3 .'li'g4+ �g7 3 4.�fl .

a n a lysis d i agram

analysis d iagram

White seems to be in trouble, but his queen and knight are working very har­ moniously and the black king is with­ out defenders: 2 7 .g3 ! ! h6 (2 7 . . . bIxf1 + 2 8 .\tg2 �xc l loses to 2 9 . 'li' g 5 ! �f1 + 3 0 . � g I ) 2 8 .�d2 ! (it is important to force Black to take with his rook) 2 8 . . . �xf1 + 2 9 . �g2 ttJc4 ! (otherwise 3 0 . iY d 7 ) 3 0 . 'li' e 2 ! (the black rook appears t o be very clumsy here) 30 .. J �tb l 3 1 .iYe8+ \ith7 3 2 .'iVa4! ttJxd2 ! (or 3 2 ... ttJxd6 ? 3 3 .'li'c2 + ; while if 3 2 . . J ib2 ? , the black rook is trapped after 3 3 .�c I �a2 3 4.'li'b3) 3 3 .'iVxa6 and Black is fight­ ing for a draw;

4:

2 5 .�xa5 ! (the best winning attempt; if immediately 2 5 . iYh5 , then 2 5 . . . bte4) and now: A) The active defence 2 5 . . . �d4 ! ? looks good, but would not work after 2 6 . \th I ! ! bxa5 2 7 .'iVh5 �e4! (only move; 2 7 . . . h6? loses to 2 8 .ttJxh6+ gxh6 2 9.'iYg6+ �g 7 30 .'iVh7+ \tf7 3 1 .�g6+ \tf6 3 2 .�h5 ! ; while 2 7 . . . g 6 ? allows 2 8 .�xg6! hxg6 2 9.iYxg6+ �g 7 3 0 .ttJg5) 2 8 .�xe4 dxe4 29Jlab I 'iYxfl 3 0 .�f1

Due to the poor coordination of his pieces and his exposed king, Black will lose the pride of his position - the pawn on fl . B ) 2 5 . . . bxa5 ! 2 6 . ttJd6 ! ( 2 6 .'iVh5 i s a draw after 2 6 . . J lxe l + 2 7 . l:lxe l g6 2 8 . �x g 6 hxg 6 2 9 .'iVxg 6 + �g7 3 0 .ttJh6+ \th8 3 1 .ttJf5 (the only win­ ning attempt) 3 1 . .. �f6 3 2 .iYh6+ �g8 3 3 .'iVg6+ �h8) and now:

analysis d i agram

22 .

...

'iVe5xa1 ??

This huge blunder is a losing mistake. The queen retreat to c7 is also bad, los­ ing after 2 2 . . . 'iYc7 ? 2 3 .ttJh6+ \th8 2 4 . iYf5 l:Ie4 2 5 J :lac l ! 'iVb8 (or 25 . . . ttJc4? 2 6 . �xe4) 2 6 .�xe4 dxe4 2 7 .�b4! .

ana lysis d i agram

3 0 . . . e3 ! 3 1..J:.xfl exfl 3 2 .ttJh6 + ! and now:

B l ) 2 6 . . . �d4 ? loses to 2 7 .ttJxe8 �xfl + 2 8 . \th l �xa l 2 9 .'iVe6+ �h8 3 0 .�xa l d4 3 1 . �g l and White is a piece up ; B2) Or 2 6 . . J:ixe 1 + 2 7 . 11xe l with the double threat of 2 8 .'iVe6+ and 2 8 .I;ib I ; 249

Chap ter

W i n n i n g C h e s s M i d d l e g a m es

B3) 2 6 . . . �c8 ! ! (creating a mating threat and saving Black) and now: B3 1 ) It is not too late to go wrong and walk into the mating net with the 'logical' 2 7 . 'iYhS ? ?

I n this relatively short game, Black suc­ cumbed under the pressure and his blunder abruptly ended the game. However, the middle game position, typical for this structure, is a beautiful learning example, showing a multitude of dynamiC possibilities and strategic ideas for both sides. Perhaps I have given too many lines and the reader got lost in the trees of varia­ tions. But my advice is to analyse the comments to this game thoroughly, be­ cause it will improve the reader's un­ derstanding of these positions.

a n alysis d i agram

QI 4.8 (E I 2 ) 2 7 . . . 'iYxf2 + ! ! 2 8 . �xf2 �d4+ 2 9 . �g3 �f2 mate; B3 2) 2 7 .'iVf3 ! �xe 1 + 2 8 . l:1xe 1 g6 2 9 .'llixdS + �h8 and a draw is the likely outcome of this tactical mess , for exam­ ple : 3 0 .ttJxc8 �d4 3 I . .t:rfI �xf2 'iYf2 3 2 . �h 1 ldxfI + 3 3 . �xfI 3 4 .iVd 8 + �g 7 3 S . 'iVd 7 + �h6 3 6 .'lli h 3 + �g 7 3 7 .�d7 + . 23.

Re1 xa 1

�f6xa1

GAME 43

Alisa Galliamova Sergey Tiviakov Elista President's Cup 1 9 98 ( 1 ) 1.

ttJg1 -f3

2.

c2-c4

b7-b6

3.

ttJb1 -c3

�cS-b7

4.

d 2-d4

e7-e6

S.

a2-a3

d 7-dS

6.

c4xdS

ttJf6xdS

7.

e2-e3

ttJ bS-d7

S.

�f1 -d3

�fS-e7

9.

0-0

ttJdSxc3

b2xc3

c7-cS

11.

e3-e4

0-0

1 2.

iVd1 -e2

1 0.

ttJgS-f6

The pawn structure is the same as in our previous game Kasparov-Najdorf. with the difference that Black's knight is now developed on d7 instead of c6. The advantage of this is that it can be trans­ ferred to f6 or, in some lines (should . . . e6-eS , d4-dS materialize) , to cS , the c-file remains open and the rook can sometimes enter via c3 . Moreover, White's d4-dS break does not come with tempo. The disadvantage is that White's d4 pawn is not attacked. 1 2.

...

'iYdS-c7

Tiviakov does not have the best experi­ ences with Galliamova in this line. In their mini-match in Elista, in the previ­ ous game where Galliamova played White, Tiviakov chose 1 2 . . . 1J.c8 and let's see what happened: 1 3 .�b2 (in this line White sometimes also decides to develop his bishop to the h2 -b8 diagonal with 1 3 .�f4. The move order with 1 2 . . . 'iYc7 excludes this possibility) 1 3 . . . 'iYc7

4:

P a wn M a j o r i t y i n t h e C e n t re

variably prove that his rooks had been put on the best squares from the very beginning. In this particular situation I believe that White's f1 rook should stay on f1 in or­ der to the pawn push f2-f4, while the only question, in my opinion, remains what is the best square for the other rook. However, you should take my opinion with a grain of salt, since two considerably better players than my­ self believed that the f1 rook should be placed on d I ! Let's see a few examples: 1 4 . ttJd 2 . A natural move, preparing f2 -f4. A) One of my own games was rather instructive and continued 1 4 . . . ttJf6 ! ? A I ) Now, executing the commenced plan by trying to build an attack using the earlier discussed plans with e4-eS would not work for White: 1 5 .eS ttJdS 1 6 . 'iV e4 g6 1 7 . 'li' g4 cxd4 1 8 .cxd4 'li'd7 1 9 .ttJe4

a n a lysis d i agram a n a lysis d i agram

24.

ttJfSxg 7 !

Now it's over. 24.

...

2S.

�d 2-h6

Black reSigned. 250

�a 1 xg 7

It is always difficult to decide where to put the white rooks in such positions. Should it be f1 I d 1 , or e l l d 1 , or d 1 I a I , or maybe d l le I ? Former World Cham­ pion Anatoly Karpov, in his prime, had a fantastic natural feeling for these deci­ sions. The course of the game would in-

1 9 . . . fS ! (with the white bishop on b 2 , this move i s a solution t o the problem) 2 0 .exf6 ttJxf6 2 1 .ttJxf6 + �xf6 and Black has an excellent game. The b 7 bishop will move t o dS , defending the e6 weakness , while the f6 bishop pro­ tects the king and is hitting White's weak d4 pawn; 25 1

Chapter

W i n n i n g C h es s M i d d l eg a m e s

A2) I S .f4 bS ! and now:

A2 3 ) O n the other hand, a move like 1 6 .g4! ? definitely deserves attention and could be important from a theoret­ ical point of view; A24) 1 6 .1:[f3 g 6 ! (a good prophylac­ tic defence; if 1 6 . . Jlfd8 , White takes the initiative with 1 7 J::rh3 ! h6 1 8 .l:tfl ) 1 7 .�h l tiJhS 1 8 J :lafl .

derstand, so let's see how the game con­ tinues after the 'lesser' move : I s . . . tbf6 ( I S . . . tbb8 was Karj akin 's choice against Svidler in Foros 2 0 0 8) 1 6 .a4 cxd4 1 7 .cxd4 itb4 1 8 .tiJf3 �e7 1 9 .h3 h6 2 0 .�e3 as 2 1 .1:ldc l tbe8 2 2 . tbeS tiJf6 2 3 . tbf3 tbe8 24. tbeS .

a na lysis d i agram

A2 1 ) The bS pawn is taboo, because of the simple 1 6 .itxbS ? cxd4 1 7 .cxd4 'iYc2 ! ; A 2 2 ) The idea to start rolling the e­ and f-pawns looks promising, but does not bring more than a draw after 1 6 .eS c4 ! (locking in White's b 2 bishop) 1 7 .itc2 tiJdS 1 8 .fS

analysi s d i agram

1 8 .. .f6 ! (often a standard reaction, and one to take note of!) 1 9 .fxe6 (it is advis­ able for White to force a draw, since 1 9 .exf6 ? is bad due to 1 9 . . J�xf6 20.fxe6 tiJf4; while after 1 9 .nae l fxeS 2 0 .'iVxeS �d6 ! 2 1 .�xe6+ �h8 2 2 .tbf3 tbf6 all black pieces are working harmoniously and White will come under attack) 1 9 . . .fxeS 2 0 .'iVhS g6 2 1 .itxg6 hxg6 2 2 .'iYxg6 + \t>h8 2 3 .�h6 + ; 252

analysis d i agram an alysis d iagram

And now Black fights for the central squares with a plan already shown on a few occasions; among others, in the comments to our previous game Kasparov-Najdorf: 1 8 ... c4! (first closing the diagonals that lead towards the black king) 1 9 .�c2 fS ! . The white at­ tack is stopped, while the b2 bishop has been reduced to a mere pawn. I had to struggle to ultimately save the draw from a worse position in Sokolov­ Hracek, Calvia Olympiad 2 0 04. B) 14 . . J ::lfd8 is another common move. B 1 ) 1 5 J lfd 1 is a strategy that I do not agree with, although it has been em­ ployed by no less players than Kasparov and Svidler. It looks to me that the rook is indeed perfectly placed on fl and the correct way for White to proceed is I S .f4! with initiative on the kingside. On the other hand, there may be some­ thing in this position that I do not un-

White is slightly better, but it is difficult to improve his position : 24 . . . tiJf6 2 S .tbc4 llc6 2 6 .'ifg3 ttJhS 2 7 .'iYe3 ttJf6 2 8 . !:tc2 .tIdc8 2 9 . �ac l 'iVd8 . White has somewhat more space, but Black is holding all the vital squares, Kasparov-Kramnik, Linares 2 0 04. B2) I S .nad l cxd4 1 6 .cxd4

4:

P a wn M a j o r i t y i n t h e C e n t r e

cellent post on cS - one of the points of the development of this knight to d7 . This strategy often works for Black) 1 7 J :lc l 'iVb8 1 8 . Ilxc8 �xc8 ! (if 1 8 . . . �xc8 , 1 9 .tbf3 (not 1 9 .itbS ? ! .tIc2 ! 2 0 .ita l 'iVc7 ) 1 9 . . . exd4 (now, on 1 9 . . . itd6 ?! White has 2 0 .itbS) 20.ttJxd4 ttJeS 2 1 . ttJfS and the white knight has reached its dangerous attacking post on fS) 1 9 .dS (now, in case of 1 9 .ttJf3 , Black's bishop is well placed, controlling the fS -square, and Black gets a good game after 1 9 . . . exd4 20.tbxd4 ttJeS 2 1 .itbS itcS) 1 9 . . . tbcS 2 0 .itc4 bS ! . Seizing the initiative o n the queenside. It should be noted that as a conse­ quence of the . . . e6-eS , d4-dS transac­ tion, the newly created pawn structure in the centre makes it more difficult for White to attack Black's kingside, since both white bishops are locked in at the moment. White's main idea is to in­ crease the pressure on Black's centre by playing f2 -f4, hoping that Black will have to take . . . eSxf4, after which White will get strong mobile pawns in the centre (dS , e4) and his bishop pair will start working again. 2 1 .ita2 ttJa4

analysis d i agram

1 6 . . . eS ! ? (Black allows White to get a protected ed pawn in the centre, but on the other hand, he takes control of the dark squares, so his knight will get an ex-

a n a lysi s d iagram

2 2 . ita I ! (correctly sacrifiCing a pawn) 2 2 . . . itxa3 2 3 .f4! (White is executing his main strategic idea) and now: 253

Chapter

W i n n i n g C h ess M i d d l eg a m e s

B 2 I) A defence that would demand from White to show the best of his at­ tacking abilities is 2 3 . . . exf4 ! and we have a very sharp tactical struggle ahead, for instance : 24.�h l ! (the im­ mediate 24.e5 ? fails to 24 . . . 'iVb6 + 2 5 stt h l 'ife 3 ) 2 4 . . . �b4 ! (preparing . . . �b4-c 3 , to cut off White 's dark-squared bishop) 2 5 .e5 �c3 .

