Secrets of attacking chess [1st ed.] 9781904600305, 1904600301

Table of contents :
Symbols......Page 5
Bibliography......Page 6
Introduction......Page 7
1 The Logical Course of the Game......Page 9
2 The Elementary Stages of the General Plan......Page 44
3 Advantage in Development......Page 63
4 Play on Two Wings......Page 113
5 The Secrets Behind Morphy's Successes......Page 133
6 The Paradox of the Century......Page 153
7 A Chess-Player's Best Friend?!......Page 180
Index of Openings......Page 193

Citation preview

Secrets of Attacking Chess

Mihail Marin

[e)AI��IBIITI

First published in the UK by Gambit Publications Ltd 2005 Copyright © Mihail Marin 2005 The right of Mihail Marin to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out or otherwise circulated in any form of binding or cover other than that in which it is published and without a similar condition including this condition being im­ posed on the subsequent purchaser. A copy of the British Library Cataloguing in Publication data is available from the British Library. ISBN 1 904600 30 1 DISTRIBUTION:

Worldwide (except USA): Central Books Ltd, 99 Wallis Rd, London E9 5LN. Tel +44 (0)20 8986 4854 Fax +44 (0)20 8533 5821. E-mail: [email protected] USA: Continental Enterprises Group, Inc., 302 West North 2nd Street, Seneca, SC 29678, USA. For all other enquiries (including a full list of all Gambit chess titles) please contact the publishers, Gambit Publications Ltd, 6 Bradmore Park Rd, Hammersmith, London W6 ODS, England. E-mail: [email protected] Or visit the GAMBIT web site at http://www.gambitbooks.com Edited by Graham Burgess Typeset by John Nunn Cover image by Wolff Morrow Printed in Great Britain by The Cromwell Press, Trowbridge, Wilts.

10 9 8 7 6 5 4 3 2 1

Gambit Publications Ltd Managing Director: GM Murray Chandler Chess Director: GM John Nunn Editorial Director: FM Graham Burgess German Editor: WFM Petra Nunn

Contents

Symbols

4

Bibliography

5

Introduction

6

1

The Logical Course of the Game: The Spark of the Attack

8

2

The Elementary Stages of the General Plan

43

3

Advantage in Development

62

4

Play on Two Wings

112

5

The Secrets Behind Morphy's Successes

132

6

The Paradox of the Century

152

7

A Chess-Player's Best Friend?!

179

Index of Players

192

Index of Openings

192

Symbols

+

check

++

double check

#

checkmate

!!

brilliant move

!? ?! ? ?? Ch

1-0 112-112

0-1

(D)

g ood move interesting move dubious move bad move blunder championship the game ends in a win for White the game ends in a draw the game ends in a win for Black see next diagram

Bibliography

B.Kajic, M.Yudovich, Matchi pretendentov na pervenstvo mira, Belgrade, 1969 I.Neishtadt, Nekoronovanye chempiony, Fizkultura i Sport, Moscow, 1975 I.Darnsky, Grossmeister Geller, Fizkultura i Sport, Moscow, 1976 M.Tal, I.Darnsky, V ogoni ataki, Fizkultura i Sport, Moscow, 1978 T.Petrosian, Strategia nadezhnosti, Fizkultura i Sport, Moscow, 1985 S.Soloviov, Mikhail Tal Games, Chess Stars, 1994-1997 V.Panov, 300 partide alese ale lui Alehin, Editura Tineretului, Bucharest, 1957 Y.Korchnoi, R.Wade, L.Blackstock, Korchnoi's 400 Best Games, Batsford, London, 1978 64 Shakhmatnoe obozrenie, Moscow, 1976-81 A.Kotov, Shakhmatnoe Nasledie Alekhina (2 volumes), Fizkultura i Sport, Moscow, 1982 S.Gligoric, Igraiu protivfigur, Fizkultura i Sport, Moscow, 1983 P.Keres and J.Nunn, Paul Keres: The Quest for Perfection, Batsford, London, 1997 R.Spielmann, Teoria zhertvy, Folio, 1995 V.Vukovic, The Art ofAttack in Chess, Everyman, London, 1998 V.Turov, A.A.Alekhin, 100 luchshikhpartii, Fizkultura i Sport, Moscow, 2002 G.Kasparov, My Great P redecessors, Everyman, London, 2004 Informator series, Belgrade, 1965-2004 Schacknytt, Gothenburg, 1999-2004 Chess Extrapress, Bucharest, 1998-2001 Mega Database 2004, ChessBase, 2003

Introduction

This is only partly a logical continuation of my previous book for Gambit, Secrets of Chess De­ fence. True, after investigating several aspects of the chess fight from the defender's point of view, the desire to have a look from the opposite side was natural. And yet, the method of investigation cho­ sen is quite different from the previous case; in spite of the apparent similitude (or rather symme­ try) of the titles, the thematic universe is of quite a different nature. In Secrets of Chess Defence I focused on a series of concrete methods of dealing with inferior positions. For the new book, I have chosen a much broader approach. In the first place, I have aimed to investigate the smallest details that can create the premises for a successful attack as well as the way the objective evaluation of a certain position persists along a sequence of moves in the case of optimal play from both sides and in spite of the successive changes of character of the fight. Al­ though my previous work also contained such discussions about chess logic, the main difference is that here this will be a central theme. One could say that Secrets of Chess Logic would also be a suitable title for this book. However, all the games included here involve complex attacking opera­ tions, so the title chosen is the most appropriate one. Chess is much more complicated than what can be comprised in the notion of common sense. T herefore, a dissertation about the logical character of our royal game needs a huge amount of ana­ lytical work in order to be convincing. Keeping this in mind before starting the concrete work on this book, I had to decide which method of work would be most appropriate. One natural solution was to use the computer's help. However, as an active player I had my doubts about it. Relying too much on computerized assistance can lead to a decrease of one's prac­ tical strength, causing the human brain a pathologic laziness. There are also other negative aspects of such an approach, which I have discussed in the final chapter of the book. I soon reached the following paradox: I certainly needed to have a computer in my 'working team' but wanted to spend most of my time analysing on my good old three-dimensional wooden board. Finally, I adopted a combined solution. I asked my friend 1M Valentin Stoica, who is a specialist in computer-assisted analysis, to help me. I suggested to him the following plan: we would analyse together virtually all the material that was going to be published, but I was going to concentrate on the chessboard only, while he would split his attention between the chessboard and the screen. However, the whole conception of the book and the literary part remained wholly my job. Being a devoted researcher, he gladly embarked on this ambitious project. The reader will notice that the book contains a relatively small number of games (slightly less than 30). Can this be enough to elucidate at least part of such a major subject? I am firmly con­ vinced that it can. Each of these games is enormously complicated and contains a whole universe of situations, both of objective and psychological natures. To the human mind, chess has virtually un­ limited complexity, but Vali and I sought to analyse these games until we felt we had reached the very essence of the key issues. It would definitely have been easier to include 150 or 200 lightly an­ notated games, but I very much doubt that I could have drawn any reliable conclusions on the basis of such material. I must confess that the intermediate result of our joint work was a bit scary: a collection of bushy trees of endless variations, reminiscent of dragons with seven heads from fairy tales. There was still

INTRODUCTION

7

a lot to be done. By adorning the variations with detailed verbal comments and cutting some of the excessively developed branches, I have tried to make the final result easy to read and understand.

I should now say a few words about the general structure of the book. The first chapter presents some general topics that are typical for the research process, both dur­ ing the game and in the laboratory.

In Chapters 2 to 4 I have focused on three concrete patterns that can make an attack possible. The final three chapters have a more general background. They are built up around three of the most intriguing themes from the whole of chess history. Each of these subjects of investigation is taken from a different century. There is of course a limit to what can be covered within one reasonably sized volume. Neverthe­ less, I hope that I have managed to uncover some of the most interesting Secrets ofAttacking

Chess.

Mihail Marin Bucharest, May 2005

1 The Logical Course of the Game: The Spark of the Attack

Attacking games can embrace a huge variety of

decisive importance in the games of top-class

try to define the two extremes of

players. Virtually every game included in this

forms. I shall

this wide spectrum.

book belongs to the second category.

I am sure most of us have been captivated by

In this introductory chapter, I shall focus on

the cape and sword movies of the 1960s, fea­

two essential topics. First of all, I shall try to

turing such fine actors as Errol Flynn. During

highlight the very special moment of the game

the final and decisive fight between the main

when, as if by magic, an attack becomes possi­

heroes (the good one and the bad one), each

ble and sometimes even necessary so as not to

movement was of an attacking nature and the

lose the advantage. At the same time, I shall

whole scene was highly spectacular. In a cer­

pursue as closely as possible the logical course

tain sense, this kind of fight was typical of

of the game. Once one of the players has ob­

chess games played in the romantic period,

tained an advantage and launched an entirely

starting with Greco and finishing with Steinitz.

justified attack, his superiority should persist

People played for their and the spectator's fun.

for a long time in case of optimal play from

Players were enormously self-confident in

both sides and irrespective of changes in the

those times and the general feeling was that

character of the position.

the final outcome was determined only by the

I have to make an important clarification re­

skilfulness of the players. The question about

garding this latter point. Chess is such a com­

the soundness of an attack did not bother any­

plicated game that proving the truth over the

body seriously.

board is not always that simple. The best move

Recently I had the occasion to watch on TV a

in certain positions can be in such contradiction

martial arts contest. For most of the time, ap­

with our common sense that we cannot resolve

parently nothing happened: the two contestants

to play it. This frequently generates mistakes

were carefully watching each other without un­

even in the games of strong players. From the

dertaking any concrete action. This continued

point of view of chess logic, mistakes are quite

until the moment when the concentration of

important; there is nothing more interesting

one of them dropped for a fraction of second.

than to follow how the evaluation of a position

His opponent instantly started a lightning at­

can alternate in case of imprecise play.

tack. The fight was basically over and they both

I have chosen as a model for this chapter a

knew it very well. Everything looked quite sim­

great player of the golden era of chess, Efim

ple and convincing to most of the spectators,

Geller. What made Geller so distinctive to

but I am sure that it was the virtuosity of the

anoint him in this book as noble Knight of the

winner that caused this false sensation.

Attack? After all, he never qualified for a

The games of chess I like most are those

world-title match and one could immediately

played more or less according to the second

name several of his contemporaries whose re­

scenario. True, the final attack is usually more

sults were significantly better.

time-consuming than in the combat mentioned

We should not forget that this book is dedi­

above, but it is always instructive to follow

cated to chess as a science. Had Geller dedi­

how the least noticeable mistake can prove of

cated slightly more attention to the sporting and

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ATTACK

9

psychological side of his chess personality, he

order to make the a5-knight useful in some

might have obtained more tournament and

way.

10e4 (D)

match victories than he did. However, Geller was in the first place a researcher. When he felt that he was close to a mine of gold, he forgot about anything else and didn't give up until he found it. This explains why he frequently spent

B

more than an hour on a single move. This caused him numerous painful losses in severe time-trouble but also enabled him to obtain many brilliant wins, which are entered forever in the world chess treasury. Geller's games included in this chapter have certain points in common with each other. How­ ever, they present Geller in three different set­ tings. My aim has been to highlight not only the hidden beauty and depth of our royal game, but also the practical difficulties awaiting he who dedicates his efforts to the search for the truth. We shall start with one of Geller's most bril­

10 'tJg4?! .••

Only this move is inaccurate. Black is com­

liant victories. The remarkable thing is that the

mitting himself too much on the kingside, with­

position was for a while so complicated that for

out even having started to put the a5-knight to

decades the annotators (including Geller) were

work. It will be instructive to follow how after

unanimous in claiming that a correct defence

this move Black will inevitably get into trouble.

would have yielded Black a draw.

Just as in martial arts: Black's position loses its harmony for just a moment and this will prove

Geller - Velimirovic Capablanca mem, Havana 1971

of almost decisive importance.

11 b3 White continues his development while also

IlLlf3lLlf6 2 d4 g6 3 c4 i.g7 4 g3 0-05 i.g2 d6 6 0-0c5 7 lLlc3 lLlc6 8 d5 'tJa5 These lines involving the placement of the

consolidating the c4-square in order to isolate the enemy knight on a5.

11...f5 12 exf5 e4?!

knight on a5 contain an elevated dose of strate­

Geller rightly criticized this move and sug­

gic risk. True, Black has managed to force his

gested 12 ...gxf5!? instead. However, this would

opponent to block the centre (something White

have hardly solved Black's problems. With his

does not willingly do in the Fianchetto system)

a5-knight placed on virtually any other square

but if he is unable to generate active counter­

of the board, Black would doubtless have a

play on both sides of the board, the knight will

most promising position, but as it is, he cannot

remain passive until the end of the game. Such

sustain his e5- and f5-pawns in a satisfactory

plans as those based on ...a6, ... �b8, ...b5 and

way. White could play 13 il.b2 followed by

... e5, ... 'tJg4, ... f5 have to be combined with

'iVc2 and liae l , preparing 'tJdl and f4 in order

great accuracy, since if carried out separately

to open the kingside, where he will practically

White can face them rather easily.

play with an extra knight. True, Black could

9lLld2 e5

prevent this plan by advancing his f-pawn to f4,

Nowadays, the move-order starting with

but then White's knights would occupy the e4-

9 a6 is considered to be more accurate, but

and f3-squares with complete comfort. Here is

Black can probably play it on the next move as

an illustrative game for this latter situation:

well. The basic idea is that Black should put

Marin-Fedorowicz, Internet rpd 2000 1 c4'tJf6

pressure on the c4-pawn as soon as possible, in

2 'tJc3 g6 3 g3 i.g7 4 iLg2 0-0 5'tJf3 d6 6 0-0

...

SECRETS OF ATTACKING CHESS

10

ctJc6 7 d4 a6 8 dS ctJaS 9 ctJd2 c5 10 'ii'c 2 eS 11 b3 tDg4 12 e4 fS 13 exf5 gxfS 14 .tb2 f4 IS �ael b6 16 tDce4 .tf5 17ctJf3 l:ta7 18 h3 tDh6 19 g4 .tg6 20 'ii'e 2ctJf7 21 h4 h6 22 .th3 'itth 8 23 �h2 'fie7 24 l:lgl ctJb7 25 g5 hS 26 tDfd2 g8 24 "fixh8+ \ti>xh8 25 .id2 ! , when White will win one o f the black minor pieces and will have overwhelming c ompensa­ tion for the exchange. We can conclude that 19 . .. h6 only makes things worse for Black. b) One year after the game, Igor Z aitsev sub­ jected the game to thorough analysis and found a better defence for Black, n amely 19 . . . "fi h8 ! ?

B lack entirely adequate coun terplay; for in­ stance, 25 ... \ti>e5 26 f4+ \ti> d4 27 h4 a6 28.i dS b5 with very sharp play) 24 e7 25 "fie 1 + 'iWe5 26 'iWxe5+ dxe5 27.ixa8. White ' s chances are s omewha t better, because of his superior p awn- structure and the thematic superiority of the bishop over the knight. b I 2) H owever, the surprising 2 1 ...!txe6 ! i s even stronger. The point is that 22 'iWxh8 i s im­ possible in view of 22 ... lieI + 23 .ifl i:1xfl +! 24 \ti>xfl .i h3+ followed by 25 ... !txh8. There­ fore, White has nothing better than 22 dxe6+ .ixe6 23 'iWf3+.if5 (D) and now: ..•

(D).

w

w

The idea behind thi s move i s far from obvi­ ous. Anticipating a bit, I will mention that Black relies on the fact th at in a certain line the pin on the long diagonal i s weaker than along the h4d8 dia gonal. White has two main con tinuation s a t his disposal : b I ) Without a deeper look, it i s not entirely clear how Black has p arried the threat 20 "fic3 b6 2 1 !t e6. b 1 1 ) A partial answer is given by the line 2 1 . . . .ixe6? ! 22 .ixf6 (this exchange needs to be in serted because after 22 dxe6+ the rook is not pinned an y more and Black can play 22 ... fue6 ! ) 2 2... "fixf6 23 dxe6+ \ti>xe6 2 4 .i dS+! (this move, suggested by V alentin S toica, is better than 24 "fi xf6+ \ti>xf6 25 .ixa8, when Z aitsev believes that Black has problems holding a draw, but in fact the king is active enough to guarantee

b I 2 I ) The ending arisin g after 24 'iYxa8? ! "fixa8 25.i xa8 i s dangerous for White because after 25 . . ..ibi his queenside i s vulnerable. b I 22) Against 24 g4 Z aitsev recommends 24 ... 'iYai + 25 .tfl and now c ontinues 25 . . . 'iYe5 26 'iYxa8 .ie4 27 'iWd8 liJc6 "with the more pleas ant position for B lack" but this is not en­ tirely clear. 25 . . . Itg8 ! 26 gxf5 gxf5 27 'iYxf5+ \ti>g7, as suggested by Stoica, is probably better, intending to hide the king on h8 and start coun ­ terplay along the g-file. In this c ase, White' s compensation for the exchange looks problem­ atic. b I 23) White should force a draw by repeti­ tion with 24 'iYdS+ .ie6 25 'iYf3+ .i f5 (Zai­ tsev) . b2) Once we understan d that the above lines are not dangerous for Black, we should tum our attention to the plan employed by Geller in the game, n amely 20 l:re3. Just as in line 'bI', G eller entirely agrees with Z aitsev's variations

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ATTACK

and conclusions. I n fact, th ey both consider this move to be dubious because of 20 . . . h6 2 1.lixf6 'iVxf6 22 Itf3 .lif5 23 'iVxf6+ (Zaitsev evaluates the position arising after 23 'iVe l Ite8 24 'iVxa5 'iVai + as favourable for B l ack, but this is an ex­ aggeration since after 25 .in Ite l White has a perpetual with 26 'iVc7+ !i!>g8 27 iVb8+ !i!>g7 28 'iVc7+ ! - but not 2 8 'iVxb7+? !i!>f6 !) 2 3 ... !i!>xf6 24 g4 Ite8 (D) "and B lack stands better". In spite of some minor inaccuracies, every­ thing looks quite logical and the overall correct­ ness of the analysis is quite difficult to question. This would mean that 1 9 . . . 'iVh8 is an entirely satisfactory defence , when White has no way of proving an advantage. I was quite puzzled by this conclusion. S omething was definitely wrong from a logical point of view. Geller had attached a question mark to Black ' s 1 2th move in both his book of selected games and his notes for Informator. My general feeling was that he was right and besides I also had s ome doubts about the 1 0th move. White answered with 1 3 f6 and 14 �e4 which received an exclamation and a double excl amation mark respectively. And all of a sudden, all that Black needs in order to equalize is one good move ( 1 9 ... 'iVh8 ! ) . Geller indirectly trie s t o explain this paradox by mentioning the fact that such a defence as the queen's move to the c omer would be virtu­ ally impossible to find during the g ame. Or, re­ formulating everything, White' s 1 3th and 1 4th moves had a purely practical value. And yet, I could not believe that such a searcher for the truth as Geller would have simply b luffed in such an interesting position. Therefore, I started being concerned about the commentators' honesty. Looking at the game from a neutral position, Zaitsev was no doubt sincere, but could it be that Geller, h aving been directly involved i n the whole matter, while an impeccable artist and researcher, was at the same time slightly tricky as a sportsman? Mter all, there was still the chance that s omebody would play the whole variation against him again. It took me some time to understand that what I had to check were not the concrete variations but their c onclusions . And finally, I discovered

13

that I had to dig precisely in the position where B lack was reported to be better, instead of look­ ing for an advantage in the supposedly equal positions . Specific ally, I could not agree with the evaluation of this last position.

w

White will soon re-establish material equal­ ity and then pure positional factors should be taken into account. Just as in the game Marin­ Fedorowicz, Black would indeed have a nice position if his knight were placed on virtually any other unattacked square on the whole board. The pawn-structure favours him in principle and w e could talk about the knight dominating the bishop. However, the way it is, the knight still needs a lot of time to become active ( ... a6 and . . . b5 or . . . b6 and . .. llJb7-d8-f7-e5) while the theoreti­ c ally 'bad' bishop will be an extra bishop for practic al purposes, at least for a while. The sig­ nificance of thi s detail will b e increased by the fact that gxf5 will create some weaknesses in Black ' s kingside structure. White' s first concern should be to prevent the enemy rook from being activated along the first rank, as . . . Ite 1 -a l xa2 would free the knight in a most convenient way. Therefore, 25 !i!>n ! is more or less forced, when Black must decide on which wing to start his counterplay: b2 1 ) 25 ...h5 ! ? is aimed at winning space on the kingside and preventing the white rook from penetrating on thi s side of the board. However, White seems to h ave a way to break through: 26 gxf5 g5 (26 . . gxf5 ? ! is pointless because of 27 l:h3 !i!>g5 2 8 l:1g3 + followed by h4, �g5 and, i f .

14

SECRETS OF A TTACKING CHESS

need be, .t f3) 27 h4 ! ? (this is a better solution than 27 �h3 h4 28 f4, when after 28 . .. � xf5 29 fxg5 � xg5 30 �f3 a6 31 It f7 � e3 Black has ad­ equate counterplay) and now: b2 11) Black c annot keep the position closed with 27 . . . g4 because after 28 � f4 he i s not in time to trap the white rook with 28 ... � e5 in view of 29 � e4+ when it is the black rook that is lost! Otherwise, after 28 . .. : e5 (preventing the activation of the white rook by : e4) 29.t e4 a6 30 f3 White manages to consolidate his extra pawn, maintaining the better chances. b2 1 2) 27 . . . gxh4 28 � h3 �xf5 29 : f3+ �g6 3 0 : f4 h3 (trying to distract the bishop from the strong e4-square ; the alternative is 30 . . . a6 3 1 i.e4+, when after 3 1 . . . � g7 3 2 I1 xh4 or 3 1 . . . �g5 3 211 f5+ � g4 3 3 f3+ Black's doubled pawns are very vulnerable) 3 1.t xh3 a6 32 ..t e6 (D) and White continues being the first one to create threats.

B

For instance, 32 . . . �b8 3 3 �f3 b5 34.t f7+ � g5 35 cxb5 axb5 36 : g3+ � f5 (after the pseudo-active 36 . . . � h4 White c ontinues with 37 � g2 planning.t e6, f4 and It h3#, thus forc­ ing Black to give up material) 3 7 ..t xh5 � e5 3 8 iLf7 c4 39 � e2lL'l b7 4 0 b4 and White keeps ev­ erything under control, maintaining his extra pawn as well. b22) Since B lack fails to prove adequate re­ sistance on the kingside, starting the queenside counterplay with 25 . .. a6 makes a lot of sense. Play may c ontinue 26 gxf5 gxf5 (26 . . . g5 fails to block the kingside because of 27 l1 h3 , when af­ ter 27 . . . � g7 28 f6+ � g6 29 ..t f3 White' s �xtra

piece makes itself felt, or 27 . . . � h8? ! 28 �e3 ac­ tivating the rook with decisive effect) 27 l:ih3 � g7 28 � g3+ �f6 (D). Is there a way White can increase his pressure now?

w

Yes, there i s : 29 h4 ! . With the knight still out of play for a while, such relatively slow plans as the advance of the pawn to h5 followed by I1 g6+ c an prove quite unpleasant. 29. . . 1i e7 (threaten­ ing to neutralize White' s initiative with . . JI g7 ; after 29 ... b5 3 0 cxb5 axb5 3 1 h5 Black i s still a few tempi away from activating his knight, since the immediate 3 1 ...c4 is strongly met with 3 2 b 4 followed b y f4, in order to restrict the en­ emy king, and It g6+) 30 I1 g8 ! (an important zwischenzug, opening new horizons for the rook) 30 . . . b5 3 1 cxb5 axb5 32 h5 c4 (32 . . . : g7 is less effective now; after 33 I1 h8 � g5 3 4 f4+ �xf4 3 5 I1 xh6 lL'lb7 3 6 lI g6 l:.h7 37 h6 the black king is cut off from the main theatre of action and the threat of lI g7 is quite unpleasant) 3 3 bxc4 (now that the rook has left the third rank, the thematic 33 b4? ! fails to 3 3 . . . c 3 ! 34 bxa5 lI c7, when the pawn suddenly becomes very dangerous) 33 . . . lL'l xc4 (or 33 . . . bxc4 34 f4, when White remains in control ; in the long run, the outside passed pawn and the weakness of the black pawns might tell) 34 IX g6+ � e5 3 5 lI xh6 lI a7 3 6 lI e6+ � d4 37 h6 and the h6pawn is quite dangerous while Black' s counter­ attack has hardly started yet. What is the conclusion? First of all, we should note that even in case of Black's best defence, as suggested by Igor Zaitsev, White retains a long-term initiative.

