ENDGAME

STRATEGY1

M. I. SHERESHEVSKY
ENDGAME STRATEGY
By
M. I. SHERESHEVSKY
Translated by
K. P. NEAT

CADOGAN CHESS
LONDON, NEW YORK
 Cadogan Books
Distribution
UK/EUROPE/AUSTRALASIA/ASIA/AFRICA
Distribution: Grantham Book Services Ltd, Isaac Newton Way,
Alma Park Industrial Estate, Grantham, Lines NG31 9SD.
Tel: 0476 67421; Fax: 0476 590223.

USA/CANADA/LATIN AMERICA/JAPAN
Distribution: Macmillan Distribution Center, Front Sc Brown
Streets, Riverside, New Jersey 08075, USA.
Tel: (609) 461 6500; Fax: (609) 764 9122.

English translation copyright © 1985 K. P. Neat

All Rights Reserved. No port c f this publication may be

reproduced, stored in a retrieval system or transmitted in any
form or by any means: electronic, electrostatic, magnetic tape,
mechanical, photocopying, recording or otherwise, without
permission in writing from the publishers.

First published 1985 by Pergamon Press

Reprinted 1988
First Cadogan Books edition 1994

British Library Cataloguing In Publication Data

A catalogue record for this book is available from the British
Library.

ISBN 1 85744 063 3

This is a translation o f the Russian edition, published in 1981 by

Polimya, Minsk.

Published by Cadogan Books pic, Letts House, Parkgale Road,

London SW11 4NQ
Cover design by Brian Robins
Printed and bound in Great Britain by BPCC Wheatons, Exeter
 CONTENTS

Foreword to the English edition (A. Yusupov) . . . vii

Introduction ............................................................. . LX

Basic principles of endgame play (S. Byelavyenets) 1

Centralization of the king ....................................... 4

The role of pawns in the endgame .......................... 12

The problem of exchanging ..................................... 19

“Do not hurry” ......................................................... 41

Schematic thinking .................................................. 55

The principle of two weaknesses ............................. 67

The struggle for the initiative ................................. 80

Suppressing the opponent's counter-play ............. 96

Positions with an isolated d-pawn .......................... 108

The two bishops ....................................................... 121

The 3—2 Queen-side pawn majority ...................... 143

Complex endings ....................................................... 152

Index of players ....................................................... 215

Index of m aterial....................................................... 217

 FOREWORD TO THE ENGLISH EDITION

M. Shereshcvsky’s book Endgame Strategy was published in the USSR in 1981

in an edition of 50,000 and was immediately sold out.
The author, one o f the strongest players from Byelorussia, has worked extensively as
a trainer. In his lessons with young players Shereshevsky has made use o f the end game
teaching methods o f one o f the country’s leading trainers, M. Dvoryetsky, and has
worked out a definite system.
This book contains an interesting selection of endings. Along with classic ex­
amples there are endings both of leading modern grandmasters, as well as of less well
known players. The author aims in the first instance to explain the course of the
struggle, penetrate into the psychology of the players’ actions, and to focus the
readers’ attention on the turning points and characteristic mistakes. In contrast to
the majority of works on the endgame, the book is divided into chapters not accord­
ing to material, but according to the playing methods which are most characteristic
of the given group of endings. The names of certain chapters have an unusual ring:
“Do not hurTy", “The problem of exchanging” . “The principle of two weaknesses”,
and so on. For this English edition the author has added a number of endings played
in recent times, as well as endings from games by the strongest English players.
The main value of the book, in my opinion, lies in the fact that it contains
specific advice and recommendations on how to improve endgame technique, for
which the practical player will sometimes search in vain when studying multi-volume
reference books on the endgame.

A. Yusupov
International Grandmaster
 INTRODUCTION

From the practical point of view, the endgame is the least well studied stage of
chess. Chess literature contains very few works on the endgame, and in the main
these are reference works, in which theoretical and not practical positions are
analyzed.
The present book is an attempt to study and systemize certain basic practical
principles of the playing of chess endings.
The necessity for a systematic approach to the study of chess endings occurred
to me mainly as a result of my teaching experience. It is no secret that, in the pre­
paration of young players, many trainers and teachers devote most attention to
the study of numerous opening systems and the forms of middlegamc resulting
from them. The endgame is always allotted very litde time.
Some trainers give their pupils the most elementary conceptions of the end­
game, assuming that with the general development of a player his mastery of end­
game play will also rise. Others demonstrate long and complex analyses from refer­
ence books, although the probability of such positions being repeated in a practical
game is slight. It is evident that both approaches are a long way from the truth: the
mastery of a player is directly dependent not so much upon his amount of theo­
retical knowledge, as upon his understanding of the general principles of conducting
chess endings.
In 1976 I happened to be the second of international master Mark Dvoryetsky
during the USSR Championship 1st League in Minsk. Dvoryetsky adjourned his
game with grandmaster Taimanov in a superior position. In one of the lines of
analysis a rook ending with f- and h-pawns was reached Dvoryetsky referred to a
book on rook endings, and began studying the appropriate chapter. I was surprised:
after all, Dvoryetsky is a great expert on the endgame. To my question he replied
that he knew the basic principles of playing such endings, but did not even attempt
to remember lengthy concrete analyses. Later during the tournament we frequently
discussed the question of how to study the endgame. Dvoryetsky considers it
essential to know the classics, to analyze complicated practical rather than theo­
retical endings, and to find general rules and principles of play in complex endings.
And in theoretical endings it is sufficient to know whether the ending is won or
drawn, and to have a rough impression of the plan of play.
Of course, every trainer has his own style of working, and his own system for
preparing players. But it is worth recalling that Mark Dvoryetsky, an Honoured
Trainer of the Russian Federal Republic, has prepared three Junior World Cham­
pions. And all three - Valery Chekhov, Artur Yusupov and Sergey Dolmatov - are
very strong in the endgame.
O f course, the role o f exact knowledge in the endgame should not be under­
estimated. The five-volume series Comprehensive Chess Endings, edited by grandmaster
Yuri Averbakh, is the most detailed reference work on this aspect of the game. And

IX
„ INTRODUCTION \

even so, in the introduction to this series it is emphasized that a sure indication of a
strong player is good playing technique in complex endings.
The present book studies such basic principles of play in complex endings as
centralization of the king, schematic thinking, prophylaxis, and the principles of
“do not hurry” and of two weaknesses. The majority of these were formulated with
amazing precision and conciseness in an article by a talented Soviet master who was
killed during the Second World War, Sergey Byelavyenets, an extract from which is
given after this introduction. Also examined are typical endgame positions with the
advantage of two bishops, an isolated d-pawn, and a 3—2 Q-side pawn majority.
Some examples are given to study the problem of exchanging, and ways of battling
for the initiative in the endgame. In conclusion we give a number of complex end­
ings, in which the various principles expounded in the previous chapters are put into
practice.
The knowledge of many rules, and the choice of a specific plan based on them, is
mainly of a psychological nature. Therefore in certain examples, especially where
Byelorussian players are involved, I have laid particular emphasis on the competitive
situation in which the game was played.
Of course, the rules and recommendations given in the book cannot be regarded
as unshakeable and universal endgame laws. Chess is too complex and diverse for
that. Latent or manifest in each position are its rules, principles and regularities,
many of which a player will often sense intuitively.
Without pretending to offer universal recommendations, the author has aimed
mainly to help players to be better oriented in endings, and to be more correct and
accurate in taking the necessary decisions in practical play.
 CHAPTER 1

BASIC PRINCIPLES OF ENDGAME PLAY

S. BYELAVYENETS

During the fierce battles of the middle- things are different. Only in rare, so-
game, passions arc aroused. Sacrifices called combinational endings, must the
and striking combinations are in the air, attention be focused on calculation, on
and each of the players watches intense­ tactics. In the overwhelming majority of
ly for tactical blows, clever traps, and endings it is essential to think in terms of
subtle unexpected moves. Then sudden­ plans. Variations play a secondary role.
ly, mass exchanges take place, the heated The main role belongs to schematic
combinational skirmishes come to an thinking, and the possibility of setting
end, and a prosaic endgame ensues. up this or that position is checked by
Sometimes the transition into the end­ calculating variations. We plan the de­
game occurs at the will of one of the ployment of our own pieces that we re­
players, who assumes that here it will quire, taking account, o f course, o f what
be easier to exploit his advantage. the opponent may do. Then we check by
In the endgame, technique becomes of a calculation of variations whether it is
primary importance. First of all a player possible to achieve this position. For
must retune his thinking and his mood. example: in the following position from
One can virtually forget about‘brilliancy’ the game Capablanca—Ragozin, Moscow,
and tactics. I would advise every player, 1936, White formulated his tasks very
if time on his clock permits, to spend concisely.
several minutes on ‘calming the passions
aroused in him’. Subsequently this loss
of time will without fail be justified,
since the player will be examining the
position correctly, from the ‘endgame’
point of view. What does this mean?
This question must be dwelt on in
some detail, since multi-tome endgame
books, with their countless examples
and positions, do not devote sufficient
attention to the course of a player’s
thinking in the endgame.
In the middlegame his thoughts are
mainly occupied by the calculation of
variations, which are subordinate to Capablanca writes about his thoughts
some aim. The main things that a player during the game:
is occupied with in the middlegame are “White’s plan is to prevent the ad­
the checking of all kinds of tactical vance of the c-pawn (after which the b-
blows, and the calculation of combina­ pawn could become weak) and to con­
tions and variations. In the endgame trol the entire board up to the fifth rank.

1
 Endgame Strategy

This is achieved by moving the king to ness, sometimes bordering on tedium,

e3, and by placing the rook at c3, the they realize an advantage.
knight at d4, and the pawns at b4 and The repetition of moves in the end­
f4. After he has attained such a position. game plays an important role. Disregard­
White will be able to advance his Q-side ing the fact that it gains time for think­
pawns.” ing, it can be mentioned that, by repeat­
As we see, variations did not interest ing moves, the active side acquires
Capablanca, and he was not even inte­ certain psychological gains. The defender,
rested in the time and speed with which whose position is inferior, often cannot
the planned position would be attained. stand it, and creates a further weakening
The main thing was that the required which eases his opponent’s task. In addi­
type of position had been selected, and tion, repeating moves enables the posi­
the subsequent play followed according tion to be clarified to the maximum ex­
to plan. tent. We know that many upholders of
‘pure’ chess will severely criticize us for
1 Nd4 Rb7 this advice. But we cannot refrain from
2 b4 Bd7 advising players: you should sometimes
3 f4 Ke7 repeat moves in the endgame! In the
4 Kf2 Ra7 struggle every chance has to be exploit­
5 Rc3 Kd6 ed, and there is nothing ugly or unethical
6 Rd3 Ke7 in repeating moves.
7 Ke3 Ra4
8 Rc3 Kd6 11 Ne2 g6
12 Rd3+ Kc6
The set-up planned by White is com­ 13 Kd4 Ra6
plete. He is now faced with a new 14 Re3+ Kdé
problem — that of advancing his Q-side 15 Nc3 f5
pawns. To do this he must first take his 16 b5
king to the aid of the pawns which are
to be advanced. The pawns begin their advance and
the white pieces are able to support
9 Rd3 Kc7 them. It should be noted that Capa­
10 Rc3 Kd6 blanca began advancing his pawns only
when his pieces had occupied their
A few words about repeating moves. strongest possible positions. Of course,
A basic rule of the endgame is: do not now Black cannot capture on a3 due to
hurry! If there is a possibility of advanc­ 17 Ne4+.
ing a pawn two squares or one, advance
it first one square, look carefully around, 16 . . . Ra8
and only then advance it a further 17 Kc4 Be6+
square. Of course, you should not hurry 18 Kb4 c5+
in quiet positions, whereas in combina­ 19 bxc6 Bg8
tional endings things are different. To 20 Nb5+ Kxc6
many the rule of “do not hurry” may 21 Rd3
seem paradoxical, but in fact it is seen in
practically all the endings of games by White can now manage without the
great masters of the endgame. Look care­ advance of his a-pawn. Black’s K-side
fully at the endings o f Capablanca and pawns are weakened, and one of them
Flohr, and you will see with what slow­ falls.

2
 Basic Principles of Endgame Play

21 ... g5 associated with . .. f3.