(3 0 . . . �d 7 ? loses to 3 1 .'iVg5 ! ) 3 1 .'iVxg4 - the black king is still exposed , his a2 knight is stranded, and White's initia­ tive continues ; B2 1 2) 2 6 . . . �xe6 ! (this ingenious so­ lution is probably Black's best defence) 2 7 .tLle4! (it is important for White to keep his attacking potential alive. The logical 2 7 . dxe6 would lose after 2 7 . . .Ilxd2 2 S . exfl + �fS 2 9 .'iYe4 �xa2 3 0 .'iVxh7 nxa l ! (not 3 O . . . 'it>xfl ?, be­ cause Black loses his a2 rook after 3 1 .iYh5 + g6 3 2 .'iVd5 +) 3 1 .'iVg S + �e 7 3 2 . �xa 1 'lifS 3 3 . Id d 1 tLlc5 !) 27 ... �xa l 2 S .dxe6 and now:

3 2 . tLlg5 and White wins) 3 0 . 'iVd3 ! 'iV e 7 3 1 . �xf4 �fS (if 3 1 . . . �f6 , 3 2 .'iYfl Wins) 3 2 .'iYxb5 . White has re­ trieved all his material , while his attack continues ; B 2 1 2 2 ) 2 S . . . tLlc3 ! (exchanging a stranded knight for a well-placed one) 2 9 . exf7 + 'it>hS 3 0 . tLlx c 3 �xc 3 3 1 . �xf4. With his strong pawn on f7 White is better, but probably he does not have enough pieces to finish off. So, Black is likely to escape with a draw. Back to the position after 2 2 .�a 1 �xa3 2 3 . f4.

ana lysis d i agram

With only one bishop working, It IS considerably more difficult for White to make use of his strong, mobile central pawns : 2 6 .e6 ! (2 6 .d6? is an obvious mistake after 2 6 . . . �xa l 2 7 J ixa l tLlc3 2 S . 'iVf3 tLlxa2 2 9 . �xa2 �e6) and now: B2 1 1 ) If 26 . . . �xa 1 , White proceeds with 2 7 . exf7 + �fS 2 S J �Ixa 1 ( 2 S .'iVh5 ! ? is an option) 2 S . . . tLlc3 29. 'iVh5 ! tLlxa2

analysis d i agram

B2 1 2 1 ) The attacking queen central­ ization with 2 S . . . 'iYe5 , in general a good idea, would not help here : 2 9 .exfl + WhS (or 2 9 . . . �fS

analysis d i agram

a n a lysis di agram

3 0 . �e I ! and Black is forced to return his extra piece with 3 0 . . . �g4 !

3 0 .'i¥e l ! �d4 and now trouble comes from the other side : 3 1 .'iYh4 ! �b8

254

ana lysis d i agram

B 2 2 ) 2 3 . . . �c5 + (this check is the be­ ginning of Black's troubles and it was definitely better to wait with it. Since now White achieves his strategic obj ec­ tives, the position - even if it is very complicated - will be considerably eas­ ier to play from the white Side) 24.�h l and now: B2 2 1 ) 24 . . . exf4? is not good, since after 25 .e5 , compared to the previously commented positions, White has both of his bishops working excellently behind his powerful , mobile centre pawns : 25 . . . b4 (or 2 5 . . . �e3 2 6 .e6) 2 6 .e6 ! fxe6 (or 2 6 . . . 'iVc7 2 7 .exf7 +

4:

P a wn M a j o r i t y in t h e C e n t r e

'iVxfl 2 S .d6 ! 'iYxa2 2 9 . 'iVe 7 ) 2 7 .dxe6 �e7 2 S .'iYh5 �fS 2 9 . �b l g6 3 0 .�xg6 and Black is mated; B2 2 2) 24 . . . �d4 2 5 .fxe5 �xa l

a n a lYSis d i agram

2 6 . 'iYh5 ! (an important zWischenzug, probably missed by Black) 26 . . . �fS ( 2 6 . . . g6 does not work due to 2 7 .'i¥g5 ! �eS 2 S .d 6 ! �e6 2 9 .�xe6 fxe6 (not 29 . . . �xe6?? 3 0 . d7 and White queens) 3 0 .�xa 1 and White's protected ed pawn on d6 , in combination with Black's weak king, is worth much more than Black 's two ers on the queenside) 2 7 .�xa l tLlc3 2 S .�b3 and now:

a n a lYSis d i agram

B2 2 2 1 ) 2S . . . 'iYb 6 ? (this natural move is a losing mistake, because Black will always be ' one move too late' until the end of the game) 29 .�fl 'iYe3 255

Chapter

W i n n i n g C h ess M i d d l eg a m e s

3 1 .e6! (not 3 1 .tbf1 ? 'iYd4 ! 3 2 .�c2 g6 with 33 . . .tbxe4 to follow, and Black wins) 3 1 . . . 'iYxd2 3 2 . exf7 + �h8 (3 2 . . . thf7 ? ? 3 3 .d6) 3 3 .'iYeS ! (3 3 .'iYh4? loses to 3 3 . . . gS ! 3 4.'iYh6 l:Ixf7)

analysis diagram

3 0 .e6 (pushing the other pawn would also have won : 3 0 .d6 �e6 (or 30 . . . g6 3 1 .'lWh4! 'iYxd2 3 2 .'iWf6 �fs 3 3 .e6 ! and Black is soon mated) 3 1 .�xe6 fxe6 3 2 .d 7 ! and White wins) 30 ... g6 (or 3 0 . . . 'li'xd2 3 1 . e 7 ) 3 1 . 'iVf3 'iYxd2 (3 1 . . . �xf3 loses to 3 2 . �xf3 tbe2 3 3 .e 7 lie8 34.d6) 3 2 .'iVf6 ! (even more brutal than the immediate 3 2 .e 7 ) 3 2 . . . aS 3 3 .e 7 1 -0 , Galliamova-Tiviakov, Elista President's Cup ( 1 . 2 ) 1 9 9 8 ; B 2 2 2 2 ) The only way t o stay i n the game was 2 8 . . . aS ! , threatening . . . a4, in order to remove White's bishop from the dangerous diagonal : 29 JlxaS . Black has sacrificed a pawn, which is rather irrelevant here, but he has won time, which will save his life. The following entertaining line ends in a perpetual check: 29 . . . 'iYb6 3 0 . �a l 'iWe3

1 4.

Black wants to provoke the weakening g2-g3 in order to give more impor­ tance to his b7 bishop. It is question­ able, however, whether this was indeed worth two tempi here. 1 6.

92-93

'iVf4-c7

1 7.

h4-h5

h7-h6

1 8.

tLJf3-h21

lla 1 -d 1

3 3 . . . �h3 ! (only move) 3 4 . gxh 3 (3 4.�g 1 is a draw after 34 . . . tbd 1 3 S .�xd l �xg2 + 3 6 J:rxg2 'lixd l + 3 7 . I:ig l 'iYf3 + 3 8 . It g 2 �f1 +) 34 ... 'iVd3 3 S .'iffs tbxe4 1 4.

3 6 Jda8 'iYb l + 3 7 .�g2 'iYb 2 + 3 8 .�g l 'ifc l + 3 9 .�g2 'ifb2 + with a draw. 1 3.

256

P awn M aj o r i t y in t h e C e n t r e

analysis d i agram

analysis d i agram

an alysis d i agram

(also good is I S .�c l as 1 6 .�b l �a6 1 7 .eS b4 1 8 .axb4 axb4 1 9 .�gS tbb6 2 0 .'lWe4 g6 2 1 .'iVh4 �a7 2 2 .�f6 �xf6 2 3 .exf6 tbdS 2 4.�e4 with a huge ad­ vantage in Carlsen-Aronian, Mexico City 2 0 0 7 ) I S . . . aS 1 6 .�c l �a6 1 7 .eS 11£b8 1 8 .tbgS tbf8 1 9 .'lWhS �xgS 2 0 .�xgS �b 7 2 1 .h4 'iVc6 2 2 .f3 'iVe8 2 3 .'iVg4 �dS 24.�f6 tbg6 2 S .hS gxf6 2 6 .exf6 with a big advantage for White in Sokolov-Leko, Wijk aan Zee 2 0 04.

4:

�c1 -b2

lla8-d8

Advancing the queenside pawns with 1 3 . . . c4 ? 1 4.�c2 bS would be a strategic error, since it takes the pressure off White's centre and hence gives White a free hand to concentrate on the kingside attack, for example: 1 5 JHb 1

...

tLJd7- b8

Black transfers his knight to c6 in order to exert pressure on White's d4 pawn. This is a well-known strategy, seen in a number of games. On the other hand to be honest, I have never understood this kind of strategy. It may well be that the black knight indeed belongs on c6, hitting White's d4 pawn, but then the question is: what was it doing on d7 in the first place? Wouldn't it have been better to develop it to c6 immediately, saving two tempi - since White's moves have been logical, and the placing of the knight on d7 has not forced White to make any unnatural decisions? 1 5.

1 8.

...

c5xd4

1 9.

c3xd4

tLJb8-c6

20.

�d3-b1

tLJc6-a5

The counterplay attempt with 20 . . . eS would not work well :

h2-h4

This move is always useful in these positions. 1 5.

This type of knight transfer is not often seen in such positions. However, it cor­ responds to the needs of the position and makes sense here, the only draw­ back being that it temporarily removes the knight from the control of the cen­ tral squares d4 and eS .

...

'ii c7-f4 1?

an alysis d iagram

A) The exchange sacrifice 2 1 .dS tbd4! 2 2 .�xd4 looks interesting, but it is ques257

Chapter

W i n n i n g C h es s M i d d l eg a m e s

tionable whether it gives White enough compensation: 2 2 . . . exd4 2 3 .�xd4 (Black is better after 2 3 .ttJf3 �cS 24.ttJxd4 fS !) 2 3 ... �xa3 24.ttJg4 (if 24.'iYg4, 24 . . .fS ! (White has enough compensation after 24 . . . f6 2S .eS ! fxeS 26 .iYg6) 2 S .exfS 'iYe7 ! 26 .f6 ,g:xf6) 24 . . . �c8 ; B ) 2 1 .iYc2 ! 'iYd6 2 2 .dS ttJaS 2 3 .ttJg4 and White is better. 21 .

follow, White gets a devastating attack, which is not surprising, since all the white pieces are harmoniously partici­ pating in the attack, while Black's ttJaS and �b 7 are completely out of play. 24.

...

25.

e4-e5

e6xd5

J:!:d 1 -c1

21 .

...

Wic7-d 7

22.

tLlh 2-g4

ltd8-c8

28.

...

Wid7-a4!

Good , active defence. Black is ready to meet 2 6 .iYd3 with 2 6 . . . iY'e4. 26.

d4-d5!

White decides that the time has come for a thematic central pawn sacrifice. It was also possible to calmly continue building the attack with 2 3 . 'li'd3 ! f6 (if 2 3 . . . b1xc 1 24.ldxc 1 fS , White is better after 2 S .ttJeS 'iYe8 2 6 .'iYe2) 24.ttJe3 (With the idea of a thematic follow-up with 2 S .dS exdS 2 6 .eS) and Black's de­ fence is far from easy. 23.

...

24.

iilb2xc1

l:tf1 -e1

With 2 6 .iYf3 , in order to threaten 2 7 . 'iY fS , White would have enough compensation, but Black also has his re­ sources for an active defence. The fol­ lowing entertaining line illustrates some pros and cons of the position : 2 6 . . . �c8 ! (only move) 2 7 .ttJe3 �cS ! 2 8 .ttJxdS ( 2 8 . ttJfS is an alternative) 2 8 . . . iYd4 2 9 .�f4 �e6 3 0 .ttJc7 �c4 3 l .l:Id 1 'fHb 2

J:c8xc1

White is eyeing the h6 pawn. However, the alternative 24 . .sxc l ! looks stronger, since after 24 . . . exdS (ignoring White's pawn on dS and trying to exchange more pieces with 24 . . J�c8 is probably a better defence) 2 S .eS with 2 6 . iY'd3 to

tbg4-f6+!

If 2 8 . ttJe3 , the rook pOSItIOn on e l gives Black a n important tempo : 2 8 . . . 'iYc3 2 9 .'iYe2 'iYxeS 3 0 .�b2 iYgS 3 1 .'iYd3 fS 3 2 .'iYxdS + \ii h 8 .