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ArrACK

The question whether this i s enough for a w i n i s beside the point, because at th e moment h e started his c ombination, Geller h a d a n un­ doubted, but by no means winning, advantage. Secondly, it is quite remarkable that the suc­ cessive modifications in the character of the po­ sition do not significantly alter its obj ective evaluation. F inally and, from a c ertain point of view, most intriguing are the practical diffi culties in pursuing the truth in the jungle of variations. One whole year was needed for Black's best de­ fence to be found, but even so, only the tip of the iceberg was revealed. Anyway, after the move played in the game, White' s task is considerably easier.

20l:te3! (D)

B

15

iVh 8 ! is less clear) 22 i.xh6 b5 (this way, B lack' s previous play makes some sense) 23 cxb5 ':xb5 24 .i.g5 (renewing the threat of g4) 24 ... ttJ c4 ! ? (the only way to avoid immediate disaster; 24 J:lb4Ieads to a hopeless ending af­ ter 25 g4 i.xg 4 26 hxg4 �xg4 27 .i.xf6 iVxf6 28 iVxf6+ �xf6 29 �e6+ �f5 3 0 l:txd6, when 30 . . . c4 is too slow in view of 3 1 :a6 winning) 25 .i.xf6 (before- c apturing the knight, White has to release the tension; 25 bxc4? loses the queen to 25 . ,U,b l +) 25 .. .'iWxf6 26 'ii' xf6+ �xf6 27 bxc4 J:b l + 28 � h2 l:[ c 1 (Black has m anaged to obtain some counterplay, but with accurate play White can keep things under control) 29 g4 i. d7 3 0 l:a3 lhc4 3 1 It a6 ! ? (aiming to decoy the king to a more exposed position) 31...�e7 (better than the apparently active 3 1 . . .�e5 , when after 3 2 :'xa7 Jib5 3 3 :te7+ the d6-pawn is lost) 32 lha7 with a clear advantage for White, although B lack still seems to be in the game. ..

. .

21 :f3 .i.f5 22 g4 'ii'h8 (D)

w

We are already familiar with the idea behind this move. White will soon restore the material balance under much more favourable circum­ stances than in the lines examined above.

20 b6 •.•

Resignation. Velimirovic accepts the inevi­ table and prepares to bring the knight back into play in the slowest possible way. G eller in­ tended to m eet 20 ... i.f5 with 2 1 h3 ! - again un­ hurried moves , in spite of being a whole rook down. As we have already seen, B lack' s ac­ tions need a lot of time as well. For instance, 2 1...h6 ! ? (probably the best practical chance; G eller only analysed 2 1 . . .h5, when after 22 g4 ! hxg4 23 hxg4 i.xg4, 24 i.h3 ! leaves B lack un­ able to prevent the penetration by the white rook via either e6 or the h-fIle; instead 24 :te4

23 .i.xf6! The attempt to increase the pres sure with 23 i.h3 c an be answered by 23 . . . Itg8 ! , when some problems along the g-file might arise.

23 "'xf6 24 iVxf6+ �xf6 25 gxf5 gxf5 26 �e3 •••

The occupation of this vital open file guaran­ tees White a virtually decisive advantage. After the hurr i ed 26 .ih3 ? ! Black can still defend with 26 .. . 1:.g8+ 27 �f1 It g5 .

2 6 ttJb7 2 7 l:e6+ �f7 (D) 28 .if3! •••

SECRETS OF A TTACKING CHESS

16

Geller considers 40 b6 a more technically ac­ curate move, giving the following line: 40 l:1b4 4 1 b7 hS 42 a4 h4 43 as h3 44 a6 h2, when White is just in time to hold the enemy pawn with 4S �h7, after connecting his own a- and b-pawns. ...

w

40 J:th3 41 f3 Wd4 42 b6 �h2+ 43 wet :hl + 44 Wf2 :h2+ 45 Wg3 l:b2 46 b7l1xb3 47a4 c4 48 as c3 49 a6 c2 50 �c71-0 An impressive game ! Had Velimirovic played against some other opponent, his inac curacies might have passed without being noticed. (In fact, his sharp and risky style of play had yielded him a 1 00% score in the tournament crosstable before meeting Geller in the 6th round.) How­ ever, I assume that a logical mind like Geller' s had little trouble finding the right (though hid­ den) path at the critical moments. Indeed, even if it was impossible to c alcu­ late the final consequences of the sacrifice, the passive position of the as-knight was obviously going to offer long-term compensation for the rook. Prob ably the most important thing to foresee was the pawn sacrifice prepared by Velimirovic against the natural 14 .ltb2. After that, the method of elimination was rather easy to apply, resulting in the fantastic 14 liJdxe4 ! !. However, there are certain situations which can prove much more difficult to handle from a practical point of view even for such a perfec­ tionist as Geller. True, the following game was played 20 years before the one we have just seen but we can hardly try to explain Geller's failure to find the right path by his lack of form or ex­ perience' since he shared 2nd p lace in this event. ••

In view of the threat of .lthS+, Black cannot guard the e-file and defend his fS-pawn at the same time.

28 J:tg8+ 29 Wfl Wf8 As Geller pointed out, 29 ... �gS does not offer a reliable defence because of 30 h4 driving the rook away from the control of the hS-square. 30 iLh5 �g5 3 1 �e8+ Wg732 :e7+ Given the mutual time-trouble, the transition to a rook ending with a material advantage is the simplest way of converting the positional advantage into something concrete. White has taken full advantage of the passive position of the knight and this seems like a good moment to eliminate it, since the knight has fmally be­ come u seful, by offering some protection to the black queenside. 32 Wh6 33 l'ixb7 l:Wt5 34 �xa7 �h2 35 �d7 Wg5 36 ':xd6 Wf4 37We2 b5 (D) ••

•••

Geller - Keres USSR Ch, Moscow 1951

w

38 cxb5 �e5 39 �d7 l:1h4 40 a3

As in the previous example, one of the main is­ sues will be a relatively misplaced as-knight. However, we shall notice two significant differ­ ences. First of all, Keres's knight will have complete freedom of action and its passivity will have a temporary character, which must have placed p sychological pressure on Geller. B esides, White was given a choice of principle (which he didn't in the previous game): either to focus on the loose position of that knight in order to create a double attack or to ignore that

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ATTA CK

side of the board completely by launching a devastating attack on the other wing.

1 e4 eS 2 lb fJ lbc6 3 .lib5 a6 4 iLa4 lbf6 5 0-0 .lie7 6 11e1 b5 7 .lib3 d6 8 c3 0-0 9 h3lbaS 1 0 i.. c 2 c5 1 1 d4 "ifJc7 1 2lbbd2 cxd4 13 cxd4 i.b7 14lbfl l:ac8 (D)

w

17

Later, Keres started employing another, more subtle, method based on 1 5 ... lbd7 against both 1 5 i.d3 and 1 5 i.bl. In case of a natural devel­ oping move, such as 1 6lbe3 , he would release the tension in the centre with 1 6 . . .exd4 in order to activate all his minor pieces. This method of play looks especially effective against 1 5 .libl, when White still has to justify the placement of his bishop. The critical test probably consists of 16 d5, when Black can stir up huge complica­ tions with 16 ... f5, undermining the d5-pawn and activating his b7-bishop.

w

So far, both sides have played quite logically. Even today, this is considered to be one of the main lines of the Ruy Lopez. For instance, ECO classifies this position under e99, in a certain sense meaning 'the ultimate' variation of this an­ cient opening at the time the code was created.

1 5 i.. b l At the time when the game was played, this slightly strange move was an attempt to im­ prove over 1 5 .lid3 , when after 1 5 ...d5 1 6 exd5 Black can play 1 6 ... e4 ! 17.lixe4 lbxe4 1 8lhe4 i..x d5, with excellent compensation for the pawn. By retreating the bishop to a less exposed square, Geller avoids the fork with ... e4 and keeps his attacking potential intact. The obvi­ ous drawback of this move is that the aI-rook is cut off from play for a long time, but, as we shall see, Black needs to display great ingenu­ ity to take advantage of this detail. I S...d5 (D) At the time I am writing, theory on this whole variation is rather well developed. How­ ever, until very recently, I have never played the closed systems of the Ruy Lopez. And yet, my feeling is that this move is an inaccuracy. Black opens the diagonal of the white light-squared bishop too easily.

1 6 exd5 exd4? ! Since 1 6 . . . e4 seems less effective now in view of 1 7lbg5, Keres adopts a rather simplis­ tic solution. This attitude could have cost him dearly. Later, he c ame to the conclusion that this line should be continued a bit further with 1 7 . . . 11fd8 1 8 lbxe4 lbxd5. Indeed, Black has definite compensation for the pawn, in view of his excellent mobilization and White' s prob­ lems developing his queenside (to which the bI-bishop belongs now). As for a definitive evaluation of this position, tournament practice might some day give an answer. After Keres's move, White has a very promis­ ing position. With the exception of the aI-rook, all his pieces are well situated. His minor pieces are targeting the enemy kingside, the el-rook keeps the e7-bishop under pressure, while the queen is ready to j oin an eventual attack. Black has some problems with his coordination. His a5-knight is quite far from the king and can hardly be of any use in the eventuality of a

SECRETS OF ATTACKING CHESS

18

lightning attack. Even the seemingly active ma­ j or pieces doubled on the c-file do not threaten anything serious. With his next move, Geller causes his oppo­ nent big problems with recuperating the dS­ pawn.

problem is that B lack has already gone wrong and there are no completely satisfactory options available to him any more. Curiously, contem­ porary theory is not too well settled here:

1 7 �g5 ! (D) w

B

a) ECO gives 18 'fid3 as winning, but Black seems to be able to survive with 18 . . c;t>f8! (ECO only mentions 18 . .. g6? when after 19 l:1xe7 'fixe7 20 'fixd4 White obtained an overwhelm­ ing attack in Ernst-Hedman, Avesta 1 993), ac­ knowledging that the inftltration of the queen to h7 is not dangerous in itself, if it does not lead to an immediate mate. For a similar move, see the game Anderssen-Steinitz from Secrets of Chess Defence, Chapter 4 (The King as a Fighting Unit). After 19 �xf6 (19 ct:Jxd4 can now be met by 1 9. . h6 !?, followed by the capture on dS) 19 .. .i.xf6 we suddenly see the drawback of re­ treating the bishop to b1: the lack of coordina­ tion between the rooks and the weakness of the first rank prevent White from obtain anything c oncrete. For instance, after 20 'ti'a3 + (the most ambitious move but not the best; White should probably admit that he has no advantage with either 20 'ti'xh7 :Xel 21 ct:Jxel .txdS or 20 l:1xe8+ l:1xe8 21 'ti'xh7 �xdS; in fact, I would prefer Black in both cases) 20 . .. �e7 2 1 l:1xe7 (this is intended to be a strong tactical blow) 2 1 . ..'fixe7 22 'ii'x aS White has managed to win material, but after 22 . . . l:1c 1 23 'ii'd 2 'i¥cs his queenside is completely p aralysed. Once again, the usefulness of the move . . . cJif8 is obvious: White cannot take on h7 with check. b) 18 'fixd4 seems to be a better solution. After 18 . . .ct:JxdS 19 .tfS ':cd8 2 0 llac 1 ct:Jc6 2 1 .

This pin is quite unpleasant. The main point is that 17... ct:JxdS? leads to disaster in view of 18 �xe7 ct:Jxe7 19 �xh7+! cJixh7 20 ct:JgS + cJig6 2 1 ct:Jg3!, as given by Keres. This is a superb il­ lustration of the force displayed by White' s mi­ nor pieces in combination with the queen. The bishops were employed to eliminate important defenders and to create disharmony in Black's camp, while the knights offer the queen reliable support in the hunt for the king. The immediate threat is 22 'fihS+ cJif6 23 ct:Jh7#, while after 21. . .l1h8 22 'fig4 ! the threat of ltJe6+, hitting both g7 and c7, is impossible to parry adequately since 22 ... fS allows 23 rie6# while 22 . . .cJif6 is strongly met by 23 h4!!, placing the gS-square under control and creating decisive threats such as l:1e6+ or ct:Jh7+ followed by 'figS#. Keres took what l ooked like the most reason­ able decision:

1 7 h6? The idea is to interpolate this move before capturing on dS. H owever, Keres c orrectly points out that this natural move should h ave been the losing mistake. However, he is too op­ timistic when considering Black's p osition as entirely satisfactory after the developing move 1 7 .. . l:1fe8 (D) . This is a good move, but the ...

.

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ATTA CK

\Vhite his deeh hhpment and tains strong pressure. This has occurred in some correspondence games. now relLEn 17 . . . h6"'

w

19

might asked lf whether here wasn�t any simpler solution available. Then he must have discovered the variation actually played in move and decided that hhp to the the this better decisioi1� baing com� pletely fIsk-free. If my supposition is right, Geller would h ave needed just an effort of will to follow his instinct. since Keres considers the of the with to be the evil, he agree:h after i.g5 �xd5 20 'iVd3 g6 2 1 h4 White has a nice attacking position. Although these variations instructieo� most imereting phasc� are of thc:� pine is still nome. 18 ...lbxdS 19 'iVd3 (D)

B

1 8 �h4? This is the turning point of the game. White's and onn quite far has the danger louming Over I am pretty sure that Geller was looking for a tac­ tical crowning of his superior strategy and that the sacrifice I S �xh6 ! didn't escape his attenlhlter 1 S.��n the 'quiet' unsolvahle problems. gives again, the combined action of the queen and the minor pieces has a devastating effect. Keres the follO\v1 analysis 5" fd8 the kinh hi the WhH activate rook via the fifth rank; both 1 9 ... �xd5 20 'iVxh6 � xf3 2 1 lbg3 ! followed by lbf5 or, in some cases� lbh5 and 1 9 �97 20 lbh3 lbgS 2 1 lbf5+ lbxe7 I Hhpeless Black) 20 Keres indhchttes, 2 1 lbh5 2 1 �e4h less clear in view of 2 1 ... �fS ! 22 'iVxf6 i.g7 23 'iVh4 d3, when White suddenly has problems his queensihe while thn black king (eliminhhh "h5 22 safety! the h5-rook, whIch IS the key defender of the king) 22 . . . lbxg4 23 'iVxh5 lbf6 24 'iVg5+ �fS 25 'iVh6+ c;t>eS (or if 25 ...�\ S then 26 lbg5 foland 17+) and by iLh7 compktinh development sirnple and threatening 'iVhS+ or �e 1 i s decisive. However, calculating everything over the thOrne here and is not ��

nixt few 1/1(>\es, Whit:: Here hend on d4. HovveLvv it is pev� could captured fecdy understandable that such an approach, obviously aiming for equality, didn ' t appeal to Geller� At the same entirel e it is no clear White In� cClualize deed� b l -bis nv", Hill not Dosition easy to improve, while the a5-knightjust needs tojump once (to c4) to j oin the rest of the well­ black "

dO �g3

A

2 1 �xdti

It is quite telling that after 2 1 lbe5 lbc4 ! Black's minor pieces are perfectly coordinated for an attack agai n�t the enemt For 'iVxg6+ ��;hxg6? stancc� case of -vJiig 7 B1hLte cannot pieDc his with 24 'iVe6+ c;t>hS 25 i.xd6 due to 25 . . . lbf4 (as indicated by Keres).

2

,!�;vcd6 22

20

SECRETS OF A TTACKING CHESS

from a tactical point of view, as the comments based on Keres's variations (see below) clearly

\;ig2

This was the position Geller had been aimfor. The delivers double and 'practLalld forced' ";;lc4, Wall excellent ;lttacking with 23 'iVxh6, threatening 4Jg5 followed by l1e6 or maybe h4. We cannot really blame Geller being by simple lIlion when rejectinh Ole combi 1 8 .i.xlt« \\ho knows , if he had had such a choice in his game against Velimirovic, maybe chess history would had one rnasterpiecc ;nlike counger Keres ckarly ringing as a sIgn that heard the ' 'Ite bishop has been on b l too long and seized the opportunity to change the course of the game radically: 02 4Jf(L What a shock! Black leaves a whole knight en p rise for the sake of a direct kingside attack. still immedi Il()U­ preventlnd the bishop rom eing cut 011 Irom the play with . . . d3, Imt he was not psychologically ready to switch to pas­ sive defence . Keres writes that he intended to «ontinue Wi )L.4Jc4, the pieces almost sormnetrical rather than 23 ... 4Jc6, when 24 4Je5 offers White some compensation for the pawn. •••

)3 'iWxa3'1! lcof3 24 C;1'lxh3+ !lor the w hite will ew move;;, be defended only by a knight and two doubled pawns, obviously a poor match for the tandem of queen and knight. B esides, Black needs just lude hh; in thl' l1;lck. tempo wonder I1verythin) works out

White feels like driving the knight back. In case of 25 h8 3 2 i.xf7 'ii'xf7 and Black' s strong mass o f pawns offers him adequate compensation for the ex­ change. B efore deciding which plan to choose, White should fITSt capture on c4: 25 bxc4 ! (D).

B

Once again, this i s the move that involves the smallest degree of commitment. Now: a) If Black captures back with 25 . . . dxc4 the advance of the e-pawn with 26 e6 ! i s more ef­ fective. a l ) 26 . . . lLldf6 continues in the same way as in the previous variation, but now after 27 exf7 + Ihf7 28 liJe5 i.d6 29 lLlxf7 l:txf7 White can

36

SECRETS O F ATTACKING CHESS

safely play 3 0 j£,e6 with a clear advantage. First of all, B lack' s queenside structure is less com­ pact after the previous exchange on c4. Sec­ ondly, White has coped without the exchange on f6, thus preserving his strong bishop and leaving B lack with a passive knight on e 8 . a 2 ) Black can also tr y 26 . . . fxe6 but this is hardly an improvement. After 27 .itxe6+ h8 28 ttJd5 White takes over control of all the cru­ cial squares and is ready to launch a devastating attack. Black has no time to establish commu­ nication between his pieces. 28 iH6 (with better piece coordination, Black might be suc­ cessful with this attempt to neutralize the pres­ sure along the a l -h 8 diagonal, but here this move does not save the day; however, other moves are no better: the neutral but rather slow 28 . . . .itd6 gives White the time for 29 hS ! fol­ lowed by ttJh4, while 28 . . . bS ! ?, relying on the poorly defended f3-knight, is strongly met by 29 ttJgS ! , in both cases with a devastating at­ tack) 29 ttJf4 ! (the combined pressure from White' s minor pieces and queen will soon be­ come unbearable; of course, White is not inter­ ested in capturing on either f6 or c7, because this helps Black activate his knight) 29 ... .itxb2 (the more c autious 29 . . . ttJd6 prevents the sub­ sequent fireworks, but fails to provide an ade­ quate defence in view of 30 ttJgS ! ) 3 0 ttJg6+ (this move involves an important commitment, because most of the pieces will be hanging now; this means that White should c alculate the ensuing attack right to the end before giv­ ing the check on g6) 30 . . . h7 3 1 ttJgS + ! xg6 32 j£,fS+ ! ! (an elegant move, clearing the e­ file { and in particular the e6-square } for the queen) 32 .. JhfS (otherwise ii'e6 would mate) 33 ii'xe8+ f6 and now, the knight that had previously forced the king out of his shelter and offered support to the queen in some side­ lines unexpectedly mates with 34 ttJh7#. b) B lack's other possibility is 2S . . . d4, aim­ ing to keep White' s dark-squared bishop tem­ porarily out of play, but allowing his opponent to install a strong knight on d5 with 26 ttJd5

B

. . .

(D) . Play acquires a slower character than i n line 'a' but the general assessment is more or less the same: in spite of his material deficit, White has

the advantage. For instance: 26 . . . ttJb8 (Black i s practically forced t o remove this knight from the rather exposed d7 -square, since after 26 . . . .lid8 27 ttJxc7 ii'xc7 28 e6 ! he loses at least an ex­ change, without obtaining adequate compensa­ tion) 27 e6 (even without creating immediate threats, this move remains very strong) 27 ... .lid6 (placing the bishop on a relatively safer square and clearing the seventh rank; 27 . . . fxe6? loses outright to 28 ii'xe6+ followed by ttJeS, while 27 . . . f6 is answered by 28 hS and ttJh4 with con­ tinuously increasing pressure) 28 hS ! ? ( since White' s play is oriented towards invading the kingside light squares, this move is of higher priority than some other natural options such as lie l ; more specifically, the point is to discour­ age the capture on e6 because of .itxe6+ fol­ lowed by ttJh4 - this is a first typical moment illustrating the hidden strength of provoking . . . h6 right after the opening) 28 . . . ttJc6 29 lie 1 (but now, this move is well-timed, putting addi­ tional pressure on the e-file and preparing the activation of the dark-squared bishop; against the immediate 29 ttJh4 Black c an consider playing 29 ... ttJe7 aiming to cover the fS- and g6-squares and exchange the central knight) 29 . . . fS ! ? (aiming to defend in a more or les s blocked position; if Black waited a few more moves, then White' s pre ssure would become decisive after .lic 1 and ttJh4; now that White' s maj or pieces are doubled o n the e-file, 2 9. . . ttJe7 certainly looks less appealing than in the previ­ ous line, and White would play 3 0 exf7 + lllf7 3 1 �e6 with an overwhelming advantage) 3 0 ttJh4 ttJe7 3 1 .lig2 ! (White i s not interested in

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ATTA CK

37

winning the exchange b ack, because his minor pieces are definitely stronger than the passive black rooks) 3 1 . . :iVa6 3 2 ltJg6 (D).

his light squares and therefore we can conclude that objectively Horvath' s 23 . . . .:d8 was a better defence.

B

24 ltJe3 (D) Following the policy initiated with 23 .th3 . 24 e6 is still premature in view of 24 . . . ltJdf6 25 ltJe5 'tid6 with unclear consequences.