22 Rd6+ Kb7
23 fxg5 hxg5
24 Rf< Rf8 26 ... Rc8
25 Rxg5 f4 27 Rg7+ Kb6
26 Nd4 28 Rg6+ Kb7
29 Nb5 Rf8
A very important move. Capablanca 30 Nd6+ Kb8
deprives his opponent of any chance 31 h4 Resigns.
 CHAPTER 2

CENTRALIZATION OF THE KING

In the transition of a game into the

endgame it is the role of the king that
changes most of all. In the opening and
the middlegame the king endeavours to
find a safe shelter and to observe the
battle from afar, without as a rule taking
any direct part in it, but in the endgame,
when the probability of a mating attack
is greatly reduced, the king is transform­
ed into an active fighting unit. Often an
experienced player, anticipating the
transition into an ending, will in advance
bring his king closer to the centre, so as
then to obtain a playing advantage. position too there are vulnerable points:
Centralization of the king is one of his pawn formation is not flawless. Were
the main principles of endgame play, his rook’s pawn at h2, all would be
' but, paradoxically, many players some­ dear. As grandmaster Bondarevsky aptly
times disregard it. Centralization of the put it, the white pawn at h4 is that hook
king in the endgame is hardly ever in­ which, by holding on to. Black creates
correct. It can only be inopportune. counter-play. This is a classic ending and
Very often, when one side has several it has been annotated by many authors,
apparently equally good plans, the but the most correct and accurate ana­
correct one will be that in which the lysis is that given by Bondarevsky in
main factor is the centralization of the Sbakbmatny Bulletin, 1973 No. 1.
king. We can therefore advise the reader:
if the game has gone into an endgame 1 ... g5!
and you are considering what plan to 2 hxg5
adopt, never forget about the king!
White is forced to fix the K-side pawn
structure, otherwise Black, by exchang­
Capablanca—Reshevsky ing on h4, will give White a weak pawn.
2 h5 came into consideration.
Nottingham, 1936
2 ... fxg5
(See next diagram) 3 Nb4

White stands better. Black, apart from After 3 f4 gxf4 4 exf4 d4! Black
his isolated pawn, has a whole complex would in time lose his d-pawn, but
of weak dark squares on the Q-side, would activate his bishop and have good
especially c5. The winning procedure drawing chances in view of the limited
seems fairly straightforward: it should be material remaining.
sufficient to transfer the king to d4 and
the knight to c3, when Black will practi­ 3 ... axb3
cally be in zugzwang. But in White's 4 axb3 Bb7

4
 Centralization o f the King

5 g4 Here 14 Nc7 no longer works: 14

. . . Bxf5! 15 Kxf5 Kf3.

14 . . . Be8
15 Kd47

A mistake, which was noticed only by

Bondarevsky. The grandmaster shows
that White could have won by 15 f6!
Kf3 16 N f4g3 17 Kf5 Bd7+ (other-
wise e 4 -e 5 -e 6 ) 18 Kg5 Bc6 19
Nxe6!!g2 20 f7gl=Q + 21 Kf6.

15 . . . Kf3
16 e5 g3
The position has clarified. Black’s 17 Ne3
only counter-chance is . . . h5. There are
two fundamentally different ways of
preparing this advance. The first, which
occurred in the game, is to prepare the
advance of the rook’s pawn using the
king. The second is to centralize the king
to defend the weaknesses and to support
. . . hS with the bishop.

5 ... Kg7
6 Ke2 Kg6
7 Kd3 h5
8 gxh5+ Kxh5
9 Kd4 Kh4
10 Nxd5 Kg3 17 . . . Kf4?

Black’s idea becomes clear — he has The decisive mistake in a drawn posi­
counter-play on the K-side. tion. Alekhine showed that 17 . . . Bd7
would have lost to 18 e6 Bc8 19 e7
11 f4 g4 Bd7 20 f6B e8 21 Nf5, but Bonda­
revsky found a draw by 17 . . . Bf7!!
Black loses after 11 . . . Bxd5 12 18 e6 Bg8 19 e7 (or 19 b4 Kf4!)
Kxd5 g4 13 f5K h3 1 4 f6 g 3 15 f7 19 . . . Bf7 20 f6K f4!, when White is
g2 16 f8=Q gl=Q 17 Qh8+ Kg2 18 unable to improve his position. By the
Qg8+. brilliant manoeuvre 17 . . . Bf7!! Black
gains a tempo for playing his king to f4.
12 f5 Bc8 But now the rest of the game is clear
without any explanation.
Not 12 . . . Bc6 13 Nc7. 18 e6 g2
19 Nxg2+ Kxf5
13 Ke5 Bd7 20 Kd5 Kg4
14 e4 21 Ne3+ Kf4
 Endgame Strategy

22 Kd4 8 Nbl Rdl 9 g7 Rgl Black wins.

In the game Black played
Black resigned, since there is no de­
fence against 23 e7 followed by the 1 ... Rh8
transfer of the knight to c7.
We will now analyze the second possi­ with the threat of giving check and
ble plan: 5 . .. Kf7 6 Ke2 Kf6 7 Kd3 winning the knight.
Ke5. Now, as Bondarevsky pointed out,
White has to reckon with the possibility 2 Ne3
of . . . d4 followed by . . . Kf4, there­
fore: 8 Nc2 Bc6 9 Nd4 Be8 10 Kc3 Forced, since 2 Nc3 merely strength­
h5 11 gxh5 Bxh5 12 Kb4 g4 13 ens the effect of . . . aS.
fxg4 Bxg4 14 Kxb5 Ke4 IS KcS
Kxe3, and Black gains a draw. 2 ... Re8+
Of course, in a practical game it would 3 Kd4 Rd8+
be impossible to calculate all the above 4 Ke4!
variations. Therefore it is the general
assessment of the position which One of the most difficult moves in
assumes primary importance. Had this ending. It would seem that Emanuel
Reshevsky chosen the second path, he Lasker evaluated the position intuitively.
would have gained a draw. Centralization After 4 Kc3 Black does not have a
of the king does not always give the de­ forced win, but the probability of defeat
sired result, but in the majority of cases for White would be considerable. Black,
it is necessary. for instance, could play 4 . . . Rd6. If
now the knight moves from e3, the black
rook reaches d l, while if the white
Em. Lasker—Ed. Lasker pawns advance the black king transfers
behind the back o f the rode to e7. The
New York, 1924 main drawback to White’s position after
4 Kc3 is that his king is cut off from
the K-side, and he therefore has no real
counter-play.
Emanuel Lasker chooses a plan which,
though risky, is the only correct one.
He keeps his king in the centre, from
where it can easily be switched to either
wing. By threatening the advance of his
pawns. White forces Black to p la y . . .a5,
and he hopes to be able to cope with the
one black pawn. This plan demands cool­
ness and an accurate appraisal of all the
subtleties in the position. Had the black
rook been on the seventh rank, White’s
Black is the exchange up. Annotating play would not have succeeded.
this game in the tournament book,
Alekhine showed that after 1 . . . Rd7!
2 Ne3 (2 N f6 Rd8! 3 gSaS 4 bxaS 4 ... a5
b4 5 g6 b3) 2 . . . a5 3 bxa5 b4 4 5 bxa5 b4
g5 Kc5 5 Nc2 b3 6 Na3 b2 7 g6 Kb4 6 a6!

6
 CcntraHzadon o f the King

This is the tactical basis o f White’s Kb2, forcing the win o f the pawn by
play. He would have lost after 6 g5 bJ Nc5.
7 Nc4 Kc5 8 Nb2 Rd2 9 Nd3+ Kc4
10 Ne5+ Kc3. 18 Na4 Kf5
19 Kb4 Kxf4
6 ... KcS 20 Nb2 Ke4
21 Na4 Kd4
6 . . . b3 does not succeed: 7 Nc4 22 Nb2 Rf3
Kb5 8 Nb2 Kxa6 9 Ke3 Kb5 10 gS 23 Na4 Reí
Kb4 11 g6Kc3 12 Na4+ with a draw. 24 Nb2 Ke4

7 a7 b3 The final attempt to break through.

8 N dl
25 Na4 Kf3
The pawn a t a7 makes this move 26 Ka3!
possible.
Now 26 . . . Ke2 is met by 27 Kb2.
8 ... Ra8
9 g5 Rxa7 26 ... Ke4
10 g6 Rd7 27 Kb4 Kd4
11 Nb2 Rd2 28 Nb2 Rh3
12 Kf3! 29 Na4 Kd3
30 Kxb3 Kd4+
White is rescued by the centralized
position of his king. Black is forced to Drawn. A very difficult ending, in
retrace his steps, since after 12 . . . which the white i king saved his army
Rxb2? 13 g7 White wins. from defeat. '

12 . . . Rd8
13 Ke4 Kd6

Now White’s king rushes across to the

Y Fyedorov—Chemikov

Krasnodar, 1974
Q-side to the black pawn, sacrificing his
own passed pawns.

14 Kd4! Rc8
15 g7! Ke6
16 g8=Q+ RxgS
17 Kc4 Rg3
After 17 . . . Rb8 Black again loses
his pawn: 18 Kc3 Kf5 19 Nd3 Rb6.
Now 20 Kb2 is bad due to 20 . . .
Ke4, but if White coolly waits with
20 Nb2 Kxf4 21 N a4Rb8 22 Nb2,
Black has no possibility o f winning. As
soon as the black king reaches e l with Wc give the commentary to this game
the white king at c3 and knight at b2, by grandmaster Averbakh in Sbakbmaty
there immediately follows Na4 K dl; vS S S R , 1978 No. 9:

7
ÍS -»
 Endgame Strategy

“White has a powerful initiative. His White plays 4 Rf2 and takes his king
knights have broken into the enemy across to the Q-side, winning.
position and with the support o f the Why did Fyedorov, an undoubtedly
rook have created the grounds for talented master, fail to find this quite
various tactical blows. Black’s pieces straightforward forcing manoeuvre?
are restricted, and only his bishop is dis­ It seems to me that he played his king
playing some semblance o f activity, by to f2 without much thought. A pawn is
attacking the eJ pawn. Fyedorov played attacked, it has to be defended, and the
king is better placed in the centre.”
1 Kf2 The position demanded thinking in
terms of schemes, and the centralization
following the good old rule that in the o f the king should have been deferred
endgame the icing should head for the for the moment.
centre. But this natural move gave Black
a respite, and by
Dvoryetsky—Smyslov
1 ... Nc6
Odessa. 1974
he began exchanging o ff White's attack­
ing pieces. It is true th at after

2 Nxc6 bxc6
3 Rcl

Fyedorov won the c-pawn, but Black

gained the opportunity to activate his
rook, and in the end, not without the
help of his opponent and . . . time
trouble, he gained a draw.
Yet in the diagram position White had
the possibility of an elegant three-move
manoeuvre which would have tied the
opponent hand and foot: 1 Ng4! Nc6 White's K-side pawn majority is more
(2 Rxd8 was threatened) 2 Rfl! Nd8 of a reality than Black’s on the Q-side.
3 h4!, and Black’s position is hopeless. Black has to decide which pawn form­
ation to leave White with on the K-side:
e4/e3/f3/g3, or (after 1 . . . Qd4)
e4/f3/f2/g3.

1 ... Kc8?

Black should have concentrated all his

forces, including his king, on the K-side
to parry the opponent’s onslaught. Per­
fectly reasonable was either 1 . . . Qd4
2 Rxd4+ Ke8, or 1 . . . Bd6 2 fxe3
Ke7.

2 fxe3 g6?i

8
 Centralization of the King

This allows White to seize the initia­

tive completely on the K-side. 2. . .Be7
is preferable.

3 e5! Bg7
4 f4 f6

Odierwise after 5 Ne4 Black is in a

bind.

5 exf6 Bxf6
6 e4 h5
7 Kg2 Bxc3
8 bxc3 centralization.

After this exchange the difference in 1 ... Kg7

the positions of the two kings is especi­ 2 b4?!
ally noticeable.
White’s king is too far from the centre,
8 ... bS and any transition into a knight ending
9 e5 aS will favour Black. He should have played
10 Kh3! 2 Kfl.

Black has no way of opposing the 2 ... Ne7

break-through of the white king. 3 a4?!

10 . .. b4 Not 3 Qe5+ QxeS 4 Nxe5 Kf6 5

11 Kh4! Re8 Nf3 Ke6 6 Nxd4+ Kd5 7 Nb5 Kc4
12 Kg5 Re6 8 Nd6+ (after 8 Nxa7 Kxb4 the knight
13 Kh6 is lost) 8 . . . Kxb4 with advantage to
Black, but White should have activated
Black resigned, since against Kg7—f7 his king and only then advanced his
there is no defence. \ pawns.
—•TXf
* ./ 3 ... Nd5
-----' Roizman—Mikhalevsky 4 g3?

Minsk, 1979 White still thinks that he has the ad­

vantage, and he parries the threat of
(See next diagram) 4 . . . Nf4 with a transition into a
drawn queen ending. But the black
Black’s passed pawn is blockaded, pieces are already pretty active. After
while White has all the preconditions for this move it is difficult for White to
creating an outside passed pawn on the avoid the exchange of queens, and in the
Q-side. In the event of the knights being knight ending, thanks to the activity of
exchanged the game should end in a his king, Black gains the advantage.
draw, while after the exchange of queens
much will depend on the mutual placing 4 ... Nc3
of the kings. Black therefore begins 5 Qe8

9
 Endgame Strategy

Risky is 5 Qa5 Qe6 6 Qxa7 Qe4

7 Qa6 Qf3, while after 5 Qc5 the
difference in the positions of the kings
becomes important — Black is the first
to reach c4.