25 .

258

P a wn Ma j o r i t y in t h e C e n t re

3 2 .'iYe4! 'li'xf2 + 3 3 . �h l l:1b8 ! (3 3 . . .l:: k8 ? may look more natural as it attacks the white knight , but leaves Black's rook exposed and loses after 34 .�xh6 ! iY'xg3 (on 3 4 . . . gxh6 both 3 S .iYg4+ and 3 S .iYh 7 + win) 3 S .iYfS ! I;:Ib8 (or 3 S . . . 'iYh4+ 3 6 .�g2) 3 6 . e 6 ! fxe6 3 7 .'iYh 7 + �f8 3 8 .�xg 7 + 'fHxg 7 3 9 .l::t fl + ! �xfl 40 .ttJxe6 +) and now:

This move has its advantages, but the rook was also well placed on d 1 . The immediate 2 1 .ttJ g4 100ks promising.

23.

4:

an alysis d i agram

analysis d i agram

A) 34.�xh6? is not good because of 34 . . . iYxg 3 ! (34 . . . gxh 6 ? still loses to 3 S .iYh 7 + �f8 3 6 .iYxh 6 + � e 7 3 7 . 'iV g S + �f8 3 8 . 11d 8 + ,gxd8 3 9 .iYxd 8 + �g7 4 0 . h 6 + �xh6 4 1 .iYh8+ �gS 42 .iYh4 mate. With the white queen on e4, the text now threat­ ens mate) 3 S .iYh 7 + �f8 3 6 .'ifh8 + �e7 and now White's own king is un­ der attack; B) 34.e6 ! fS ! 3 S .iYxfS 'iYf3 + 3 6 .\th2 'iYf2+ 3 7 .�h3 �fl + 3 8 J:lxfl 'iYxfl + 3 9 .�g4 'iYd l + 40.�h3 'iVh l + and White cannot escape the perpetual check. 26.

...

'iVa4-c4

27.

'iVe2-f3

iilb7-c8

White correctly judges that all his pieces are already in perfect positions and a better moment to sacrifice is not likely to come. The time to strike is now!

28.

...

29.

e5xf6

g7xf6

White has correctly sacrificed her knight, but with the best defence it still does not promise more than a draw. However, the position is complicated and under such circumstances , the best defence mostly does not materialize in practice.

29 .

...

iile7-d8?

Black wants to keep the f6 pawn under attack, but this will turn out to be irrel­ evant. The text is a losing blunder. A) 2 9 . . . .ig4?? runs into a forced mate : 3 0 . fxe 7 �xf3 3 1 .exf8iY+ \txf8 3 2 .�xh6+ �g8 3 3 . l:le8 mate; 259

Chapter

W i n n i n g C h ess M i d d l eg a m e s

B) It was important to stick to active defence with 29 . . . �cS ! . With an active bishop on cS Black would have counterplay, hitting the f2 -square and having threats of his own. Also, last but not least, White's queen transfer via e3 is now impossible. B l ) 3 0 .�xh6 is a forced draw after 3 0 . . . �g4 !

a n a lysis d i agram analysis d iagram

2 60

P awn M a j o r i t y in t h e C en t re

good due to 3 1 . . . 'iYxf2 + 3 2 .'iYxf2 �xf2 + 3 3 .�xf2 �xfS 34 . .t1eS l::t d 8 35 . �xfS lLJc4 and the black d-pawn will be fast) 3 1 . . J �txc8 3 2 .�xh6 lLJc4 (bringing the knight back into play and preventing 3 3 . l:t e S ) 3 3 . 'iff5 (only move) 33 .. :�Yxf2 + 34.'iYxf2 �xf2 + 3 S .'iitxf2 d4. The bishop on f8 cannot be taken because of 3 2 ... �xf8 3 3 .'ifd6+ 'iit g 7 34.h6+ ! .

3 1 JleS ! - only move : B l l ) 3 1 . . . 'iYc l + 3 2 .�xc l �xf3 is a draw after 3 3 . ..th6 nd8 . The winning attempt 3 3 . . . �e4 ? loses after 34.�xe4 dxe4 3 S . �xf8 <;i(xf8 ( 3 S . . . �xf8 ? 3 6 J Ie8 lLJc4 3 7 .h6) 3 6 . l::t xe4 �xa3 3 7 .g4! and the white pawns cannot be stopped; B 1 2 ) 3 1 . . . �xf3 3 2 J l g S + �h8 3 3 .�g 7 + h8 3 2 .�xh6 � g 8 3 3 .�g 7 + l::t x g7 3 4 .fxg 7 + 'iit x g 7 3 S .h 6 + ! �xh6 3 6 .'iYf6 + h7 3 7 :iYxf7 + gS 3 9 :ifg6 mate) 3 1 .�xc8 (3 L �.e3 ? blunders to 3 1 . . . iY'xf6; while 3 1 .�xh6?! is also not

4:

This sharp ending seems to be equal, for instance : 3 6 .g4 l::td 8 (if 3 6 . . . d3 , 3 7 J:ie4! d2 ? 3 8 .11d4) 3 7 .gS d3 3 8 . g 6 ! d2 3 9 J :t e 7 d l 'iY 4 0 . gxf7 + � h 8 4 1 . 11e8 +
�c1 xh6

This move is good enough for the win. However, considerably stronger was 3 0 :iYe3 ! �xf6 3 1 .'ifxh6 and Black can­ not prevent the mate. 30.

...

'if c4-g4

If 3 0 . . . 'ifd4, 3 1 .�g7 ! and the march of White's h-pawn cannot be stopped: 3 1 . . :iY'd2 3 2 .l::td l 'ifgS 3 3 .h6 �xf6 3 4.'�xf6 'iYxf6 3 S .�xf6 ; or 3 0 . . . �g4 3 1 .Vi'e3 �xf6 3 2 .�xf8
'iVf3xd5 !

�dSxf6

Or 3 1 . . . �b 7 3 2 .iYd3 . 32.

�h 6xfS

�cS-b 7

33.

.l::t e 1 -eS!

'ifg4-a4

34.

�b1 -h7+

�gSxh7

35.

'iVd5xf7+

1 -0

Being a l .d2 -d4 player, I have obviously had a considerable number of 'pawn majority in the centre' structures - and things have gone right many times , and wrong many times. Below I shall give two interesting examples where things went right.

GI 4. 1 (D8 S) Ivan Sokolov

GAME 44

David Howell London 2 0 0 6 (5) ttJgS-f6

1.

d2-d4

2.

c2-c4

g 7-g6

3.

ttJb1 -c3

d 7-d5

4.

c4xd5

lLlf6xd5

5.

e2-e4

ttJd5xc3

6.

b2xc3

�fS-g7

7.

'iVd1 -a4+

The purpose of this check is to slightly disrupt Black's regular Griinfeld devel­ opment. However, given the fact that White is also losing time with his early queen moves in the opening, the main idea behind the move is to avoid the di­ rect, sharp lines of the Griinfeld.

7.

...

'iVdS-d7

The other main move is 7 . . . lLJd7 . Then the idea of 7 .'iYa4+ becomes clear. With a knight on d7 , just like in our previous game Galliamova-Tiviakov, it would be more difficult for Black to ex­ ert pressure on White's d4 pawn. S.

'iYa4-b3

0-0

9.

ttJg1 -f3

b7-b6

This is the natural Griinfeld place for Black's light-squared bishop. 1 0.

�c1 -e3

ltcS-b7

11.

�f1 -d3

c7-c5

1 2.

0-0

ttJ bS-c6

1 3.

l:ta1 -d 1

After a logical sequence of developing moves on both sides, we get a position which, as regards pawn structure, re­ sembles our Kasparov-Naj dorf example (Game 42) , with the important differ­ ence that Black's bishop is now fianchet­ toed on g7 instead of developed on e7. This has some pluses: White's d4 pawn is easier attacked and also the black king is better protected against different sorts of direct attacks. The minus is that, should White be on time with the pawn push h2 -h4-hS - and most of the time he is - Black's kingside may qUickly be­ come vulnerable to attack. 1 3.

...

ttJc6-a5

1 4.

'ifb3-b1

c5xd4

1 5.

c3xd4

261

Chapter

W i n n i n g C h ess M i d d l eg a m e s

change it) 2 2 . . J:txc4 2 3 .'iYe3 'iYc3 24 .!Id3 'iYb2 2 S .�h2 !Ic3 2 6 .ned 1 h5 2 7 . lIxc3 �xc3 2 8 .'iYgS 'iVc 7 + 2 9 .f4 (Black is unable to develop any counterplay to counter White's king­ side attack. Note the miserably ive position of Black's bishop on b 7 ) 2 9 . . . .tIe8 3 0 .ltJfS . 1 5.

...

e7-e6

The main idea behind this move is to prevent d4-dS for the time being. The most popular move is actually l S . . . lIac8 . White then indeed pushes 1 6 .dS and play usually continues 1 6 . . . ltJc4 1 7 .itd4 itxd4 1 8 .ltJxd4, and now: A) If Black does not fix the centre and opts for imaginary activity with 1 8 . . . 'Ii'a4 ? ! , things can qUickly go wrong, since the black king is left vul­ nerable and without defenders, while Black's activity on the queenside does not generate nearly adequate counter­ play: 1 9 J :tfe 1 nc7 2 0 .h4 ! (whenever possible, this is always a useful move for White in the Grunfeld Indian Defence) 20 . . . 'iY as ?! 2 1 . 'Ii' c I ! (transferring the queen to the kingside) 2 1 . . . �fc8

2 2 .itxc4! (the black knight is well placed on c4 and it is a good idea to ex262

P awn M a j o r i t y in t h e C en t re

B 2 ) Or 2 1 . . . ita6 2 3 .�c 1 bS

2 2 . b!xc4 itxc4

a n alysis d i agram

2 4 . 'ifa 1 ! (an important tempo) 24 .. J l e 8 2 S .ltJd2 nc8 2 6 . a4 a6 2 7 .axbS axbS 2 8 .'�' a 6 'ifcs 2 9 .h4 and Black will soon have problems defend­ ing his vulnerable king; B 3 ) 2 1 . . . llfc 8 2 2 . �xc4 llxc4 2 3 .'Ii'd3 �cS

analysis di agram

White had a strong attack and soon won in Milov -Krasenkow, Warsaw Ech 2005. B) Two rounds later i n the same tour­ nament , Krasenkow refrained from 1 8 . . .' ii'a4 and fixed the centre with 1 8 . . . eS . This is a standard idea here.

analysis d i agram

analysis d i agram

Black is that the knight on c4 has almost reached its perfect position : on d6. A potential problem for Black is that his king can become vulnerable. I experimented with this line from the white side in 2 0 0 6 and had an interest­ ing game against Krasenkow. After 1 9 .ltJf3 , I had expected that Krasenkow would follow his previous game and proceed with 1 9 . . . 'iYe 7 . It may be inter­ esting for the reader to know what I had prepared for that event.

4:

As shown in similar positions in Galliamova-Tiviakov, this type of plan is positionally sound and slows White down. Here, an additional plus for

analysis d iagram

The main idea of both 1 9 . . . 'ife7 and 1 9 . . . f6 is to defend the eS pawn, so that Black can place his knight on d6. In my opinion, if the black knight reaches the d6-square, it is worth more than White's bishop on d3 . Also, I think that if White gets queen + knight versus queen + bishop in this pawn structure, he will have the somewhat better chances. So my idea was to proceed with 2 0 .�xc4 ! (the earlier Krasenkow game continued 2 0 J �c 1 ltJd6 2 1 .g 3 f6 2 2 .ltJd2 l:1xc 1 2 3 . 11xc 1 lIc8 24Jixc8 + �xc8 and a draw was soon agreed in Naer-Krasenkow, European Champion­ ship, Warsaw 2 0 0 5 ) 20 . . . �xc4 2 1 .1dc 1 and now: B 1 ) White is better in the case of 2 1 . . J db4 2 2 J�k7 ! 'iYxc7 2 3 .'iYxb4;

a n a lysis d i agram

24.'ifa3 ! and now: B3 1 ) If Black attempts to obtain counterplay with 24 . . .fS , after a rela­ tively forced sequence of moves White will be better in the endgame : 2 S . 'ifxa7 ! fxe4 2 6 . ltJ g S ! 'iYxg 5 2 7 .'ifxb 7 e3 2 8 .fxe3 'iYxe 3 + 2 9 . �h 1 'iYf4 ! 3 0 . 11g 1 ! 'iYf8 3 1 . 'iYxb6 (or 3 1 .'iVd7 'iff? 3 2 .'�' d 8 + iVf8 3 3 .'ifxb6 llxdS , transposing) 3 1 .. J�xdS 3 2 .a4. 263