B

It becomes obvious that Black has problems maintaining the blockade on the dark squares, mainly because of his weaknesses on the light squares. Now: b l ) 32 . . . ltJxd5 helps White activate his Cata­ lan bishop without loss of time with 33 .i.xd5 . For instance: 3 3 . . ..:t"6 34 e7+ h8 29 h5 ! liJdf6 30 liJh4 or 27 . .. i.f6 28 i.xe6+ 'it>h8 (28 . . . 'it>h7 29 fVd3+ g6 is even worse because of the reply 3 0 h5) 29 liJf4 with a devastating attack in both cases. b32) Therefore, 26 . . . liJdf6 should b e pre­ ferred, when things are slightly less clear than in the similar lines examined above. For in­ stance, 27 exf7+ lIxf7 28 liJe5 d4 (this is the last possible moment for B lack to close the dan­ gerous diagonal; after, for instance, 28 . . . .td6 29 liJxf7 'it>xf7 White play s the familiar 30 .tg2 ! and if 30 iH8 aiming to protect the cru­ cial d5-pawn, then 3 1 g4 ! c ontinuing to under­ mine it) 29 .te6 fVe4 (the centralization of the queen is an important tactical element, when ei­ ther attacking or defending ; 29 . . . liJd6 fails to defend the f7-rook securely because of 30 liJf5 !) 30 .txf7+ 'it>f8 3 1 f4. Black will soon win his piece b ack, but White' s position looks prefera­ ble in view of the w eaknes s of the g6-square and the strong outpost on e5 . However, there is still a lot of play left. For instance, after an

39

exchange of queens on e3 in case of an eventual .tg6, White will have to lose some time to win the p awn back. c) The other possible way of changing the course of the game by closing the a l -h8 diago­ nal is the immediate 24 ... d4. In case of either 25 liJdS or 25 liJf5 B lack intends to play 25 ... ..tf6, thus returning even more material than just a pawn. The resulting position would set White' s pair o f bishops against a rook and two p awns, which means approximate material equality by middlegame standards (in the endgame the value of the pawns tends to increase). In principle, White' s further play is easier to carry out and he is probably entitled to claim some advantage. First of all, the central pawns will not be easy to promote, because the d4-pawn i s under pressure. Secondly, White' s minor pieces are relatively stable and could produce a kingside attack in the future. However, the position would remain quite complicated. 25 liJf5 :reS (D) 25 . . . lIde8 has no independent value, being likely to transpose after 26 liJxg7 ! 'it>xg7 27 .txd7 lId8 to note 'b' to Black ' s 27th move.

w

...

S o far, Black has defended with the most nat­ ural moves and his light squares seem to be more or less securely covered. As it often hap­ pens, the tactical explosion will take place on a dark square.

26 liJxg7! Thi s i s White ' s first concrete achievement: the black king ' s position i s irremediably weak­ ened. However, this cannot be regarded as a real

SECRETS OF AITACKING CHESS

40

sacrifice. In fact, White will soon recover all his previous material investment.

26 �xg7 27 �xd7 d4 •••

Once again the safest defence. By closing the long diagonal, Black c onsiderably reduces White' s attacking potential. There i s also a cer­ tain abstract logic behind this move: Black's main achievement s o far i s of a material nature and his strong centre is an important part of it. Therefore, in order to make use o f his whole potential and thus more or less maintain the balance, he has to benefit from his pawns in one way or another. Other moves are not too appeal­ mg: a) 27 .. Jhd7 allows White to recover most of his material, keeping a strong attack after 28 e6+ �f6 29 .ll.xf6+ �xf6 30 iYe5+ �e7 3 1 exd7+ �xd7 32 'iif5+ followed by bxc4 and 'tJe5. b) 27 . . . :Sf8 100ks quite passive. After 28 e6+, 28 . . . c3 29 .ll. x c3+ transposes to note 'b2' to Black's 24th move, while after 28 . . . d4 29 'tJxd4 cxd4 30 .ll.xd4+ f6 3 1 'iig 4+ �h7 32 'iUf5+ �h8 33 'iUg6 Black's pieces, cramped on just two ranks, can do very little to parry White's numerous threats. In case of 2 8 . . .f6 (D) White has at least two promising ways to carry out his attack:

followed by lLlli4-f5 and 'iUg4, when rather soon White will be attacking with all his pieces) 30 iYxc4 'tJe8 (removing the knight from the c-file in order to prevent 'tJxd4) 31 'tJh4 f5 3 2 'tJg6 with strong pres sure. b2) The more dynamic solution is 29 'tJe5 ! ? sacrificing the knight but keeping the dark­ squared bishop alive. Play may then continue 29 . . .fxe5 30 .ll. xe5 + .ll. f6 3 1 e7 ! (underlining Black's vulnerable pieces) and now: b2 1 ) After 3 1 . . . .l'ixd7 3 2 'iUg4+ �f7 3 3 exfSiY+ �xf8 3 4 iYf4 g8 3 5 iYxf6 Black's only defence i s 3 5 ... 'iif 8, when 36 iYg6+ �g7 37 iJ..x g7 wins material for White, while Black doesn' t really h ave time to create any true counterplay with his queenside pawns, mainly because of the weakness of his king; for in­ stance, 37 . . . 'iUxg7 38 'iUd6 c3 39 l:tel c2 (other­ wise lr.e7) 40 lr.c 1 ! winning the pawn and very soon the game. b22) 3 1 . . . �xe5 (relatively best) 3 2 exf8iY+ lhfS 33 'iUxe5+ lr.f6 34 bxc4 and White re­ establishes material equality, keeping the better chances due to his superior activity and safer king position.

28 �xe8 lr.xe8 (D)

w w

b l ) With Black' s pieces so passively placed, a slow plan such as 29 h5 is perfectly playable. White intends to occupy the weak light squares with his pieces, starting with 'tJh4. For instance, 29 . . . d4 (29 . . . c3 fails to restrict the bishop, be­ cause after 30 .ll. x c3 d4 it can be transferred to another, equally effective, diagonal with 3 1 .ll.d2

After eight moves of sharp play, the posi­ tion i s about to c alm down. With his next move, White c an re-establish material equality by capturing on c4. How could we describe the events we have just witnes sed? In a slightly worse position, Black tried to solve his problems by releasing the tension in

THE LOGICAL COURSE OF THE GAME: THE SPARK OF THE ATTA CK

the centre. Obj ectively speaking, his decisions were probably no better or worse than 1 9 . . . lDc7 or 20 . . . lDc7, but from a practical point of view it required from White a certain ingenuity. 2 1 dS was the spark that c aused the explosion of the previously accumulated tension, c au sing the play to become highly dynamic. The knight sacrifice on g7 was the c ulmination of the pre­ vious play and caused irreparable damage to B lack' s structure, in the same way a proj ectile uses its energy to make a big hole in any wall that happens to be in its way. The similarity goes a bit further too. Once the damage has been done, the projectile becomes an inert ob­ j ect. Two moves after 26 lDxg7 a new phase of the game is about to start. Once the dynamic potential has been converted into something of a more stable nature, positional play is ex­ pected to start again. Indeed, White cannot cre­ ate any new threats in the near future, since both his minor pieces left on the board still need some time to become active. From a logical point of view, we would ex­ pect the obj ective evaluation of the position to be more or less the same as before the tactical phase, since, based on the detailed analysis pro­ vided above, we do not consider that either player made any significant mistakes in this phase of the game. True, some alternatives were worth considering at certain moments, but they were by no means stronger than the moves played in the game. Since psychology is another important di­ mension of the game of c he s s , I shall als o try to describe my s tate of mind around this mo­ ment. The tournament (a strong GM round­ robin event) had gone badly for me so far and all of a sudden I was given the possibility of all these tactical fireworks. B eing surrounded by a crowd of onlookers (including GMs) while sac­ rificing large amounts of material is quite re­ warding, of course, but the desire to achieve something of a more c oncrete nature on the board itself can be a rather stres sful element. Finally, eight moves after the start of the whole operation, I was given the possibility to re­ establish material equality, keeping my chances for an attack intact in view of the weakened po­ sition of the enemy king. This caused me to

41

relax prematurely and I completely failed to realize that this was another critical moment, of no lesser importance than that before the 2 1 st move.

29 bxc4? ! A s far as I remember, I played this move rather quickly. The desire to block the poten­ tially threatening queenside maj ority is quite natural, but I should have been equally c on­ cerned about the other half of the board, on which all my hopes were based. White ' s basic plan is to regroup with llel , .i.c 1 and lDh2-g4 . However, this plan i s rather time-consuming and B lack can organize his defence. Knowing the further c ourse of the game, I can s ay that . . . f5 is an important defensive resource, aiming either to block the p osition or to activate several of Black's pieces in case of exf6+. If White could prevent the advance of the f7-pawn for a while, he would have every chance to relaunch his kingside attack. Without these preliminary considerations, I c ould not find the apparently paradoxical move 29 'iVxc4 during the game. Black c annot enjoy the tempo spent by his opponent on c apturing the pawn, since the threat of 30 lDxd4 has to be prevented somehow. The most natural way to continue is 29 . . . lDe6 (29 . . . il.d8 is strongly met by 30 b4, aiming to weaken, blockade and eventually win the enemy pawns, while 29 . . . b5 prematurely weakens the c5-pawn - after 30 'iVc2 Black should prevent the pawn-break b4 with 30 . . . lDdS but then 3 1 iVe4 wins time for building an attack with llel , il.c 1 , lDh2, etc.). However, the knight' s stability on e6 is only rel­ ative. In the long run, f4-f5 will be very strong. B esides, with the prospect of a further opening of the e-file, the knight would certainly be in the way of its own pieces. Once White has c ap­ tured the pawn and induced a slight lack of co­ ordination in Black ' s camp, he can transfer his queen back to the kingside starting with 30 iVd3 . This move prevents . . .f5 for a while ( ... f6 is simply answered with llel ) and will be fol­ l owed by the standard plan: llel , �c 1 , etc. This is a logical conclusion of the previous compli­ cations and B lack would have faced a difficult defensive task. 29 iVb7 •••

42

SECREI'S OF A ITA CKING CHESS

29 . . .'iVc8 ! ? aiming to place the queen on ei­ ther g4 or f5 is also interesting; for instance, 3 0 'ii'e 4 f 5 3 1 'i1ec6 lbe6 with a c omplicated game. 30 lbh2 (D)

B

3 1 l:.e l should be preferred. However, my eval­ uation that White keeps a clear advantage was too optimistic. After, for instance, 3 1 . . .'ifc6 32 Jic 1 'ife6 (or 32 . . . 'i¥g6 ! ? aiming to provoke 3 3 h 5 before placing th e queen o n c 6 and thus get the gS-square for the bishop), B lack keeps a s olid position. Later on, he c ould also consider some queenside activity, based on . . . a6 and . . .bS. Obj ectively, White cannot claim any ad­ vantage. 31 i.xf6 32 'ifg4+ �h8 White has allowed the centre to be opened at a moment when two of his pieces (the rook and the bishop) are rather misplaced. Black's posi­ tion is now to be preferred, mainly because of his c ontrol of the only open file and the long di­ agonal. However, from this point on, the game was marked by B lack's tiredness after a diffi­ cult defence and by his acute time-trouble. Al­ though from a purely chess point of view such external factors are of no relevance, from a practical perspective we could also say that the objective reality from the board has been trans­ ferred into another dimension. I managed to win only after several mistake s by both sides, which explains why I prefer to refrain from giv­ ing the rest of the game in a book mainly dedi­ c ated to the truth. However, the tournament committee later awarded me the beauty prize, apparently under the impression of the first half of the game and failing to understand the nature of its second half. •••

30 f5! In spite of his time-trouble, 10zsef understood that he could not cope without this move and that the sooner he played it, the safer it would be for him. After, for instance, 30 . . . ikc6? ! 3 1 ge l 'iie 6? ! 3 2 �c 1 threatening ltJg4 with a strong attack it i s too late for 3 2 . . . f5 ? ! because of 3 3 exf6+ i.xf6 34 �xh6+ ! , when 34 . . . �xh6 is impossible due to 35 �d2+ winning the queen. 31 exf6+? ! In my comments for Informator I c orrectly pointed out that this move is inaccurate and that •••

2 T h e E l e m enta ry Stages of t h e G e n e ra l P l a n

About ten years ago, during one of the training sessions of the Romanian Olympic team (some­ thing not very common in our country, 1 must confess), our captain showed us the following position with the purpose of testing our strate­ gic abilities.

B

Ahues - Alekhi n e Bad Nauheim 1 936 All of us knew that the position was taken from a famous game played by Alekhine and remembered more or less the further course of the game. (Alekhine obtained a crushing attack after 1 5 . .. .txg3 1 6 hxg3 CfJe7 1 7 b4 'iVd7 1 8 CfJc2 CfJed5 1 9 CfJa3 b 5 20 axb5 axb5 2 1 'iVe2 c6 { the consolidation of the queenside has been completed and Black can concentrate his efforts towards the other wing } 22 CfJc2 'iVf5 23 'ufc 1 h6 24 'ua5 'uac8 25 CfJal CfJg4 26 Wfl 'ue6 27 'uxb5 'uf6 and White ' s position soon collapsed. ) 1 believe that m o s t of us considered that several paths would lead to Rome here, since White already had a difficult position, with his passively placed e l-knight and his 'bad' bishop.

Therefore, the minutes of silence that followed after the moment the position was set up on the board were mainly generated by politeness rather than by our inability to solve ' the test' . And yet, trying to look at the position with fresh eyes 1 discovered that 1 6 fxg3 could have spared White a lot of trouble. 1 thought that B lack's main problem is that he cannot keep both wings under control simultaneously. Spe­ cifically, White threatens to improve his posi­ tion with either 1 7 g4 or 17 b4. His king is definitely safer than in the game while the half­ open f-file could offer him some chances for counterplay. Therefore 1 considered the move 15 ... aS instead, taking the b4-square under con­ trol, in order to prepare . . . .txg3 followed by . . . 'iVd7. At some moment, one of my team­ mates considered that we had dedicated suffi­ cient time to ' solving' the puzzle and we started discussing the position. We soon concluded that 1 6 fxg3 wouldn' t have completely solved White ' s problems in view of the strong answer 1 6 . . CfJdS . In the middle of the debate, Mircea Pavlov, who was the assistant captain and had been si­ lent until that moment, s ai d something that caused me a deep revelation: "Personally, 1 would try to apply Alatortsev ' s method of thinking and ask myself in which stage of the s trategic construction B lack finds himself. The answer is that he should connect rooks and the best way to do it i s 1 5 . . . 'iVd7." Although Pav­ lov's statement was met by the audience with a total lack of enthusiasm (I even remember that somebody ironically asked with a French ac­ cent "A la Who? Never heard about him ! "), 1 understood immediately what he meant and that the implications were of a very deep nature. The move perfectly suited my intentions to .

44

SECRETS OF A ITA CKING CHESS

keep both wings under control: the queen places the g4-square under observation in advance, while b4 is prevented by keeping the bishop on the board for the moment. The decision be­ tween . . . ..txg3 and ... a5 will be taken later de­ pending on the concrete circumstances. According to Vukovic (in The Art of Attack in Chess), this i s indeed the correct move-order, since the move involving the least commitment is made fIrst. Thus, several systems of thinking had the same point of convergence, something that is more typical for mathematics . And yet, the reader might ask, isn't all this just much ado about nothing? After all, B lack had a signifi­ cant advantage in the diagram position anyway. Isn ' t it out of place to make such abstract con­ siderations here? Maybe it is. B ut I have related this small in­ cident because it had a strong influence over my way of thinking for a long period. While considering their next move, how many of us haven' t been carrie d away by calculating count­ less moves, without any precise orientation? Sometimes, simple thinking can both spare a lot of energy and time and show the correct di­ rection for further investigation. Here is a case where I made full use of the 'Alatortsev method' .

Marin - Krasenkow Olympiad, Elista 1 998 1 liJf3 dS 2 g3 ..tg4 3 i.. g 2 liJd7 4 c4 e6 S cxdS exdS 6 0-0 liJgf6 7 liJc3 c6 8 d3 ..teS 9 h3 ..thS 10 a3 as 1 1 e4 dxe4 1 2 dxe4 0-0 (D) During the first half of the Elista Olympiad I had been feeling quite ill. Therefore, after 1 0 rounds I had played only 4 games with the dis­ couraging score of just 1 point. By the time Pavlov, who had been promoted to captain in the meantime, told me that I was going to play on top board in the difficult 1 1 th-round match, I already felt considerably better physically but not so confidently psychologically. This partly explains my choice of opening. I was hoping for a slow course of events, in order to get into a fighting mood little by little. How­ ever, when we reached the diagram position I started reproaching myself for having chosen a

w

variation with which I was not too familiar for such an important game. Fortunately, I saw Mircea' s silhouette around and remembered Alatortsev. "Well, I shall follow general princi­ ples by developing the bishop and connecting the rooks next," I decided. 13 i.f4 �e8 14 'iVc2 However, things became more concrete after Michal ' s next move. 1 4 h6? ! I already suspected that this was intended not as a prophylactic move, but as the start of an ambitious plan, aiming to win space on the kingside. Not being entirely prepared to calcu­ late concrete variations, I decided that the next stage of mobilization should consist of concen­ trating the rooks on the central lines. IS liad1 gS? ! (D) •••

w

But when this move was played, I started feeling like a complete patzer. "Now I see it

THE ELEMENTARY STA GES OF THE GENERAL PLAN

clearly: thi s is the difference between an aver­ age GM who mechanically applies the general principles and a really strong player, who judges the position more concretely." Indeed, in case of any retreat of the bishop Black has every chance to take over the initiative after, for instance, . . . 'fIc7, . . . liad8, and . . . i.g6. 1 6 �d6 is also not too appealing since it just helps Black complete his development with 16 ... il.xd6 1 7 lixd6 'fIe7. Just when I had resigned myself to this sad conclusion, I suddenly remembered Pavlov's loud voice advising: "Always trust the princi­ pies ! " I looked at the position again and noticed that, apart from the right to move, White is two whole steps ahead in his development, namely the connection between the rooks and the occu­ pation of open files. (The activity of the e8-rook barely matches the white rook's activity along the d-file.) This meant that I should find a forced way to obtain a big advantage. After a few more minutes , I played: 1 6 eS! (D)

45

why Michal refrained from this move and can­ not remember what I had planned against it. Anyway, it would have been quite difficult for White to find the correct route to an advantage. And it certainly was more difficult to deal with this task when analysing the position for this book. Pavlov never told me anything about the difficulties in proving the principles in practice, but I have learned it on my own. The next move is practically forced: 1 7 liJxdS ( 1 7 i.c 1 achieves nothing in view of 1 7 . . . i.. g 6) 1 7 . . . cxd5 (D) but now the choice becomes much wider than White would like:

w

B

Now it was Michal' s turn to sink into a deep think. 1 6 gxf4 From a practical point of view this is not the most challenging reply. Black aims to destroy the white king' s residence but he will achieve this only at the cost of material losses. The less committal 16 ... lDd5 should be preferred, keep­ ing the threat against the bishop in reserve. This is the only way to compensate to a c ertain ex­ tent for the delay in development. I do not know .•.

a) Retreating the bishop allows Black to con­ solidate his position with . . . i..g 6. b) 18 i..x g5 hxg5 1 9 l:'ixd5 100ks like a rea­ sonable way to prevent . . . gxf4xg3 but after 1 9 . . . il.xf3 20 i.xf3 ':xe5 2 1 :tfd l iixdS 22 ':xd5 B lack gets an important tempo for his de­ fence with 22 . . . 'fIc7 hitting g3 (the variation 22 . . . 'fIe7? ! 23 il.g4 i.d6 24 i.. x d7 'fIxd7 25 ':xg5+ cJtf8 26 'fIh7 cJte7 27 'fIh4 cJte6 28 ':f5 ! ! threatening 'fIf6# is suggestive of the dangers facing Black in this kind of position). c) The only move left for White is 1 8 llxd5, when after 1 8 . . . gxf4 White has the nice trick 1 9 e6 ! (at first sight, 1 9 gxf4 100ks promising for White, because of his strong central pawns and the disorganized black minor pieces; however, after the accurate 1 9 . . . 'fIe7 20 %:tfd l liad8 2 1 'fId2 i.. xf3 2 2 i.. xf3 lDxe5 ! Black simplifies to a draw). After the virtually forced variation 19 ... i.. g 6 ( 1 9 . . . lixe6 20 ':xh5 leaves Black in a worse position without any compensation) 20

SECRETS O F ATTA CKING CHESS

46

'iVd2 fxg3 2 1 exd7 i.. x f2+ 22 �h l l1e6 (D) the positi on might look not entirely clear.

w

17 ... i.. g 6 doesn' t have too much independ­ ent value since after 1 8 'iVd2 Black is basically forced to play 1 8 . . . 'iVxf6 (both 1 8 . . .llJxf6 1 9 'iVxf4 and 1 8 . . . llJe5 1 9 'iVxf4 are simply bad for . Black) 1 9 'iVxd7 fxg3 and now White has the accurate 20 �h 1 inviting Black to decide with which piece he will c apture on f2, transposing either to the game or the note to Black' s 1 8th move. 18 'iVfS (D)

B

True, White has a strong pawn on the sev­ enth rank, but Black is still a pawn up and occu­ pies some very advanced posts in the vicinity of the white king. However, after 23 llJe5 ! threat­ ening llJg4 the g3-pawn is in serious danger. B lack has two ways of placing the g4-square under c ontrol: c l ) 23 ... h5 allows White to weaken the en­ emy king' s position with 24 llJxg6 fxg6 25 11d6 l1xd6 26 'iVxd6 which ensures him a dangerous attack after 26 . . .'iVg5 27 l1d l l1d8 2 8 'iVe6+. c2) The alternative is 23 . . . �h5, when White c an continue 24 11c 1 ! (threatening to win by 25 l1c8 11xc8 26 dxc8'iV 'iVxc8 27 l1d8+) 24 . . . �h7 25 'iVf4, when it is not at all easy for Black to keep his position together; for instance , 25 ... l1f6 26 'iVh4 11c6 ! ? (this little trick seems to be the only option; in case of 26 . . . i.. g 6 there c ould fol­ l ow 27 llJg4 l1e6 28 'iVxh6+ �g8 29 llJxf2 eliminating B l ack's main trump) 27 'iVe4+ Jig6 28 11xc6 ! ! Jixe4 29 i..x e4+ �g7 30 l1xh6 ! and with approximate material equality White has a most dangerous attack. Black ' s defensive re­ sources are restricted by the strong white pawn. It is also remarkable that the g3-pawn and the f2-bishop offer a perfect shield to the white king.

17 exf6 The material is equal now, but Black's minor pieces are hanging. There is no way back for him.

1 7 fxg3 •••

Attacking all the enemy' s minor pieces.

18 gxf2+ .••

1 8 ... i.xf2+ enables Black to obtain a more stable dark-square control, but at the same time the pawn is safely blocked. With the energetic continuation 1 9 �h 1 (but not 1 9 11xf2? gxf2+ 20 �f1 { or 20 �xf2 'iVb6+ } 20 . . . l1e1 + ! 2 1 llJxel Jixd 1 ) 1 9 . . . i.. g 6 2 0 'iVxd7 'iVxf6 2 1 11d6 ! White keeps the better chances; for instance, 2 1 . . .'iVe7 (2 1 . . .l1ed8 22 l1xf6 l1xd7 allows 23 llJe5, while after the passive 2 1 .. .'iVd8 White has 22 llJd2 followed by llJde4) 22 'iVxe7 l1xe7 23 l1fd 1 11ae8 24 i..f 1 ! followed by �g2 with better chances. Sooner or later, the g3-pawn will fall.

19 �h 1 i.. g 6 19 ... i.xf3 10ses to 20 nxd7 although 20 'iVxf3 i s also s atisfactory. 20 'iVxd7 'iVxf6 (D) From a formal point of view, Black has man­ aged to maintain material equality. However, his kingside has been seriously weakened while the f2-pawn is rather vulnerable. If White man­ ages to win this p awn, he will get a practically won position.