5 ... Qe6!
6 Qxeb fxeó
7 a5 Kf6
8 Kfl

White cannot prevent the advance of

the e-pawn by 8 f4 d u e to 8 . . . Kf5
9 Nc5 e5, and after the exchange the the other critical square cl. The position
black king penetrates to c4. is drawn. But if White moves his knight
to another square, e.g. d2, the black
8 ... e5 king acquires a choice of critical squares,
9 Nc5 and White loses: 2 Nd2 Kc2 3 Nc4
Nbl 4 Ne3+ K el 5 N c 4 N a 3 !(5 ...
Here to o 9 f3 does not help: 9 . . . d2+ is also good enough).
Ke6 10 Kel Kd5 11 Nb2 e4. L et us now continue our analysis of
the game after the possible 12 Nf6+
9 ... e4 Kc4 13 Nxh7 Kd3 14 Ng5 (or 14
10 Kel Ke5 Nf6 Kc2 15 Ng4 e3! 16 fxe3 d3
11 Nd7+ Kd5! 17 Nf2 d2+ 18 K f l Ne4) 14 . . . Kc2
15 Ne6 Nb5! 16 Ng5 Nd6! (it was
The main thing is to activate the king! essential to vacate c3 for the black
king) 17 Ne6 d3 18 Nd4+ Kc3 19
12 b5 Ne6 d2+ 20 Ke2 Nc4, and wins.

12 Nf6+ should also be examined. 12 . . . Nxb5

Before looking at any variations, we 13 Nf6+ Kc4!
must dwell on the principles of playing
such endings. Black wins if his king can
control the queening square of the d-
pawn, i.e. stand on one of the critical
squares c2 or cl without being pur­
sued by the white knight. The defender
must attack the king in such a way that
he can control one of the critical squares
and check the king when it steps onto
the other. We will consider an example.

(See next diagram)

Black is in check, and on 1 . . . Kb2

there follows 2 Nd4! when the knight 14 Kd2?
controls one of the critical squares c2,
and can attack the king if it returns to After this Black wins very easily.

10
 Centralization o f Che King

White had two other possibilities: Ne6 Ne4 19 Nc7 (19 Nf4 loses to
(a) 14 Nxe4 (the weaker alternative) 19 . . . Nf2+ and 20 . . . Nd3+) 19
14 . . . Kd3 15 Nd2 Kc2 16 f4d3 . . . Nxg3 20 Nd5+ Ke4 21 Ne7 (21
17 g4 Nd4 18 h4 Kcl 19N e4N c2+ Nf6+ Kd3 22 <¡6 Ne4 23 Nd7 Ke3
20 Kf 2 d2, with an easy win. 24 Ne5 d3 25 Nc4+ Kd4 26 Nb6
(b) 14 Nxh7! (attempting to create axb6 27 a7 Ke3 28 a8=Q Nf2+)
an outside passed pawn on the K-side) 21 . . . g5 22 Kd2 Nfl+ 23 Kcl
14 ...K d 3 15 Ng5 (15 N f6N c3 16 Ne3 24 a6 Nc4 25 Kdl d3 26 Kcl
Ng4Kc2 17 N eSd3 18 K 'c4Nbl 19 Kd4 27 K dl d228 Ke2 (if 28 Kc2
Ne3+ K c l 20 Nc4 Na3!) 15 . . . Nd6! Ke4) 28 . . . Kc3 29 Nd5+ Kc2 30
(not 15 . . . e3 16 b4). Here White Nb4+ Kb2 31 Kdl Kc3 32 Nd5+
again has two possibilities. We give some Kd4 33 Nb4 Ke3 34 Nd5+ Kf2 35
sample variations: Nc3 Kf3 36 Kc2 Ke3 37 Nd5+ Ke2
(bl) 16 Ne6 Kc3 17 K d lN c4 (n o t 38 Nc3+ (were f4 available to the
17 . . . d3 18 Nc5 with a draw) 18 white knight, the game could end in a
Ng5 (after 18 Nc5 White loses in a draw) 38 . . . Kel. When the black
pawn ending: 18 . . . Nb2+ 19 K el pawn is at d2 and the white king at d l,
Nd3+ 20 Nxd3 Kxd3 21 K dl e3 22 the critical squares are d3 and e3,
fxe3 Kxe3 23 g4 g5! - not 23 . . . Kf3 while with the white king at c2 the
24 g5 Kg3 25 Kd2 with a draw - critical squares became e2 and e l.
24 <¡6 d3 25 K el d2+ 26 K dl Kd3 The game actually concluded:
27 b4 gxh4, and mates) 18 . . . Kd3!
19 Ne6 (if 19 b4e3) 19 . . . Nxa5!
20 Nc5+ Kc4 21 Nxe4 Nb3 22 h4 14 . . . Nd6
a5 23 Kc2 d3+ 24 Kb2 a4 25 g4 15 Nxh7 e3+
Kd4 26 Nd6 a3+ 27 Kxb3 d2 28 16 fxe3 Ne4+
Nb5+ Kd3 29 Nc3 a2. 17 K dl d3
(b2) 16 Kdl e3 17 fxe3 Kxc3 18 White resigns.

ll
 CHAPTER 3

THE ROLE OF PAWNS IN THE ENDGAME

In the endgame the main task is not 10 f4!?

usually the immediate mating of the
opponent’s king, but the queening of a A piece sacrifice, the consequences of
pawn. Therefore, in comparison with which had to be judged accurately.
the middlegame, in the endgame the
value of the pawns increases. This must 10 • •• Bxc3
be taken into account when solving ex­ 11 bxc3 d5
changing problems. Right in the open­ 12 Bb3 f6
ing stage o f the game examined below 13 fxe5 fxg5
White earned out a complex and deep 14 Rxf8+ Kxf8
combination, as a result of which play 15 Qf3+ Kg8
went into an ending, by-passing the 16 Rfl Nc7
middlegame. For the sacrificed piece 17 Qf7+ Kh8
White gained three pawns. 18 exd5 cxd5
19 Qf8+ Qxf8
20 Rxf8+ Ng8
21 Nf3
Schlechter—Duras
As a result of his combination White
San Sebastian, 1911 for the moment has only one pawn for
the piece, but the dominating position
1 e4 e5 o f his rook enables him to acquire two
2 Nf3 Nc6 further pawns.
3 Nc3 Nf6
4 Bb5 Bb4 21 . . . Be6?
5 0 -0 0 -0
6 d3 d6 Probably the decisive mistake. Black
7 Bg5 Ne7 incoirecdy approaches the exchanging
8 Nh4 c6 problem. The advance o f White’s passed
9 Bc4 Ne8 pawns will be best supported by his long-
range bishop, and the black cavalry will
clearly be unable to cope with it and the
three enemy infantrymen. After the
correct 21 . . . Bd7 22 Rxa8 Nxa8
23 BxdS Bc6! (leaving the opponent
with a knight, not a bishop) 24 Bxc6
bxeó 25 Nxg5 Ne7 a hard struggle
would have been in prospect. It should
be added that on 21 . . . g4 White can
continue 22 Nd4! Bd7 23 Rxa8 Nxa8
24 Bxd5, obtaining three pawns for a
piece in a favourable version.

22 Rxa8 Nxa8

12
 The Role o f Pawns in the Endgame

23 NxgS Nc7 ' Lukov—Syemkov ^ K v ^ _

24 Nxe6 Nxc6
25 Bxd5 Nd8 Bulgarian Championship, 1977 ^ o .
26 d4

Black is unable to prevent the advance

of the central pawns.

26 . . . Ne7
27 Bb3 Nec6
28 Kf2

the pawns.
White has sacrificed a knight, obtain­
28 . . . Na5 ing three pawns in return, one of which
29 c6! Nac6 has reached the seventh rank. But with
his last move . . . Nb8—c6 Black has
Not 29 . . . Nxb3 30 e7. simultaneously threatened the white
rook and passed pawn. In the event of
30 d5 Ne7 1 Rxc5 Rxe7 2 Bd5 Nb4! 3 Bxb7
31 d6 R el+ 4 Rcl Ral+ the white king is
mated. Nevertheless White has a way to
The pawns sweep away everything in maintain his initiative.
their path.
1 Rxc6!
31 . . . Ndc6
32 dxe7 Nxe7 Now Black has an extra rook for only
three pawns! But in the endgame there is
The rest is dear. a difference between the absolute and
relative values of pawns. Follow the ad­
33 Kf3 Kg8 vance of the white infantry, which the
34 Kc4 Kf8 opponent’s superior forces is unable to
35 Ke5 Ke8 prevent
36 Bd5 b6
37 Be4 h6 1 ... Bxc6
38 Kd6 Kd8 2 f6 h6
39 Bd3 h5
40 h4! b5 The threat was 3 Rg5+ followed by
41 Bxb5 Nf5+ Bxf7.
42 Ke5 Nxh4
43 Bd3! 3 Re3 c4!

44 Be4 is threatened, and on 43 . . . A clever defence. By sacrificing a

Nxg2 there follows 44 Kd6. Black fourth pawn, Black opens the c-file and
resigned. gains counter-play.

13
 Endgame Strategy

4 Bxc4 Rcc8 Miller—Wdtma nde r

5 b3!
Izhevsk, 1949
Reinforcing the bishop and opening
an escape fo r the king. Not 5 Rg3+ Kh8
6 Bxf7??Be4+.

5 ... Ra5

To be able to answer 6 Rg3+ with

6 . . . Rg5.

6 g4! Bc8
7 f4

The pawns advance in strict battle

formation.
Black has some compensation for the
7 ... Rca8 exchange in the form o f his superior
8 Rd3 pawn formation, his well-placed pieces
and his extra pawn. But it is White to
The d8 square is put under attack. move, and he carries ou t what appears to
be a winning combination.
8 ... Kh7
1 Nxd5! KxdS
Rook checks do not achieve anything. 2 R fl

9 Rd6 There appears to be no defence

against 3 g4, but Black finds a counter-
Not allowing the black king to come combination.
out: 9 __ Kg6?? 10 Bd3+.
2 ... a5!
9 ... Rc8 3 g* Nxc3!!
10 h4 Ra7 4 Rxc3 b4
11 Kb2 5 Ref3 Bxc2!
6 Rxf6 b3
White can improve the position o f his
king, since the advance of his pawns can­ (See next diagram)
not be prevented.

A curious situation. The black bishop

This hastens the inevitable. and « " ° Pawns Provc no weaker than
the white rooks.

12 Rxd7 Bxd7 / 7 Ra6 Kxd4

13 Bxf7 Rb8 >~ 8 RxaS b2
14 g5 hxg5 9 Rb5 c3
15 fxg5 Be8 V 10 Rb8 Bd3
16 g6+ Resigns. 11 R el Kc5

14
 The Role o f Pawns in the Endgame

which involves sacrificing his bishop.

1 c5!

All White’s hopes are pinned on this

pawn.

1 ... Rxe3

As shown by Keres, 1 . . . dxc5 2

Bxc5 Nd8? gets Black nowhere due to
3 fxe5 fxe5 4 Bd6!, but 2 . . . g5!
came into consideration.
Black prepares to block the b-file with
his bishop. 2 Rxb2 g6?

12 Kf2 Bb5 In time trouble, Keres makes a mis­

13 Re5+ Kd6! take which leads to a draw. He should
14 RexbS c2 have answered with a counter-sacrifice of
15 Rxb2 cl=Q a piece to eliminate the enemy passed
Drawn pawn in the variation 2 . . . h5! 3 Rb8+
Kh7 4 R f8dxc$! 5 Rxf7 exf4 6
gxf4 Rxe4, and Black must win the
Botvinnik—Keres \JKWC. rook ending. White is not saved by 3 c6
Rc3 4 Rb7 Nh6 5 Rd7 Ng4 6 Rxd6
19th USSR Championship ' J R cl+ 7 Kg2 Rc2+ (pointed out by
Moscow, 1951 Botvinnik).