C h a p t e r 4: P a wn Ma j o r i t y in t h e C en t re

W i n n i n g C h es s M i d d l e g a m es

White's ed a-pawn is strong, while Black's king remains exposed. Please note that plenty of rook endings are lost for Black; B3 2 ) 24 . . . �c8 2 5 Jlc 1 nc2 ! 2 6 .'ife3 'ifc7 27 Jlxc 2 'iYxc 2 2 8 .h3 f6 .

lems, since White seems to be better af­ ter 3 0 .�xh6 'iYxe4 3 1 .g5 ! ; B3 2 3 ) 2 9 . . . 'iYxa2 . I f White is to have the initiative, Black may at least have a pawn to compensate for his troubles such is the logic behind this pawn snatch. Now: B3 2 3 1 ) The immediate 3 0 .g5 does not bring more than a draw after 3 0 . . . iYa 1 + (not 3 0 . . . f5 ? 3 1 . 'ii' c 1 ! ) 3 1 .
an a lysis d i agram

This is the kind of position I wanted to reach. White's d5 pawn is strong, well-defended and not easy to block; instead of a strong knight on d6 , Black may have a lousy blocker in the bishop on d 7 . Furthermore, Black may have problems defending his e5 pawn should the f6 pawn be exchanged, and Black's king is very vulnerable; as will be shown in a number of lines, the king has no defenders at all, while White's queen + knight are a deadly attacking duo. My - probably correct - assess­ ment was that this position should favour White. My analysis continued 2 9 .g4! (intend­ ing 3 0 .g5 in order to create a black weakness on e5) and now: B3 2 1 ) The easiest way to stop White's plan is 29 . . . g 5 ?, but this loses to 3 0 .tLl xe5 ! fxe 5 3 1 . 'ifxg 5 + �f7 3 2 .'ifh5 + �f8 3 3 . 'iYxe5 �f7 3 4.'ifh8 iYb 1 + 3 5 . cJJ h 2 ; defence ingenious B3 2 2 ) The 29 . . . h6 ! ? does not quite solve the prob264

analysis d i agram

3 3 . . . �b 7 ! 34.ctJxe5 (or 34.'iVxe5 �xe4 3 5 . 'ife 8 + �g 7 3 6 . 'ife 7 + draw) 3 4 . . . 'iYxd 6 ( 3 4 . . . �xe4+ ? loses to 3 5 . f3 ) 3 5 .'iYc4+ 'It>g7 3 6 .'iff7 + 'It>h8 3 7 . ctJc4 'if c5 (White has some winning chances after 3 7 . . . �xe4+ 3 8 .f3 ) gxf5 3 9 . exf5 'iYxc4 3 8 . 'ifxb 7 40.'iYb8+
3 2 . ctJxe5 fxe4 3 3 .ctJg4! the black king finds itself in a mating net)

Black that the e5 pawn can be a weak­ ness now) 2 2 . 'iVb2 �c7 2 3 ."iYa3 ! f5 ! (this standard break is obligatory) 24.exf5 gxf5 2 5 .'iYxa 7 �xd5 2 6 . 'iYxb6 �xa2 2 7 . l:lfe 1 (strangely enough, this position resembles some kind of Sveshnikov Sicilian. White has a small advantage and Black should be careful) 2 7 . . . e4

analysis d i agram

Now the black king has no defenders at all and White wins nicely with 3 2 .gxf6 ! 'iYxf3 3 3 .'iYc6 ! and Black is mated. One may well imagine that after spend­ ing quite some time to come up with the analyses shown above, I was quite disappointed to see a beautiful possibil­ ity, along with all my homework, being thrown out of the window when Krasenkow surprised me with 1 9 . . . f6. So I was left on my own, looking for over-the-board solutions. I decided that taking on c4 would be less efficient now, but that, on the other hand, White might have a nice attacking manoeuvre in ctJh2 -g4.

a n a lysis di agram

2 8 .�b 1 ! �xb 1 2 9Jixd6 (the knight should be preferred to the light-squared bishop here, so this was a good trade) 29 . . Jlb7 3 0.'ifd4 'iYg7 3 1 .ctJfI 'iYxd4 3 2 . llxd4 �d3 3 3 .ctJg3 l:lb5 ? ! 34. �c 1 �e5 ? ! 3 5 .f4! l:re6 3 6 . .tId5 and Black was in serious trouble in Sokolov-Krasenkow, England tt 2 0 0 5 / 0 6 . 1 6.

1 6.

a n a lysis d i agram

2 0 .h3 ctJd6 2 1 .ctJh2 'iVe 7 (if 2 1 . .. f5 , then 2 2 . ctJf3 or 2 2 Jlfe 1 would remind

h2-h4

This standard plan is a good idea here as well. ..

.

f7-f5!?

This move is similar to plans shown in the comments on Game 42 - Kaspa­ rov-Naj dorf (see Vaganian-Razuvaev) and has a sound strategic idea behind it. White has to find an energetic way to stop Black from executing his strategic idea, which is to remove White's e4 pawn and take control of the d5 -square. One of the earlier games in this posi­ tion continued 1 6 .. J :tac8 1 7 .h5 gxh5 ! ? (an interesting solution to the problem; 265

Chapter

W i nn i n g C h es s M i d d l e g a m es

Black is not afraid to weaken his kingside) 1 8 . d5 lLJ c4 1 9 . .id4 f6 2 0 J�k l exd5 2 1 .exd5 lLJd6 (it is not clear to me why Black refrained from taking on d5) 2 2 .'iYb3 ld.xc l 2 3 .l:txc l Ik8 24.11d l with a complicated game in Milov-Rowson, tt 2 0 0 5 .

1 9.

...

�g 7xe5?

Taking a second pawn is a mistake. Black's dark squares are now very weak, while lLJa5 and .ib 7 are totally out of play for the moment. Necessary was 1 9 . . . 'ife6 ! and after 2 0 . .if4 White has a number of posi­ tional trumps (ed e-pawn, a nice d4-square for his knight, attacking pos­ sibilities on the kingside, etc.) that en­ sure his compensation for the sacrificed pawn, but there is a complicated struggle ahead. 20.

ttJf3xe5

21 .

�e3-d4

�d6xe5

This bishop is a monster and Black is falling under a deadly attack. 1 7.

�d3-b5!

This move prepares a standard pawn sacrifice, already seen in several earlier examples. If 1 7 .h5 fxe4 1 8 . .ixe4 .ixe4 1 9 . 'iYxe4 'iY d5 , Black gets what he wants and has a good game. 1 7.

...

�d7-d6

1 8.

d4-d5!

e6xd5

The pawn had to be taken. In the event of 1 8 . . . fxe4, White gets a strong attack after 1 9 .dxe6 'iYxe6 2 0 .lLJg5 'if g4 2 1 . 'ifb4 with 2 2 J:rd7 to follow. Also interesting is 2 1 . .id7 'iYxh4 22 . .ie6 + �h8 2 3 .g3 'iYh5 24.�g2. 1 9.

266

e4-e5!

21 .

...

'iYe5-e4

22.

'iYb1 -c1

'i¥e4xh4

2 8 J :re6 with 'ife5 to follow, and Black's position collapses. 23.

...

1:lf8-f7!

24.

'iYc7-e5

l:lf7-e7!

4:

P a wn M a j o ri t y in t h e C en t r e

30.

�e1 xe7 !

'iYh4xe7

31 .

'i¥g7-h6

�b7-c6

Preparing the king 's retreat. 32.

�d4-c3!

This original bishop manoeuvre I had seen when I decided to play on with 3 0 J:te 7 . 32.

...

33.

�c3- b4 !

�d6 -c7

The point behind White 's idea. This bishop is to be sacrificed in order to get to the black king. The sad news is that it does not promise more than a draw. Here I started to realize that things were not the way they seemed, but the next two checks had to be given anyway, so I played: 25.

'iYe5-h8+ �g8-f7

26.

'iYh8-g 7+ �f7-e6

Here I thought for almost one hour try­ ing to find something concrete, and I couldn't believe my eyes when I saw that the black king was safe in the mid­ dIe of the board.

23.

'iYc1 -c7?

When I played this move, I thought that Black's resignation would come in a few moves, completely missing Black's 24th move. There was no need to hurry. The last white piece not yet in play is the fl rook, and it would have been clever to . Je 1 ! , after which I proceed with 2 3 . a think that Black's position cannot be de­ fended, for example : 2 3 . . Jif7 24.'iVe3 ! 'iYe4 2 5 .'ifh6 'iVg4 2 6 .f3 'iYh5 2 7 .'iVf4 2 7 . . J laf8 2 8 . 'if e 5 ) (threatening

27.

llf1 -e1 +

�e6-d6

28.

�d4-e5+

�d6-c5

29.

�e5-d4+

�c5-d6

It is difficult to agree to a draw by per­ petual check in a middlegame where your opponent's most active piece is his king !

33 .

...

'iYe7-d7?

In time-trouble Black makes a losing mistake. 3 3 . . . 'if f7 ? should also lose after 34 . .ia6 ! f4 3 5 . ld.d4 ! . However, Black had to keep cool and take the sacrificed piece : 3 3 .. .' iYxb4 ! 3 4 . 'iYxh 7 + and now: �b 7 3 5 . 'iYg 8 + A) 3 4 . . . �b8 ! 3 6 .'iVf7 + \itb8 and White has nothing better than to give perpetual check; B) Black can also courageously move his king to the centre with 3 4 . . . \t>d6 ! ? 3 5 . 'iYxg 6 + �c5 3 6 . .ixc 6 �d8 ! ( 3 6 . . J :t c 8 ? ? loses to 3 7 . 'ii e 6 ; or 3 6 . . . lLJxc6 ? ? 3 7 .�c 1 +) 3 7 . .ie8 (an al­ ternative is 3 7 .l:Ic 1 + ! ? lLJc4 3 8 . .ib 7) 37 ... 'iYe4 L 267

C h a p t e r 4 : P a wn Ma j o r i t y i n t h e C en t r e

W i n n i n g C h e s s M i d d l eg a m e s

3S.

ii.aSxbS

�d 7xbS

analysis diagram

The black king looks very strange on c5 , but it may well be that White has noth­ ing better than the forced drawing line 3 8 J:k l + �b4 3 9 .a3 + �xa3 40.'ifg 3 + tt:Jb3 4 1 .'iYc3 Wa2 ! 4 2 . �c2 + �a3 43 .'ifb2 + �b4 44.'ifc3 + �a3 . 34.

�bS-aS!

Now Black's pieces are clumsy and his position soon collapses. 34.

...

3S.

ii.b4xaS !

�bSxaS

Exchanging all the minor pieces should, on general grounds, be a good idea for Black, but his king is much too exposed and will not get out of the mating net. 37.

'iYhSxh7+ �c7-cS

3S.

1:td1 -c1 +! �cS-bS

ii.cS-bS

If 34 . . . tt:Jb 7 , 3 5 .'iff4+ tt:Jd6 3 6 . l::t c I , and Black is subject to mortal pins. Black also collapses under the pins after 34 . . . �b 7 3 5 .'iff4+ �c8 3 6 .l::r c l + tt:Jc6 3 7 .�b5 ! a6 3 8 .�a4 b5 .

39.

'iYh7-b7!

Cutting off the black king's escape route.

an a lysi s d i agram

The pin on the a4-e8 diagonal has been solved, but new and bigger problems are just around the corner: 3 9 .�a5 ! bxa4 40 .'ti'e5 ! 'iff7 4 1 .'iYd6 and Black gets mated, either on d8 or on c 7 . 268

39.

...

40.

I:rc1 -b1 +

�bS-c4

41 .

'iYb7-cS+

�c4-d3 �d3-e4

1:taS-hS

42.

l1b1 -d1 +

43.

'iYcS-eS+ �e4-f4

44.

lld1 -d4+

Black resigned. He will be mated in two : 44 . . . �g5 45 .'ife 7 + �h6 47 . �h4.

Throughout my career I have regularly employed the Petrosian Variation with white as my main weapon against the Queen's Indian. In the beginning of the year 2 0 0 2 I spent some extra effort studying this line and the middlegame positions ariSing from it. Such efforts usually pay off and indeed, in the 2 0 0 2 - 2 0 0 5 period I scored well with this line and played a number of inter­ esting games. In my comments to this game I shall also give a few other interesting examples.

QI 4. 1 7 (E I 2 ) Ivan Sokolov

GAME 45

7.

...

tLldSxc3

S.

b2xc3

ii.fS-e 7

The move order with 7 . 'if c2 allows Black, as mentioned in the above com­ ments , to neglect the development of his kingside and attack the white centre immediately with 8 . . . c5 ! ? 9. e4 tt:Jc6.

Judit Polgar Hoogeveen 2 0 0 3 (5) 1.

d2-d4

tLlgS-fS

2.

c2-c4

e 7-eS

3.

tLlg1 -f3

b7-bS

4.

tLlb1 -c3

�cS-b7

S.

a2-a3

d 7-dS

S.

c4xdS

tLlfSxdS

7.