THE ELEMENTARY STA GES OF THE GENERAL PLAN

47

after 25 . . . 'ii'g 6 26 ttJxe8 �e4 with strong pressure) 24 . . . 'ii'e 7 25 'ii'x a7 �xa7 26 ttJxe8 �e4, when the spectacular 27 !td7 is inoffensive in view of the c alm 27 ... 'ii'e 6. In spite of his mate­ rial advantage, White is certainly not better, be­ cause of his weakened king position and the lack of coordination between his pieces. 22 .i.xdl 23 ttJxdl (D)

w

•••

B

2 1 'ii'x b7 ? ! A familiar syndrome. Just as in the game against Horvath, once the c omplications have come to an end, I played a rather natural move, letting my advantage slip away. B y prematurely capturing the pawn, I hoped to open new hori­ zons for my g2-bishop, but failed to notice that my queen will be cut off from my kingside for a while. Had I understood the right order of pri­ orities, I would have started the fight against the f2-pawn immediately with 2 1 ttJh2, when White can count on an advantage; for instance, 2 1 . . . �h5 (2 1 . . Jie7 22 ttJg4 'ii'f4 is met with 23 'ii'd 2 winning the f2-pawn soon ; here we see how important it is to have the queen relatively close to the kingside) 22 ttJg4 'ii'f4 (22 . . . �xg4 23 'ii'x g4+ 'ii'g 6 10ses to 24 ttJe4) and now 2 3 'ii'd 2 ! looks even stronger than in the previous note since after 23 ... 'ii'x d2 White has the strong zwischenzug 24 ttJf6+ and if 24 . . . f2 cxd5 and Black has every chance of winning.

23 iLh3! This simple move proves that White has not made any progress. 24 ..t g2 This move surrenders a pawn without put­ ting up any resistance . From a practical point of view, 24 l1b l (D) is more stubborn but only de­ tailed analysis will highlight whether objec­ tively speaking it is stronger than the game continuation. •••

B

The radical changes in the position haven' t altered the situation with respect t o how far the . two sides have fulfilled their plans. White still needs to secure his back rank and defend his b2-pawn. The weakness of the c6-pawn is com­ pensated by the slightly less obvious weakness on e2. For instance , after 23 h4 ':xb2 24 lIxc6 Black maintains his material advantage with 24 . . . li xe2. With his next move, B arcza tries to solve both problems at once.

23 ..tf3 As pointed out by Neishtadt, 23 i.f1 is no better. After 23 . . . l1 xb2 24 l1 xc6 l1 d2 Black wins a pawn, since 25 e3 only makes things worse in view of 25 . . . l1 d l (threatening . . . ..t h3 followed by mate) 26 � g2 ..te4+ 27 f3 li d2+ (even better than 27 . . . ..t xf3+ since it enables Black to preserve his active bishop) 28 �g l i.xf3 and B lack is close t o winning. Neishtadt recommended 23 f3 as the best practical chance, giving the line 23 . . . l1 xb2 24 e4 dxe4 (24 . . . ..t g6 25 l1 xc6 l1 d2, not mentioned by Neishtadt, certainly deserves attention) 25 fxe4 iL g6 26 d5 "with chances for a draw",

What should the elements of Black' s further plan be? The first thing that springs to mind is the centralization of the king to d6. This i s not only in accordance with general endgame prin­ ciples but also pursues the concrete aim of de­ fending the c6-pawn and thus making the threat of . . . ..t f5 followed by . . . lixb2 quite real. a) However, after the immediate 24 . . . '\t> f8 White is able to free his position starting with e 3 and b4 (in either move-order) ; for instance, 25 b4 �e7 26 e3 '\t>d6 27 ..td l '\t> c7 28 ..tb3, when after 28 ... 11d2 29 i. a4 in view of the threat of li c 1 Black has nothing better than re­ turning the rook to a2. b) We can conclude that there is an even more imperative element in Black's plan, that

THE ELEMENTARY STAGES OF THE GENERAL PLAN

of preventing White from freeing his position. This can be achieved with the strong move 24 . . . g5. The difference can be seen in the lines 25 b4 ..tf5 26 11 e l g4 ! 27 i..g 2 'u'xe2 and 25 e3 i.. f5 26 'u' e l g4 27 i.. d l ,U,xb2, as given by Neishtadt. By means of . . . g4 Black forces the enemy bishop to occupy an unfavourable posi­ tion, thus reiterating the weakness of the back rank and making the further capture on c6 im­ possible . Facing the threat of the king' s transfer to d6, White should react quickly. The only way to activate the bishop is to undermine the enemy centre. Neishtadt believed that White could ob­ tain reasonable drawing chances with 25 e4 (D) preventing . . . i.. f5 and planning b4-b5 .

B

(see above). Similarly, the move to improve upon is 26 . . . � e7 . B lack can combine in a more effective way the two facets of his plan by play­ ing 26 ... g4 ! ' The white bishop is forced to retreat to an unfavourable position with 27 i..h l (any other retreat loses the e4-pawn while 27 exd5 fail s to create any re al threats after 27 . . . gxf3 ! 28 dxc6 �e7 29 b5 i..f5 30 l:. e l � d8 3 1 b6 l:.b2, when the white pawns are not dangerous and will soon become weak), when play may con­ tinue 27 . . . l:.d2 28 exd5 exd5 29 i..g 2 (this move, solving the problem of the back rank, might be the best practical chance, similarly to that cho­ sen by B arcza in the game; 29 b5 is simply bad because of 29 . . . cxb5 30 f4 b4 3 1 i.. x d5 l:.xd4 with a safe extra pawn, while 29 f3 fails to free White' s position after 29 . . . Ihd4 30 fxg4 ..t xg4 { threatening to consolidate B lack' s advantage with either . . . l:.d l + or . . . ..t e2 } 3 1 b5 cxb5 32 l:.xb5 kI d l + 33 � g2 ..t f5, winning the bishop) 29 . . . i.. x g2 30 � xg2 ,U,xd4 offers Black reason­ able chances of converting his extra pawn into victory because of his space advantage on the kingside. Thus, we can conclude that 24 lib 1 gives Black more difficult practical problems but is of more or less equal value to the text-move from an objective point of view. 24 i..x g2 25�xg2 kIxb2 26 e3 'u'b6 Since there are many pawns still on the board, Black has reasonable chances to convert his material advantage into victory. Keres man­ aged to win after 22 further moves, but the rest of the game is not so relevant to our topic. Let us try to give a description of the events over the l ast 9 moves. In the initial position, both sides had an equal number of tasks on their agenda. Black was in a slightly advantageous position because it was his tum to move but this should have materialized in a rather insignifi­ cant extra pawn had White been more careful and less greedy on his 2 1 st move. After his mis­ take, he fell a further tempo behind in the race to fulfil the elements of the two sides' plans. From that point on, he had no way of avoiding the loss of a pawn. To put it in other words, being one step ahead on a temporal scale was equivalent to a pawn on the material scale. Clearly, it is dangerous to toy ..•

His analysis continued 25 . . . �f8 26 b4 � e7 (? as we shall see) 27 exd5 exd5 28 b5 cxb5 29 i.. x d5 11d2 30 f4, when White's position is far from appealing but probably better than it looks. The whole sequence looks rather logical and choosing the right place for an improvement is not easy. First of all, I was curious to check whether 27 . . . cxd5 would manage to keep the bishop out of play, but reached a negative an­ swer after 28 b5 � d8 29 b6 � c8 30 ..th5 f6 3 1 i.. e 8 ..t f5 3 2 i.. d7+ ! (this tactical trick enables White to attack the e6-pawn while keeping the pawn on b6 in order to prevent the defensive move . . . l:. a6) 32 . . . �b7 33 l:.e l with a likely draw. Finally I understood that Black's mistake in Neishtadt' s line consists of the one-sided atti­ tude towards the two stages of the main plan, as was the case in my line starting with 24 . . . �f8

51

SECRETS OF A TTA CKING CHESS

52

with such notions, but one should take into ac­ count that in positions where the possession of the initiative is important, the equation I step = 1 pawn c ould offer a general orientation in the j ungle of v ariations . Of course, it would be an exaggeration to consider that a tempo is always equivalent to a p awn. This i s just a draft rule valid in certain types of open positions. If the centre is closed, temporary factors tend to lose part of their significance. I will refrain from giving any examples of that kind since this book focuses on dynamics and attack. Instead, I shall illustrate the other extreme case. Here is an example where for the sake of one tempo White gives up his most precious piece, the queen.

w

Lommer Die Schwalbe, 1 9 65 White to play and win

1 �e3+! The first move is natural: the king approaches the battlefield with tempo. 1 'iVxh5 ? 'iVxh5 2 �xh5 c 2 leads to a draw.

1 �b3! ...

Now the natural c ontinuation i s 2 'iVxh5 'iVxh5 3 :t xh5 , when it looks as if White c an create a mating-net around the enemy king, based on the fact that the pawns take away some important squares from his maj esty. Indeed after 3 . . . c2 4 �b5+ White wins in all lines: 4 ... � a2 (4 . . . � a4 allows 5 lLl c3#, while after 4 . . . �c4 5 lLlxa3+ White eliminates both enemy pawns) 5 lLlc3+ � a 1 6 � d2 and it's all over.

However, after the correct 3 . . . a2 4 i:.b5+ � c2 B lack avoids the mate while keeping the enemy king out of play at the same time. It would be so much better if B l ack weren' t given the possibil­ ity of advancing any of his pawns , thus keeping the mating-net intact. How can White win that decisive tempo? The answer is quite elegant:

2 lb:h5 ! ! 'iVxf7 There i s nothing better since the queen i s pinned. 3 li bS+ Now White mates in all lines.

3 �a2 Or: 3 . . . � a4 4 lLl xc3#; 3 . . . � c4 4 lLlxa3#; 3 . . . � c2 4 lLlxa3+ � c 1 5 lib1#. 4lLlxc3+ �al 5 libl# There are, of course, some simple explana­ tions for the necessity of investing such a big quantity of material for the sake of just one tempo. First of all, only the structure with the pawns on c3 and a3 m akes the mating ideas possible and therefore should be preserved at any cost. Secondly, the pawns are quite close to their promoting square s and any step forward (if allowed) would increase their relative value. ...

So far, we have examined only situations where elementary stages of the main plans were involved. We have seen that as a rule even an apparently insignificant temporal advantage of one or two tempi can be converted into some other form of superiority of a more stable na­ ture. However, since the final purpose of a game is to mate the enemy king, we can expect that at a certain moment the current stage of the plan has to take more decisive forms such as the cre­ ation of a mating-net or, as an intermediate step, gaining significant amounts of material. In hard-fought games, which are the main sub­ j ect of this book, the opponent will also gener­ ally be quite advanced with the fulfilment of his plan. If the mutual plans involve a similar de­ gree of danger, then the final outcome of the battle is decided by temporal factors. B y far more interesting are situations where the characters of the two sides' threats are dif­ ferent. What is more important? To be more ad­ vanced on the temporal scale or to create the stronger threats? These questions are impossible

THE ELEMENTARY STA GES OF THE GENERAL PLAN

to answer with a general formula; fortunately, chess i s more c omplicated than that. During a g ame, it is important to evaluate in advance the chances of success of one ' s own plan in con­ frontation with the enemy' s and, if necessary, switch to defence while still possible. I shall try to illustrate the two extreme situations of the balance between speed and intensity.

B

53

material advantage or, on the contrary, overesti­ mated the dangers. In any case, after 26 l:. xh7 b5 ! (Black has to activate his knight via b 6 ; the check on f3 would now prove a simple waste of time; for instance, 26 . . . 'iVf3+ 27 lD d3 and now 27 . . . 'iVe2 28 l:. l h4 ! , followed by l:. f4 with a strong attack, or 27 b 5 28 l1 1 h4) 27 l:. l h4 'iVd l B lack ' s counterplay suffices for a draw, but probably not more than that. For instance, 2 8 lD g6+ 'it' d8 29 l:.xn as 30 l:. h8+ 'it' c7 3 1 l:. hh7 b4+ 32 axb4 axb4+ 3 3 'it'xb4 'iVxd4+ with perpetual check. 26 'iVf3 27 lDd3 b5! Again, the only w a y t o bring new forces into the attack. 28 ltel ! (D) It is important to activate this rook before the kni ght is installed on c4. In case of the immedi­ ate 28 lt xh7? lDb6 followed by . . . lDc4+ B lack creates decisive threats before White's attack has even started. . . .

•••

Ivkov- Korchnoi USSR-Yugoslavia, Leningrad 1 957

B

The position is quite sharp, although we are still 1 5 moves away from the culmination point. B oth sides have doubled rooks on half-open files and are about to eliminate an i mportant pawn from the enemy' s defensive system. Black is to move, but in compensation, the capture on c3 will not actually lead to a material advantage for him.

23 J:txc3 24 'iVxc3 l1xc3 25 'it'xc3 'iVg4 Korchnoi rightly refrains from the passive 25 . . . lDf8, when, after a consolidating move such as 26 l1 1 h4 Black has no chances to create any dangerous threats with the queen alone and i s forced to remain passive. After the text-move, both sides will try to create threats against the enemy king in the most effective possible way (meaning with the maximum speed and degree of danger). 26 'it'd2 Ivkov spends a valuable tempo on this pro­ phylactic move, which aims to take c ontrol of the e2- and d l -squares. It is difficult to s ay whether he was under the mirage of his minimal .•

28 'it'd8! Black i s on his guard. Mter the hurried move 28 . . . lDb6? White answers 29 l:. f6 followed b y l:. c7+ with a strong attack, since 29 . . . ltJ c4+? i s simply met with 30 Itxc4 winning material. While carrying out one' s own plan, it is always useful to slow down the enemy' s attack. One tempo spent on prophylactic moves is an excel­ lent investment if it causes the opponent the loss of two or more tempi. 29 l:.c6 ! B oth players conduct this phase of the game very well. Although the rook move does not ...

54

SECRETS OF ArrA CKING CHESS

threaten anything decisive yet (maybe simply :a6 ! ?) it actually puts his opponent in a kind of zugzwang. Indeed, Black' s only way to im­ prove his position consists of placing the knight on b6, thus leaving the f6-square undefended and allowing lH6. Moreover, it looks like most of Black' s other moves would simply worsen his position. For instance, 29 . . . a5 will leave the b6-square without support and after 30 kLa6 only White can create threats against the enemy king. The immediate 29 �xh7 ? clearly loses the race after 29 . . .l2Jb6 3 0 !th8+ �d7 3 1 :b8 12Jc4+ 3 2 �c3 a6 ! ?, when B lack defends his b5-pawn and maintains unpleasant threats such as . . . liJxa3 or . . . liJb2. White ' s uncoordinated rooks fail to generate any counterplay.

the b4-square, allowing 37 . . . 'iVb2+ 38 �d3 'iVd2#) 37 ... 'ifa3+ 3 8 g l l:. d l +, while the rather passive 24 "iYe l is also unsatisfactory because of 24 . . . "iYxb2 (better than 24 . . . liJe3+ ? ! 25 � g l , when 25 . . . "iYd4 is parried simply by 26 liJb5), when 25 tDxd5 is strongly met with 25 . . �xe2 ! . .

B

ADVANTA GE IN DEVELOPMENT

2 4 lZ'lxf4 ? !

67

25'iYf2

•••

Even such a romantic as Spielmann seems to have shown respect to material from time to time. This natural move, winning a pawn and thus creating some premises for long-term com­ pensation, in fact slows down the rhythm of the attack. 24 . . . lZ'le3+ ! is more dangerous, severing the feeble communic ations between White' s wings. For instance: a) 25 c;t> f2 keeps the bishop well defended and c onnects the rooks, but invites an irritating pin along the second rank. 25 . . . 'iYxb2 26 ::te l (it i s essential for White to defend the knight; 26 lZ'l d l i s impossible because of the simple 26 . . . lZ'lxd l +) 26 . . . l::t d 3 ! and White will not be able to defend his pieces for long. b ) 2 5 c;t>g l offers the bishop some mobility, but the king still obstructs the coordination of White' s remaining pieces. Play may continue 25 . . "iVxb2 26 i£.b5 (the most consistent move; 26 lZ'l d l lZ'l xd l 27 .txd l ::txel + 2 8 "iVxe l �d4+ 29 c;t>n �xd l and 26 lZ'lb5 lZ'l c2 do not need any special accuracy from Black) 26 ... ::td2 ! (more ambitious than 26 "iVxc3 27 � xe8 ::txe8 , when White can still survive with 28 h3 followed by c;t>h2) 27 ::txe3 (forced, owing to the threat of mate in two on g2) 27 . . . l:.xe3 2 8 �xe3 �e l + ! 2 9 � n �xg2+ 3 0 c;t>xg2 'iYxe3 (D). .

. . .

The main merit of Spielmann' s previous move is of a practical nature. Not being under immediate pressure any more, White faces a difficult choice between several more or less equivalent continuations. a) True, a ' neutral' move such as 25 "iVf3 ? loses by force to 25 . . . lZ'lxe2 26 lZ'l xe2 l:.e3 ! (an important zwischenzug, forcing the white queen to occupy a less favourable position; the imme­ diate 26 l:. d6? is answered with 27 �c3 ! ::tf6+ 28 � g l , when I can only see how B lack could force a draw with 28 . . . 'iYxc3 29 bxc3 li fe6 30 �2 ktf6+) 27 'iYg4 (in c ase of 27 'iYf2 the rook­ lift 27 . . . :t d6 is decisive this time) 27 . . . l::td e8 28 h4 (what else?) 28 ... kt 8e6 with a decisive at­ tack. However, besides the game continuation Mieses could have considered two other rea­ sonable moves. b) The most natural of them is 25 .1l. f3, aim­ ing to release the pressure by exchanging one pair of rooks. Indeed, after 25 .. J b.e l + 26 'iYxel 'iYxb2 White gets some time to complete his de­ velopment by clearing the h2-square for ' his king with 27 h4 (other moves are worse: 27 "iVh4 ::t d2 with dangerous back-rank threats ; 27 jte4 g6 secures Black ' s own back rank and threatens either . . ::t d2 or . . . � e8 ; or 27 a4 c4 28 lZ'l d l 'iYf6, when the combined threats on both wings are difficult to deal with) 27 . . . � f8 ! (an elegant multi-purpose move: the king not only approaches the centre in view of the possible exchange of queens, but also places the e7square under control , restricting White ' s coun­ terplay in the middlegame; the straightforward 27 . . . lZ'l d3 28 lZ'l d l 'iYxa2 i s met with 29 "iVe7 ! but after the text-move the threat of . . . lZ'l d3 be­ comes rather annoying) 28 lZ'l d l �xa2 29 � g l (unlike his black counterpart, the white king is forced to move to a passive square so as to al­ low the activation of the rook; from a formal point of view, Black has a material advantage already, but with such weak pawns he should also be concerned about piece activity to pre­ vent the position from becoming double-edged) 29 . . . 'iYa4 (29 . . . lZ'ld3 30 'iYc3 'iYa4 is a worthy al­ ternative) 30 �h2 (this is probably better than 30 lZ'lf2, when after 30 . . . 'iYd4 3 1 �h2 � e8 White . . .

.

From a materialistic p oint of view, White' s position might look acceptable, but h i s coordi­ nation continues to be rather poor. After, for in­ stance, 3 1 lZ'l e2 a5 he is completely paralysed. The a5-pawn will most probably cost him a mi­ nor piece in the future.

68

SECRETS OF A TTACKING CHESS

must remain passive) 30 . . . liJd3 3 1 "iYe4 (as usual in such situations , the centralization of the queen is the best warranty against immediate trouble; B lack would consolidate his advantage after 3 1 "iYg3 "iYd4 followed by . . . c4) 3 1 . . . l! d4 32 "iY a8+ "iYe 8 (the position becomes double­ edged after 32 . . . r;J;; e7 33 g 3 , when the king feels a bit insecure on e7) 33 "iYxe8+ r;J;; xe8 3 4 r;J;; g3 f5 (D).

under observation ; 30 . . . "iYc2 is less accurate in view of 3 1 \t> g l ) 3 1 l!hfl (now 3 1 \t> g l is met with 3 1 . . J �e8 creating a deadly pin) 3 1 . . . l! d4 with an ever-growing initiative. We return to the position after 25 "iYf2 (D):

B

w

25 J!d4 It i s Black's tum to have a difficult choice be­ tween several tempting continuations. For in­ stance, 25 ... liJxe2 is probably only enough for a draw after 26 l'he2 "iYc 1 + 27 l!e l l! xe1 + 28 "iYxel "iYf4+ 29 r;J;; g l "iYd4+. A draw is also likely after 25 . . . l! d6: 26 l! d l "iYxb2 (26 . . . 'iVxe2+? does not work because of 27 "iYxe2 liJxe2 28 l!xd6) 27 l! xd6 cxd6 28 "iYxf4 "iYxc3 29 "iYxd6 (eliminating an important pawn and creating some potential back-rank threats; 29 r;J;;f2? is bad because of 29 . l!e6, keeping a strong attack with approximate mate­ rial equality) 29 . . . "iVb2 30 "iYd l c4 3 1 \t> f2 and now Black should force a perpetual before it' s too late: 3 1 . . . "iYb6+ 3 2 \t> g3 'iVg6+ 3 3 \t> f2 "iVb6+. After Spielmann' s move, White has to chal­ lenge the stability of the black pieces as soon as possible. For the time being, Mieses followed the right track: 26 g3 (D) As Spielmann pointed out, 26 l!d l ? loses to the elegant 26 . . . 'iVxe2+ 27 "iYxe2 liJ xe2 2 8 l! e l l'lf4#. 26 liJh3? ! B oth players must have been tired by this point. This natural move is in fact a further con­ cession, not only giving up all the remaining ..

How should we assess this position? Black's chances are probably better while there is basi­ cally no risk for him, but White can put up re­ sistance. I believe that from a practical point of view 25 J. f3 i s the best continuation. c) 25 liJb l is slightly worse though not en­ tirely bad, aiming to force the black queen to step back with 25 . . . "iYd4 and then return to c3 with 26 liJ c 3 . Black is then by no means forced to accept the repetition of moves. He can rein­ force his attack with 26 .. . l!e6 ! 27 J. f3 l! g6 ! (it seems that B lack cannot get anything more than a perpetual by playing 27 . . "iYc4+ 28 r;J;; g l "iYd4+ 2 9 r;J;;f1 "iYc4+ 30 \t> g l since 3 0 . . . liJ d3 3 1 l! d l "iYd4+ 3 2 \t> f1 l! de8 3 3 liJ e2 leads to a rather unclear position) 28 "iYh4 (pinning the knight and leaving the f2-square available for the king; if 2 8 "iYf2 then 2 8 . "iYc4+ forces 2 9 M e2 l:! e6 with strong pressure, since 29 r;J;; g l ? allows 2 9. . .liJh3#) 2 8 . . ."iYd3+ 29 liJ e2 l!f6 ! (still preparing the decisive blow, since 29 . . . liJxg2 30 J. xg2 l! f6+ can be answered by 3 1 "iYxf6 gxf6 3 2 l! g l r;J;; f8 3 3 \t> f2, when White c an sur­ vive) 30 \t> f2 (after 3 0 "iYf2 l! e8 White c an hardly move) 3 0 . . . "iYd2 ! (keeping the e l -rook .

..

. .

•••

AD \

L 1 GE IN LyE iELOPMELE

69

White has managed at least to group all his pieces around his king. flb4) w,; (with In signifinant the of BlackE b:lck rank is. B e sides, the threat against the f4-knight is not real yet. Therefore, I believe that the best v;lflLage i s for an decide Oil althoua h n o t ea\V move during a practical game.

advantage but also offering White the possibility of choOSLll.s botween safe wav:� reach­ mg appTlixirnAe equalitLl i� emoving knight from f4 was probably the last thing Spielmann wanted to do. Alternatives : rnove as nf such a) :iVx l ear to b ' " might hLL\ li6 more so. him, and practical problems than 27 gxf4, which should lead to approximate equality after 27 . . . 'iVxc3 28 31 ligl "iVdb i.. f3 lixx hO "iVxe l �xel li g3 , look black like a reasonable equivalent for the bishop) Black faces a difficult choice. ?! fl ot work Llncause of al) ': li for 29 c with an h8 xd White. a2) The spectacul ar 27 .. :iVxe2+? ! also ap­ pears insufficient : 28 li xe2 (more ambitious lDxe2 hxg3 lDxg 3 than 28 lixe l + li d3 :a3, Black is not worse) 28 ... li xdl + 29 lie l li dxe l + 30 'iVxe 1 1he l + 3 1 �xe l and Black has to fight for a dr:111 , .. "iVb7 a3) solm � eems to maintaining the tensi ]H, instal £1:, "iVf3 lbd5 (28 . "iVb4 ! ? is also interesting, but after 29 lDc3 lDd3 30 ilxd3 lixel + 3 1 �g2 the the back and �pawn weakne1;< moment t o retreat c>recisely torces he seemed to be so close to his goal : 3 1 . . . lie8 3 2 li f1 lif8 3 3 li f2 with unclear play) 29 � f2 30 i..d 3 than 29, 'iVb4 cerwhen 30 :hf the but tainly has compensatIOn for the ..