3 c6 Rc3
4 Rb7!! Kg7

Botvinnik showed that Black could

also hardly have hoped to win after
other moves, for example: 4 . . . Kf8
5 Rb8+ (5 c7 Ke8) 5 . . . Kg7 (if
5 . . . Ke7, then 6 Rb7+ Ke8 7 Rb8+
N d8 8 Rc8! with the threat of 9 c 7)
6 Rb7. Black has gained a tempo, but
it is difficult to make use of it, since on
6 . . . f5 there follows 7 fxe5 dxe5 8
exf5 gxf5 9 Rd7, threatening c6—c7
Black has a decisive positional advant­ and d5—d6.
age. His passed b-pawn, supported from
behind by his rook, ties down White’s 5 c7
entire forces. The white bishop is attack­
ed, and if it moves the game will be con­ (See next diagram)
cluded by Black bringing his knight over
to the Q-side via f7—d8—b7—c5 or a5. The players have exchanged roles.
Botvinnik finds the best practical chance. Black’s passed pawn has been eliminated

15
 Endgame Strategy

On the previous move the queens were

exchanged A brief glance at the position
shows it to favour Black. Formally White
has sufficient compensation for the ex­
change — two pawns, but it is difficult
for him to deal with the enemy pawn at
f3. White’s only hope lies in his Q-side
pawn armada.

1 Bd4 Rfe8
2 Kg4 R fl
3 Kf4

at the cost of the white bishop. White’s 3 . . . Re4+ was threatened

passed pawn has reached the seventh
rank and is worth no less than the black 9 • • • Re2
knight. Not one black piece has freedom 4 Be3 Rxh2
of movement. The rook cannot move off 5 b5 f2
the c-file, the knight is restricted by the 6 Kf3 Rcl
threat of c8=Q+, and king moves are 7 Bxf2 R fl
met by Rb8+ or c8=Q. A positional 8 c4 Rfxf2+
draw! The game continued: 9 Ke4 Re2+

5 . .. Rc2 The ending after 9 . . . Rxd2? 10

6 Kgl hS Rxd2 Rxd2 11 b6 favours White.
7 h4 Rc4
8 Kg2 Rc2+ 10 Kd4 Rxd2+
9 Kfl Rc4 , 11 Kc5 Rb2
Drawn .„\ ■’
In Informator N o. 23, grandmaster
Marjanovic gives 11 . . . Rxd5+ 12
Kxd5 Kg8 13 c5 Kf7 14 Kd6 Ke8 15 Kc7
Timman—Gligoric as winning for White. But in Shakhmaty v
SSSR (No. 12 1989) V. Lukmanov points
Bad Lautcrberg, 1977 out that after 12 . . . Rb2! 13 Kc6 g5!, it
is Black who wins. White should
therefore meet 11 . . . Rxd5+ with 12
cxd5 R c2+ 13 Kd6 R b2 etc., when both
sides must be content with a draw.

12 Kc6 Rhc2
13 c5

As a result of an almost forced series

of moves Black has an extra rook for just
one pawn, but White’s passed pawns on
the Q-side, supported by his king, are a
formidable force.

16
 The Role o f Pawns in the Endgame

13 ... Rc3 23 b8=Q Rxb8 24 Rxb8 Rxb8 25

14 Kb6 hS Kxb8 h3 26 c 7 h 2 27 c8=Q hl=Q
1$ c6 Rbc2 White must give perpetual check by
14 Kc7 Kh7 28 Qg4+. The game continued:
17 Rd6 Rxg3
18 b6 20 • ♦♦ Rb5
21 Ra4 R bl
Black is now a whole rook up, which 22 Rd4 R lb3
he would be glad to give up fo r the 23 Ra4 Rb2
opponent’s passed pawns, but die pawns, 24 Rd4 R5b3
having reached the sixth rank, arc worth 25 Ra4 R bl
more. Drawn

By skilful play Glogiric has stopped

the enemy pawns at the last line, and has
created a passed pawn of his own. It White is a knight up for only a pawn.
appears that White is faced with having Apart from his central pawn, which can
to suffer in the queen ending after 20 easily be blockaded, Black has no far-
Rd8 h4 21 b8=Q Rxb8 22 Rxb8 advanced passed pawns. But the position
Rxb8 23 Kxb8 h3 24 c7 h2 25 is of a concrete nature, and White’s
c8=Q hl=Q, but Timman finds a brilliant apparent well-being is deceptive.
defensive idea.
1 ... e3
20 Rd4!! 2 Rd6+

It transpires that the situation on the Leonid Kubbel, now an acknowledged

board is a positional draw. If Black plays classic of chess composition, gives the
20 . .. Kh6, White replies 21 Rd8, and following variations after 2 Re2:
Black cannot go into the queen ending, ” . . . Black wins by the study-like 2 . ..
since after queening his pawn White wins b3 3 Nd3 Rc3 4 Kfl (or 4 g4) 4
by a check at h8. In the event ot the . . . Rxd3 (Black gives up the exchange
exchange of the f- and g-pawns after and obtains two connected passed pawns
20 .. . g6 21 fxg6+ Kxg6 22 Rd8 h4 against a rook) 5 cxd3 a4 6 Rb2 Kb5

17
 Endgame Strategy

etc., or 3 c4 Rd8 4 R el (4 g 4R dl + Nevertheless!

5 Kg2 Rd2, and after the exchange of
rooks the b-pawn queens) 4 . . . b2 5 4 Nd4+ Kc5
Nxf7 (attacking the rook; if 5 Nf3 e2) 5 Nxc2
5 . . . Rd2 etc.”.
5 Rc6+ Kxd4 6 Rxc2 b3 is com-
2 ... Kb5 pletely hopeless.
3 Nf3
5 ... Kxd6
6 Kfl b3
There is no other way of defending 7 Na3 Kc5
the c-pawn. 8 g4 Kb4
9 Nbl a4
3 ... Rxc2! White resigns. f
 CHAPTER 4

THE PROBLEM O F EXCHANGING

As material is reduced the problem

of exchanging becomes of primary im­
portance. In the opening or middlegame
the consequences of an incorrect piece
exchange can sometimes subsequently
be repaired, but in the endgame such a
mistake can be fatal. Of course, in the
majority of cases an experienced player
will easily determine which exchange
favours him. But situations often arise
where an exchange which seems plaus­
ible on general grounds turns out to be
routine and not in accordance with the
demands of the position, whereas a knight at f4 — the pride of his position.
decision which is at first sight para­ In the game there nevertheless followed:
doxical proves to be the only correct
one. We will also consider cases where 1 ... Nxd3!
there is only one reply to the question
“to take or not to take?”, but there are Capablanca gives up his splendid
several different captures leading to play knight for White’s bad bishop, demon­
of a different nature. In addition, this strating a subtle assessment o f the posi­
chapter contains examples where one tion. It would seem that in the joke pre­
of the sides is faced with loss of material valent among chess players “die worst
We will see how difficult it can be to bishop is better than the best knight”
choose the least out of several evils. there is a grain o f truth. The black
Simplification is often the best way of knight at f4 occupies an ideal position,
realizing a material or positional advant­ of course, but how can any real advant­
age. The outcome of a game may depend age be extracted from this? White’s
mainly on the ability of a player to solve bishop seems to be bad, but it holds to­
correctly the problem o f exchanging, gether his K-sidc pawn formation, and
whether to make a timely simplification has fair prospects in the event o f possible
of the position or, on the contrary, play on the Q-side. Black aims to take
maintain the tension. play into an ending where his bishop
may prove stronger than the white
knight, taking account of the K-side
Klein—Capablanca pawn formation.

New York, 1913 2 Nxd3 Bc6

3 R dl Rcd8
(See next diagram) 4 b3 Nf4
5 Ng2?!
As yet this is far from being an end­
game position. Black has a big advant­ A strange move: 5 Nf5 or 5 Nxf4 is
age: the better bishop and an ‘eternal’ more natural.

19
 Endgame Strategy

5 ... Nxd3 26 Nxf7 Bxt'7

6 Rxd3 Rxd3 27 Qg5+ Kf8
7 Qxd3 Rd8 28 Qh6+ Ke7
29 Qg5+ Ke8
Possibly White was hoping for 7 . . .
Bxg4? 8 Nxh4gxh4 9 Rg2. The checks are at an end, and White
resigned.
8 Qe2 h3! It is curious that in his book M y Chess
9 Ne3 aS Career Capablanca does not even com­
ment on 1 . . . Nxd3! For him such a
Black sets about creating weaknesses plan was the natural continuation.
on the Q-side. The advantage o f bishop
over knight is obvious.
Flohr—Spielmann
10 R fl a4
11 c4 Bled. 1931

Now the d4 square is weakened, but

11 bxa4 is even worse due to 11 . . .
Qf4! followed by . . . Ra8.

11 ... Rd4
12 Nc2 Rd7
13 Ne3 Qd8
14 R dl Rxdl+
15 Nxdl

After 15 Qxdl Black’s 15 . . . Qd4

is again very strong.

15 ... Qd4 On the Q-side Black has a weak c-

16 Nf2 b5 pawn and a badly placed bishop at b7.
17 cxb5 axb3! He makes a desperate attem pt to escape
18 axb3 Bxb3 from the vice.
19 Nxh3 Bdl
1 ... f5!
Black’s passed b-pawn and the weak­
ness of White’s K-side decide the out- Preparing the freeing break . . . cS.
come of the game.
2 Nd6!!
20 Q fl cxb5
21 Kg2 b4 A solution which is convincing in its
22 Qb5 b3 simplicity. The white knight is forced to
23 Qe8+ Kg7 move, but where to? 2 NcS can im­
24 Qe7 b2 mediately be discarded, since after
25 Nxg5 Bb3 2 . . . NxcS 3 dxcS White loses the
greater part o f his advantage. On 2 Nc3
White’s threats arc easily parried, or 2 Nf2 there follows 2 . . . cS!,
while the black pawn is about to queen. when the worst for Black is over. That

20
 The Problem o f Exchanging

only leaves moves to g3, g5 and d6. 11 Rc6

In the first two cases it is unfavourable
for Black to play . . . c5, e.g. 2 Ng5 The black pieces are tied to the de­
c5? 3 dxc5 Nxc5 4 Rxc5 Rel+ 5 fence of the a6 pawn. To win White
Kf2 R xbl 6 Rc7+ with decisive needs to create one further weakness in
threats. But at g3 the knight is badly the opponent’s position.
placed, and 2 Ng5 can be met by
2 . . . h6, driving the knight to the edge 11 ... Kf8
of the board and obtaining counter-play 12 h4 Kg7
after . . . Rbc8. Flohr does not try to 13 KÍ2 Khó
ding to the advantage he has already 14 Bb3
achieved, but converts it into a different
form. The immediate 14 g4 is also possible,
but since Black is deprived o f the slight­
2 ... Re6 est counter-play. White operates accord­
3 Nxb7 Rxb7 ing to the principle “do not huriy!’’.
4 dS! cxd5
5 RxdS 14 ... Kh5
15 Kg3 Rd7
Black’s weak c-pawn and his bad bis­ 16 Kh3 Kh6
hop have disappeared from the board, 17 Kg3 Re7
but his position has not improved. After 18 Kf4 Kh5
the series of exchanges White’s advantage 19 g3
is not as obvious as it was before, but it
is of a stable nature. All the black pawns Black is in an unusual form o f tug-
are on white squares, the a6 pawn being zwang. His knight has no moves due to
especially weak. The bishop is clearly the loss of his a6 pawn, and his rook
superior to the knight, which has no must not allow the white king to e5.
strong points, and White's rooks are Therefore his king must retreat from h5,
significantly more active than the oppo­ allowing g3—g4.
nent's. By the following manoeuvre
Flohr improves the deployment of his 19 . . . Kh6
forces, exchanges one pair of rooks, and 20 Bg8!
then sets about implementing his stra­
tegic plan of giving Black a second weak­ A continuation of the previous tactics.
ness on the K-side. The bishop will return to a2 and will
stand better than at b3, being defended
5 • • • Nf6 by the b2 pawn against a horizontal
6 Rdc5 Rbe7 attack by the black rook.
7 Kfl Ne8
8 Ba2 Re2
9 R5c2 Rxc2 20 . .. Kg7
10 Rxc2 21 Ba2 Kh6
22 g4! fxg4
With the exchange of one pair of 23 fxg4 Kg7
rooks the white king can advance with­ 24 g5
out fear.
The noose tightens. There is practi­
10 . . . Nc7 cally nothing that Black can move.

21
 Endgam: Strategy

24 . . . Kf8
25 Kf3 Kc8

Black’s game is lost. White has many

ways to win, e.g. 26 Rf6 followed by
the exchange of rooks at f7 at an
appropriate moment. Flohr chooses a
different, more consistent plan: he com­
pletely breaks up the black pawns on the
K-side, giving his opponent a second
weakness. In passing White sets a clever,
camouflaged trap, into which Black falls.

26 h5! ? gxh5 obvious mistake. After the correct 1 Nf4

27 Rh6 b4 it is unlikely that Black would have been
28 Rxh5 Re5?i able to realize his insignificant positional
29 Kf4! Rxa5 advantage.
30 g6!i
1 ... Bxd4!