'iVd1 -c2

This is the other main move in the topi­ cal Petrosian Variation. 7 .e3 we have studied in Kasparov-Najdorf (Game 42) and in Galliamova-Tiviakov (Game 43 ) . White's idea is simple - he wants to play e2 -e4 in one move. However, later on the white queen mostly moves to e 2 and normally we get either a transposi­ tion or positions very similar to those ' after 7 .e3 . One of the reasons that 7 . 'V/iIc2 is more frequently played than 7 .e3 nowadays, is that white players have problems to find an advantage in the line 7 . e3 g 6 . 7 .'ifc2 , however, gives Black an extra pOSSibility in the line with 8 . . . c5 - see Sokolov-Kramnik in the comments to 8 . . . �e 7 .

analysis d i agram

This black set-up is not logical. His king is still stuck on e 8 . White has possibili­ ties to open up the centre with d4-d5 , play on the black king stuck in the mid­ dle of the board , and an unpleasant check or pin along the a4-d8 diagonal may come any moment. That said, so far white players have sur­ prisingly failed to find an advantage in this currently popular line. White can develop his dark-squared bishop to ei­ ther e3 or b 2 . A ) The main variation i s 1 0 .�b2 �c8 1 1 .�d l cxd4 1 2 .cxd4 269

Chapter

W i n n i n g C h es s M i d d l eg am e s

1 5 . . . fS ! (this familiar strategic plan, shown in a number of examples in this book already, works excellently here) 1 6 .exfS ildS ! 1 7 .'iYb 1 exfS + 1 8 .�fl 0-0 1 9 .ilxfs

analysis d i agram

1 2 . . . ild6 ! . This move by Joel Lautier has resurrected the whole line. For a number of years, the variation was con­ sidered not playable for Black on the grounds of the game Kasparov-Murey, Moscow Interzonal 1 9 8 2 , which con­ tinued 1 2 . . . a6 (the idea of this move is to rule out the unpleasant check on a4. As we are about to see in Petursson­ Lautier, it was not necessary to prevent this check) 1 3 .'iVd2 ! tLJaS 1 4 .dS exdS I S . exdS ild6 (otherwise White 's d-pawn continues its j ourney) 1 6 .ilxg7 'ife 7 + 1 7 .ile2 rlg8 1 8 .iVh6 fS 1 9 .ilf6 and White soon won. The point behind 1 2 . . . ild6 ! is that: A I ) In the case of 1 3 .'iYb 1 , the game Radj abov-Grischuk , Corsica 2 0 0 3 , beautifully illustrates Black's possibili­ ties: 1 3 . . . 'iYe 7 ! 1 4.'iVa2 tLJaS I s .ild3

analysis d i agram

270

P a wn M a j o r i t y in t h e C e n t r e

(Black correctly first finishes develop­ ment qUickly. 1 3 . . . cxd4 ? ? was a losing mistake, due to 1 4 .ilb S 'if c7 1 5 .ilf4 \lfc8 1 6 . tLJ xd4 ilcs 1 7 . tLJ fS 0 - 0 1 8 .'iYc3 f6 1 9 .iVg3 g6 2 0 .ilc4+ �h8 2 1 .ilh6)

analysis diagram

1 9 .. J:rxfS ! 2 0 . 'iYxfs ilc4+ 2 1 .�g l n.f8 2 2 .'iVgS ( 2 2 .'iYc2 ? ? is a horrible blun­ der due to 2 2 . . Jixf3 ) 2 2 . . . 'iYe2 2 3 . �b l ilf4 24 .'iYg4 and now only the fact that this was a rapid game can explain that, after playing a model game up to here, Grischuk missed the simple 24 . . . 'iYc2 , winning easily; A2) 1 3 .dS exdS 1 4.exdS 'iYe 7 + ! (this check is essential) 1 5 .ile 2 tLJaS 1 6 . 'iVa4+ �d8 ! '

analysis d i agram

a n a lysis d i agram

in general, White's pieces are poorly coordinated. 1 7 .tLJd4 (trying to castle. If White can manage that, the assessment of the posi­ tion will change drastically. If 1 7 .ilxg 7 l:lg8 1 8 .'ifd4, then 1 8 . . . tLJb3 ! 1 9 .�b2 ldxg 7 2 0 . iVxg 7 1:lc2 ) 1 7 .. J �c4! 1 8 .iVbS lde8 ! 1 9 .�f1 l:tcs 2 0 .'iVd3 1:lxdS and Black soon won in Peturs­ son-Lautier, Marseille 1 9 8 8 . B ) 1 0 .ile 3 ! ? ile 7 . This allows a dangerous pawn break. A safer alternative is 1 0 . . . cxd4 1 1 .cxd4 l:tc8 1 2 .'iYa2 ild6 1 3 .ild3 0-0 1 4.0 -0 tLJe7 I s Jifc l tLJg6 1 6 .g3 'ife7 1 7 .a4 ilb4 1 8 .h4 �fd8 1 9 .hS tLJf8 2 0 .h6 .

4:

Surprisingly enough , Black is better here. His king is relatively safe on d8, while the white king cannot castle and

a n a lysis d i agram

It seems that White is taking the ini­ tiative but the unorthodox reply ' 20 . . . gxh6 ! often solves Black 's prob­ lems. See for a similar solution Milov-Rowson, in the comments to our previous game Sokolov-Howell. Players sometimes make the mistake to automatically discard such possi­ bilities ' on general grounds' . After 2 1 .�xc8 ldxc 8 2 2 . ldd l hS a compli­ cated game ensued in Ponomariov­ Anand, Sofia 2 0 0 6 . 1 1 . dS ! exdS 1 2 . ldd l d 4 (forced. White is clearly better after 1 2 . . . 0-0 1 3 .exdS tLJaS 1 4 . c4) 1 3 . cxd4 O - O !

1 4.dS ! (taking a pawn with 1 4 .dxcS does not promise anything after 1 4 .. .'iVc7 and Black gets his pawn back and has a good game) 1 4 . . . tLJd4 1 S .'�· b 2 lde8 ? This move i s the main reason for Black's trouble in this game and it is, in a higher sense, a decisive mistake. Cor ­ rect was 1 S . . . tLJxf3 + 1 6 .gxf3 'iYd6 ! and it is far from easy for White to move his centre, for example : 1 7 .f4 (or 1 7 .eS 'iYg6) 1 7 . . . 'iYg 6 ! and now, if 1 8 .ild3 ?, Black has 1 8 . . . c4 ! 1 9 .ilc2 fS ! and White's centre collapses; 2 0 .eS 'iYg2 ) . Now, after 1 S . . . �e8 ? : B 1 ) 1 6 .tLJxd4? i s bad due to 1 6 . . . cxd4 1 7 .iVxd4 ( 1 7 .ilxd4 'ifxdS ! 1 8 .exdS ?? ilb4 is the kind of mate we often see in books for beginners) 1 7 . . . ilcs 1 8 . 'ifd3 'iVh4 1 9 .ilxcs bxcS and White starts losing his central pawns; B 2 ) 1 6 .ilxd4 ! cxd4 1 7 .ilbS ilxa3 ! (Black is looking for salvation in tac­ tics) 27 1

C hapter

W i n n i n g C h es s M i d d l eg am e s

that the bishop o n b 7 has been a ive bystander throughout the game. 3 5 . . . 1::1 f8 (or 3 5 . . . 'iYc l + 3 6 .tLle l 'iYc7 3 7 .d6 ! nxd6 3 8 . 'if e5 and Black is mated; 35 . . . nc8 loses to 3 6 .d6 'ifc l + 3 7 .tLle l 'ti'xh6 3 8 .�d7 'iYh l + 3 9 .�e2 �a6+ 40.�f3 'ti'h5 + 4 1 .Wg3)

a4-e8 diagonal, Black will have to give a pawn by playing a7 -a6 ; B3) 1 3 . . .'iVc7

a n a lysis d i agram

1 8 .'iVxd4 ! (taking the bishop is wrong, since Black is better after 1 8 .'ifxa3 ? l:[xe4+ 1 9 . WfI 'ifxd 5 ) 1 8 . . . �c5 1 9 .'iYd3 �b4+ 2 0 . �fI . White has a strong pawn centre and the h I rook will be brought into play via the h-file, so the king being on fI is not a handicap after all. Black's b7 bishop is out of play, while there is no time for Black to start moving his ed pawns on the queenside : 2 0 . . J � f8 (or 20 . . . l:[e7 2 1 . tLl d4) 2 1 . tLl d4 'iVc8 (White has a massive advantage in the case of 2 1 . . . �c5 2 2 .�c6 �xd4 2 3 .�xb 7 l:[b8 2 4.1tc 6 �e5 2 5 . g 3 ) 2 2 .h4 ! it c 5 2 3 .h5 (it could well b e that 2 3 . tLlf5 ! is more exact, since in the case of 2 3 . . . g6 White has 24.tLlh6 + ! W g 7 2 5 .h 5 ) 2 3 . . . 'ifg4 2 4 . tLl f3 ! �ad8 2 5 .llh4 ! 'iVc8 2 6 .�c4. Tired of his difficult, ive position, Kramnik now sacrifices a pawn, hoping to create some sort of counterplay. However, White simply has a massive advantage here and the black position cannot be saved: 26 . . . b5 2 7 .�xb5 �b6 2 8 .h6! g6 2 9 . 'ifb3 ! 'iVc5 3 0 . 'iYb2 f6 3 1 .�f4 ! �c 7 (or 3 1 . . . iYd6 3 2 .g 3 ) 3 2 J :k l 'iY b 6 3 3 . .tt xc 7 ! 'ifxc7 3 4 J:lxf6 l:rxf6 3 5 .'iVxf6 . The white pawns are running and Black 's king is in a terrible shape. Note 272

analysis d i agram

an alysis d iagram

3 6 .'iVb2 ! J:rc8 3 7 .d6! 'iVc l + 3 8 .'iVxc l nxc 1 + (Black has reached an ending, but at what price ?) 3 9 .�e 2 (the white pawns cannot be stopped) 3 9 . . Jlc8 40.tLlg5 a6 4 1 .�a4 1 - 0 , Sokolov­ Kramnik, Wijk aan Zee 2 0 04. 9.

e2-e4

0-0

It is a bad idea for Black to attack the white centre immediately with 9 . . . c5 ?, because 1 0 .�b5 +! forces Black 's bishop to c6 , where it will be badly placed. A) 1 0 . . . tLlc6 simply loses a pawn after 1 1 .'ifa4 l:tc8 1 2 .tLle5 'iVc7 1 3 .�f4 (or 1 3 . 'iYxa7 ) 1 3 . . . a6 ! (the only move) 1 4.�xa6 �xa6 1 5 . 'iVxa6 ; B ) 1 0 . . . �c6 1 1 .�d3 tLl d 7 1 2 . 0-0 h6 (if I 2 . . . 0-0 , 1 3 .d5 !) 1 3 .�d l and now: B l ) 1 3 . . . 0-0 1 4.d5 ; B2) In case of 1 3 . . . cxd4 White has 1 4.tLlxd4 ! (much better than 1 4.cxd4 1::1 c 8 ) 1 4 . . . 'iVc7 1 5 . tLlxc6 'iVxc 6 1 6 .'ife2 ! �d8 1 7 .�b5 'ifc7 1 8 .'ifd3 , and in order to solve the pin on the

1 4.d5 ! (With the black king still on e 8 , White sacrifices a pawn i n order to open up the position. This pawn sacri­ fice resembles the one in Kasparov­ Ivkov given in the comments on Game 42 - Kasparov-Naj dorf) 1 4 . . . exd5 1 5 .exd5 1txd5 1 6. 1tb5 a6 ( 1 6 . . . �c6 would also not have solved the problem of the black king on e8 ; after 1 7 . 1tf4 ! 'iVb7 1 8 .�xc6 'iYxc6 1 9 .I:te l it is obvi­ ous that Black's dreams about castling will remain wishful thinking : 1 9 . . . f8 2 0 . �ad l and White's initiative is worth much more than a pawn)

4:

P a wn M a j o r i t y in t h e C en t r e

B3 1 ) 1 7 . . .'iYb 7 ? loses by force to 1 8 .�xd 7 + 'iVxd7 1 9 .c4 'ifg4 20 J :txd5 'iYxf4 2 1 .�e l �a7 2 2 . tLle5 ! (the most accurate win) 22 .. .Iic7 2 3 .tLlg6 fxg6 24.'iYxg 6+ �f7 2 5 J :td8 + ; B3 2 ) 1 7 . . . �xf4 1 8 .�xd 7 + Wxd 7 1 9 . .trxd5 + (the black king remains ex­ posed and the queen is very badly placed on f4) 1 9 . . . �c7 2 0 J:te l 1td6 (or 20 . . . �f6 ? 2 1 ..tre4, trapping the queen, while 20 . . J lhe8 ? loses a piece after 2 1 . .trde5 'iff6 2 2 .'iVe4) 2 1 . .trf5 ! 'ifc4 2 2 . �e4 'ifb5 2 3 J :txf7 + Wb8 24 . .tre6 .trd8 2 5 . c4 'ifc6 2 6 .tLle5 'iVc8 2 7 . 'ifb 1 1 - 0 , Kasparov-Gheorghiu , Moscow Interzonal 1 9 8 2 . 1O .