..

White hal, 1 10blems himself Loro the Slime ',xIong preSltlITt: is subjelktl Lo. Here possible continuations, illustrating B lack's at­ tacking ideas: b l ) 27 �f3 ? dlows a win bo T It::lnS � "iV xf2 27 ... xfk+ is equallb and to b2) 27 li mate after 27 ... xel + 2 8 "iVxel "iVg2#. b3) 27 lDbl makes more sense, but after 29 " iVb4 2H f1 Black c astles . Other moves are far less appealing: 2 1 . . .gxf5 (2 1 . . .'iVd7 allows the small c ombination 22 lhh7 ! ? l:txh7 23 'fVxg6+ g8 30 'flid6 with a decisive in­ filtration by the white queen. b) 26 . . . lixg3 27 'flixg3 'flif8 simplifies the position but leaves the c7-rook hanging. White wins with 28 tiJg6+ hxg6 29 'ilxc7 tiJxh5 30 'ilUxc6. c) Sadly, the thematic inclusion of the bishop in the defence with 26 . . . iLe8 also fails tactically to 27 Mxg8+ tiJxg8 2 8 'ilg3 'ilUf8 (28 . . . tiJf6 is strongly met with 29 tiJf7+ ! interfering on the seventh rank) 29 iLxe8 'ilxe8 30 tiJd7 ! ! , when Black cannot prevent both threats of mate, on e5 and g7. Clearly, 30 tiJg6+ is also sufficient. 27 l1xg8+ tiJxg8 (D) •••

White' s pressure has reached a culmination point. The immediate threat is 30 i..f7 and if 30 . . . lixf7 then 3 1 'tJg6+ winning the queen, but White has several other threats, depending on Black' s answer. 29 'tJf6 29 . . . lig7 3 0 :txg7 rJlxg7 allows 3 1 'ilg5+ followed by tiJf7+, while 29 . . . iLe8 30 iLxe8 'ilxe 8 is met with 3 1 tiJg6+ rJlg7 3 2 'tJe7+. 30 �f7! 1-0 B areev resigned in view of the loss of his queen. •••

The art of switching from one wing to the other must be one of Beliavsky' s specialities. Two years l ater he played the following game, where, in spite of the fact that the opening was the same, the scenario could be described as a perfect mirror image of the game with B areev.

w

Beliavsky- Yusupov Linares 1 989

28 'ilUg3

1 d4 e6 2 c4 fS 3 g3 'tJf6 4 iLg2 dS S 'tJf3 c6 6 0-0 �d6 7 iLf4 �xf4 8 gxf4 0-0 9 e3 'tJbd7 10 'ile2 'it>h8 1 1 tiJc3 'ile7 (D) In this kind of position, Black u sually pre­ pares the advance of the g-pawn to undermine White' s control of the c entre and create some

SECRETS O F A TTA CKING CHESS

118

w

B

attacking chances. Although White' s traditional plan is based on b4-b5, B eliavsky decides to anticipate Black' s activity for the time being. 1 2 'itth l Preparing t o double rooks o n the g-file. 1 2 l:g8 Yusupov appears to have underestimated White' s intentions . Pinning his hopes on his kingside attack will eventually lead to disaster. A more balanced strategy consists of develop­ ing the bishop to a6 in order to exert some pressure against the white queenside and thus restrict White ' s activity on that side of the board. For instance, 1 2 . . . b6 ! ? 1 3 ltJeS ( 1 3 cxd5 exd5 14 l:.fc 1 ltJe4 als o looks acceptable for Black, while against 1 3 Iiab l preparing b4, 1 3 . . . .ta6 looks strong) 13 . . .ltJxe5 14 fxeS ltJe4 I S tbxe4 fxe4 1 6 f3 (White can prevent . . . �a6 with 1 6 cxdS but after 1 6 . . . exd5 1 7 f3 exf3 1 8 l:txf3 Black c an change plans and develop the bishop to a more natural square with I S . . . .te6, when after 1 9 liafl c 5 the position looks about equal) 1 6 . . . exf3 1 7 ':xf3 .ta6 and Black has easy play. 13 cxd5 exd5 The typical capture in such situations. B lack is still underdeveloped and has to keep the posi­ tion as closed as possible. Mter 1 3 . . .cxdS 1 4 �ac 1 th e open c-file will soon b e occupied by White. 14 .t h3! (D) This move not only clears the g-file for the rooks but also gives Black some problems with his coordination by attacking the f5-pawn. 14 tbg4

To place the knight on the more natural e4square, B lack would have to defend his pawn with 14 ... g6 first. After 15 Iigl ltJe4 the thematic . . . gS is safely prevented for a while. Therefore, White c an prepare active operations on the queenside (a3 , b4) and in the centre (tbe5, f3 ) with 1 6 l:ac l (Beliavsky recommends 1 6 ,Ug2 "with a clear advantage" but this plan seems to be less accurate than in the game ; for instance, 16 . . . tbdf6 17 l:.agl iLe6 18 'iVc2 'uacS 19 tbe5 c5 with counterplay). This half-waiting rook move invites Black to continue development w ith 1 6 . . . tbdf6, whereupon his own knight jumps into the centre with 1 7 ltJeS ; for instance, 17 ... tbd7 (or 1 7 ... .te6 1 8 f3 driving the active knight away) 1 8 ltJd3 ! (White should avoid the exchange on eS for the time being because the white knight is obviously better placed on d3 than the black one on d7), when I S . . . 'lWh4 does not bother White because of the calm 1 9 .tn followed by f3 . 1 5 %:tgl ltJdf6 1 6 l:.g2 With such a concentration of black pieces on the kingside, White decides to group more de­ fenders around his king. In case of a further . . . g5, the doubled white rooks will certainly be well placed. 16 i.e6 17 %:tagl l:.af8 (D) So far, we have followed a movement of the white pieces from the left to the right. Clearly, White feels safe on the kingside, but how is he going to make further progress? 1 8 a3! ! There i s a strong similarity between thi s move and 1 7 h 3 from the previous game. After

•••

•••

..•

PLA Y ON Two WINGS

w

a sequence of moves preparing the occupation of a certain file, all of a sudden White modestly advances the rook' s pawn on the opposite wing. True, the wings are reversed here, but this only adds charm to this remarkable similarity. The main difference is that in the present game White' s previous moves had a rather defensive or at least prophylactic character, but from a practical point of view the consequence i s simi­ lar: B lack' s pieces have been attracted to that side of the board. What is the concrete aim of 1 8 a3 in this case? Once he has reached a culmi­ nation point on the kingside, White returns to his main active plan, the queenside minority attack, based on b4-bS . With all his pieces on the oppo­ site wing, B lack cannot easily prevent this plan. We now return to 1 8 a3 ! ! (D):

119

in view of the expected b4 advance. Here are some other possible continuations: a) 18 ... a5 only prevents b4 temporarily, but creates some queenside weaknesses. White can continue 1 9 ltJa4 followed by ltJcS-d3 intend­ ing either b4 or ltJeS. b) B ecause of the unnatural position of the g4-knight, 1 8 . . . ltJe4 allows White favourable simplifications with 1 9 ltJxe4 dxe4 20 �xg4 fxg4 2 1 ltJeS, when the knight is obviously stronger than the bishop, while the g4-pawn is in danger. c) The attempt to increase the pressure with the thematic queen manoeuvre 1 8 . . . Vi'e8 1 9 b4 Vihs gets Black nowhere because of the calm 20 �g3 . d) 1 8 ld7 i s a faster way to transfer the bishop to hS than in the game. Then White should remove the queen from the dangerous diagonal with 1 9 Vi'c2 (after 1 9 b4 .thS it i s not easy for White to free himself from the pin, since if the queen leaves the knight undefended, the tactical blow . . . ltJxe3 becomes possible; 1 9 iLxg4 gets White nowhere because of 1 9 . . . fxg4 ! 20 ltJeS �e6 and returning the bishop to fS in order to take control of the relatively weak light squares). After 1 9 �hS 20 .txg4 Black's ex­ tra tempo compared to the game hardly makes any significant difference. 19 b4 (D) ...

. . .

B

B

19 �e8 Black can open the a-file with 1 9 . . . a5, dis­ tracting the queen from the kingside. However, after 20 Vib2 axb4 2 1 axb4 the further advance ...

1 8 iLd7 Caught by surprise, Black makes a neutral move, placing the bS-square under observation •••

120

SECRETS OF A TTA CKING CHESS

of the b-pawn would damage the black structure more seriously than in the case of Black re­ fraining from . . . as . On the other hand, the activ­ ity along the a-file offers some compensation. For instance , 2 1 . ..':a8 (once again, 2 1 . . .ttJe4 leaves White with the better pawn-structure af­ ter 2 2 ttJxe4 fxe4 23 iLxg4 iLxg4 24 I!xg4 exf3 25 b5) 22 b5 (in view of the conclusion of the main line, this thematic move might be prema­ ture; White should probably spend some more time preparing it) 22 . . . cxb5 (in case of 22 . . . il.e8 White takes full advantage of B lack' s lack of coordination by 23 ttJe5 ! ) 23 ttJxb5 il.xb5 24 'iUxb5 'iUe4 (or 24 . . . l::t a 2 25 �xg4 ttJxg4 26 I!f1 ttJf6 { defending the dS-pawn } 27 'iib I ! and suddenly B lack loses a pawn) 25 J.xg4 ttJxg4, when 26 'iUe2 threatens to trap the queen with h3, ctJe5 and f3, which forces B lack to seek the initiative with 26 . . . l:gc8. Mter 27 h3 (changing the move-order with 27 ttJd2 'fIe7 28 h3 is of no help because of 28 . . . 'fIh4) 27 .. J�c2 28 ttJd2 l:aa2 ! 29 'fIxg4 ! fxg4 30 ttJxe4 dxe4 3 1 hxg4 ':xf2 32 ':xf2 lW'2 3 3 11b l White ' s advantage will probably prove insufficient to win. How­ ever, weighing up the static and dynamic fac­ tors during a game is not an easy task and we can understand Yusupov for following the gen­ eral principle of not moving his pawns on the wing where he was going to be attacked. We now return to 1 9 . . . �e8 (D) :

could leave the kingside vulnerable. B esides, B lack can do very little to prevent the standard minority attack and it doesn' t matter if the plan is successfully carried out immediately or only a while l ater. However, Beliavsky wrongly at­ taches an exclamation mark to this move. The idea of exchanging on g4 is correct, becau se White ' s strategy i s c entred on the control of the dark squares all over the board. However, a more accurate move-order i s 20 'iib 2 J.h5 and only now 2 1 �xg4 ! ctJxg4 (after 2 1 . . .fxg4 22 ctJe5 the bishop is obviously misplaced on h5) 22 l1g3 transposing to the game. 20 ctJxg4 With the bishop still on e8, B lack should play 20 . . . fxg4 ! ? 2 1 ttJe5 .i.d7, preparing to take control of the light squares; for instance , 22 'fIc 2 g6 23 b5 �f5 with counterplay. 2 1 :g3! White rightly refrains from the tempting advance of the h-pawn. 2 1 h3 only restricts White' s manoeuvring possibilities ; for instance, 2 1 . . .ctJf6 22 ttJe5 �h5 23 'f!Ib2 ttJd7 24 ':g3 ttJxe5 25 dxe5 b6, when the bishop i s stable on h5 because the h3-square is unavailable. 2 1 .i.hS (D) •••

..•

w

w

22 'fIb2

20 .i.xg4 ? ! B eliavsky continues t o switch his play be­ tween the two wings. This i s the correct ap­ proach, since hurry ing with the queenside attack

A natural and consistent move, probably played without hesitation. The desire to remove the queen from the dangerous diagonal is un­ derstandable. However, it is even better to press a bit more on the kingside with 22 11h3 . Then: a) In case of the passive 22 . . . 'fIe8 White c an continue 23 'fIc2 (keeping an eye on the

PLA Y ON Two WINGS

f5-pawn and thu s discouraging the activation of the black rook to h6) 23 .. .lU6 24 llg2 (over­ protecting the f2-pawn) followed by �a4-c5 and a4, in order to switch back to his queenside plan. b ) After 22 . . . �f6 White has apparently achieved very little, since ... j.xf3 seems un­ avoidable. However, he can take advantage of the opposition of his queen and the black bishop himself with 23 �e5 ! ' For instance, 23 . . . 'iVxe5 (23 . . . j.xe2? ? allows 24 �g6#, while after 23 . . . ii'e8 24 f3 the bishop remains out of play) 24 iixh5 'iVe6 (24 .. :iVe8 defends the g6-square but loses a pawn to 25 'iVxf5) 25 'iVg6 followed by llg5 with a very strong initiative for White. After the text-move Black gets some chances for counterplay, which he fails to seize. 22 �6 Instead of signalling the retreat, Black can play on the kingside with 22 . . . :f6 ! ? 23 b5 IDl6, when White's queenside actions are more re­ stricted - indirectly - by the strong kingside pressure. This counterplay would not have been possible if White had adopted the more accurate move-order suggested in the previous note. 23 �e5 (D) •••

B

121

26 ... f4, when the white queen and knight are suddenly misplaced. 24 f3 ! Yusupov might have overlooked this strong move. 24 ...�xe5 A sad necessity. 24 . . . �xe3 ? 25 'ife2 lbc4? 26 �g6+ costs Black his queen. 25 dxe5 The correct recapture. 25 fxe5? ! allows Black strong counterplay with 25 . . . f4. 25 h6 The significant difference compared to the line suggested in the notes to Black's 23rd move i s that Black has been forced to exchange on e5 and after 25 ... g5 26 �e2 ! the king feels rather unsafe in the comer. Black suddenly has to release the tension he has created, but this is far from easy. For instance, 26 . . . g4 (26 . . . gxf4? loses to 27 e6+, while 26 ... h6 fails to consolidate the g5-square in view of the immediately deci­ sive 27 :h3) 27 �d4 (a perfect square for the knight) 27 . . . j.g6 (27 . . . gxf3? allows 28 �xf5 ! ) 2 8 'iVc2 (putting more pressure o n the f5-square and threatening 29 fxg4 fxg4 3 0 f5) 28 . . . h5 29 'iVc5 with perfect control on the dark squ ares. In this variation, the long-ranged white queen clearly made a superb j ob of playing on two wmgs . 26 �e2 (D) •••

B

23 �g4? Black insists on exchanging White's active knight. The p seudo-active 23 . . . �e4? also gets Black nowhere because of 24 :h3 g6 25 f3. With hindsight, we can state that 23 . . . �d7 ! 24 f3 g5 ! ? is a better way to seek counterplay. Accepting the s acrifice with 25 fud7 'iVxd7 26 fxg5 allows Black to activate all his pieces with •••

26 b6 Since his kingside plan seems to have become bogged down, Black at least tries to prevent White's queens ide attack. After 26 .. .'it>h7 27 •••

1 22

SECREI'S OF A TTA CKING CHESS

liJd4, pressing on with 27 . . . g5 only creates new weaknesses and White obtains a clear advantage with 28 l:Ih3 ! . Against the passive 28 ... 'ife8 White plays 29 'ifc2 attacking f5, while the attack initiated with 28 . . . gxf4 29 lIxg8 .l:'txg8 3 0 .l:'txh5 'ifg 7 fails t o 3 1 'iff2 fxe3 32 'ifg 3 , keeping the material advantage. The more restrained 27 . . . g6 is answered with 28 'ifc3 , planning 'ifc5 and, in case the black queen moves away, b5. 27 liJd4 c5 Black cannot refrain from this move, since 27 . . . 1le8 10ses a pawn to 28 'ifc2 'ifd7 29 lIe I .

W

28 liJb5 (D)

B

Black has prevented the minority attack in its classical form, but has allowed the knight to d6. 28 \ith7 Throughout the whole game, Black has not managed to find the right moment to advance his g-pawn. Maybe this was the last possible occasion to carry out this p awn-break under reasonable c ircumstances, by 28 . . . g5 ! ? Then: a) Yusupov might have refrained from this because of 29 .l:'th3 ? ! but after 29 . . . gxf4 ! (D) Black obtains strong counterplay. The main dif­ ference c ompared to the line given in the note to Black' s 26th move is that the knight is one move further from the kingside. Here are some illustrative lines : al ) Rej ecting the sacrifice with 30 exf4 does not prevent the attack: 30 .. Jhg l + 3 1 �xg l .l:'tg8+ 32 \itfl and now Black can take advantage of the unstable position of the white knight by 32 . . . 'ife8 ! 3 3 liJd6 'ifa4 with strong counterplay. Now it is Black who is playing on both wings. •••

a2) 3 0 lhh5 .l:'txg l + 3 1 �xgl lIg8+ 3 2 \itfl 'ifg7 3 3 'iff2 c xb4 ! (this anti-positional move, giving up control of the d4-square, is fully j usti­ fied tactically: Black opens the c-file for his pieces in order to strike from the other side too; as always, combining play on both wings should be done with care - changing the move­ order with 3 3 . . . 'iVg6 ? ! 34 l:Ih4 cxb4 fails to 3 5 liJd6 fxe3 36 'iVg3 , when White repels the at­ tack and keeps his material advantage) 34 exf4 (White has to consolidate his kingside before anything else; 34 axb4? is completely wrong because of 34 . . . 'iVg6 3 5 lIh4 fxe3 36 'iVg3 'iVc6 and the switch to the queenside i s decisive thi s time; after 34 liJd4? fxe3 Black h a s t o o many pawns for the knight; or 34 .l:'th4? b3 ! followed by . . . lIc8 , when, besides an attack, B lack has a dangerous passed pawn) and now Black c an force a draw by 3 4 . . . 'iVg6 (or 34 . . . .l:'tc8 3 5 'iVb2 lIg8 with the same result) 35 �h4 'iVc6 36 liJd4 'iVe I + 37 \ite2 'iVb2+ 3 8 \ite3 'iVc3+ with a per­ petual. We can conclude that in several lines Black's attack has been facilitated by the unstable posi­ tion of the knight. b) Therefore, 29 liJd6 ! suggests itself as an obvious improvement. In this case, White is able to keep control of the position after 29 . . . gxf4 (29 . . . �h7 transposes to the game) 3 0 exf4 'iVe6 (30 . . . 'iVh4 fails to question White's stability because of the calm 3 1 'iVc l ! , defending the f4pawn and hitting c5; after 3 1 . . .c4 3 2 'ife3 White has a strategically winning position) 3 1 bxc5 bxc5 32 'iVc3 c4 and now the transfer of the knight to the perfect blockading d4-square

123

PLAY ON TWo WINGS

starting with 3 3 ttJb5 ensures White a clear and long-lasting advantage. 29 ttJd6 g5 Too late, but 29 ... c4 only prolongs Black's agony after, for instance, 30 'iYd4 'iYe6 3 1 l:1h3. 30 'iYc2 (D)

B

The triumph of the play on two wing s . The pressure against the f5- and c5-pawns keeps Black ' s counterplay under control while ensur­ ing further queenside progres s . 30...'iYe6 3 1 l:.h3 With all the white pieces in optimal positions, this thematic move is the quickest way to win. 3 1 'iYg6 3 1 . . .J.e8 is completely hopeles s because of 32 bxc5 bxc5 33 'iYxc5 with a safe extra pawn and an overwhelming positional advantage . The familiar attack starting with 3 1 . . . gxf4 3 2 l:.xg8 l:.xg8 fail s t o 3 3 Zhh5 'iYg6 3 4 'iYxf5, when 3 4 . . . fxe3 is best answered by 35 e6, pre­ paring to stop the e3-pawn with 'iYxg6+ and :e5 . 32 fxg5 B y this stage of the game, choosing the win­ ning path is a matter of taste. 32 bxc5 bxc5 3 3 'iYxc5 i s equally good, o f course . 32 hxg5 (D) It i s remarkable that Black is subj ected to de­ cisive pressure precisely on the wing where he was supposed to attack. This is the result of White' s strong prophylactic kingside moves, right after the opening, as well as the switch to the queenside at the right moment. 33 e6

Or 3 3 ttJxf5 . 33 h6 34 ttJf7+ ,l;txf7 35 exf7 'iYxf7 36 bxc5 bxc5 37 'iYxc5 l:.g6 38 'iYd4 llg8 39 llet 'iYe6 40 .:tg3 g4 41 'iYf4+ 1-0 •••

I was deeply impressed by these games (for both their individual value and their similarity). Therefore, I felt quite proud when soon after analysing them I was given the possibility of putting my knowledge into practice. Although we shall notice certain similarities with the pre­ vious games, I believe that here the combina­ tion of simultaneous threats on both wings is illustrated in a clearer form. In B eliavsky' s games th e switch from one wing t o the other occurred only once per game, while here it is more of an oscillation of the main action be­ tween the opposite sides.

w

•••

••.

Mari n - Narciso Andorra 2000 We immediately notice certain strategic el­ ements from the previous examples . White' s advantage consists o f his ' better' bishop, the strong outpost on e5 as well as the potential weaknes s of the c 6-pawn. However; the closed character of the position and Black's reason­ able development make his position difficult to break. The introductory moves are quite natural. 1 9 .:tgl �d7 Another possible continuation is 19 . . . ttJg6 20 'iYe3 'iYf6 2 1 iLh3 and now Black's attempt to activate the bishop along the f l -a6 diagonal

124

SECRETS OF A TTA CKING CHESS

by 2 1 . . .j,a6 22 :tg3 b4 is met first with the strengthening to the maximum of the kingside pressure by 23 krag 1 a4 24 liJg5 liaf8 and then the thematic opening of play on the other side with 25 a3 ! followed by the infiltration by the queen with �d2, e3, �aS-c7, etc. We shall see more about it in the game. 20 �h3 liJg6 2 1 'ife3 'iff6 22 :tgS :tg7 23 nhS (D) Acknowledging that 23 !tagl leads to noth­ ing concrete in view of 23 ... liag8, White in­ tends to create some threats by means of :h6 and liJg5 .

However, having simplified the position and taken over control of the g-file, Black has a much easier task to conduct the defence than in the game.