A deeply-conceived exchange. In the

King’s Indian Defence, which was played
in this game, the black-squared bishops
are especially valuable. Fischer ex­
changes his fianchettoed bishop, foresee­
ing a forcing manoeuvre, as a result of
which Black remains with two rooks and
a knight against White’s two rooks and
a bishop.

2 Rxd4 Bb5
3 R el
Black resigns.
3 R fdl Nb3 4 Rb4 Be2 is bad for
White.
Szabo—Fischer
3 ... Nb3
Buenos Aires, 1970 4 Rb4 Nxcl
5 RxbS
(See next diagram)
All the pawns are on one wing, which
Black has slightly the more active slightly favours the side with the knight
position, but a draw is the most likely But how is this advantage to be trans­
outcome. But the Hungarian grand­ formed into something real? After 5 . . .
master plays Nd3 6 R fl Black does not have any
serious advantage. The American grand­
1 Nd4? master finds a fine knight manoeuvre.

committing a serious, but by no means 5 ... Ne2+!

22
 The Problem o f Exchanging

6 Kfl?! followed by the centralization o f his

king. Blade would have had every right
This natural move is the decisive mis- to count on a draw. But instead he
take. White would have retained some played
drawing chances after 6 K hl, leaving
f l free for his rook. 1 — Bf6?

6 ... Nc3! Black should have aimed for the ex­

change of knights, and not bishops, since
The knight has taken up an ideal posi* in a knight ending a spatial advantage is
tion. The manoeuvre .. . . Rd8—d2 can- often a decisive factor.
not be prevented.
2 Bxf6 Nxf6?!
7 Rc5 Rd8
8 Bh3 Rdd2 Now the white king breaks through in
9 Rc8+ Kg7 the centre, ahead of Black’s. It would
10 Re3 Ndl! have been better to allow the spoiling of
his pawn formation, but restrain the
Ten moves ago it was impossible to white king, by capturing on 56 with the
imagine such a turn of events. The loss paw a After 2 . . . gxf6 3 Ke2 Kf8 the
of two pawns is inevitable for White. The pawn sacrifice 4 Kd3 Nxf2+ 5 Kd4
finish was: does not achieve anything due to S . . .
N dl, and White would have had to waste
11 Rf3 Rxf2+ time or. driving the black knight from e4.
12 Rxf2 Rxf2+
13 Kgl Re2 3 Ke2 Kf8
14 Bg4 Rxe4
White resigns. 3 . . . Ne4 is dangerous due to 4 Kd3!
Nxf2+ 5 Kd4, and here 5 . . . N dl is
well met by 6 e4, if there is nothing
Mikhalyevsky—Akopov better.

Rostov, 1977 4 Kd3 Ke7

5 Kd4 Nd7
6 Nd2!

White avoids the unclear complica­

tions which could have arisen after 6 e4
Kd6 7 e5+ Kc6 8 Ng5 KbS, and pre­
vents Black from activating his king,
since 6 . . . Kd6 can be met by 7 Nc4+
Kc6 8 e4, cramping Black still further.

6 ... Nb6
7 e4

The pawn ending after 7 Ne4 Na4

The position looks roughly equal, and 8 Nc5 Nxc5 9 Kxc5 Kd7 is most pro­
after the exact 1 . . . Nd6 2 Nd2 f6 bably drawn.

¿ S -C
23
 Endgame Strategy

7 ... Na4 20 . . . fxg4

21 fxg4 N dl
After 7 __ Kd6 8 e5+ Kc6 9 Ne4 22 g5?!
Kb5, apart from 10 Nd6+ with un-
dear complications. White has the simple
10 Nc3+, retaining all the advantages of This move should not have been
his position. made, since Black acquires counter-play
by . . . Kf7—g6. 22 a4! was good,
8 e5 f5 when a possible variation would be 22
. . . Nb2 23 a5 b5 24 Kc5 g5 25
Instead of the backward f7 pawn, hxg5 hxg5 26 Ng6+! Kf7 27 Kb6
Black acquires a weakness at e6. Nd3 28 Kxa6 Nxb4 29 Kxb5 Nd5
30 a6 Nc7+ 31 Kb6Nxa6 32 Kxa6
9 Nc4 Kd7 Kxg6 33 Kb6, winning.
10 Nd6 b6
11 f3 Nb2
12 h4! 22 . . . hxg5
23 hxg5 Nb2
White aims to weaken the opponent’s 24 g6
K-side, which Black is unable to prevent.

12 ... Na4 White cramps Black’s position to the

13 Nf7 Ke7 maximum, but uses up his reserve
14 Ng5 h6 tempo. However, after 24 Kc3 Ndl'*'
15 Nh3 Kd7 25 Kd3 Nb2+ 26 Kd4 Na4! (26 . . .
16 Nf4 Ke7 N d l is bad due to 27 a4) 27 Nd3
17 Kc4?! Kf7! 28 Kc4 Kg6 29 Kb3 b5 30Nc5
Kxg5 31 Nxa6 Black has serious
By subtle play White has gained the counter-play.
better position. He should now have shut
the black knight out o f the game by 17
Nd3, when Kc4 followed by b4—b5 is 24 . . . Na4
decisive, e.g. 17 Nd3! Kd7 18 Kc4 25 Nd3 Kd8
Kc6 19 Nf4! Kd7 20 Nh5. 26 Kc4 Kd7??

17 ... Nb2+ A fatal blunder. After 26 . . . b5+

18 Kc3 Ndl+ 27 Kd4 Kd7 28 Nc5+ Nxc5 29 Kxc5
19 Kd4 Nb2 Kc7 White does not have the tempo
20 g4?! move g5—g6, and the game ends in a
draw.

It would have been simpler to give the

opponent the move by 20 Kc3 Ndl+ 27 b5 axb5+
21 Kd3 Nb2+ 22 Kd4, when 22 . . . 28 Kxb5 Nc3+
Na4 is bad due to 23 Nd3, while 22 29 Kxb6 Nd5+
. . . N dl is very strongly m et by 23 a4!, 30 Kb7 Nc7
e.g. 23 . . . Nb2 24 a5 b5 25 Nd3, 31 a4 Nxg6
with a won pawn ending. 32 a5 Resigns.

24
 The Problem o f Exchanging

Najdorf—Averbakh 7 a5?l b5

Candidates Tournament All White’s pawns are on dark squares,

Zurich, 1953 which emphasizes the unfortunate posi­
tion o f his bishop at e l.

8 Rc3 Rc8
9 Rxc8+ Nxc8
10 f3 Ne7
11 Bf2 Kf7!

After 11 . . . Rb2? 12 Rcl White

would have seized the c-file. Black has
no reason to hurry.

12 R bl Nf5
13 K fl Nd6
14 Rb3 Nc4
White’s Q-side pawns are weak, and he 15 Kg2 f5
has the inferior bishop and a ‘hole’ at c4.
The c3 pawn is attacked, and the Zugzwangl 16 f4 or 16 e4 fails to
Argentinian grandmaster plays 16 . . . Nd2, king moves are impossible
for the same reason, and 16 Rd3 is
1 Bd2? decisively met by 16 . . . Rb2.

An incorrect decision. The knight 16 R bl Nxe3+

should have been retained at all costs, 17 Kgl f4!
since after its exchange there is nothing 18 gxf4 Nf5
with which to defend the white squares 19 Kfl g6!
on the Q-side. After 1 Nd3! Nc4 2 R fcl 20 Rb3 Kc7!
followed by the approach of the king to 21 R bl Kd7!
the centre, White would have had hopes
of saving the game. White resigns, since after . . . Rc4
Black wins a pawn in the most favour­
1 ... Nc4 able situation. By simple moves and with
2 Bel Bxb4! inexorable consistency Averbakh real­
ized his advantage, without allowing his
The decisive exchange. White’s back­ opponent the slightest chance.
ward pawn on the c-fflc is removed, but
the file itself comes under the command
of the black rooks. Panno—Bronstein
3 cxb4 Candidates Tournament
3 Rxb4 is also unpromising. Amsterdam, 19S6

3 ... Na3! (See next diagram)

4 Rb3 Nb5
5 e3 Rc2 Black is a pawn up with a good
6 a4 Nd6 position, although there are oppositc-

25
 Fndgamc Strategy

that after 4 . . . KxdS White can set up

a white-square blockade a t e4. Bronstein
could hardly have imagined that, a pawn
up and with the white bishop badly
placed, he would be unable to win the
game. Only after a painstaking analysis
did it transpire that by this move Black
threw away the win. After 4 . . . KxdS
White is seemingly unable to organize a
defence, e.g. 5 Rd2 eS 6 Re2 Rg7
(the rook cannot be moved off the b-file,
since then the white bishop will come
into play by Bb5) 7 R dl Re7 8 Re4
coloured bishops on the board. It is his Rb6 9 Rd2(d3) Rg6 10 BbS Rg4,
move and, exploiting the poor position winning a second pawn. But now White
of the enemy bishop, he sets his central unexpectedly acquires counter-play.”
pawns in motion.
5 Rd2! Bf6
1 ... d 5! 6 Re2! Rg7
2 cxd5 Kd6
3 Kf3 Bd4 On 6 . . . Bxh4 there follows 7 Be8!,
4 Ba4 while if 6 . . . Rxh4 7 Rxh4 Bxh4
8 Be8.

7 Be8 Re7

R hel was a possible th reat

8 Rxe7 Kxe7
9 BxhS

White has regained his pawn. He can

meet 9 .. . Rh8 with 10 Kg4.

9 ... c4
10 bxc4 dxc4
Black is faced with an exchanging 11 R cl c3
problem. Bronstein makes a natural 12 Bg6 BeS
move 13 h5 Kf6
14 Rc2
4 ... exd5?
A draw is now inevitable. The finish
and. .. throws away the win. Here is the was:
commentary on this move by the Soviet
master Goldberg in the tournament 14 ... Kg5
book: 15 Re2 RbS
“At the board it is unlikely that any­ 16 Rc2 RcS
one would have resisted the temptation 17 Rcl RdS
to connect his pawns, since it appears 18 Rel Bd6

26
 The Problem o f Exchanging

19 Rgl+ Kh6 2 Kd3

20 P el Bc5
21 Rc2 Kg5 Events now develop by force. Hope*
22 Rc2 Bb4 less for White is 2 a3 Kd4 3 Kd2 aS.
23 Re2 Bd6
24 Rc2 2 ... Kb4
Drawn 3 Kc2 Ka3
4 Kbl a5
This ending shows how an insignific­ 5 K al a4
ant mistake when exchanging can re­ 6 bxa4 Kxa4
duce to nought the fruits of a player’s 7 Kbl
previously excellent play.
7 Kb2 b4 would not have changed
anything. When he went into the pawn
Svyeshnikov—Kasparov "1° ending Svyeshnikov may possibly have
thought that he would reach this posi­
47th USSR Championship ' • J T ' tion with him to move, when White is
Minsk, 1979 saved by Kc2! Ka3; Kc3. This ending
shows how seriously the problem of ex­
changing must be approached when
going into a pawn ending.

7 ... Ka3
8 Kal b4
9 Kbl
White resigns.
ll
Son—Khorovyets

Tashkent. 1978

Four of White’s pawns are on squares

of the same colour as his bishop, and the
black king occupies a dominating posi­
tion in the centre. Nevertheless it is not
clear whether or not this advantage is
sufficient for Black to win, since he has
no pawn breaks on the K-side. With his
last move Black offered the exchange of
bishops, and White, after an insufficient­
ly deep analysis of the position, agreed
to the exchange.

1 Bxc5?
Black’s passed pawn is much more
1 Bel is o f course correct. dangerous than the opponent’s, and in
addition her king and knight occupy
1 ... Kxc5 ideal positions.

27
 Endgame Strategy

1 .. ■ h4! almost a pawn ending, where a win is

possible only by exceptionally subtle
This move was sealed by Black, setting play,
her opponent a difficult choice.
11 Qxd5+ KxdS
2 Kb5 12 Kf6 Nd2

White gives up a piece, but 2 h3 Kc5

3 Nc3 Kb4 was equally cheerless.

2 ... Na3+!
3 Kb6 Nxbl
4 a6 d2
5 a7 dl=Q
6 a8«Q Qb3+
7 Kc7

As a result o f a practically forced

series of moves, Blade has a material
advantage and an easily won position.
Now 1 . . . Qf7+ 8 Kb6 Qf6+ 9 Kc7 13 KfS!
Nc3 would have won quickly.
The routine 13 Kg$? would have
1 ... Qc4+ lost after 13 . . . Nf3+ 14 Kg4 Ke4, or
8 Kd6 Qc5+ 14 Kf4 Nxh2 IS KgS Nf3+ 16 Kf4
9 Ke6 Qe5+ Ng5. Bad for White is 13 h3 Ke4 14
10 Kf7 Kg5 Nf3+. We again see how dangerous
the transition into a pawn ending can be
(here the play develops in analogy with
pawn endings).
13 . . . Nf3
There is nothing better.
14 h3!