�f1 -d3

c7-c5

11.

0-0

iV dS-cS

One of the main moves here. The white bishop on d3 is a dangerous attacking piece and Black wants to exchange it. Moreover, in that case the black queen will be transferred to the a6-fI diago­ nal, controlling the light squares. 1 1 . . . 'iYc8 is also a tempo move, since Black threatens 1 2 . . . cxd4. White is forced to move his queen to e 2 . Alternatives are 1 1 . . . 'iYc7 , 1 1 . . . tLl d 7 or 1 1 . . . tLlc6. 1 2.

'iVc2-e2

1 3.

�f1 -d1

1 3.

...

�b7-a6

an alysis d iagram

1 7 .�f4! . Again this tempo move is cru­ cial. Now:

ttfS-dS 273

Chapter

W i nn i n g C h es s M i d d l eg a m e s

Black first wants to place a rook on d8 , hitting d4, and then decide where to develop his b8 knight. However, given the fact that the b8 knight will be devel­ oped to d7 anyhow, it would make more sense to first develop the b8 knight and then see what to do with the f8 rook. Main theory agrees and the game continuation is considered to be a sideline. Most of the games continue with 1 3 . . . �xd3 1 4.�xd3 cxd4 I S .cxd4 ttJd7 and here, the main theoretical discus­ sion concentrates on 1 6 .eS , with which White employs a standard plan (seen in many of our examples) of surrendering the dS -square in order to gain more space on the kingside. Here the absence of light-squared bishops makes the de­ fence easier for Black. Back in 2 0 0 3 , preparing for the Hoogeveen tournament, I analysed the standard attacking move 1 6 .h4, which keeps all the options regarding the cen­ tral pawns open. Since Polgar played 1 3 . . . �d8 , I was not able to test this plan (which would have been a novelty at the time) in Hoogeveen, but I got my chance a few months later in Reykj avik, against Stefansson : 1 6 . . . 'iVa6 (plaCing the black queen on the fl -a6 diagonal and freeing the c8-square for a rook is the most logi­ cal continuation here. On a6 , the queen is better placed than on c4 , since now Black often plans ,gac8 fol­ lowed by trc3 or gc4. Moreover, the queen being on c4 could cost Black a tempo after 1 6 . . . 'iYc4 1 7 . h S h 6 I 8 .�f4 �ac8 1 9 .1dad l ttJf6 2 0 .ttJ e S , with some advantage for White in Sokolov-A. Kovacevic , Bosnia tt 2 0 04) 1 7 .�g S 274

a n a lysis d i agram

A) Now 1 7 . . . �xgS ! 1 8 .hxgS l:.ac8 was something Black should consider, in order to follow Black's plan in the game Sokolov-Bologan, shown in the next comment. This solves all Black's problems and is the main reason why I stopped employing this line with white. Compared to my game against Bologan, Black would be a tempo up hence he would play . . . trf8 -e8 at once. I was confident that White had the better chances here, because in my analYSis I failed to notice Black 's plan of . . . 'iVa6 -a4, .. Jdc8 -c4, . . . lIf8 -e8 and with 'li'a4 and �c4 indirectly pressuriz � ing White's e4 pawn, Black threatens . . . e6-eS . In the game mentioned later on it will become clear what I mean . B) 1 7 . . . �fe8 1 8 . nad 1 Z;Iac8 1 9 .'iVe3 ( I 9 .�xe 7 �xe 7 2 0 . eS is definitely an alternative here. White gets his kingside attack) 1 9 . . .f6 (like Polgar, Black prefers to keep the dark-squared bishops on) 2 0 .�f4 bS ! (preparing counterplay with . . . ttJb6 -c4) 2 1 .hS ttJb6 and now: B 1 ) The knight transfer to g4 (like in Game 43 Galliamova-Tiviakov) is inter­ esting, but does not promise White more than a draw after 2 2 . ttJh2 ! ? ttJc4 2 3 . 'iV g 3 ttJ b 2 2 4 . ttJ g4 and now : 24 . . . �h8 ! ? (24 . . . ttJxd3 2 S .ttJh6+ �h8 ! (2 S . . . �f8 ? 2 6 .btxd3

4:

P awn M a j o r i t y in t h e C e n t re

analysis d i agram

analysis d i agram

with 2 7 .'ifxg 7 + ! �xg 7 2 8 . ldg 3 + to follow, and White wins) 2 6 . ttJf7 + Wg8 2 7 .ttJh6 + with perpetual check) 2 S .h6 . B I I ) Here, 2 S . . . g6? would be a mis­ take that allows an immediate tactical solution:

3 I . . . ttJxeS ! 3 2 . l:txb 7 ttJxf3 + 3 3 .�xf3 �xb 7 34.'li'f6 + �g 7 3 S . 'iYxe6 and in the end we get an unusual position on the board, which is difficult to assess. White's pawns are dangerous, but Black has enough resources. I would be sur­ prised if White gets more than a draw here, while he may easily get in trouble; B I 2 2) 2 6 . .tIf3 gxh6 (the greedy 2 6 . . . ttJxd I ? loses to:

analysis d i agram

2 6 .ttJxf6 ! �xf6 2 7 .�eS Ilf8 2 8 .I;If3 ttJxd I 2 9 . lixf6 �xf6 3 0 . 'iff4 ! �f8 3 1 .�xf6+ �g8 3 2 .'iYc7 !tf7 3 3 .�e7 ! �f8 34.�d6 �f7 3 S .'iYd8 + and Black is mated. B 1 2) 2 S .. J:tg8 and now: B 1 2 1 ) If White is ambitious , he will have to opt for 2 6 . ttJ e S ! ? fxe S (instead, 2 6 . . . ldcf8 ? loses nicely to 2 7 .ttJg 6 + ! hxg6 2 8 . 'iVxg 6 , with �h3 to follow) 2 7 .hxg 7 + l:Ixg 7 2 8 .�xe S �f8 2 9 . 1I f3 ! 'li'b7 3 0 . l:tb I ! ttJc4 3 I . l:txb S

a n a lysiS d i agram

2 7 . ttJeS ! l:Icf8 2 8 . ttJ g 6 + ! hxg 6 2 9 .'iVxg6 with � g 3 o r !th3 t o follow) 2 7 .�eS ! (2 7 JIb 1 ? is wrong due to the simple 2 7 . . . hS 2 8 J �xb2 �xg4) and now: B I 2 2 I ) 2 7 .. J::rcf8 2 8 . �xf6 �xf6 29.iVh4 lhg4 30.hf6+ hf6 3 1 .'iVxf6+ Wg8 3 2 .'ifd8+ Wg7 (Black must be care­ ful, since 3 2 . . . Wf7 ? loses to 3 3 . l:Ic I ttJc4 34. nc3 ) 3 3 .'iYe7 + with a draw; 27S

Chapter

W i nn i n g C h ess M i d d l e g a m es

B 1 2 2 2 ) 2 7 . . . fxes 2 8 . iYxes + l::r g 7 2 9 . t2Jxh6 t2Jxd l 3 0 . t2J f7 + \t g 8 3 1 .t2Jh6+ and again a draw by perpet­ ual check. B 2 ) 2 2 . t2Jd2 ! �f8 2 3 . 'iY g3 'iYb 7 24.h6 (an alternative was 24.1:Ib 1 , moving one rook away from the t2J b 2 fork, which will happen a few moves later) 24 . . . g6 (24 . . . g s ? ? is a terrible blunder due to 2 s .�xgs )

B 2 2 1 ) Black i s simply a pawn u p after 3 0 .�xes l::tx es (3 0 . . . t2Jxfl is a draw: 3 1 .�xf6 t2Jxd3 3 2 . iYh8 + �f7 3 3 . 'iff6 + ) 3 1 .'iYxd 1 �xh6 3 2 .t2Jf3 l::f. e 8 ; B 2 2 2 ) 3 0 . 'iYxd 1 ! exf4 3 1 . 'ifb 3 + 'iVe6 3 2 .iYxbs �xh6 3 3 .lid7 and White will likely win back his pawn and make a draw - but there is not more in it for him. B2 3 ) 2 6 . . . J:Icd8 ? (a decisive mistake) 2 7 .es ! t2Jb2

analysis d i agram

2 s .ds ! ? (this standard solution is some­ what premature here. However, over the board it was difficult to correctly assess the ensuing complications. An alterna­ tive was still 2 5 J :tb 1 ) 2 5 . . . t2Ja4 ! (2 s . . . exds ?! 2 6 .exds is good for White, since Black would then not have . . . e6-es with tempo, closing the positi­ on at the critical moment) 2 6 .d6 ! and now: B2 1 ) In the case of 2 6 . . . t2Jcs White has an initiative after 2 7 .es fxes 2 8 .�gs ! ; not 2 8 . �xe s ? due to 28 . . . t2Jxd3 2 9 .'iYxd3 l::t c s . B 2 2 ) Correct is to immediately fork White 's rooks with 2 6 . . . t2Jb2 ! , and af­ ter 2 7 . d 7 t2Jxd l 2 8 . dxc8'iY �xc8 ! ( 2 8 . . . l:txc 8 ? loses to 2 9 . ifg4 ! ) 2 9 . 'iVf3 es ! (White has compensation in the case of 29 . . . t2Jb2 3 0 . l:rc3 ) and now : 276

1 4.

h2-h4 1

Given the fact that my preparation had originally been based on the plan with h2 -h4 (albeit in a slightly different po­ sition) , this was almost an automatic decision. Alternatives are 1 4.i�J4 or 1 4.�b 2 . 1 4.

...

c5xd4

1 5.

c3xd4

�a6xd3

1 6.

l::r d 1 xd3

ttJ b8-d 7

1 7.

�c1 -g5

P a wn M aj o r i t y i n t h e C e n t re

a n alysis d i agram

1 7.

...

f7-f6?!

This move weakens Black's kingside pawn structure and will be the source of Black's problems. Better was 1 7 . . . �xgs and now: A) 1 8 .t2Jxgs h6 1 9 .'iih s looks prom­ ising, but in reality it only leads to a forced draw after 1 9 . . . hxgs 2 0 .hxg s g 6 ! (the best defence. White gets a win­ ning attack after 20 . . . t2Jes ? 2 1 . l::th 3 t2Jg6 2 2 . 'iYh 7 + �f8 2 3 . l::t f3 W e 7 24.'iYxg 7 �f8 2 5 .ds !) 2 1 .'iVh6.

analysis d i agram

2 8 .�d4 ! (when playing 2 6 .. J lcd8 ?, Black likely overlooked this. t2Jd2 -e4 is now a terrible threat) 2 8 . . . .txd6 (or in­ deed 2 8 . . . t2Jxd 1 2 9 . t2Je4) 2 9 . exd6 t2Jxd l 3 0 .t2Je4 �f8 3 1 .'iYf3 ! 'iff7 3 2 . l:f.xd 1 gs 3 3 .�e3 'iYg6 34.�xa 7 1 - 0 , Sokolov-Stefansson, Iceland tt 2 0 04.

4:

It seems that White has a small advan­ tage and Black has very limited counter­ play, but Black has an active plan: 20 .. .'iVa4! 2 1 .'iYf4 ne8 ! (with the idea .. J �c4 and . . . es L Note that, as men­ tioned earlier, Black has wasted a tempo on . . . �d8 and could have had the same position with an extra tempo !) 2 2 . nc 1 (if 2 2 . g3 , then, indeed, 2 2 . . J lc4 ! 2 3 .�g2 es ! and Black is better) 2 2 . . . l:tc6 2 3 J :tdc3 11ec8 24. l::tx c6 llxc6 and now: B 1 ) 2 s J �txc6 'iYxc6 2 6 .t2Jes t2Jxes 2 7 .'iVxes

a n alysis d i agram

It seems that nh3 and 'iYh8 mate are in­ evitable, but Black has a nice defence: 2 1 . . . 'iic 4! 2 2 . l::t ad 1 and now 22 . . . 'iYc2 ! 2 3 . I1 1 d2 (2 3 . .t:rh3 ? ? is a terrible blun­ der after 23 . . . �xd 1 + 24. \th2 �xd4) 2 3 . . . 'iic l + 24.l:ld l (2 4.�h2 ? runs into 24 . . . t2Jes ! 2 s . dxes I:txd3 2 6 . l::t x d3 iYf4+) 24 . . . �c2 with a draw; B) 1 8 .hxg s 'if a 6 1 9 . !lad 1 l:lac8 2 0 . 'ife3 .

an alysis d i agram

2 7 . . . h6 ! (now Black gets an easy draw) 2 8 .g 3 (or 2 8 . gxh6 'iVc 1 +) 2 8 . . . hxg s 2 9 . 'ii b 8 + �h7 3 0 .�xa 7 'iVxe4 3 1 . �xb6 �e 1 + 3 2 . \tg 2 'iVe4+ 3 3 . �h2 'iVc2 34.�g2 draw, Sokolov­ Bologan, Poikovsky 2 0 04; B2) Better was 2 s .g3 �f8 ! (bad for Black is 2s . . J:txc 1 + ? 2 6 .'iYxc 1 t2Jf8, 277

Chapter

W i n n i n g C h es s M i d d l eg am e s

when White wins with 2 7 .tDeS ! 'iYxd4 2 8 .'iYc7 'iVa I + 2 9 .�g2 'li'xa3 3 0 .'ti'xf7 + Wh8 3 1 .tDc4 'iYcs 3 2 .eS , fol­ lowed by tDd6) 2 6 .Wg2 ! (threatening .i:th l and forcing Black to take on c l , giv­ ing White control of the c-file) 2 6 .. Jlxc 1 2 7 . 'iVxc 1 as (in the absence of active moves it is often a good idea to start pushing pawns closer to their promotion squares) 2 8 .'iYc8 + rJ;;e 7 29 .'iYc7 �e8

With all his pieces harmoniously devel­ oped , White has set the stage for direct operations. 21 .

e4-e5!

f6-f5

22.

d4-d5 !

e6xd5

White's dS pawn had to be taken. The protected ed pawn on d6 and Black's irreparably weak e6 pawn would be worth much more than an exchange after 2 2 . . . tDcS 2 3 .d6.