24 a3! Again this modest advance of a rook's pawn on the opposite side from the main theatre of action . . . The exclamation mark is not so much for the move itself (I could have played it a little bit later as well) , but for the general idea behind it. After reaching a favourable kingside config­ uration, White will try to attack the as-pawn in order to force ... a4, when both the queen or (in case of mass exchanges) the king will have free access to the black queenside via b4-a5. 24 J1.e8 25 :gS :aa7 (D) It appears that Black has misplaced some of his pieces for nothing, since now it i s not easy to release the tension along the g-file without making a significant positional concession. For instance, after 25 . . . lhg5 26 fxg5 White gets a mobile kingside maj ority as well as an excel­ lent outpost on f4. •••

B

w 23 liJf8 ? ! Black decides t o prevent lih6 and t o activate his bishop via e8, but in doing so he breaks the communication between his rooks for a few moves, giving White the time to consolidate his domination. We have already seen such a motif in B eliavsky-Bareev. B etter is 23 . . . :ag8 and if 24 lih6 then 24 . . . 'iff8 ! . With this far from obvi­ ous move, Black prepares to transfer his queen to b8, in order to control the e5-square and at­ tack the f4-pawn, thus forcing events. Play may continue 25 :tgl (25 liJe5 liJxe5 26 'ifxe5 gets White nowhere due to 26 . . . �8 ! ) 25 . . .�8 26 liJe5 liJxe5 27 lhg7 lhg7 2 8 fxe5 'ifg8. Clearly, White preserve s a long-term advantage, be­ cause of the large number of pawns fixed on light squares. In the near future, he needs to im­ prove his position with 'iff2, �g2-f3 , h3, h2 and then try to provoke either . . . a4 or . . . b4, in order to get routes of infiltration for the king and thus be able to offer mass simplifications. •••

26 liagl h6 27 llSg3 lixg3 From a practical point of view, this exchange is most welcome for White, who gladly im­ proves his pawn-structure. However, it is hard to criticize Black's move, since in case of a more neutral move such as 27 ... liJg6 28 �g2 :tg8 29 :h3 lih7 White can proceed with his plan by playing 30 �d2 'ifd8 3 1 e3 followed by 'ifel and liJd2-b3, when Black does not have sufficient forces available to defend the as­ pawn and so will have to advance it.

PLAY ON Two WINGS

28 hxg3 (D)

B

125

f6-square with 29 . . . lZ'ld7 could lead to problems after 30 'iVd2 'iVd8 3 1 g4 fxg4 32 i..x g4 i..f7 3 3 fS ! , when th e position opens t o White' s advan­ tage. 30 'i>tgl i.. d 7 3 1 'i>tf2 i..c 8 32 i1 h2 lih7 33 i.. g 2 �d7 Black has regrouped according to his plan and now invites his opponent to undertake con­ crete action. 34 'iVc3 Having achieved an optimal kingside re­ grouping, White turns his attention to the other wmg. 34 :iVd8 35 lZ'ld2 I intended to prevent . . . hS by means of J.. f3, and then transfer the knight to d3 via b3-c 1 , forcing on the way the advance of the a-pawn. 35 h5 (D) •.

Now that the structure has changed, White will have to adjust his kingside plans. He will transfer his king to the safer f2-square and his rook to the h-file. Although g4 is a potential threat, he should not play it too soon, since this would allow Black to activate his bishop along the b l -h7 diagonal. Provoking . . . hS could be considered a significant achievement, since the knight would get the gS-square, but at the same time White should try to prevent further simpli­ fication with . . . h4. For Black it will prove diffi­ cult to keep both his rook's pawns on their current squares . 28 i1 g7 29 i1g2 (D)

••.

w

•••

B

29 lZ'lg6 Black places his knight on a rather passive square in order to put pressure on the f4-pawn and thus safely prevent g4. Trying to reach the ••.

Black feels the danger and prepares kingside counterplay. If White continued with 36 lZ'lb3, then 3 6 . . .a4 followed by 3 7 . . . h4 would help Black free his position to a certain extent. How­ ever, there is no need for such a hurried ap­ proach. By threatening to attack the as-pawn White has indirectly provoked a weakness on the other wing. Therefore, the best idea is to change plans for the time being and restrict Black' s kingside activity: 36 lZ'lf3 ! 'i>tg7 36 . . . h4 loses material to 37 lZ'lgS . But now, Black renews the threat by clearing the h8square for the rook and placing the f7-square under control.

126

SECRETS OF A TTA CKING CHESS

37 'iic 1 ! Again a move o f a prophylactic character. 37 . . . h4 is prevented in view of 3 8 gxh4 lbxh4? 39 'iih l , winning material. 37 'iif6 38 e3 It might look as if White is defending. He safely protects the f4- and d4-pawns in order to prepare the transfer of the queen to hI without fearing . . . h4. 38 i.e8 Black is now ready to defend his h5-pawn with the bishop if necessary, but the e6-pawn i s left without defence for an instant, which al­ lows White to make further progress. 39 'iie l ! (D)

B

•.•

•••

B

39 a4 Finally, Black has to give in, since 39 . . 'iid 8 loses material to 40 lbg5. With his rook on h8, B lack could have probably refrained from this new concession, but even in positions with a closed character chess i s sometimes 'the trag­ edy of just one tempo' . 40 lbg5 lih6 4 1 'iia S 'iie 7 B ecause of the strong threat of 'iic 7 +, Black has no time to advance his h-pawn yet. ••.

.

42 'iia 8! (D) Psychologically, it was quite hard to refrain from the elegant 42 'iic 7, when after 42 ... Wf6 43 'iib 8 e5 ! Black considerably frees his posi­ tion and 44 dxe5+ ? lbxe5 even leaves him on top all of a sudden. 42 e5 A desperate attempt to get counterplay. In case of a waiting move such as 42 .. .'�f6 play •••

could continue with 43 Jif3 h4 44 gxh4 Ihh4 45 %:lxh4 lbxh4 46 iYb8 (threatening a decisive check on e5) 46 . . . lbg6 47 Jih5 . If White ' s pres­ sure proves insufficient to break through on the kingside, the king would march in to b6 com­ pletely unhindered, when B lack' s position will soon collapse. The immediate 42 ... h4 is quite similar; for instance, 43 gxh4 l:.xh4 44 ':xh4 lbxh4 45 'iic 8 i.d7 46 'iic 7 (planning .tfl -e2-h5) 46 . . . l2Jxg2 47 Wxg2 followed by marching the king to b6 via f2 and e l . 43 dxe5 lbf8 This is equivalent to resignation. Black prob­ ably relied on 43 ... lbxe5 44 fxe5 'iix e5 but then s aw that after 45 'iid 8 'iix b2+ 46 �f3 'iie 5 47 ':h4 White manages to consolidate and pre­ s erve his material advantage. Undermining the f4-pawn with 43 ... h4 44 gxh4 d4 45 exd4 lbxf4 also fails to 46 li'c8 . In these variations w e can see that the white queen is just in time to defend the kingside, which it abandoned several moves ago. Communication between the wings is cru­ cial for such situations; otherwise Black's COUll­ terplay would become too dangerous. 44 lbf3 lbe6 The attempt to give perpetual check with 44 . . . li'xc5 45 li'xe8 li'c2+ fails to 46 Wfl ! (46 WgI li'di + gets White nowhere) 46 . . .'iid 3+ 47 WeI iYbI + (or 47 . . . li'xe3+ 48 �dl li'd3 + 49 lbd2, winning) 48 �e2 li'xb2+ 49 lbd2 and the checks are over. Black's counterplay also comes too late after 44 . . . i. d7 because of the calm 45 li'a7 defending the c5-pawn and threatening the decisive lbd4 .

PLA Y ON Two WINGS

45 liJh4 The rest i s easy and needs no c omment. 45 ..:iVf8 46 i.h3 lDxcS 47 liJxf5+ Wg6 48 liJd6 l:th8 49 i.f5+ Wh6 50 g4 liJe4+ 5 1 ..txe4 dxe4 52 liJxe8 1-0

A Tri bute to the Class i cs For a long time, 1 mentally associated my game agains t Narciso with B eliavsky ' s 'reversed twins' . Recently, 1 was rather surprised when 1 looked at the following game, played by the fourth World Champion, Alexander Alekhine. 1 knew that Alekhine is regarded as the first great player who employed the method of playing on both wings simultaneously, and also found a considerable number of inaccuracies i n this game, but some strong similarities with my own game gave me food for thought. 1 had studied Alekhine's games as a child but forgot most of them in the meantime. And yet, certain things must have accumulated in my mind at a sub­ conscious level. The relative ease in handling the previous game (1 remember that 1 played the technical part rather quickly) must have had something to do with having seen Alekhine' s game 3 0 years earlier. . .

15 h4! i.f6 16 h5 liJrs 17 g3! ? Alekhine follows the dictum festina lente, preferring to prepare a safe square on g2 for the king instead of rushing in with 1 7 liJeS . 17 :aa7 The generally desirable exchange of the light-squ ared bishops with 1 7 . . . h4 1 8 liJd l J.. a6 does not free Black from problems after 1 9 .txa6 11xa6 2 0 liJe3, when White keeps the ini­ tiative. 1 8 liJdl (D) ••

B

White continues to refrain from liJeS, and brings the other knight into the attack first. 1 8 i.g7 The computer-like continuation 1 8 . . . gS ! ? 1 9 liJeS ( 1 9 fxgS i.xgS achieves nothing for White) 1 9 . . . l:.c7 avoiding the opening of the h-file is probably better, but thi s would be too much for a player from the 1 920s . . . Anyway, Alekhine could have prevented this change in the course of events by exchanging on g6 at an earlier stage. 1 9 liJe3 f5 This line of defence prepared by Black will prove vulnerable. 20 'iYe2 (D) 20 a4? This is a significant mistake, immediately bringing B lack to the verge of defeat. The weak­ ness of the b4-square is even more significant than it was in my game, since White will be able to install a knight on this square, thus tying Black down to the defence of the c6-pawn. And yet, it is precisely this moment that has a logical •••

w

•.•

Alekhi n e

-

Chajes

Karlsbad 1 923 The strategic elements are already familiar to us. White starts a kingside offensive, aiming to open a file for his rooks.

12 7

128

SECRETS OF ATTA CKING CHESS

B

B

c onnection with my whole play starting with 24 a3 . Why i s this? I have checked the c omments available in chess literature regarding this critical moment that I probably read in my childhood. The most interesting of them belongs to Alexander Kotov. He writes that had B lack not advanced this p awn willingly, Alekhine would have had a dif­ ficult task provoking this weakness. Alekhine himself wrote in the notes to this move that he would have transferred the e3-knight to e5 , via c2-e l -d3 , after the preliminary �f1 , i.bl - this is a more modest approach, of c ourse. How­ ever, in his notes to the 3 3rd move he mentions the transfer of the king to the opposite wing as part of his plan, in order to be able to invade the enemy camp in case of simplification into an ending. In other words , I played as if somebody had told me the c orrect plan ; this 'somebody' was my subconscious . . . Anyway, B lack' s most active c ontinuation i s 2 0. . . �a6, aiming t o exchange the passive bishop and if 2 1 a4 then 2 1 . . .Vi'a8 with strong counter­ play against the b2- and d4-pawns in case White captures the pawn on b5 . 2 1 'D e2 (D)

fxg5 �xg5 26 f4 soon followed by g4) 25 Vi'xe5 lbg6 26 Vi'e3 gxf4 27 gxf4 �h8 2 8 �d2 followed by ':cg l -g5, l:thg l and �e2-h5, with continuously increasing pressure. b) 23 . . . i.xe5 24 dxe5 ! (curiously, the cap­ ture with the queen is less adequate now be­ cause after 24 Vi'xe5 lbg6 25 Vi'e3 B lack can take advantage of the fact that the c6-pawn is not yet attacked and regroup with 25 . . . �ae7 ! ? threatening both . . . e5 and . . .gxf4 followed by . . . Vi'c7) 24 . . . lbg6 (24 . . . g4 does not really block the position, since White can open it by means of f3) 25 Vi'd2 d4 ! ? ( aiming to prevent lbd4 with total control of the position) 26 ttJb4 with a stable advantage for White, although the slightly weak c5- and f4-pawns restrict his active possi­ bilities. 22 �f1 i.f6 23 'DeS �xe5? ! Giving up the control o f the dark squares too easily. Some waiting move like 23 . . . Vi'c7 is better.

21...:i:iae7 A possible attempt to complicate matters i s 2 1 . . .g5 ! ? with the point that 22 'Dxg5 i s met b y 22 ... e5 with strong counterplay. However, White can preserve his advantage with 22 h6 �f6 23 'De5 ; for instance : a ) 23 .. Jic7 2 4 'Db4 �xe5 (24 . . . �b7 places the bishop on a very passive square, increasing the effect of the opening of the kingside after 25

24 Vi'xeS The careless 24 fxe5 ? allows Black to keep the position closed with 24 . . . g5 ! . However, 24 dxe5 is entirely possible, in order to get the d4square for the knight. 24 Vi'e7 Black could have considered 24 . . . l:tg7 ! ? in order to be ready for ... g5 followed by . . . e5 at the right time. •••

25 Vi'f6 An important moment. Alekhine decides to keep the queens on the board for the time being. He aims to provoke some additional weaknesses or concessions of some other nature before

PIAY ON Two WINGS

trading queens. Although 25 hxg6 'iVxe5 26 dxe5 hxg6 27 lDd4 followed by the transfer of the king to b4 certainly looks promising for White, he can hope for more. For a definitive verdict about this decision see also the comment on B lack' s 49th move. 25 lif7 26 'iVh4 'iVe7 27 bxg6 lDxg6 (D)

129

W

•••

w

Alekhine's apparent indecision i s in fact a rather unpleasant method of play against an op­ ponent who completely lacks counterplay. 42 .liec7 43 lih2 �g6 Reinforcing the h7-square and the h-file in the safest possible way. 44 'iVe3 Wc8 45 lichl Finally, all Black's weaknesses (c6, e6, h7) are subj ected to strong pressure. 4S Wb7 46 Wd2 1le7 47 lDd3 lDd7 (D) ••

This i s the first concrete achievement gener­ ated by the presence of queens on the board. B ecause of the mating threat on h8, Black has to spoil the compactness of his structure. In fact, Black's pawn configuration is almost identical to Narciso's. 28 'iVh5 Vi'f6 29 �e2 The start of a major regrouping. The queen has done its job on the kingside and has to re­ turn to the centre, in order to control the vital dark squares. The bishop will replace the queen on h5. 29 J:tg7 30 'iVf3 lDf8 3 1 Vi'e3 llee7 32 lDb4 �d7 33 �h5 lDg6 34 lDd3 �e8 35 We2 Finally, the white king approaches the queen­ side. 35 WfS 36 Wd2 llb7 37 i..f3 We7 B eing afraid of the possible resulting end­ games, B lack also rushes his king to the queen­ side, in order to prevent his rival from marching in. However, his situation will not be much s afer than on g8. Besides, the kingside will cer­ tainly be weakened by the departure of one of its main defenders. 38 �he1 lDf8 39 lDb4 Wd8 40 Wd3 Ilge7 4 1 "iVd2 1la7 (D) 42 1lhl

•••

w

••

.••

Black tries to guard all the entries but. . . 48 �h5! There i s a little paradox here. Optically, one could not imagine a ' worse' bishop than the black one, since all the black pawns are on light squares. And yet, in order to make further prog­ ress, White has to exchange this bishop, which is safely defending the kingside. However, if the infiltration by the white rooks does not prove decisive, Black will have considerably

SECRETS OF A TTA CKING CHESS

130

freed his position as a consequence of this ex­ change. 48 �a8 49 �xg6 (D) •••

B

49 hxg6? The s econd paradox. This natural move, sig­ nificantly improving the black structure, is in fact the decisive mistake. After the strategically less desirable 49 . . . 'ii'x g6 White will have seri­ ous problems making significant progress if Black defends carefully. For instance, 5 0 �h6 'ii'g8 5 1 �c2 (clearing the d2-square to give the queen access to the queenside; 51 'ii'e 2, planning to intensify the attack against the h7-pawn, leads nowhere after 5 1 . . .�f8 52 'ii' h 5 �ff7) 5 1 . . .�ae8 52 'ii'd 2 'ii'g 7 and now: a) After the structural modifications induced by 53 ltJe5 ltJxe5 54 fxe5, the queen will have acces s to both wings from the key d2-squ are, but this will not prove of decisive importance yet. 54 . . . �b8 55 � l h4 (after 55 'ii'f4 B lack can consider sacrificing the h7-pawn with 5 5 . . . 'ii'g4 ! in order to get counterplay against the f2-pawn in case of 56 'ii'x g4 or against the f4-pawn af­ ter the immediate 5 6 �xh7 ; 55 'ii'a 5 i s less dan­ gerous than it l ooks, since after 55 . . . �c8 5 6 iib6+ �a8 one cannot s e e a decisive blow yet) 55 . . . l:.b7 56 :!f6 b4 ! ? (preparing counterplay along the b-file and, more importantly, denying access to B lack' s position) 5 7 axb4 'fl..e e7 and Black survives, although his life is not exactly rosy. b) 5 3 ltJel ! ? (the knight is aiming for g5 , in order to attack both the h7- and e6-pawns) 53 . . . ltJf6 (planning to meet ltJf3 with ... ltJg4) .••

54 'ii'a 5 �c8 (the only move; against 54 . . . ltJd7? White ' s idea i s crowned with success after 5 5 ltJf3, when the secondary threat of ltJ e 5 { under­ mining the b6-square } i s an important comple­ ment to the main threat, ltJg5 - a real triumph of simultaneous play on two wings) 55 ltJd3 (re­ turning to b as e ; neither 55 iib6+ �a8 5 6 ltJd3 ltJg4 nor 55 ltJf3 ltJg4 ! 56 1hh7 'ii'x h7 57 �xh7 'fl..xh7 58 ltJg5 'fJ.e7 offers White anything con­ crete) 55 . . . ltJd7 56 'ii'c 3 (defending the d4-pawn for the moment and preparing a new wave of at­ tack; the hurried 56 ltJe5 ltJxe5 57 fxe5 offers B lack some counterplay after 57 ... f4 ! 5 8 gxf4 'ii'g 4) 5 6 . . . �b8 57 ': l h4 ':f8 (Black should take c are; after 57 . . . ltJf6? 58 'ii'e l threatening 'ii'e 5+ Black is suddenly in trouble, because af­ ter any knight move the e6-pawn is hanging) 58 'ii'a 5 �b7 ! (the only move again; the greedy 58 . . . 'ii'x d4? loses to 59 'fJ.xh7, and the same goes for 5 8 . . . l:.c8 ? 59 lixh7 'ii'x h7 60 1hh7 :xh7 6 1 ltJb4, when Black cannot defend the c6-pawn any l onger) 59 ltJe5 'fl..a 8 and White still needs to prove a winning plan. Generally, White ' s technical problems are connected with the vulnerability of the d4-pawn and the impos­ sibility of finding a 1 00% ideal square for the king. On a2, the king is far from the access route c 3-b4-a5 that it needs to use in case of simplifications. If this position is impossible to win, then Alekhine' s 25th move must have been mistaken. In the position resulting after the exchange of queens he would have had the h-file at his disposal while the light-squared bishops would still be on the board. 50 �h7 First things first. The infiltration by the rooks becomes more problematic after 50 ltJe5 ltJxe5 5 1 fxe5 'ii'f 7. 50 'fl..a e8 Black loses immediately in case of 50 ... lIxh7 5 1 'fl..xh7 �c8 52 ltJe5 ltJxe5 5 3 fxe5 followed by 'ii'h 6 or 'ii'g 5 with a decisive attack. 5 1 ltJe5! Finally, the position is ripe for this thematic knight jump. 5 1 ltJf8 A sad necessity. The seventh-rank pin proves decisive after 5 1 . . .ltJxe5 5 2 fxe5 'ii'f8 5 3 'ii'g 5 'fJ.g7 54 'ii'x g6 ! llee7 55 'ii'h 6. ...

•••

PLAY ON Two WINGS

52 J::t h 8 J::tg 7 53 lbf3 (D)

131

59 lbh7 wins easily, but Alekhine intended to finish the game in true 'play on two wings ' style. 59 �a7 60 �c3 �a6 6 1 liJf7 'tl.a8 62 liJd6 'tl.eb8 (D) ••.

B

w

White clears the e5-square for the queen while transferring the knight to the equally at­ tractive g5-square. Black c annot prevent the new wave of invasion because his knight is pinned. 53 J::t b 8 54 lbg5 'tl.e7 55 'iVe5! 'iVxe5 56 fxeS Black's complete passivity makes his posi­ tion hopeless. 56 �a8 57 J::tg 8 b4 "Hoping for 5 8 axb4 'tl.eb7, with some chances" (Alekhine) . 5 8 J::th h8 'tl.ee8 5 9 axb4 ...

.••

Can White make any further progres s on the kingside? 63 �hl ! He does not need to ! The rook threatens to deliver the decisive blow to the enemy king with �al xa4#. Alekhine' s feel for a mating at­ tack in what is basically a technical position is impressive. 63 liJd7 64 �al l-0 ••.

5 T h e Sec rets B e h i n d M or p hy ' s S uccesses

I have never hidden m y addiction t o the classics of chess. However, along the years I have not been very sure where a line should be drawn when it comes to the concrete usefulness of studying classic games. Should I start with the golden era of the S oviet chess school, or maybe also include some outstanding players from the first half of the 20th century? Recently, I simply gave up the very idea of drawing a line. It was a sudden revelation; many ideas that I had carri ed in my mind for a long time received a logical classification. It all started when I opened the book Nekoronovanye Chempiony (Uncrowned Champions) written by Neishtadt. I read the historical part with interest and saw surprisingly (for me ! ) many in­ teresting games played by Philidor, La Bour­ donnais , S taunton, etc. However, I was in a bit of a hurry: I wanted to reach as soon as possible the last chapter, dedicated to Morphy, consid­ ered by the authorities to be one of the most brilliant players of all time. Life is, however, full of surprises; for a long time I wasn't able to finish reading the chapter about Morphy. It was like pursuing for a long time the idea of finding a famous mine of gold, but encountering an oasis of happiness just be­ fore reaching the initial goal. The last-but-one chapter was dedicated to Adolf Anderssen, the winner of what is con­ sidered the first international chess tourna­ ment (London, 1 85 1 ) . The name was familiar to me; my first trainer had shown me, when I was 9 years old, the ' Immortal' and the 'Ever­ green' games. From the very first game I played through on the board, I was bewitched. S o many ideas that I had come acro s s during my c areer were there, i n pure form, in games played by an amateur chess-player (Anders sen

taught mathematics in Breslau) more than a century before the ChessBase generation ! When I finished the Anderssen chapter, I fi­ nally started to read the section about Morphy. I read the biography and played over some of the games, but didn' t succeed in finishing the chap­ ter. Every time I opened the book, the first thing to do was to step again and again into the newly discovered ' oasi s ' . When the first wave of enthusiasm was con­ sumed, I started asking myself some questions. Morphy ' s supremacy over his contemporaries was considerably more impressive than An­ ders sen' s in his golden time s . B esides, the match that opposed the two champions ended in a total fiasco for Anderssen. B eyond my subjective taste, such historical facts certainly needed some explanation of a technic al nature. I referred to the opinion of such outstanding commentators and analysts as Euwe, Neishtadt and, more recently, Kasparov, who unanimously agreed on the fact that Morphy' s play repre­ sented the first step towards modem chess, while Anderssen ' s games, though spectacular, lacked depth. Here are some fragments which can be found in the books of all three authors mentioned above: "the old combinative school led by Anderssen proved powerles s against the modernized technique of Morphy . . . which was based on a sound positional foundation." "[Morphy] spontaneously established the three main principles of opening play : 1 ) the rapid development of pieces 2) the seizure of the centre 3) the opening of lines". Neishtadt goes a bit further and, regarding the first point above, makes the following in­ teresting specification: for the sake of devel­ opment, Morphy' s pieces never (or rarely) retreated when attacked. He also specified that

THE SECRETS BEHIND MORPHY'S SUCCESSES

while Morphy' s strategic skills were clearly su­ perior to Anderssen' s, he wasn' t weaker than him on tactical territory either. With such a clear guide, I started examining the games played between the two heroes, ea­ ger to find the practical reflection of the theo­ retical knowledge I had acquired. I soon found the following game (which was in fact played after the official match and formed part of a se­ ries of games dedicated to the King' s Gambit) that matched perfectly the points 1 , 2 and 3 mentioned above.