10 . . . Qd5+??

The problem o f exchanging! There

was a straightforward win by 10 . . .
Qf4+ followed by . . . Nc3. With the
queens on the board the extra knight
quickly decides matters, whereas after
their exchange the position reached is

28
 The Problem o f Exchanging

Here is die result of the incorrect Aronin—Smyslov 4-

queen exchange. White threatens to
drive away the knight by 15 Kf4 and 19th USSR Championship
then attack Black’s only pawn. Moscow, 1951
It is true that Black has a study-like
way to win. We suggest that the reader
himself should try to find the win from
the diagram position, and only then play
through the game continuation.

14 • • • Kd4
15 Kf4 Nel!
16 Kg4 Ng2
17 Kf3 Ne3!
18 Kf4

Black has found the only moves to White has a decisive advantage. In com­
defend her pawn, but how can the white parison with the white pieces, Black’s
king be forced out of opposition? occupy pitiful positions. The fine knight
at f5 is greatly superior to the black
18 . . . Kd3 bishop obstructed by its own pawns,
19 Kf3 Nf5 while the white rook holds sway in the
enemy rear. In addition to all this it is
Repeating moves to gain time on the now White to move, and he can calmly
clock. pick up the c6 pawn, retaining all the
advantages o f his position. To Aronin’s
20 Kf4 Ne3 misfortune, he had to seal the next move.
21 Kf3 Kd2!
1 Rg8?!
Black drives the white king, which is
forced to maintain the opposition, a This move in itself is strong enough,
little further from the h4 pawn. but the question mark is attached be­
cause it is made with the faulty idea of
22 Kf2 Nd5! transposing into what appears to be an
easily won pawn ending.
The decisive manoeuvre.
1 ... Kh7
2 Rxg7+?
23 Kf3 Ne7!
24 Kg4 Ng6 A mistake which shows how easily a
25 Kg5 Ke3 certain win can slip away, due to an in­
26 Kxg6 Kf4! correct evaluation of an ensuing pawn
ending. In Aronin's defence, it has to be
said that it was very difficult to foresee
It was still possible to go wrong: 26 Smyslov’s brilliant defensive idea. Be­
. . . Kf3?P 27 Kf5 Kg3 28 Ke4, with sides, a player who is faced with defeat
a draw. But now White resigned. mobilizes all his strength and clutches

29
 Endgame Strategy

at the slightest chance, however im­ The outside passed pawn also gains
probable, like a drowning man at a straw. the opportunity to become protected.
The player who is winning, on the other
hand, is reluctant to seek saving chances 6 h4 c5
for the opponent, especially if they are 7 Ke2 Kh7!
of a very difficult nature. Even great 8 Kd3 Kh6
players have been known to relax in
such a situation, and chess history knows Smyslov’s brilliant plan begins to re­
o f numerous similar examples. White veal itself. On the natural 9 Kc4 White
would have won most easily by 2 Re8!t even loses after 9 . . . f5! 10 exf5 e41,
answering 2 . . . Kg6 with 3 Re7. when the black pawn queens. No better
is 10 Kd3 f4 11 gxf4exf4 12 Ke2
Kh5 13 e5 Kg6, when the black king
2 ... Rxg7 eliminates White’s passed pawns. The
3 Nxg7 Kxg7 game continued
4 g4
9 c3 a5
10 cxb4 axb4

and the players agreed a draw.

On the other hand, the transition into
a pawn ending, where all the nuances
have been correctly worked out to the
end, can be the quickest way to win.

It appears that Aronin’s idea should

guarantee him an easy win. With his last
move White has deprived Black of
counter-play on the K-side involving
. . . f5 and . . . g4. He has in mind the
following variation: 4 . . . Kf7 5 Ke2
Ke6 6 Kd3 Kd6 7 Kc4 a5 8 f3 Kd7
9 Kc5Kc7 10 c3 bxc3 11 bxc3 Kb7
12 Kd6 Kb6 13 c4 Kb7 14 cS, and
White wins. But Smyslov has prepared a
series of surprises.

4 ... hxg3!
Black is tw o pawns up, but since both
In addition to all his other advantages. are doubled one gains the impression
White gains the prospect of creating an that the winning path will be long and
outside passed pawn. hard. But the Soviet grandmaster finds a
forced win, which involves going into a
5 fxg3 g41! pawn ending.

30
 The Problem o f Exchanging

1 ... h4! White resigns.

2 gxh4 Bxd4!
3 Rd2 e5 It is difficult to solve the problem of
4 Bxd4 exchanging when a transition into a rook
ending is in prospect. In rook endings
there are increased drawing tendencies,
and we will see how even the best grand­
masters in the world can go wrong when
transposing into a rook ending.

Gufeld—Dolmatov

Daugavpils, 1978

What is Black to do now? In the event

of 4 . . . exd4 White, of course, does
not fall into the trap 5 Rxd4+? KcS!,
but plays 5 Rf2!, after which Black’s
winning chances in the rook ending are
highly problematic. If Black plays 4 . . .
Rxd4+, the pawn ending after 5 Rxd4+
exd4 6 Kxd4 must end in a draw, e.g.
6 . . . h5 7 a4 a6 8 Kc4! Ke5 9 Kc5
Kf4 10 Kb6 Kg4 11 KxaS Kxh4 12
Kxaó Kg5 13 Kb7, with a draw. But With his two strong bishops, outside
Stein had envisaged in advance passed pawn, superior pawn formation
and well placed pieces, White has an un­
4 .. . Ke61! disputed advantage.

by which Black gains a decisive tempo. I Bxb5

5 h5 Rxd4+ Gufeld decides to take play into a

6 Rxd4 exd4 rook ending which appears highly pro­
7 a4 Kf6 mising for White. As became clear on the
8 Kxd4 Kg5 conclusion of the game, 1 Bel would
9 Ke5 have been stronger, retaining the advant-
age in a complicated ending.
In the variation 9 Kc5 Kxh5 10 Kb5
Kg4 11 Kxa5 h5 12 K a6h4 13 1 ... RxbS
Kxa7 h3 14 a5 h2 IS a6 hl=Q fo ra 2 Nc3 Rb3
draw White is short of just one tempo. 3 Rxd6 Rxc3
The game continued: 4 Rxc6 Rxa3
5 Rc8+ Ke7
9 ... Kxh5 6 Rc7+ Ke6
10 Kf5 a6 7 a7
 Endgame Strategy

The strongest continuation. After 14

Rg7 Kd4 15 Rxg6 Rxa7 1 6 R f6 K d 3
Black has good counter-play.

14 . . . Kb4
15 Kb2 Ra5!

15 . . . c3+ 16 Kc2 Kc4 17 Rc7+

Kd4 was bad due to 18 Kb3.

16 Re7

After 16 Rg7 Black gains counter-

As a result of forcing play, White has play by 16 . . . Kc5!, e.g. 17 Kc3
reached the position for which he was (17 R xg6R xa7 18 Rf6 Kd4 19 g6
aiming in carrying out the exchanging Kd3) 17 . . . Ra3+ 18 Kd2 Ra2+ 19
operation. Black’s position is difficult. K el R al+ 20 Kf2 Ra2+ 21 Kg3 c3
White threatens 6 g5, after which his 22 Rxg6 Rxa7 23 Rh6 (23 Rg8 Rc7!)
king approaches the c-pawn, and Black 23 . . . Ral! 24 Rh8 Ra6! 25 Rc8+
ends up in zugzwang. Rc6 26 Rxc6+ Kxc6 27 g6 c2 28
g7 cl=Q 29 g8=Q Qf4+ 30 Kh3
7 ... Ra2+! Qh6+! (but not 30 . . . Qxf3+ 31 Kb4
8 Kel! Qxe4?? 32 Qa8+), and Black gives per­
petual check.
White sees through his opponent's
camouflaged positional trap. On the 16 . . . c3+!
natural 8 Ke3? there would have 17 Kc2 Ra2+
followed 8 . . . c4!, and if 9 g5 fS, 18 Kbl
with the threat of mate by . . . f4.
On the natural 18 Kd3 Black should
8 ... g5! not lose after 18 . . . Rd2+ 19 Kc3
9 hxg5 Rd8 20 Rb7+ Kc4 21 Rc7+ Kb3
22 f4 exf4+ 23 Kxf4 c2 24 c5 Ra8
After 9 h5 White is not threatening 25 e6 Rxa7 26 Rxc2 Kxc2 27 Kc5
to place his opponent in zugzwang, since Kd3 28 Kf6 Ke4 29 e7 Ra8 30
the black king acquires the additional Kxg6 Kf4, and the black king succeeds
square f6. in latching on to the ’tail’ of the white
pawns.
9 ... g6
10 K dl c4 18 ... Ra6
11 K cl Kd6! 19 Rxe5 Rxa7
20 Rc6 Kc4
In a rook ending the possibility of 21 Rxg6
active counter-play often proves more
important than a material deficit. (See next diagram)

12 Rxf7 Kc5 White has now won a third pawn, yet

13 K bl Ra4 the game ends in a draw. Tartakover
14 Rd7! was probably right when he said: “ If it

32
 The Problem o f Exchanging

I t transpires that White does not

achieve anything by 26 Rf6+ Kxe4 27
Rf7 Rc6, when the position is a draw.
The finish was:

26 Re8 Kg5
27 g7 Rxg7
28 Kxc3 Kf6
29 Kd4 Rxg4
30 Kd5 Kf7
Drawn.

In the transition to a rook ending the

comes to that, a rook ending can be won stronger side must weigh up everything
only thanks to the quality of the pawns, ‘for’ and ‘against’ just as carefully as in
but not the quantity". the transition to a pawn ending.
We will now analyze a famous ending
21 . . . Kd3 between Capablanca and Alekhine,
22 Rd6+ Ke3 which played, in Alekhine’s words, a
23 Kc2 very important role in his subsequent
battle for the World Championship in
On 23 e5 there would have followed his match with Capablanca.
23 . .. Kxf3 24 e6 Re7!

23 . . . Kxf3
24 Re6 Rc7
25 g6

One gains the impression that White Black’s position is difficult, practically
will nevertheless win. If 25 . . . Kxg4, lost He has five pawn *invalids’, inferior
then 26 Rf6 Kg5 27 Rf7, winning. minor pieces, and unco-ordinated rooks.
But Dolmatov, who has conducted a Alekhine's opponent was the then World
difficult defence splendidly, crowns it Champion, die great master of endgame
fittingly with a brilliant king move. technique, Capablanca, and so no one
was in any doubt that Black would lose.
25 . . . Kf4!! Only Alekhine himself was not yet ready

33
 Endgame Strategy

to lay down his arms, and he was able comment:

to demonstrate that, even on the sun, "White overestimates his chances in
spots can occur. the resulting rook ending. He should
have aimed not for the exchange of
1 ... Bc6 bishops, but for the exchange of at least
2 Rd4 one pair of rooks, since this would have
practically eliminated his opponent’s
White tries to gain a material advant­ only counter-chance — the advance of
age as soon as possible, but allows Black his a-pawn. He should have played, for
to co-ordinate his forces. 2 NbS! was example, 9 Rd2 a5 10 Bd3 a4 11
prrferable. R bl R xbl 12 Bxbl Rb8 13 Bd3 a3
14 Be2, with the threat of Rd3, tying
2 ... Ng6 the black rook to the a-file, after which
3 Bd3 Nh4 it would not have been difficult for
4 Bfl Ng6 White to realize his advantage on the K-
5 Ne2 Ke7! side."
6 R el Rgb8!
7 Nxf4+ Kf8 9 ... Rb2+
10 Re2 Rab8
Black has lost a pawn, but has signifi­ 11 Be4
cantly improved his position.
Otherwise White cannot defend his a2
8 Nxg6+ pawn.

Capablanca connects his opponent's 11 « • « Rxe2+

pawns, but gains the prospect of creating 12 Kxe2 Bxe4
an outside passed pawn on the K*side. 13 fxe4 Ke7
14 Rd2 Ke6
8 ... hxg6 15 Ke3 c6!