23 .

...

�d8-f8!

After landing in a difficult pOSitlOn, Black finds the best defensive moves. 2 3 . . . tDcs ? would lose by force to 2 4 . tD xfS . �f8 2 S .'iVg4! tDxd3 2 6 .tDh6+ Wh8 2 7 .tDf7 + �g8 2 8 .'li'e6. 24.

'iVe2-f3

The position is sharp and it is easy to blunder, for instance : 24.tDe 6 ? tDcS ! and Black is better. 24.

...

25.

h5-h6!

tLld 7-c5

This move prepares the scene for a piece sacrifice. Naturally I could not cal­ culate things till the end, but I did see a lot of promising lines and with Black's exposed king, his limited number of defenders and the black queen stranded on a6 , the logic of the attack should be on the white side. 25.

analysis d i agram

23.

3 0 .dS ! 'iYxe4 3 1 .d6 'iYa8 3 2 .Wg l ! 'iVd8 3 3 .tDd4 tDcs 34.'iVc6 + �f8 . White has obviously more than enough compensa­ tion for the sacrificed pawn. Whether he can win is another question. 1 8.

�g5-f4

'iVc8-b7?

This will turn out to be a tempo loss. Better was the immediate 1 8 . . . 'iVa6 . 1 9.

h4-h5

lla8-c8

20.

l:la1 -d1

'iYb7-a6

278

tLlf3-d4 !

The white knight goes to its natural post , aiming for its best attacking square - fS . White is planning to decide the game with a direct attack on the black king. He had safer options at his disposal. In general, it is difficult to de­ fine right and wrong in such situations; the decision is often a matter of style. White had a safe advantage with 2 3 .'iVa2 , but now a typical story where the better side plays it safe and ends up in a better, but still drawn ending, may easily materialize after 2 3 . . . 'iVc4 ! (Black has to force the queen swap ; 23 . . . tDf8 ? loses to 24J:rxdS JdxdS 2 S .'iYxdS + Wh8 2 6 .'iYf7 'iYa4 2 7 .h6! 'iYxd l + 2 8 . �h2 ) 2 4 . 'iYxc4 axc4 2 S .l:IxdS llxf4 2 6 . 11xd7 I::r x d7 2 7 .11xd7 �cS ! (2 7 . . . �xa3 ? is a blunder, lOSing to 2 8 .tDgS) 2 8 J�ha7 h6 and Black has good drawing chances.

...

4:

P a wn M aj o r i t y in t h e C en t r e

3 0 .tDf7 + ! �g8 3 1 .tDxd8 �fxd8 (or 3 1 . . Jlcxd8 3 2 .'ii g S + �f7 3 3 .'iff6+ �e8 34.'iYe6 mate) 3 2 .e6 ! and Black is soon mated. 28.

Ird 3-f3!

Precision is needed. 2 8 . tDxfS only leads to a draw after 2 8 . . . l:Ixfs ! 29 . 'iVxfS tDxd3 3 0 . 'iYe 6 + �f8 3 1 .'ii f6 + �e8 3 2 .'iYe 6 + . 28.

...

29.

tLld4xf5

itf4xe5

Finally the white knight reaches its best attacking square !

g7-g5!

Undoubtedly Black's best defence, since it forces White to find an exact execu­ tion. In the case of 2S . . . g6, White would have an easy continuation in 2 6 .'iYxdS + �h8 2 7 .tDe6 tDxd3 2 8 .Jdxd3 , with a clear and risk-free advantage.

26.

�f4xg5 ! !

�e7xg5

27.

'iVf3-h5

itg5-f4

2 7 . . . �d8 loses to 2 8 .tDxfS �h8 2 9 .tDd6 ! tDxd3 (or 2 9 . . J�c7 3 0 .1:If3 )

29 .

...

'lWa6-b7

Black finally brings his queen to the de­ fence of the kingside, but the position can no longer be saved. Other moves also lose : A) 2 9 . . J �ce8 loses to 3 0 .tDd6 �xd6 3 1 .'iYgS + �h8 3 2 .'tWg7 mate ; B) Or 2 9 . . . h8 3 0 .tDe7 ! �b2 (the only available square for the black bishop on the long diagonal) 3 l . ab 1 ! and now: B 1 ) 3 1 . .. axf3 loses to 3 2 J:txb2 ncf8 3 3 .gxf3 bS (bringing the black queen back into play) 3 4 . 'li'e S + 'iYf6 3 5 . 'iYxf6 + �xf6 3 6 . 1lxb S tD e 6 3 7 .�b 8 + and the ending i s easily won for White. The black bishop has run out of squares on the diagonal; B2) 3 1 . . . �d4 10ses t0 3 2 .'iYxdS ; 279

W i n n i n g C h ess M i d d l eg a m e s

B3 ) 3 1 . . . 'if c4 3 2 . ttJxc8 l:[xc8 3 3 .'iff7 ! �g8 (otherwise 34.'iVf8 +) 3 4.�g3 l::t c 8 35 J:tg7 'ifc2 3 6 . .t1e 1 (it's never too late for a terrible blunder, spoiling a beautiful game and resulting in endless, sleepless nights : 3 6 J ixb2 ? ? 'ifc 1 + 3 7 .
nd 1 xd5 !

1:tcS-eS

31 .

'iYh5-g5+ �gS-hS

32.

l:td5xe5

Here I was spending most of my time checking whether my planned tactical idea worked. And it does work indeed. The aesthetic effect of my 3 3rd move also played a role for me in that I did not look for other solutions. Even better was 3 2 .ttJe7 ! ! �b2 3 3 . ld.eS ! , winning on the spot. 32 .

...

The only move that prevents mate, see 3 S . . . ttJe6 3 6.'iVg7 + ttJxg7 3 7 .hxg7 mate; or 3 S . . . �xg3+ 3 6 .fxg3 ttJe6 3 7 .iYd2 ! l:Ie4 3 8 .ifd3 ! �g4 3 9.'iVe2, winning. 36.

'iYg5xe7

tbc5-e6

Black is forced to give her knight as well. 37.

'iYe7xe6

'iYbS-f4

Black has prevented an immediate on­ slaught but her problems are far from ' solved. White is a sound pawn up, and the black king is still under threat. 3S.

a3-a4

l::I e Sxe5

3S .

...

a 7-a6?

This blunder in time pressure simply loses a pawn and concludes the game. 3 8 . . . ifxf'2. ? ? would run into a mate af­ ter 3 9 .�g8 + ! �xg8 40 .'iVe S + ; the only way to struggle on was with 3 8 . . . �f7 . 39.

'iY e6xb6

;etfS-bS

Or 3 9 .. J�:r6 40.'iVb2 ! 'iYxh6+ 4 1 .
tbf5-e7 ! !

l:le5-e1 +

34.

�g1 -h 2

'iYb7- bS+

35.

IU3-g3

l:re1 xe7

40.

New I n Chess Code System

'iYb6-e3

'iYf4-h4+

41 .

J:tg3-h3

'iYh4-f6

42.

'iY e3-c3

1 -0

White stands slightly better Black stands slightly better White stands better Black stands better White has a decisive advantage Black has a decisive advantage balanced position

=+ ± =+= +-+

unclear position compensation for the material strong (sufficient)

00

weak (insufficient)

<

better is weaker is good move excellent move

2

bad move blunder interesting move dubious move only move with the idea attack initiative lead in development counterplay mate correspondence

280

;!;

>

�

?? !? ?!

D !J.

t tt x

281

W i n n i n g C h e s s M i d d l eg a m e s

I ndex of Players The numbers refer to pages. A

1 09 Abramovic 1 4, 5 1 Adams 96 Akobian 70 Alburt Alekhine 1 99 , 229 1 2, 53, 7 1 , 79, 88, Anand 1 3 6- 1 3 7 , 1 4 1 , 1 4 7 - 1 48 , 1 5 5 , 1 5 7 , 1 6 3 - 1 64, 1 6 7 , 2 7 1 Arencibia 1 27 Aronian 13, 67, 72, 85, 89-90, 2 5 7 240 Arutinian Asrian 71 1 26 Atalik

23 69

Csom Cvitan D

Dizdarevic Dominguez Donner Dreev

1 2, 50 1 66 1 32 234

E

Ehlvest Elianov Ermenkov Euwe

1 72 90 69-70 229

208 Babula 216 Bacrot 63 Balogh 31 Bannik 1 9, 1 38 Bareev 1 2 8 , 1 64 Barsov Beliavsky 1 2 , 4 1 , 5 1 , 7 0 , 8 7 , 1 1 2, 1 2 1 - 1 22, 1 2 5 - 1 2 7 , 1 7 6- 1 7 7 , 1 8 2 , 2 0 5 , 2 1 3 157 Bentzen 69 Berg Bologan 1 2 , 47 , 2 7 4, 2 7 7 1 7 5 , 208 Bonsch Botvinnik 1 3 , 2 9 , 6 1 , 7 1 - 7 2 Brenninkmeijer 131 Brinck Claussen 32 Bronstein 1 2 - 1 3 , 2 6 , 5 8 , 1 3 2 Browne 242 1 07 Brunner 1 66 Bruzon

Fine Fischer Flohr Ftacnik Furman

229 2 3 , 1 7 9- 1 8 1 , 224 1 83 1 1 3, 127 2 9 , 1 8 5 , 1 8 9 , 24 1

G

Galliamova 2 2 1 - 2 2 2 , 2 24, 2 3 7 , 2 5 0- 2 5 1 , 2 5 6 , 2 6 1 -2 6 2 , 2 69 , 2 7 4 Gelfand 141 , 163 Geller 1 7 2 , 1 7 6, 1 8 6, 1 89 Georgiev,Kir 88, 1 74 273 Gheorghiu Gligoric 1 1 - 1 2 , 2 1 - 2 2 , 24, 3 8 , 86, 94, 1 2 7 - 1 2 8 270 Grischuk 242 Groszpeter 233 Gulko 1 76 Gusev 3 2 , 1 42 Gutman Gyimesi 71

c

Capablanca 1 99 Carlsen 257 Cebalo 86- 8 7 , 1 0 6 , 1 09 - 1 1 0 , 1 32 1 3 , 6 1 , 7 1 -72 Chekhover Christiansen 1 07 1 94 Comas Fabrego

22 1 -2 2 3 , 241 , 2 6 1 , 27 1 252 Hracek Hubner 1 2 , 2 8 , 8 7 , 1 3 6 , 1 6 3

Ibragimov 1 50 Illescas Cordoba 8 5 - 8 8 , 9 3 , 9 5 , 1 0 3 , 1 2 1 , 1 46 , 1 5 7 , 1 7 1 , 182 Ivanchuk 1 3 , 7 1 , 7 9 , 8 5 , 8 9 , 1 57, 2 1 6 1 2, 38, 239, 273 Ivkov J

F B

Howell

H

Halkias Hansen,C Hellers Hergott Hort

215 1 56, 222, 227-228 1 98 1 42 151

Jaracz Jirovsky Johansen

8 6 , 99 1 38 1 1 , 14

K

1 3 , 76, 1 39 Kaidanov Karjakin 253 242 Karolyi Karpov 44, 4 8 , 6 7 - 6 8 , 8 0-8 1 , 8 7-88, 1 1 2 , 1 1 7 , 1 1 9, 1 4 1 - 1 42 , 1 5 5 , 1 64, 1 7 3 - 1 7 5 , 1 8 5 , 200 Kasimdzhanov 2 0 0 , 2 1 4- 2 1 5 Kasparov 1 3 , 67 -6 8 , 79 - 8 1 , 86- 8 7 , 94-9 5 , 9 9 , 1 0 7 , 1 1 2 , 1 1 7 , 1 2 1 , 1 32, 1 3 7, 1 40- 1 42 , 1 5 7 , 1 7 2 , 1 8 5 , 2 2 1 -22 4, 2 3 7 - 2 3 9 , 242 , 25 1 -25 3 , 2 6 1 , 265, 269 - 2 7 0 , 2 7 3 Kavalek 1 85 1 2 , 3 1 , 3 5 , 229 Keres 1 1 8, 1 7 5, 1 79, Khalifman 1 92 , 202, 2 3 5 22 -2 3 , 1 0 7 Knaak 207 Kolev Kortchnoi 1 0 3 , 1 7 6 , 2 2 6 , 23 1 Kotov 1 2, 3 1 274 Kovacevic ,A Kramnik 7 1 , 8 5 , 87 -8 9 , 9 3 , 99, 1 03 , 1 1 7 , 1 2 1 , 1 3 2,