1 33

4 dS! Morphy doesn' t spend time parrying the threat. Instead, he opens the diagonal for his queen' s bishop, at the same time preparing an outpost on e4 for the knight - just like in Nei­ shtadt' s description. S �b3 Mter S exf6 dxc4 Black's better possibilities for further development as well as the bishop­ pair compensate for his pawn weaknesses. 6 'iVe2+ is not dangerous ; for instance, 6 . St.e6 7 'iVeS 'iVd5 ! (but not 7 lbc6 8 fxg7 !) . S lbe4 The knight occupies an active position while the bishop has been forced to retreat to a more passive square. 6 lbf3 J.g4 7 0-0 lbc6 (D) •••

..

...

•..

Anderssen - M orphy Paris 1 858 1 e4 eS 2 f4 exf4 3 .i.c4 Although by that time it was theoretically established, this is the first move in slight con­ tradiction to the recommendation to develop knights and only then bishops. True , this ax­ iom was stated several decades after the mo­ ment the present game was played, but it i s remarkable that Morphy u sually played the opening in accordance with it, thus being sev­ eral steps ahead of his generation. The ex­ posed position of the bishop will soon cause White some trouble. 3 lbf6 4 eS? ! (D)

w

••.

B

With each move, Black develops a piece, at the same time creating concrete threats based on strong pressure in the centre. 8 �a4 A good illustration of the fact that under cer­ tain circumstances a development advantage tends to increase. Quite sadly, White was de­ prived of the most natural way to defend the eS-pawn and prevent . . . .tcS, namely 8 d4?, which fails painfully to 8 lbxd4 ! , but instead of moving his bishop for the third time, An­ derssen could have tried to keep Morphy' s de­ velopment advantage under control with 8 'iie l ..tcS+ 9 h l . True, after 9 .i.xf3 1 0 ':xf3 ( 1 0 gxf3 might be better although after 1 O . lbgS followed by . . . lbe6 Black consolidates his ma­ terial advantage and keeps a solid position) ...

Anderssen was seduced by this early action in the centre, hoping to obtain the initiative. After B lack's accurate ans wer, it will become clear, however, that White has simply lost a valuable tempo for his development.

...

.

.

SECRETS OF A TTA CKING CHESS

134

1 O . . ttJxe5 1 1 �xf4 ttJg6 Black retains the extra pawn and the better development. Tactics don' t work yet; for instance, 1 2 �xf7 c;t;xf7 1 3 'iVxe4 �e8 ! . 8 g5 (D) Morphy' s first deviation (although not a very significant one) from his main strategy. Simpler is 8 . ttJg5 in order to transfer the knight to a more stable position and defend the f4-pawn without creating any weakness. For instance, 9 d4 �xf3 1 0 gxf3 ttJe6, when 1 1 c4 i s not dan­ gerous because of 1 1 . . .dxc4 1 2 d5 'iVg5 + 1 3 c;;t>h l 0-0-0 and B lack completes his develop­ ment with a very good game. Morphy' s move i s more ambitious but also riskier. .

•.•

..

1 6 ttJc3 ttJxc3 17 'iVxc3 cxb4 1 8 axb4 �xb4 19 'iVd3 as Finally, Black's development advantage has been converted into a more stable form of ad­ vantage. Black has won a second pawn, while retaining a solid position. Within a few moves, Morphy' s position attains complete harmony. 20 h4 h6 21 hxg5 hxg5 22 g3 (D)

B

w

Anderssen makes a last attempt to change the course of the game, relying on the slight weakness of the black kingside.

22 :b6! 23 :a2 23 gxf4 gxf4 24 �xf4 isn ' t too appealing ei­ ther; for instance, 24 . . . Iig6 25 c;t;f2 'iVb6 and the white king is surrounded by dangers from both wings. 23 :g6 (D) Mter the rook transfer, the black kingside is well defended, while the white king starts feel­ ing insecure. •••

9 �xc6+? ! Finally allowing Black a clear advantage. By hurrying with this unnecessary exchange of a bishop that has already consumed three tempi, White lets his opponent immediately under­ mine the centre. Anderssen might have feared that on the next move Morphy would play . . . 'iVd7 , keeping his structure intact, but he ob­ viously underestimated the dynamic strength of the doubled pawns. 9 d4 is better. 9 bxc6 1 0 d4 c5 1 1 c3 �e7 1 2 b4 Anderssen finally understood that he was on unsafe ground. With this move, he tries to pre­ vent his centre from being undermined with . . . cxd4 followed by . . . c5, but this means his queenside remains undeveloped for a few more moves. 12 cxb4 1 3 cxb4 0-0 1 4 'iVb3 �b8 15 a3 c5! The blockade has been broken anyway ! •.•

•.•

•.•

w

THE SECRETS BEHIND MORPHY'S S UCCESSES

24 �g2 ! ? "Anderssen's ingenuity i n the following p art of the game is a clear sign that his defeat in this match wasn' t caused by his bad form. The main thing is that his tactical mastery started to mani­ fest itself in a desperate position, reached as a result of Morphy' s superior strategy in the pre­ vious phase of the game." These were my fIrst thoughts when I saw this fIne game. Later I had to revise my views though. 24 iLh3 25 e6 iLxg2 25 . . .�xe6 26 �h2 .u.h6 is also reasonable. 26 exf7+ rJi;g7 27 rJi;xg2 'iVc8 (D) •••

135

The game we have j ust examined seems like a perfect illustration of the official thesis regard­ ing the new elements introduced by Morphy and that gave him such a supremacy over his contemporaries. And yet, I soon discovered that Anderssen had been the first to offer a practical lesson about development to his acclaimed op­ ponent. Let us follow the very fIrst game of the main match between these wonderful players. There are several common points between these two games, the only significant difference being that the roles were reversed.

M orphy- Anderssen Match (game 1), Paris 1 858 1 e4 e5 2 ttJf3 ttJc6 3 iLc4 iLc5 4 b4 Another popular gambit at the time. It was in fact Anderssen's favourite variation with White. 4 i.xb4 5 c3 J..a 5 6 d4 exd4 7 0-0 ttJf6 (D) A rather provocative move, instead of the ap­ proved continuation 7 . d6 8 cxd4 iLb6.

w

•••

..

w

With his next few moves, Black regroups for the final attack, while keeping several impor­ tant squares around his own king under control. 28 ttJe5 �h6 29 gxf4 �fb8 30 fxg5 Anderssen has fInally managed to create threats against the enemy king, but Morphy was ready for action j ust one move earlier. The tragedy of one tempo . . . 30 �h2+ 3 1 rJi;gl �hl+ 3 2 rJi;f2 li8h2+ 33 rJi;e3 �h3+ 34 ttJf3 �xf3+! This well-calculated combination i s the shortest way to win. "Morphy' s sparkling tacti­ cal mastery prevented many of the commenta­ tors from gaining a deeper understanding of his style of play." Once again, these initial thoughts of mine look rather naive now. 35 �f3 3 5 rJi;xf3 is no better: 3 5 . . . 'iVh3+ 36 rJi;f2 �h2+ and mate next move. 35 'iVxc1+ 36 rJi;f2 'iVe1 + 37 rJi;g2 'iVgl# ( 0-1 ) • • •

•••

8 e5 ? ! The same typical mistake as i n the previous game. With his queenside still undeveloped, Morphy starts a premature attack. One year later he demonstrated the correct plan in a blind­ fold g ame played in a simultaneous display: 8 i.a3 ! (preventing Black from castling) 8 d6 9 e5 ! (now this move i s considerably stronger) 9 . . d5 ? (this reaction isn't adequate any more; however, as the German master B erthold S uhle later demonstrated, White' s attack is very strong anyway: 9 ttJxe5 1 0 ttJxe5 dxe5 1 1 'iVb3 'iVd7 ...

.

...

SECRETS O F AITA CKING CHESS

136

1 2 �eI e4 1 3 tDd2 J.. x c3 1 4 tDxe4 �xe l 1 5 tDxf6+ gxf6 1 6 IIxeI + Wd8 1 7 iif3 f5 1 8 iif4 f6 I 9 'iVh6 winning, or 9 . . . dxe5 I O iib3 iid7 1 1 tDg5 with strong threats) 1 0 �b5 and White won in Morphy-Greenway, London blindfold simu1 1 859. S dS 9 i.b5 As in the previous game, White doesn' t have enough development to pick up the gauntlet. After 9 exf6 dxc4 1 0 fxg7 l:.g8 1 1 i.g5 iid5 1 2 lIel + i.e6 1 3 i.f6 iif5 1 4 i.xd4 0-0-0 Black manages to bring his king to safety, with a con­ siderable advantage. 9 tDe4 1 0 cxd4 1 0 J.. a 3 was indicated by Anderssen as a possible improvement. He analysed 1 O . . . �xc3 1 1 tDxd4 ( 1 1 tDxc3 might be better although I doubt that White' s compensation is sufficient) 1 1 . . .�xal 1 2 tDxc6 but here 1 2 . . .iid7 1 3 tDd4 c6, as given by Neishtadt, seems to solve Black's defensive problems, leaving him with a mate­ rial advantage. 1 0 0-0 White has achieved absolutely nothing to compensate for the material deficit. In fact, it is Black who is better developed. Morphy decides to win his pawn back, although it is obvious that this involves further positional concessions. 11 �xc6 To be compared with Anderssen's exchange on c6 from the previous game. 1 1 ...bxc6 12 iia4 i.b6 13 iixc6 �g4 (D) •••

••.

14 i.b2?! A new inaccuracy, after which White should hardly have a chance to save the game. 1 4 �a3 only helps Black improve the position of his rook with 1 4 . . . lIe8, when the manoeuvre .. J1e6-g6 puts the enemy king in serious danger. The only acceptable defence is 1 4 i.e3 in order to give the kingside some protection too . In this case, Black can play 1 4 ... f6, taking advantage of the fact that, compared to the game position, the e5-pawn isn't well enough defended. 14 ...i.xf3 15 gxf3 tDg5 By simple means, Black has obtained a prac­ tically decisive attack. The f4- and f2-squares are hopelessly weak. 16 tDd2 (D)

B

.••

w

Anderssen develops his last minor piece, sub­ jecting the enemy centre to strong pressure.

1 6 lIeS Planning to activate the rook along its third rank. Although it certainly accentuates the sim­ ilarities with the previous game, the move itself is not the only possible continuation of the at­ tack. Anderssen might have been surprised by the ease with which he had managed to achieve such an overwhelming advantage against his mighty opponent. He probably couldn ' t believe that the time had come for decisive actions and decided to increase his pressure slowly. As I later discovered, this was his preferred manner of playing, very much in contradiction with the general opinion of the commentators. Anyway, the straightforward 1 6 ... tDh3+ ap­ pears no worse than the game continuation. Play may continue 17 Wh i ( 1 7 Wg2 keeps the f2-pawn defended but makes 1 7 . . . iih4 equally ••.

THE SECRETS BEHIND MORPHY'S SUCCESSES

unpleasant due to the additional threat . . . 4Jf4+) 1 7 . . . "iVh4 (for the moment, B lack is attacking with j u st two pieces but the typical manoeuvre . l;le8-e6-g6 remains in reserve, of course) 1 8 'iVc3 (the other possible way to bring the queen back to the defence, 1 8 'iVxdS, leaves the d2knight undefended, allowing 1 8 . . . 4Jxf2+ 1 9 'litg2 'iVg5 + 20 'litxf2 'ilxd2+) 1 8 . . . 4Jxf2+ 1 9 'litg2 4Jh3 and White' s position i s i n ruins. 17 'lithl 4Jh3 18 f4 (D) .

13 7

B

.

B

A good practical chance. White is ready to give up the doomed pawn immediately in order to slow down Black's attack. At the same time, he clears the long diagonal in order to bring his queen back via d5 . 1 8 'iVh4 Anderssen carries out his attack in a method­ ical way. Once again, the simpler 1 8 . . . 4Jxf4 is entirely possible as well. White can prevent the enemy queen from emerging to h4 with 1 9 4Jf3 but after 1 9 . . . 'iVc8 20 l;lgl 'iVf5 his position re­ mains difficult. •••

the moment when it seems that this was consid­ erably worse than developing the initiative. Black's better coordination can be crowned tactically with 2 1 . . .c 5 ! . White cannot maintain his central stability. Mter 22 'ile4 (aiming to avoid problems along the d-file ; 22 4Jc4 is met by a devastating rodeo of the black knight with 22 . . .4Jxf4 23 'ilf3 4Jh3+ 24 'lith l 4Jg5 25 'ilf4 'ilxf4 26 l;lxf4 4Je6, while 22 dxc5 helps Black include new resources in the attack with 22 . . . .ixc5+) 22 . . . cxd4 2 3 Sl.xd4, White has ap­ parently managed to keep his position together but the spectacular 23 . . . 4Jxe5 ! makes full use of the black pieces ' activity. For instance, 24 .ixb6 4Jg4 ! (the careles s 24 ...4Jg6? allows 25 �f2) 25 4Jf3 'iVh6 with an extra pawn and at­ tacking chances for Black. 22 'ilf3 (D)

B

19 'ilxd5 Finally obtaining some sort of piece coordination. 19 4Jxf2+ 20 'litgl 4Jd3 2 1 .ic3 (D) •••

2 1 4Jxf4? ! Until now, whenever he had a choice between relatively equivalent continuations, Anderssen's decisions have been dictated by the aim of get­ ting the maximum piece activity, ignoring ma­ terial factors. Or, in a more general sense, he preferred abstract thinking to concrete action. Finally, he now opts for pawn-grabbing, just at •••

Suddenly, Black faces some problems with his f7-pawn.

138

SECRETS OF A TTA CKING CHESS

22 ttJh3+ 23 'iith l ttJg5 24 "iVg2 And now, his knight has somewhat lost its stability. 24 !tad8 (D) •••

W

•••

w

25 ggl ? ! Morphy misses his chance. H e could have saved the day with 25 ttJf3 ttJxf3 26 "iVxf3 . All of a sudden, White is attacking along the f-file, opened as a consequence of the elimination of the weak pawns. Maroczy gives the following main line as favourable for Black: 26 . . . i.xd4 (26 . . . lIe7 27 i.b4 :t ed7?? loses to 28 e6 as in­ dicated by Maroczy, while after 26 . . . :td7 White can develop a dangerous initiative with 27 l bel followed by !t e4) 27 "iVxf7+ �h8 28 !tf4 "iVxf4 29 "iVxf4 .i.xc3 . However, it does not really look like White should be in danger of losing here. 25 h6 26 !taft Once again, 26 ttJf3 deserves consideration although Black can probably keep the better chances by 26 . . . "iVh3 with favourable simplifi­ cations. 26 "iVh3 27 "iVc6? ! "iVd7 28 "iVg2? ! .i.xd4 29 .ixd4 "iVxd4 White should have exchanged queens when he had the chance, although this wouldn't have spared him from a long and painful endgame. 30 ttJf3 Morphy had probably relied on this move, failing to notice Black' s reply: 30 "iVd5 31 h4 ttJe6 (D) Black has kept his extra pawns and went on to win a long ending, which, however, is some­ what away from our main subj ect. ..•

•••

.••

32 "iVg4 "iVc6 33 1Ig2 :td3 34 "iVf5 :ted8 35 "iVf6 "iVdS 36 "iVf5 :tdl 37 lhd l "iVxdl + 38 �h2 lId3 39 :tf2 :te3 40 ttJd2 lIe2 41 "iVxf7+ �h8 42 ttJe4 gxf2+ 43 ttJxf2 "iVd5 44 ttJg4 "iVxa2+ 45 �g3 "iVb3+ 46 �h2 "iVc2+ 47 'iitg 3 "iVc3+ 48 �h2 "iVc6 49 h5 as 50 ttJf6 gxf6 5 1 "iVxf6+ �g8 5 2 "iVg6+ �f8 5 3 "iVxh6+ �e8 54 "iVg6+ �d7 55 h6 "iVdS 56 h7 "iVxe5+ 57 �gl ttJg5 58 h8"iV "iVxh8 59 "iVxg5 "iVd4+ 60 �ft a4 61 "iVf5+ �c6 62 "iVc8 �b5 63 �e1 c5 64 "iVb7+ �c4 65 "iVf7+ �c3 66 "iVf3+ "iVd3 67 "iVf6+ �b3 68 "iVb6+ �c2 69 "iVa7 "iVc3+ 70 �e2 a3 71 "iVa4+ �b2 72 "iVb5+ "iVb3 0-1 The strong similarity between these two games is a clear hint that Morphy didn ' t really bring anything fundamentally new regarding the abstract matter of optimizing development. And yet, what could be the explanation of his categorical win over Anderssen? I believe that a careful analysis of the following game will give us some answers about the true balance of forces between these two outstanding figures of the 1 9th century.

Anderssen - M orphy Match (game 6), Paris 1 858 1 a3 To many observers and commentators, An­ derssen' s switch in the [mal part of the match to this unpretentious move might have seemed a sign of resignation. I suggest a completely different interpreta­ tion. It could simply be that Anderssen finally

THE SECRETS BEHIND MORPHY'S S UCCESSES

understood his opponent' s strong and weak points. Morphy had a colossal practical strength in positions where play had a predominantly concrete character, especially in the opening, but Anderssen ' s greater experience could have its say in more ambiguous situations, where the pieces don ' t come to direct contact too soon. The overall score in the games started with 1 a3 was equal (+ 1 = 1 - 1 ) but the course of the games is in strong contradiction with the ap­ proved thesis that Morphy' s play was based on a sounder positional foundation. In fact, i t was Anderssen who completely outplayed his brilliant opponent on strategic territory. As we shall see, Morphy managed to maintain the balance only by means of his greater practical abilities. 1 eS 2 c4 lDf6 3 lDc3 dS 4 cxdS lDxdS S e3 (D) •.•

B

The game has transposed into a reversed Paulsen Sicilian, but at the time of this game it was most probably called an 'irregular' open­ ing. True, one year earlier Louis Paulsen had experimented with the variation that nowadays bears his name, but he managed to give it mod­ em contours only some 20 years later. S .te6 6 lDn .td6 7 J..e 2 0-0 Morphy probably hoped to get attacking chances if White castled quickly, but some­ what neglects the centre by delaying the devel­ opment of his queen' s knight. S d4! lDxc3 ? ! This helps White t o strengthen his centre. 8 . exd4 is more logical.

139

9 bxc3 e4 1 0 tiJd2 (D)

B

Remarkably, the next two games in which Anderssen opened with 1 a3 reached the same structure. Morphy seems to have been unable to sense dangers of a strategic nature, while An­ derssen remained confident in his positional feeling, despite the unfavourable result of his first game with this extravagant experiment. I believe that time has come to mark the difference in handling the opening (and the game of chess in general) between these two historic figures . B oth Anderssen and Morphy were fervent adherents of the Evans Gambit. They both reached the following position quite frequently: 1 e4 e5 2 lD f3 lDc6 3 J.. c 4 .tc 5 4 b4 J.. x b4 5 c3 J.. c 5 6 0-0 d6 7 d4 exd4 8 cxd4 i. b6 (D).

w

...

.

.

However, their further treatment was dia­ metrically opposed. Anderssen generally relied

1 40

SECRETS OF A TTA CKING CHESS

on long-term compens ation, and many of his games continued with 9 d5 ltJ a5 1 0 J. b2 ltJ e7 1 1 iL d3 0-0 1 2 ltJc3 ltJ g6 1 3 ltJ e2 c5 14 'ifd2 f6 1 5 �h l �b8 1 6 ltJ g3 followed by ltJ f5, � g l , g4, etc. This i s what I would c all a strategic ap­ proach. Morphy preferred the logical developing move 9 ltJ c3 , when White maintains possibili­ ties of advancing either of his central pawns ac­ cording to the necessities. The main merit of this continuation is that in case of the most natural devel oping move, 9 . . . ltJf6, Black suddenly gets into trouble after 1 0 e5 ! dxe5 1 1 J. a3 followed by 1 2 'ifb 3 , as proved by Morphy on several oc­ casions. This is obviously a more dynamic ap­ proach than Anderssen's. Theory has registered the two variations with the names of their outstanding adherents from the 1 9th century. They both probably deserve the right to exist, just as the Evans Gambit in general, although this time I would probably prefer the Morphy variation. However, in the kind of position resulting af­ ter Anderssen's experimental 1 a3, Morphy' s approach was simply not too well suited. There are situations when one should play according to the objective necessities and forget about one' s preferences. 1 0... f5 (D)

w

Anderssen had no intention of p lacing his king under immediate fire and continues to re­ strict the enemy pieces ' activity with his pawns. 1 1 g5 (D) A rather premature attack. With his queen­ side undeveloped, Black's chances for an attack are not too great. Better chances for counter­ play are offered by opening the centre with 1 1 . . .exf3 although after 1 2 i.xf3 followed by l:.b l , 0-0 and c4, White' s chances should be preferred. In a later game from this match, Morphy var­ ied with 1 1 . . . 'ifh4+ 1 2 g3 'ifh3 but did not com­ pletely solve his strategic problems, although he eventually drew. ..•

w

12 i.c4 A remarkable multi-purpose move. The most obvious idea behind it is to open the queen' s ac­ cess to h5 and thus force Black to clarify mat­ ters on the kingside at an earlier stage than he would like . Another immediate threat after the exchange of bishops will be to transfer the queen to b3 in order to threaten both the king and the b7 -pawn. However, Anderssen also had a deeper plan in mind, probably overlooked by Morphy when playing his previous move. 12 iLxc4 1 3 ltJxc4 gxf4 Black had no time to block the kingside with 1 3 . . . g4 due to 1 4 'ifb3 . However, the exchange on f4 has a significant drawback, probably taken into account by Anderssen: the e3-square now becomes available to the white pieces . 1 4 exf4 'ife8 1 5 0-0 'ifc6 1 6 'ifb3 'ifd5 17 lIb1 b6? ! (D) •••

Black's past few moves have been a continu­ ation of the same one-sided strategy. Morphy didn't anticipate the long-term consequences of White' s advantage in the centre. 1 1 f4

THE SECRETS BEHIND MORPHY'S S UCCESSES

Another significant positional concession. Instead of further weakening his position, Black should complete his development with 1 7 . ..tiJd7, when winning a pawn with 1 8 liJxd6 'ii'x b3 1 9 :'xb3 cxd6 20 l:[xb7 does not bring White any concrete advantage due to 20 . . . liJb6 with excellent chances for a blockade on the light squares .