9 Bd3?

The solving of an exchanging problem With his last move Alekhine had pre­
can sometimes be such a difficult matter pared veiled counter-play, hoping for
that mistakes can be made even by this very reply by White.
World Champions. Here is Alekhine's 16 c5 would not have achieved any­

34
 The Problem o f Exchanging

thing due to ló . . . Rb5 17 Rd6+ 25 Kf3 Ke7

KeS 18 Rxc6 RaS, but correct, as 26 Kg4 Kf7!
shown by Alekhine, was 16 h3! with
the idea of playing Kd4—c3 and c4—c5, Precise defence. Now on 27 Kh3
when Black would have been faced with Alekhine had prepared 27 . . . g5! and
difficult problems. if 28 Kg4 Kg6 followed by . . . gxh4
and . . . Re5-h5.
16 . . . Rh8!
17 g3 RhS! 27 Rc4 Kg7!
28 Rd4!
Capablanca obviously overlooked this
deep manoeuvre by his opponent. On Capablanca gives back his extra pawn
the fifth rank the black rook occupies and tries to exploit the remoteness of
an ideal position, and due to the possi­ the black king from the centre.
ble counter-attack by Black on the Q-
side, White is unable to exploit his extra 28 . . . RxcS
pawn and to create a passed pawn. 29 Rd7+ Kf8

18 Rh2 RaS Not 29 . . . Kh6 30 Rf7.

19 Kf4
30 Kf4 Kg8
19 g4? is very strongly met by 19 31 Ra7 Kf8
, .. Ke5. 32 a4 Kg8
33 g4?!
19 . . . f6!
“ After 33 Kc3! Rc3+ 34 Kd4
Black forestalls 20 g4, on which Rxg3 35 RxaS Kf7 36 Ra8 Black
there would follow 20 . . . g5+! would still have had to defend accurate­
ly, in view of White’s dangerous a-pawa
20 Rc2 RcS But thanks to the small amount of
material remaining, he would probably
21 c5 was threatened. have been able to draw. At any rate, this
would have been the logical continu­
21 c5 ation of the manoeuvre begun by White
on the 28th move.” (Alekhine).
“After this move, which restricts the
mobility of the black rook, the white 33 ... g5+!
rook is also tied to the cS pawn, and 34 hxg5 RxgS
Black has to reckon only with the threat 35 Ra6 RcS
of Kg4—h3 and then g3—g4.” (Alek­ 36 Ke3 Kf7
hine). 37 Kd4 Rg5
38 Rxc6 Rxg4
21 . . . Rh5 39 Rc5 Rg5!
22 Rc3 aS
In this position the players agreed a
Defending against 23 Ra3. draw, since the pawn ending after 40
RxgS fxgS 41 KeS Kg6! 42 Kd6
23 Rc2 Rc5 Kf7! 43 Kd7 Kf6 ends in a draw by
24 Rc3 Rh5 repetition of moves.
 Endgame Strategy

This game can be considered an

historic one, since it played an important
role in the subsequent battle between
Alekhine and CapaDlanca for the World
Championship. This is what Alekhine
writes about it in his book On the Road
to the World Championship:
“Despite this, I did take home with
me from this tournament one valuable
moral victory, and that was the lesson I
learned from my first game with Capa-
blanca, which had the effect of a revela­
tion on me. Having outplayed me in the
opening, having reached a won position
in the middlegame and having carried weakness is the a7 pawn, is far from
over a large part of his advantage into a easy.
rook ending, the Cuban then allowed me
to neutralize his superiority in that end­ 1 Rc6 Kf7
ing and finally had to make do with a 2 Bxd5
draw.”
Later Alekhine goes on: “I had finally Grandmaster Razuvayev makes the
detected a slight weakness in my future following w iny comment to this move:
opponent: increasing uncertainty when “ It is well known that ’rook endings
confronted with stubborn resistance! Of are never won’, but it can be assumed
that Rubinstein received indulgence
course I had already noticed Capablanca
committing occasional slight inaccu­ from Kaissa in this respect.’*
racies, but I should not have thought
that he would be unable to rid himself of 2 ... exdS
this failing even when he tried his ut­ 3 R ad Rfd8
most. That was an exceedingly import­ 4 Kfl Ke7
ant lesson for the future!” 5 Ke2 Rd6
So that the reader should not gain the 6 R6c3 R6d7
erroneous impression that the stronger 7 Kd3 Ra8
side should altogether avoid going into a 8 Rc6 Rd6
rook ending, we will analyze two ex­ 9 Ke2 Rxc6
amples in which the transition to a rook 10 Rxc6 Kd7
ending is a good way of realizing an ad­
vantage.

Rubinstein—Alekhine

Cuisbzd, 1911

(See next diagram)

White has the advantage thanks to his

control of the only open file, but to
pierce Black’s defences, whose only real

36
 The Problem of Exchanging

With the disappearance of one pair of 15 • • ♦ Kc6

rooks, White’s advantage has not been 16 R cl Kd7
reduced. But now both sides have to 17 R el RÍ7
reckon with the possibility of a pawn 18 R al Kd6
ending, and the play becomes more 19 R cl Kd7
tense and concrete. 20 Rc6 Rf8

11 f3! The illusion that Black’s position is

impregnable has been created. Rubin-
Rubinstein begins preparing pawn ad* stein now embarks on his active plan.
vanees in the centre (e3—e4) and on the
K-side ig3—g4). At the same time the 21 Ke2! Rf7
possibility of a pawn ending is eliminat­ 22 Kf2 Rf8
ed: after 11 . . . Rc8 12 Rxc8 Kxc8 23 Kg3 Re8
13 e4 fxe4 (13 . . . Kb7? 14 exf5 24 Rc3 Re7
gxf5 15 g4) 14 fxe4 dxe4 15 g4 Kd7 25 Kh4 h6
(15 . . . h5 16 f5 !) 16 Ke3 Ke6 17
Kxe4 White wins. 11 . . . a5 also does
not work, since, as shown by Razuvayev,
after 12 Rxb6 a4 13 Ra6! Rxa6 14
bxa6 Kc6 15 Kd3 Kb6 16 Kc3 Kxa6
17 Kb4 White obtains a won pawn
ending.

11 ... Re8
12 Kd3 Re7
13 g4 Rc6
14 Rcl!

Here White cannot go into the pawn

ending. After 14 Rxe6 Kxc6 15 g5 White has provoked an important
(15 e4? dxe4+ 16 fx e 4 fx g 4 ) 15 . . . weakening of his opponent’s K-side, and
Kd6 16 e4 Ke6 17 exd5+ Kxd5 18 Rubinstein now finds an excellent
Kc3 Ke6 19 Kc4 Kd6 20 d5 Kd7 manoeuvre after which Black’s defences
21 Kd4 Kd6 the game is drawn. collapse.

26 Kg3! h5
14 . . . Rc7
15 R h l As was shown by Kmoch, waiting
tactics would not have saved Black. c.g.
White's plan to improve his position is 26 . . . Rc8 27 Kf2 Rc7 28 Kc2 Rc8
to play his king to g5 via h4. But when 29 Kd3 Rc7 30 Rc6 Rc6 31 gxf5!
the opponent is deprived o f counter-play gxf5 32 Rxc6 Kxc6 33 c4, with a
and is forced to wait passively, it is use­ win in the pawn ending.
ful for the stronger side to avoid taking
positive action for a certain time, i.e. to 27 Kh4!
play according to the principle of “do
not hurry”. Such tactics often bring The white king reacts very keenly to
good results. the slightest advance o f the black pawns.

37
 Endgame Strategy

27 . . . Rh7 43 Ra8 was threatened.

28 Kg5 f*g4!
43 Kf6 R fl+
Alekhine docs not miss the chance to 44 Kg7 Rgl+
set his opponent a trap. After 29 Kxg6? 45 Kf8
g3 30 Kxh7 g2 31 R cl h4 it is Black
who wins. By a subde manoeuvre Rubinstein has
forced the enemy rook o ff the fourth
29 fxg4 hxg4 rank.
30 Kxg4
45 ... R dl
White's positional advantage has be* 46 Ke8 Rel+
come decisive. 47 Kd8 R fl
48 Rd7+ Kc6
30 . . . R hl 49 Ke8 Rf4
50 Re7 Kb5
“Passive defence is no longer possible. 51 Rc7!
If 30 . . . Re7, then 31 KgS Re6 32
Kh6! followed by Kg7—f7 and if Much stronger than 51 f8=Q Rxf8+
necessary Rc7+ followed by Kxe6." 52 Kxf8Kc4. Black resigned in view of
(Spielmann). the possible variation 51 . . . Re4+ 52
Kd7 Rf4 53 Ke7 Re4+ 54 Kd6Rf4
31 KgS R bl 55 Kxd5.

After 31 . . . Rgl+ 32
white king transfers to e5.
Karpov—Hort
32 Ra3 Rxb5
33 Rxa7+ Kd6 Tilburg, 1979
34 Kxg6 Rb3
35 f5 Rxe3
36 f6

The black king is cut o ff along the

rank and is unable to prevent the ad*
vanee of the white f-pawn.

36 ... Rg3+
37 Kh7 Rf3
38 f7 Rf4
39 Kg7 Rg4+
40 Kf6!

But not 40 Kf8? Rxd4, and the game The position is roughly equal. With his
ends in a draw. last move the World Champion offered
the exchange of queens, and Hort
40 . . . Rf4+ accepts the offer.
41 KgS R fl
42 Kg6 Rgl+ 1 ... Qxd2?!

38
 The Problem o f Exchanging

In the endgame White will have an

initiative. Preferable was 1 . . . Qf5.
with equal chances.

2 Rexd2 c3
3 Re2 exf2
4 Rxe8+ Nxe8
5 Kxf2 aS!

It was for this position that Black was

aiming when he went into the ending.
Objectively speaking, the position is
drawn, but only White can play for a
win. If the pawn structure is appraised.
White has a majority on the Q-side, and
Black on the K-siae. But there is no 19 a6? was bad due to 19 . . . Ra8.
sense in White creating a passed pawn,
since Black can easily blockade it. 19 ... Re7
White’s only active possibility on the 20 Bg2 Nb8
part of the board where he is stronger is 21 Rf4 Nd7
the advance b$—b4. 22 R fl Nb8
23 Be4 Na6
6 Ke3 Kf8 24 R bl Nb8
7 R bl Kc7 25 Bf5 Nd7
8 g4l 26 Ral

A useful move.

8 ... Kd8
9 b4 Re7+
10 Kd3 axb4
11 Rxb4 Kc7
12 R bl Nf6
13 a5 Nd7
14 Ral Nb8
15 h4 Na6

(See next diagram)

Black’s defences are successfully hold­

ing. He has no weaknesses, and Karpov A brief glance at the position is suffi­
resorts to manoeuvring tactics, with the cient to see that in the chess sense White
aim of provoking a weakening in the has achieved little. But psychologically
opponent’s position and cracking his something has been done. Black senses
defences. that his position is impregnable, and
weakens his vigilance. With his next
16 Rbl Nb8 move Hort tries to drive away the annoy­
17 Bf3 Nd7 ing white bishop, and commits a serious,
18 Ral Re8 possibly decisive, mistake.

39
ki-u
 Endgame Strategy

26 . . . g6?

26 . .. Nb8 was correct.

27 Bxd7!

In comparison with the position be­

fore the previous move, a slight change
has taken place: the g7 pawn has ad­
vanced one square. A slight change, but
one with enormous consequences. The
rook ending, which in the event of the
exchange on the previous move was
completely drawn, is now transformed A pretty breakthrough, typical of the
into one which is virtually won for endgame.
White. Karpov appreciates very subtly
the slightest change of position in the 31 hxg5
endgame. 32 h6 Kxa5
33 h7 Rd8
27 ... Rxd7 34 Rxf7 b5
28 R fl! Kb8 35 cxb5 Kxb5
29 Rf6 Ka7 36 Rb7+! Ka6
30 h5 Ka6?! 37 Rg7 Rh8
38 Ke4 Kb5
After this move Black loses by force. 39 Kf3 Kc4
His position was difficult, but he should 40 Rd7 Kd3
have tried 30 . . . gxh5 31 gxh5 Ka6. 41 Kg4 Rxh7
42 Rxh7 Kxd4
31 g5! 43 Rd7+! Resigns.
 CHAPTERS

“DO NOT HURRY”

The ability to make use o f this princi­

ple demands o f a player great experience
in the playing o f chess endings. How
many endings have not been won, mere­
ly because the stronger side tried to win
as quickly as possible, and neglected to
make simple strengthening moves before
embarking on positive action. Following
the principle of “do not huny”, it is
possible to battle for a win in positions
with a slight but persistent advantage.
Only in this way can a player achieve
weakenings in the enemy position, mask
his plans, and lull the opponent’s vigil­ position, and should switch to forcing
ance. But on no account should this action only if there is an immediate way
principle be abused. One must be ready to win. Here the transfer o f the king to
at the necessary moment to switch to b6 looks very strong, e.g. 1 Kc4 Re4+
sharp and positive action, otherwise the 2 Kb3 Re5 (3 Rxf7+ was threatened)
opponent may eliminate the weaknesses 3 Nd6 Re7 4 Ka4 Ke6 5 Rxf7 Rxf7
in his position, which are often o f a 6 Nxf7 Kxf7 7 KaS Ke6 8 Kb6 KdS
temporary nature. The ability to sense 9 hS, and wins. But White hurries into
and not miss this critical point is not taking specific action, and allows the win
something that comes easily. We will to slip from his grasp.
see in examples that in such situations
even strong players often go wrong. 1 Nd6?! Re7
The logic behind the “do not hurry” 2 Nxf7 Rxf7
principle is mainly psychological. It can 3 Rc8
be especially recommended to act
according to this principle when the This looks threatening, but not for
opponent is deprived o f active counter- nothing are rook endings regarded as the
play. most drawish. Black finds a defensive
manoeuvre which saves the game.