283

1 3 6- 1 3 7 , 1 4 1 , 1 46 , 1 48 , 1 5 5 , 1 6 3 - 1 64, 1 6 7 , 1 7 5 , 1 8 0 , 1 9 3 - 1 94, 2 0 0 , 2 0 5 , 2 1 3 , 2 5 3 , 2 69 Krasenkow 2 62-263 , 265 Kuzubov 1 3 , 63 L

Larsen 24 Lautier 8 9 - 9 0 , 1 2 7 , 1 8 0 , 2 1 3 , 2 1 5 , 2 7 0- 2 7 1 Leko 1 64 , 2 5 7 Lesiege 1 64, 1 6 6 Levchenkov 32 Ljubojevic 69- 7 0 , 8 6 , 94 Lobron 222, 22 7-228 Lputian 1 83 Lutz 1 74 M

Magomedov Malaniuk Marin Melkumyan Milov 5 8 - 5 9 , 2 6 2 , 2 66 , Moiseenko 7 1 , 207, Morozevich Murey

94 1 49 1 50 2 40 271 209 141 270

N

263 Naer Najdorf 1 2 , 2 6 , 2 2 1 - 2 24, 2 3 7 , 2 5 1 -2 5 2 , 26 1 , 265 , 269, 273 Nielsen,P.H. 7 2 , 8 6 , 9 9 , 1 5 7 Nikolic ,N 131 Nikolic ,P 1 1 , 2 2 , 6 9 , 8 7 , 1 3 0 , 1 76, 232 o

Olafsson,H Onischuk

1 1 3 , 200 1 3, 76, 2 1 5

p

Pachman 1 50 24 1 Panno Petrosian 3 1 , 86-88, 1 02 , 1 2 1 , 1 2 7- 1 2 8 , 1 4 7 , 1 8 4, 2 2 1 -2 2 2 , 2 2 4, 2 3 1 2 06 Petrosian,A Petrosian,T.L. 96, 1 2 1 - 1 2 2

284

1 3 1 , 2 7 0- 2 7 1 Petursson 207 Pigusov 90, 1 75 , 1 77 Piket 1 84 Pirc 1 76 Podgaets Polgar 5 8 - 5 9 , 2 2 1 - 2 2 2 , 2 24, 2 3 7 , 2 6 9 , 2 74 Polgar,Z 1 2 , 43 Polugaevsky 1 5 1 , 1 94, 2 3 5 Ponomariov 271 Portisch 1 9, 23, 200, 238 R

Radjabov Rainfray Razuvaev Riazantsev Ribli Rowson Ruban Rublevsky Rukavina

1 2, 5 3 , 222, 228-229, 270 90 24 1 215 225, 23 1 266, 27 1 232, 235 1 3 8 , 1 47 1 84

S

80 Sadvakasov Sargissian 9 1 , 2 0 7 , 2 0 9 , 2 1 5 Sasikiran 1 2 8, 2 1 4-2 1 5 87, 1 1 7 Schandorff Schmidt 243 Shirov 1 4 , 1 4 1 , 1 48 , 1 6 3 Short 1 9, 8 8 , 1 1 7 , 1 2 7 , 1 7 1 , 1 80, 1 82, 1 98, 206, 2 1 3 , 2 1 5 1 3 , 58 Simagin 1 47 , 1 5 1 Simutowe 1 27 Skripchenko Smyslov 242-243 Sokolov 1 1- 1 2 , 1 4, 1 8 , 43 , 47 -48 , 5 0 - 5 1 , 8 6 - 8 7 , 9 0 , 9 6 , 1 0 0 , 1 0 6- 1 0 9 , 1 1 4, 1 1 7 , 1 2 1 , 1 2 6 , 1 3 0, 1 3 2 , 1 3 6 , 1 4 1 , 1 4 7 - 1 49 , 1 5 1 , 1 5 6, 1 74, 1 8 0 , 1 9 2 , 1 9 8 , 2 0 2 , 2 0 6 , 2 2 1 -2 24, 2 3 7 , 24 1 , 252, 2 5 7 , 26 1 , 263 , 265, 2 6 9 , 2 7 1 - 2 7 2 , 2 7 4, 2 7 6 - 2 7 7 Sosonko 41 1 2, 23, 28, 35, 77, Sky 8 6 - 8 8 , 1 0 2 - 1 0 3 , 1 47 , 1 5 0 , 1 7 9- 1 8 1 , 1 84, 2 2 1 - 2 2 2 , 224 22 Spiridonov

Stefansson Stocek Sturua Svidler Szabo

276 138 1 03 253 1 84

T

Tal Timman

41 , 1 03, 1 07, 235 1 9 , 23 -24, 3 2 , 8 7 , 1 40 , 1 84, 1 8 6 Timoschenko,Gen 1 03 Tisdall 1 0 9- 1 1 0 2 2 1 - 2 2 2 , 2 24, Tiviakov 2 5 0- 2 5 1 , 2 5 6 , 2 6 1 -2 6 2 , 2 6 9 , 274 Tkachiev 1 66, 1 92 Topalov 1 3 , 6 7 , 1 1 8 - 1 1 9, 1 40 U

Uusi

77

v

Vaganian 1 3 2 , 1 7 2 , 24 1 , 2 6 5 1 92 Vaisser 1 49 Van den Doel Van der Sterren 1 1 4, 1 2 6 Van der Wiel 1 3 , 63, 1 08 Van Scheltinga 1 32 Van Wely 9 0 , 94, 1 9 3 - 1 94, 222, 228-229 Velimirovic 69 Vidmar 1 83 Vyzhmanavin 1 2 , 4 1 , 5 1 , 8 7 , 1 1 2 , 1 2 5 , 1 8 2 , 2 0 5 -2 0 7 , 2 1 3 W

Wang Yue Winants

6 3 , 69 1 1 , 1 8 , 1 85

y

Yakovich 9 1 , 1 49 Yegiazarian 1 75 Yermolinsky 1 2 6 , 1 3 9 , 1 8 3 Yusupov 44, 1 8 0 , 2 0 5 , 2 1 3 , 225, 23 1 , 233 Z

Zaitsev Zhidkoy Zhukova Zuger

1 85 1 03 127 94

W i n n i n g C h es s M i d d l eg am e s

List of Games

Chapter 3

Chapter 1 Game

1

NI 1 3 . 6 (E44)

Sokolov-Johansen

Manila 1 9 9 2

Game

2

NI 1 4 . 8 (E4 1 )

Sokolov-Winants

Netherlands tt 1 9 9 4 / 9 5 1 8

14

Game

3

NI 1 4 . 3 (E4 1 )

Gligoric-Nikolic

Novi Sad 1 9 8 2

22

Game

4

NI 1 8 . 1 3 (E 2 9 )

Bronstein -N aj dorf

Budapest 1 9 5 0

26

Game

5

NI I 9 . 7 (E 2 4)

Sky-Hiibner

Turin 1 9 8 2

28

Game

6

NI 1 8 . 1 (E24)

Kotov -Keres

Budapest 1 9 5 0

31

Game

7

NI 1 2 . 1 0 (E4 3 )

Keres-Sky

Riga 1 9 6 5

35

Game

8

NI 1 9 . 4 (E2 6 )

Gligoric-Ivkov

Bled 1 9 6 1

38

Game

9

NI 1 9 . 4 (E2 6)

Vyzhmanavin -Beliavsky

Lvov 1 9 8 4

41

Game 1 0

NI 1 8 . 1 3 (E2 9)

Polgar-Sokolov

Pardubice 1 9 94

43

Game 1 1

NI 2 4 . 1 3 (E 3 2 )

Sokolov-Bologan

Saraj evo 2 0 0 4

47

Game 1 2

VO 2 2 . 1 2 (A40 )

Sokolov -Dizdarevic

Saraj evo 1 9 9 8

50

Game 1 3

NI 1 9 . 3 (E 2 6 )

Radj abov-Anand

Monaco 2 0 0 7

53

Game 1 4

NI 1 9 . 7 (E24)

Bronstein -Simagin

Moscow 1 9 6 1

58

Game 1 5

NI 2 8 . 2 (E2 1 )

Botvinnik -Chekhover

Leningrad 1 9 3 8

61

Game 1 6

NI 1 5 . 1 (E4 1 )

Kuzubov-Van der Wiel

Groningen 2 0 0 4

63

Game 1 7

EO 44 . 8 (E 2 1 )

Topalov-Aronian

Morelia/ Linares 2 0 0 8

67

Game 1 8

NI 1 8 . 1 0 (E 2 8 )

Kaidanov-Onischuk

Chicago 2 0 0 2

76

Game 1 9

EO 1 . 1 0 (A2 9 )

Kasparov-Ivanchuk

Moscow 1 9 8 8

79

Game 3 6

Q O 6 . 6 (D 5 9 )

Fischer-Sky

Reykj avik 1 9 7 2

181

Game 3 7

SL 8 . 3 (D4 5 )

Sokolov-Khalifman

Pardubice 1 9 9 4

1 92

Game 3 8

Q O 8 . 5 (D5 8 )

Sokolov -Short

Saraj evo 1 9 9 9

1 98

Game 3 9

Q O 7 . 1 0 (D5 8 )

Kramnik-Yusupov

Dortmund 1 9 9 8

205

Game 40

QO 7 . 1 0 (D5 8 )

Lautier-Short

Pamplona 1 9 9 9 / 0 0

213

224

Chapter 4 Game 4 1

TD 2 . 9 (D4 1 )

Spas sky -Petrosian

Moscow 1 9 6 9

Game 42

QI 7 . 4 (E 1 2 )

Kasparov-Naj dorf

Bugojno 1 9 8 2

237

Game 4 3

QI 4 . 8 (E 1 2 )

Galliamova-Tiviakov

Elista 1 9 9 8

250

Game 44

GI 4. 1 (D 8 5 )

Sokolov -Howell

London 2 0 0 6

26 1

Game 45

Q I 4 . 1 7 (E 1 2 )

Sokolov -Polgar

Hoogeveen 2 0 0 3

269

Chapter 2 Game 2 0

N I 2 7 . 7 (D3 8 )

Ivanchuk-Aronian

Morelia/ Linares 2 0 0 7

89

Game 2 1

TD 4. 1 6 ( D 3 4)

Kramnik-Illescas Cordoba

Linares 1 9 9 4

93

Game 2 2

S L 8 . 5 (D4 5 )

Jaracz-Nielsen

Dresden 2 0 0 7

99

Game 2 3

TD 4 . 1 6 (D 3 4)

Petrosian -Sky

Moscow 1 9 6 9

1 02

Game 2 4

N I 1 5 . 6 (E4 8 )

Sokolov -Cebalo

Pula 1 9 8 8

1 06

Game 2 5

QO 1 0 . 1 5 (D 5 5 )

Kasparov -Karpov

Leningrad 1 9 8 6

1 12

Game 2 6

Q O 1 6 . 1 2 (D3 1 )

Sokolov -Schandorff

Reykj avik 2 0 0 1

Game 2 7

TD 4. 1 6 (D3 4)

Beliavsky -Illescas Cordoba Linares 1 9 9 0

121

Game 2 8

Q O 7 . 1 0 (D 5 8 )

Vyzhmanavin -Beliavsky

Novosibirsk 1 9 9 5

1 25

Game 2 9

Q G 1 4 . 5 (D2 7 )

Sokolov -Nikolic

Elenite 1 9 9 3

1 30

Game 3 0

QG l 1 . 1 4 (D2 7 )

Kramnik-Hiibner

Dortmund 2 0 0 0

1 36 1 40

117

Game 3 1

RG 6 . 5 (C4 2 )

Kasparov -Timman

Amsterdam 1 9 9 4

Game 3 2

QG 1 3 . 1 5 (D 2 7 )

Petrosi an -Sky

Moscow 1 9 7 1

1 47

Game 3 3

QG 1 1 . 1 4 (D2 7 )

Kramnik-Anand

Dos Hermanas 1 9 9 9

1 55

Game 3 4

Q G 1 1 . 1 4 (D2 7 )

Kramnik -Anand

Dortmund 2 0 0 1

1 63

Game 3 5

QO 8 . 1 2 (D 5 8 )

Illescas Cordoba-Short

Pamplona 1 9 9 9 / 0 0

171 285

286