141

B

w

A well-timed pawn-break. 2S cxdS 26 cxdS 'ii'd 7 26 ... liJxd5 i s impos sible in view of 27 �h5+ �f8 28 liJxf5 with a powerful attack. 27 liJc4 I believe that this moment has not received the attention it deserves. For instance, in Paul Morphy 's Games of Chess, Maroczy attaches an exclamation mark to the knight j ump, with­ out any further comment. This move marks the start of a new stage of the game, of a more dy­ namic nature. The knight was very stable on e3, blockading Black's passed pawn and ensuring White a virtually risk-free game, but Anderssen must have been frustrated by the fact that it blocked the rook' s access to the kingside. For now, I shall refrain from giving a definitive evaluation to 27 liJc4 and limit myself to ob­ serving that it leads to a type of position where Morphy' s talent finally gets the opportunity to manifest itself. Besides, the e4-pawn suddenly becomes a source of counterpl ay. From a psychological point of view, 27 'iVel looks like a wiser decision. White prepares the activation of his queen and keeps the b3-rook in reserve for the time being. Play may continue 27 ... l:g6 28 Vi'h4 h6 29 'i!Vh3, when the f5-pawn is hopelessly weak. The attempt to put pressure on the d5-pawn with 29 . . . ..tf8 fails to 30 J.e5 followed by d6, which renews the threat against the f5-pawn. B lack ' s poor coordination is un­ derlined by the fact that the natural 30 . . . ttJg4 fail s painfully to 3 1 liJxg4 followed by 32 Vi'h5 , irrespective of the way Black recaptures on g4. 27 r:i;; e 7? •••

1 8 'iVa2! The queen has done its job on the queenside and is heading for e2. The main plan is to place the knight on e3 and then create threats on both wings, but preserving the queen from exchange increases the attacking chances. 1 8 c6 Once again, 1 8 . . .liJd7 100ks more natural. 1 9 'iVe2 liJd7 20 liJe3 'ii'e 6 21 c4 ttJf6 22 :b3 r:J;f7 Facing the prospect of ..tb2 followed by an eventual check on g3, the black king abandons its residence in advance. However, with the centre unstable, f7 will not prove a much better location. 23 SLb2 liac8 24 �hl (D) I believe that Anderssen' s treatment of the position has a rather modem character. All his pieces are in optimal positions, not obstructing one another, and the signal for the attack i s likely to be heard any moment now. Both d5 and g4 are in the air. 24 l:g8 Hoping to discourage an attack along the g­ file. 2S dS ! •••

•••

•••

1 42

SECRETS OF A TTA CKING CHESS

Black's position certainly l ooks dangerous and Morphy' s desire to avoid an eventual fork on e 5 i s perfectly understandable. The capture of the d-pawn with 27 . . . liJxd5? is strongly an­ swered with 28 �b5 ! (a nice way of activating the rook; Maroczy gives 28 lid l liJ xf4 29 l:.xd6 liJxe2 30 lixd7 + e6 3 1 l:.d2 as winning, but after 3 1 . . . l:. gd8 Black maintains reasonable drawing chances because of his active pieces and the reduced number of remaining white pawns) 2 8 . . . liJ e7 (28 . . . l:. c5 allows White to sta­ bilize his pressure by 29 l:.xc5 bxc5 30 .i.e5) 29 "iYh5+ lig6 and now 30 l:.xf5 + ! liJxf5 3 1 'ifxh7+ yields a decisive attack. However, Morphy had at his disposal a rather effective way of taking advantage of White' s temporary lack of coordination induced by the knight j ump, namely 27 . . . 'ifc7 ! ? (D), not no­ ticed by commentators until now.

w

Black attacks both the knight and the f4pawn. White has a relatively wide choice of continuations, but none of them seems to offer anything concrete. For instance: a) 2 8 liJe5+ i s the most natural. In this case, B lack is balancing on the edge of the precipice but there is apparently no immediate way to make him fall down. The knight certainly looks active on e5, but it obstructs the b2-bishop, thus limiting White' s attacking potential . For the moment, it is essential to choose the right square for the king. After 28 . . . f8 ! (28 . . . e8? is bad for a far from obvious reason: after 29 l:.c 3 ! iL c5 3 0 li h3 Black cannot prepare his counter­ play along the c-file with 30 . . . .i. e7 because of

3 1 l:. xh7 ! with a winning attack) White faces the unpleasant threat of . . .'ifc2. I cannot find anything better than 29 'ife3 (now after 29 lic3 .i. c5 30 l:. h3 iL e7 White has no decisive con­ tinuation available) 29 ... 'ifc2 30 'ifh 3 e8 ! ? (again, the position of the king i s essential; af­ ter the hurried 30 . . . 'ife2? White wins with 3 1 liJ d7+ ! liJ xd7 3 2 'ifxf5+ e8 3 3 'ife6+) . Not having any forced winning continuation, White has problems coordinating his pieces. For in­ stance, 3 1 .i. d4 'ife2 32 l:.bb l and now it looks like B lack should question White' s stability in the centre with 32 ... .i. xe5 (with the bishop on d4, 32 ... lic2? loses to 33 lif2) 33 fxe5 (33 .ltxe5 allows 33 . . . l:.c 2 with counterplay) 33 . . . fud5 34 :'f2 'ifg4 35 'ifb 3 l:.d8 and Black seems to be able to hold his position together. b) The simplistic 28 liJxd6+ 'ifxd6 29 l:. h3 allows Black to coordinate his force s with 29 . . . l:.g7, when after 3 0 li d l 'ifxf4 3 1 d6 'iVg4 32 'ifn White c ertainly has enough compensa­ tion for the pawn, but probably not more. c) 28 l:!c 3 is possibly the trickiest continua­ tion, when Black must defend against attacks from all sides . Play may c ontinue 28 . . . liJxd5 29 'if h5+ e6 30 lid l (30 liJ xd6 'ifxd6 3 1 l:. xc8 lixc8 3 2 'ifxh7 wins a pawn, but after 3 2 ... liJ e3 Black's centralized position and the strong passed pawn offer him adequate counterplay) 30 . . . liJ xf4 ! (D) (only this move seems to solve the problem of the hanging pieces; 30 . . . .i c5 ? allows 3 1 lixd5 ! xd5 3 2 'ifxf5+ with a deci­ sive attack; for instance, 3 2 ... c6 33 'ifxe4+ b5 3 4 lib3+ a6 35 'ifd3 and the black king is helpless) .

w

THE SECRETS BEHIND MORPHY'S SUCCESSES

White now has two ways of winning a piece, but not necessarily the game. c 1 ) After 3 1 iVh6+ lI g6 32 Itxd6+ 'iVxd6 3 3 ctJxd6 Itxh6 3 4 ctJxc8 lIg6 3S ctJxa7 (or 3S g3 ctJd3) 3 S ... lIxg2 the reduced number of white pawns as well as Black's activity makes a draw quite probable. c2) At first sight, the forced variation 3 1 lIxd6+ 'iVxd6 3 2 'iVxfS+ h2 is no remedy in view of 23 . . . �e5+ with a decisive attack. b) While 20 �g5 was our ' human' defence for White, we later discovered that the c om­ puter suggests (among others, which tend to transpose) 20 r;i;>h l . I have just played a few moves against the silicon monster and here i s where we g o t : 20 . . . iVh4 2 1 iVd2 ( aiming to trade queens with iVg5 if allowed and thus putting pressure on B lack to do something quickly) 2 1 . . . .tg6 (clearing the f-file for the rook) 22 tiJg3 i.e5 2 3 11g 1 I1f5 (with the brutal threat of 24 . . . iVxh2+ 25 r;i;>xh2 l:h5#) 24 tiJn l1h5 25 I1g3 l::t e 8 and B lack can win his mate­ rial b ack any time he wants with . . . .txg3 (thus reaching approximate material equality with rook and pawn versus two minor pieces) but he can continue to increase the pressure for a while. Needles s to s ay. the computer is not s o delighted with White' s position any more. Only ten minutes had passed and we had al­ ready discovered two maj or flaws in Garry ' s

A CHESS-PLA YER 'S BEST FRIEND ?!

analysis. Needless to say, we were anxious to see more of it. . . The game continued: 1 6 g4 (D) ...

1 81

B

w

be separated from his computer, G arry would agree with us. The reader c an relax now: the worst i s over. I shall only mention some further minor points. 17 fHhl Garry agree s with Petrosian that 1 7 fHd3 is better. In fact, thi s might be true , but we man­ aged to di scover an interesting point. After 1 7 . . sH5 1 8 ctJde4 exf4 1 9 f3 Black is not forced to settle for a more or less certain draw by 1 9 . gxf3 20 :xf3 i.xc3 but can fight for the initiative with 1 9 . . ..txe4 20 fxe4 i.e5 , when White' s pieces are rather p oorly placed. For instance, 2 1 :a2 fHh4 22 ctJb5 (or, simi­ l arly, 2 2 ctJd l f3) 2 2 . . . f3 with a d angerous at­ tack. Black plans a gradual increase in the pressure with . . . :f7 (or . . . bi.f6), . . . g8 25 ii.e4 ! ; 2 1 . . . \i'h6 i s very passive, allowing 22 'fic5 ! liJxc6 2 3 lhc7, which takes advantage of the

1 88

SECRErS OF A TTA CKING CHESS

fact that the e5-pawn is not attacked by the black queen any more) 22 e4 'ii'h 5 (Black has managed to block the white queen's access to f4, but White's attack isn't over yet) 23 11xc7+ �8 (23 . . . lbxc6? ! makes things easier for White after 24 'ii'x c6+ f8 25 l;lxf7+ xf7 26 'ii'c 7+ mating soon) 24 'ii'c 5+ g8 and now 25 lhf7 ! underlines the lack of coordination in Black' s army. H e appears t o have n o adequate defence.

on d5 : 20 lhd5 ! exd5 2 1 'ii'x d5 c6 22 'ii'd 6 and now both 22 . . . l:tc8 23 l:td l c5 24 e6 fxe6 25 ..tb2 lnI7 26 .ie4 and 22 . . . hxg3 23 e6 gxf2+ 24 n give White a decisive attack).

w

16 iNn In fact, this isn ' t such a bad square for the queen either. It puts permanent pressure on the f6- and f7 -squares , which can generate tactical explosions in any moment of carelessness. 1 6 ctJdS I7 l:Udl (D) •.•

B

Finally, White has managed to mobilize all his forces, which cannot be said about Black. In the long run, the plan of doubling rooks on the d-file, placing the bishop on c5 and driving the knight away with e4 should win material. B e­ sides, Black must constantly t ake into account the possibility of a sacrifice on d5.

17 aS ••.

Black aims to question the bishop ' s stability along the a3-f8 diagonal. In case of the ex­ pected move of the bishop to c 5 , Black c an acti­ vate his rook via a6-c6. It is hard to say whether the other way of winning space on a wing, 1 7 . . . g5, as suggested by Graf afterwards, i s better. Th e critical position arises after 1 8 h 3 g4 ! ? 1 9 hxg4 'ii'g 5 (D) (as pointed out b y Val­ entin S toica, Black does not have enough devel­ opment to play such a move as 1 9 . . . h4? ! - not facing any concrete threat, White can sacrifice

This is in fact one of the few moments when our post-mortem analysis was inaccurate. a) We thought that White should play 20 11d4, postponing the sacrifice until Black castles. However, after 20 ... 0-0-0 2 1 l;lxd5 we failed to understand the importance of the zwischenzug 2 1 . . .hxg4 ! , allowing Black to generate active counterplay along the h-file. For instance, 22 'ii'xf7 (as we shall see, it is useful to eliminate this pawn; in case of 22 'ii'd l exd5 23 'ii'x d5 c6 24 'ii'c 5 I'!h6 ! 25 'ii'b 6 'ii'x e5 26 'ii'x a6+ b8 27 11bl l:tdh8 White should play for a perpetual with 28 'ii'b 6+ c8 29 'ii'a6+ d8 30 'ii'a 8+ c7 3 1 'ii'a 7+ rather than wait until Black's at­ tack takes on decisive contours, for instance with . . . iH5-e4) 22 . . . exd5 23 'ii'x d5 c6 24 'ii'c 5 l:th6 25 e6 ! (the only way to deviate from the drawish course of the game) 25 . . . 'ii'x c 5 26 exd7+ l:txd7 27 .txc5 l:td2 28 a3 11a2 and Black seems to have adequate counterplay. B e­ sides, 20 . 11d8 i s an important alternative to 2 0 . . . 0-0-0, when White has problems continu­ ing his attack. b) The immediate 20 lhd5 ! (D) is more dan­ gerous for Black: b l ) During our post-mortem we overlooked that the zwischenzug 20 . hxg4? ! allows the ele­ gant 2 1 'ii'x f7+ ! xf7 22 lhd7+ , when the black king has no favourable squares available ; for instance, 22 . . . e8 23 i.c6 or 2 2 . . . g6 2 3 ..

..

A CHESS-PlA YER 'S BEST FRIEND ?!

189

B

B

..te4+ Wh6 24 Wg2. There i s a slight p aradox here: as an answer to . . . hxg4, the c apture on f7 with fUxf7 does not offer White chances for a significant advantage in line 'a' , where the pawn was undefended, but here, where the pawn i s covered by the king, it leads to a strong attack for White. b2) 20 . . exd5 2 1 'iix d5 lib8 (the only play­ able move, creating the p otential threat of clos­ ing the a3-f8 diagonal with a well-timed ... b4; other moves are simply wrong; for instance , 2 1 . . .lid8 ? 22 fUc5 hxg4 2 3 ..tc6 ! followed by

'ifxd5 i s not only good but also entirely neces­ s ary. With his queenside under pressure and the rook obviously misplaced on c6, B lack faces difficult problems. The immediate threat is 'ifxc6. B lack mus t play 2 1 . . .x:te6 (2 1 . . .l'1xc5 22 'ifxc5 'iie 7 is equally bad in view of 2 3 'ifxc 7 , when B lack cannot castle under favour­ able circumstances) but after 22 axb5 ..txb5 23 'ifb7 his position i sn ' t likely to resist for long.

.

lid l or e6, winning, or 2 1 . . .c6? 22 fUc5 lic8 23 lid l and Black ' s position c annot resist for too long) 22 'iid 4 ! (now 22 fUc5 is less effective be­ cause of 22 . . . b4 23 .i.xb4 lib5, when White' s attack loses p art o f its intensity) 2 2. . .11g8 (Black gives up the possibility of castling kingside too; other moves do not help: after 22 . . .b4 23 e6 0-0 24 i.xb4 lixb4 25 'iix d7 White has more than enough c ompensation for the exchange; in case of 22 ... 'iix g4 23 fUa7 lib6 24 'iix c7 'iie 6 25 lid l his pieces are rather uncoordinated) 23 fUa7 ! with a strong initiative for White.

18 ..tc5 kta6 19 a4! (D) This move i s necess ary: White has to ques­ tion B lack' s queenside stability.

1 9 ... b4 I mus t confe s s that this move took me com­ pletely by surprise and for a few seconds I feared that I had done something wrong in the previous phase of the game (a reversed psycho­ logical situation compared to the game Tal­ Geller). I had only taken into account 1 9 . . . .l:tc6, when the exchange s acrifice 20 lixd5 exd5 2 1

This whole idea of waiting for the enemy to reach the c6-square and only then sacrifice on d5 i s not new. I had seen it i n Karpov-Kas­ parov, World Ch (game 1 4), B aguio City 1 978 . The only significant difference i s that i n that game the whole operation was carried out in a simplified position, without queens on the board and without any mating threats. Meanwhile, the apparently more s olid 19 . . . c6 allows White to return to his initial plan of pres­ sure along the d-file with the simple 20 lid4. Now we return to 1 9 . . . b4 (D):

w

190

SECRETS OF A TTA CKING CHESS

20 .i.f1 ! Surprise for surprise ! I was lucky enough to find this move rather quickly; Graf had taken into account only 20 mc4 .1£.xa4 with c ounter­ play. By this time, we b oth were experiencing some time-pressure. The psychological tension had reached its maximum. 20 �c6 Under the slightly changed circumstances, this move i s playable now. On the other hand, 20 . . . .1£. xa4 fail s to 2 1 .1£. xc4 il.xdl (2 L . lI c6 22 �xdS exdS 23 lIxdS looks equally bad for Black) 22 l1 xd l �a8 (aiming to defend the rook and the c6-square at the same time) 23 i.b5+ c6 24 lixd5 ! exdS 25 �f6 with a decisive attack. 21 �xc4 (D) •••

B

2 1 g5! ? Black's queenside play has reached a dead­ end and Graf decided to open a new front of ac­ tion, according to his general strategy. B e sides winning space, this move aims to remove the white queen from f3. Alternatives : a) While m y opponent was thinking , I en­ j oyed the prospect of the variation 2 L . �g5 22 e4 'De7 and now 23 �f6(?) 'ifxf6 24 exf6 'D c8 25 eS with what I considered to be complete pa­ ralysis for Black, failing to notice that after 2S . . . 'Db6 White has problems maintaining his domination. In fact, the simple 2 3 'ifd3 i s much stronger, and wins immediately. b) I was slightly alarmed about the possibil­ ity of 2 L .fS . Now, the consequences of the ex­ change s acrifice on dS are not entirel y c lear ...

since the enemy king has the f7 -square at its disposal, but it s eems that the simple 22 exf6 �xf6 23 'ife4 maintains a stable advantage; for instance , 2 3 . . . 'D c 3 (23 . c;.tf7 loses material to 24 � d4 although 24 �xdS , c ontinuing the at­ tack, might be even stronger) 24 ': xc3 bxc3 2S � d4 e5 (the only way of prolonging the resis­ tance) 26 .i.xeS 'ife6 27 .i.b S (or 27 ..t h3). c) In case of a neutral move such as 2 L .h4 White plays 22 e4. For instance, 22 . . . 'Dc3 (other moves are equally bad: 22 . . . 'Db6 23 ..t xb6 �xb6 24 lI cd4 winning the bishop ; or 22 . . . 'De7 23 iL xe7 �xe7 24 1Ixd7 ! 'ifxd7 { or 24 . . . c;.txd7 2S �d3+ } 25 ':xc6 'ifxc6? 26 ..tb5, although I suspect that other less spectacular ways are possible as well). Now my initially planned combination 23 ':xd7 ? ! 'ifxd7 24 lI d4 is an­ swered by 24 . . . lIxcS ! 2S lhd7 'iitxd7 26 'ifxf7+ 'iit c8 27 �xe6+ 'iitb8, when suddenly Black has achieved perfect coordination, in spite of his material disadvantage. The position is rather complicated and a draw is the most probable re­ sult. Since in that line Black's main trump is his strong queenside set-up, I decided to investi­ gate the consequences of immediately disman­ tling it by 23 lI xc3 ! bxc3 24 'ifxc3 . Black's p osition i s then as hopeless as it looks ; for in­ stance, 24 ... hxg3 2S hxg3 ! (the attempt to keep the h-file defended with 2S fxg3 ? has the draw­ back of weakening the g l -a7 diagonal, which allows Black to save the game with 2S . . J 1xc5 26 �xcS ifb8 ! 27 �x a5 �b6+ 28 �xb6 cxb6 29 ..te2 'iit e 7 3 0 lI d6 ..t xa4 3 1 l1 xb6, when White' s chances are only symbolically better) 2S . . . "fie7 (this looks like the most natural way of meeting the terrible threat of ..tbS ; the COUll­ terplay initiated by 2S . . . 'ifgS fails to the simple 26 �b l ! , when White' s attack quickly breaks through) 26 .i. xe7 ! ':xc3 27 ..t f6 11h7 and now White wins by the highly aesthetic 28 .1£. a6 ! ! with the terrible threat of .1£. c8. 22 h3 The correct decision under time-pre ssure. Allowing the queen to be driven away from f3 is a mistake; for instance, 22 e4 g4 23 ifb 3 (for a little while, I thought that 23 exdS would win outright, but then I noticed that 2 3 . . . l1 xcS ! leaves B lack out of any danger) 2 3 . . . 'D c3 (in case o f 2 3 . . . 'D e7? White wins material by 24 ..

A CHESS-PlAYER 'S BEST FRIEND ?!

J. xe7 'ilxe7 25 lI xd7 'ilxd7 26 1hc6) 24 lIxc3 bxc 3 25 'ilxc 3 'ile7 ! and things are not entirely clear. 22 ... g4 Practically forced in view of the renewed threat of e4. 23 hxg4 'ilg5 (D)

1 91

However, a t some moment Graf noticed that af­ ter 27 %i xd7 W xd7 White has the elegant and not very obvious (for a human at least) retreat 28 'ild l + ! , winning the enemy queen.

w

w

Everything i s now ready for decisive action by White. 24 e4! 1tJc3 (D) This loses outright. During the post-morter.!, Graf managed to find other defences, which, al­ though not better from an objective point of view, would have confronted White with more practical problems. Inserting 24 . . . hxg4 cannot be recommended because of 25 'ild3, creating an unpleasant pin, but Black can try 25 . . . 'ilxe5 (25 . . . lIxc5 26 :xc5 ltJb6 loses to 27 :' xc7 'ilh5 28 J. g2, when White' s attack is stronger than Black's). Grat's idea i s revealed in the variation 26 exd5 'ilxd5 27 'ilxd5 exd5 28 l,I xd5 .u. ch6, when things are not entirely clear, but 26 � d4 i s stronger, with decisive material gains . We also needed some time to refute 24 . . . ltJb6. During the game I thought that 25 �xb6 l:txc4 26 �e 3 would be the end of the story, failing to notice that Black can fight on with 26 . . . lI c 3 .

25 .u.xd7! Wxd7 26 'ilxf7 + Wc8 27 J.e3 Once again, this retreat wins material. By in­ ertia, or possibly in view of my slight time­ trouble, Graf gave up some more pieces before resigning. 27 :flxe5 28 :xc6 bxg4 29 �f4 1tJe2+ 30 ..i.xe2 'ilal + 31 lIct Avoiding the last trick: 3 1 � f1 ? � h l + with a draw by perpetual. 1-0 ••

It wasn ' t just due to a lack of modesty that I chose a game of my own to finish the book. S ince my aim was to present a hand-made anal­ ysis, I felt that I could only rely on my own ex­ perience in order to make the whole comments 1 00% sincere. The final conclusion is that there are still cer­ tain Secrets ofAttacking Chess that are accessi­ ble only to humans. Those players who have the courage to trust their own feelings, knowl­ edge and power of work will no doubt have an advantage over their rivals addicted to the help of the machine.

I n d ex of P l aye rs Numbers refer to pages. When a player's name appears in bold, that player had White. Otherwise the FIRST-NAMED PLAYER had White. AHUES

- Alekhine 4 3 ALEKHINE Ahues 43 ; Chajes 1 27 ALEKSANDROV - Marin 8 1 ANDERSSEN - Mo rph y 1 33 , 1 38 ; Morphy 1 3 5 BARCZA - Keres 49 B AREEV Beliavsky 1 1 2 B ELIAVSKY - Bareev 1 1 2 ; Yu sup ov 1 1 7 CHAJES Alekhine 1 27 GELLER - Keres 1 6, 9 3 ; Tal 22; Velimirovic 9 GLIGORIC - Petrosian 1 79 GRAF - Marin 1 85 HORVATH, J. - Marin 3 2 , 59 INKIOV - Marin 72 IVKOV Korchnoi 5 3 KERES - Barcza 4 9 ; Geller 1 6, 93 KORCHNOI - Ivkov 5 3 ; Thl 1 52, 1 60, 1 64, 1 7 3 KRAMNIK - Leko 1 82 KRASENKOw Marin 44 LEKO - Kramnik 1 82 -

-

-

-

LpUTIAN -

Marin 1 00 LUND - Nirnzowitsch 3 4 MARIN - Aleksandrov 8 1 ; Graf 1 85 ; Horvath, J. 32, 59; Inkiov 72; Krasenkow 44; Lputian 1 00 ; Narciso 1 23 MIESES - Spielmann 62 MORPHY - Anderssen 1 33 , 1 3 8 ; Anderssen 1 35 NARCISO - Marin 1 23 NIMZOWITSCH - Lund 3 4 PETROSIAN - Gligoric 1 79 SPIELMANN - Mieses 62 TAL Geller 22; Korchnoi 1 52, 1 60, 1 64 , 1 73 VELIMIROVIC - Geller 9 YUSUPOV - Beliavsky 1 1 7 -

Co m poser

-

Lommer 5 2

I nd ex of O pe n i ngs Numbers refer to pages. Codes are ECO codes. Catalan E04 1 85; E09 32

King's Pawn C 3 3 133 ; C45 62; C 5 1 139; C52 135

Dutch A90 1 12 , 1 1 7

Modern Benoni A6 1 100

EnglishIFlank Opening A l l 44; A20 138

Nirnzo-Indian E20 72; E5 2 93

French C07 164

Queen's Pawn A46 81

King's Indian E66 9; E97 1 79

Ruy Lopez (Spanish) C80 1 73 ; C83 1 52 , 1 60; C89 1 82 ; C96 22; C99 1 6