Mikenas—Spassky 3 ... Rd7+

4 Kc4
M oscow, 1955
Otherwise the black rook transfers
(See next diagram) to e6.

Black’s pieces are so cramped that he 4 ... Ke5

has practically nothing to move. In such 5 Rxc6 Rd4+
a situation “do not hurry” tactics are 6 KbS Rxg4
normally best for the stronger side. 7 hS Rg5
White should quietly strengthen his 8 Rg6 RxhS

41
 Endgame Strategy

9 c6 Kf4+
10 Kb6 R hl
11 c7 Rbl+
12 Ka6 R cl
13 Kb7 Rxc7+
14 Kxc7 h5

and Black gained a draw.

Retí—Romanovsky

Moscow, 1925
to us: Bd5, Rh4, Ke4, Rh5, h 2 -h 4 ,
g2—g4—g5 etc. Many players would
pi H have played this immediately, and would
m in i ill i possibly have won. But endgame techni­
¡ff§ M¡Ü ip ■ que is not only thinking in schemes, and
it dictates another move.
¡§ W m ■ ft u¡
aA U 1 4 a4!
§J ft H ft fü
[ft ^8 n ft ¡1 ü
Do not hurry! White exploits the
chance to improve the position o f his a-
pawn. Is the diversion worth while? Un­
B s H ¡ doubtedly. If later such a ’’trifle" is all
that is lacking to achieve a win, it will
In this position the opposite-coloured be a just punishment for disregarding the
bishops give White the advantage. The principle “do not hurry”. If at a4 the
pawns at cS and e5 create an impass­ pawn stands slightly better than at a2,
able barrier to the black bishop, and it should be moved there, and then the
White has a clear-cut plan to develop his implementation of the plan continued
initiative: transfer his rook to c4 and
king to f 3, and follow up with e2—e3 4 ... Ke7
and Bd5. The black rook will be tied 5 BdS Rc7
to the defence of the c-pawn, and White 6 Rh4 h6
will be able to switch his rook to the K- 7 Ke4 Kf6
side followed by the pawn break-through 8 RhS Rd7!
h2—h4, g3—g4 and g4—g5. Black is 9 g4
unable to hinder this plan.
White clearly overlooked Black’s
1 Rc4 Kf8 latent counter-play. He could have pre­
2 Kf3 Rc8 pared his offensive by Bc4, h2—h3 and
3 e3 Be3 g2—g4, without allowing the opponent
any counter-chances. But no one is in­
(See next diagram) sured against such oversights. In the
resulting complications the decisive role
This position is worth dwelling on in is played by the position of the white
some detail. White’s plan is well known pawn at a4.

42
 Do not Hurry

9 ■• . g6l?
10 Rxh6 Kg5
11 Rh7 Kxg4

Threatening mate.

12 Be6!

The only move to win.

12 . . . fxe6
13 fxg6 Rd8
14 Rxa7 Kg5
its exchange for the bishop, to d o which
“The bishop is exactly one tempo too he must play e3-e4.
late: 14 . . . Bel 15 a5 Bh4 16 a6 3. After improving the position of his
Bf6 17 g7 Rg8 18 Rb7 Bxg7 19 a7, king, invade with his rook into the
and now in view of the threat of Rb8 opponent’s position, attacking the weak
Black is forced to play 19 . . . Ra8, pawns at h7 and b7, then achieve the
when White takes the bishop and wins exchange of rooks and go into a won
easily" (Reti). minor piece ending.
It is interesting to follow with what
15 g7 Kh6 accuracy and artistry Flohr carries out
16 a5 Kh7 his plan.
17 a6 Rd6
18 h4 Bel 1 a5
19 h5 Bh4
20 h6 Resigns. With the threat o f 2 a6.

1 ... Rc7
2 Rh6+!
Flohr—Bondarevsky
It is such moves that reveal a mastery
Moscow, 1939 of endgame technique. To advance
a5—a6 White needs his rook at cl.
(See next diagram) Therefore he could have played Rhl
immediately, but after the check any
The advantage is with White. His reply by Black will very slightly worsen
knight is obviously stronger than the his position. Perhaps this “very slightly"
black bishop, and his central pawns are will not change anything, but neverthe­
more mobile. The mobility of Black’s less Flohr considers it necessary to give
Q-side pawn mass is highly restricted, the check in this position.
and his passed h-pawn is not very danger­
ous. White's plan is: 2 ... Bg6
1. Provoke . . . a6 and thus safeguard
himself against counter-play on the Q- Now the g-file is blocked to the black
side, while securing a post for his knight rook.
at c5.
2. Transfer the knight to c5, avoiding 3 Rhl Bf5?

43
 Endgame Strategy

Probably the decisive mistake. Black 13 Nd3 Kd6

should have played 3 . . . bé 4 R cl 14 Nc5 Rc7
Be8, retaining th e possibility of counter- 15 Rf8 Kc7
play. After 3 . . . BfS White forces
. . . a6, and Black's pawn formation be­ 16 Rb8 was threatened.
comes rigid.
16 e5 Re8
4 R cl a6
Black offers the exchange of rooks,
Forced, in view of the threat of 5 a6. because White’s pressure is increasing
with every move. The minor piece end­
5 R hl ing is lost for Black. After the loss o f his
h-pawn this becomes completely dear.
The rook has carried out its work on
the Q-side, and returns to keep Black’s 17 Rxc8 Bxe8
passed pawn under control. The rook 18 Kf4 b6
will then transfer to h4, where it will 19 Na4! bxaS
assist the advance e3 -e4. 20 bxaS Bt7
21 Nc5 Bc4
5 ... Rg7 22 Kg5 Be2
6 Rh2! 23 Kh6 Kd8
24 Kxh7 Ke7
Do not hurry! Black is deprived of the 25 Kg6 Bfl
slightest counter-play, whereas White has
a clear winning plan. This last rook move The plan White now chooses is possi­
again forces Black to worsen the placing bly not the shortest, but on the other
of one of his pieces. In such cases the hand it is the safest: he transfers his king
absence of concrete threats exerts to c5 via b4.
psychological pressure on the opponent.
It can be assumed that b y 6 Rh2! White 26 Kf5 Bh3+
gained a considerable amount of valuable 27 Ke4 Bg2+
time on the clock. 28 Ke3 Bfl
29 Kd2 Kf7
6 ... Rc7 30 Kc3 Kg6
7 Rh6+! Bg6 31 Kb4 Kf5
8 Rh4 BfS 32 Nb7 Resigns.

8 . . . KgS would have been enswered

by a check from g4. Boleslavsky—Goldenov

9 e4 dxe4 20th USSR Championship

10 fxe4 Bg6 Moscow, 1952
11 Rf4+ Ke6
12 Ke3 (See next diagram)

White centralizes his king, since the Black’s isolated pawns may become a
transfer of his knight to c5 is assured. target for attack. However, only his a-
pawn is a real weakness: his c- and e-
12 . . . Rg7 pawns are excellently defended by his

44
 Do not Hurry

control o f b5, the a-pawn may be attack­

ed by the bishop from this square, and
Black has to take this into account.

6 ... Ra7
7 Re2 Rb7
8 Kc2:

At first sight this king move is a con­

tinuation of White's tactics o f manoeuv­
ring to improve his position. In fact it is
a camouflage for a deeply-conceived
plan.
king, and also supported by his knight.
Therefore White plans the advance of his 8 ••• Rc7
pawn to h5, with the aim of giving Black 9 Bb5 Ra7
a second weakness. But this move also 10 Rd2 Ke7
has its drawbacks, since after the open­ 11 Re2 Kd6
ing of the file White's g-pawn may be­
come weak. Therefore, although he plans Black's vigilance has been weakened
this advance, White deliberately delays by the seemingly harmless and un­
it, and embarks on lengthy manoeuvring, threatening nature of White’s play. Now
with the aim of worsening the oppo­ Boleslavsky begins carrying out his long-
nent’s position and creating a favourable planned activity, a point which Black
moment for h4-h5. fails to perceive.

1 Bc4 Nd5 12 Re4! Nb6

2 Re5 Rb8 13 Bfl! Nd5?
3 Bd3 Rg8
“After this imperceptible mistake
Preventing h4-h5. Black loses his a-pawn. He should have
made any move with his rook along the
4 Kcl! seventh rank, although even then his
defence would not have been easy. For
A very strong move. The king vacates example: 13 . . . Rb7 14 Bh3 Re7
a square from which the bishop can 15 Kd3, and Black is in an unusual zug-
simultaneously attack the a-pawn and zwang. If 15 . . . Kd5 or 1 5 __ Re8,
assist the h4—h5 advance. 4 Bc2 Nb6 then White makes a favourable break­
5 Ra5 is now threatened, so that Black through by 16 f5” . (Boleslavsky).
is forced to play . . . c5, after which his
position in the centre becomes less 14 Bh3 Nc7
secure. 15 Bg4 Ra8

4 ... c5 (See next diagram)

5 Bc4 Ra8
6 Re4 16 Kcl!!

Aiming at the a4 pawn. Since the ad­ Had the king been on this square
vance of c-pawn has weakened Black’s earlier, Black would possibly have fore-

45
 Endgame Strategy

(c l) 27 . . . gxh5 28 Rxf5 h4 29
gxh4 Ke4 30 Rc5 Kxf4 31 Kc3 Kg3
32 Rxc4Rxa5 33 b4.
(c2) 27 . . . Ra2 28 h6 c3 29 Rb5
Kc4 30 Rb7 cxb2 31 Rxh7 Rxa5
32 Kxb2 Rb5+ 33 Kc2Ra5 34Rc7+
Kd4 35 Rg7 Ra2+ 36 Kb3 Rh2 37
Rxg6.

25 a4 Rb3
26 a5 Rxg3
27 a6 Kc7
28 Rb5 Rgl
seen the manoeuvre threatening him. But 29 Rb7+ Kc6
by his eighth move K cl—c2 White 30 Rxh7 R al
securely masked his plans from his oppo­ 31 Rg7 Resigns.
nent. It is interesting that earlier this
king manoeuvre provoked the important
weakening . . . c5, while now Black can­ Botvinnik--Kottnauer
not avert Bdl and the loss of his a-
pawn. Boleslavsky’s play in this ending Moscow, 1947
creates a strong impression.

16 . .. Ra7
17 Bdl Nb5
18 Bxa4 Nd4
19 Bdl Nf5
20 Bg4 Rb7
21 Bxf5

Taking play into a rook ending.

21 . . . exf5

Otherwise the e6 pawn is weak.

Black has a bad bishop, and his b-
22 Re3 Kd5 pawn is weak. Is this sufficient for White
23 Kc2 c4 to win? After all, his b-pawn also re­
24 Re5+ Kd6 quires defending. White must break up
the opponent’s K-side pawns, whereas
24 . . . Kd4 25 a4 would also not Black the transfer of his knight to c4
for
have saved Black. Boleslavsky gives the suggests itself.
following variations: 1 Kd2
(a) 25 . . . Rb3 26 a5 Rxg3 27 a6 To free the knight from the defence
Rgl 28 Ra5. of the b-pawn
(b) 25 . . . Ra7 26 a5 Ra8 27 b4
cxb3+ 28 Kxb3. 1 ... Nc6?!
(c) 25 . . . Rb4 26 a5 Ra4 27 h5, Better was 1 . . . Nd7 followed by
and: . . . Nb6—c4, but Black intends to

46
 Do not Hurry

transfer his knight to f5. 18 Kc3!

2 Kc3 Nd8 Black resigned, since after 18 . . . Bd7

3 h4 Nf7 19 Bfl he loses one of his pawns.
4 Nf4

White prepares h 4 -h5. Vaganian—Shereshevsky

4 ... Bd7 Minsk, 1972

5 Kd2 Be8
6 Bd3 Nh6
7 h5!

Earlier this would not have worked

due to . . . g5 followed by . . . h6, but
now h6 is occupied by the knight.

7 ... Nf5
8 hxg6

After 8 Bxf5 gxf5 White would not

have been able to win.

8 ... hxg6 White is the exchange up, bm if he ex­

9 Kc3 Kf7 changes the white-squared bishops this
10 Be2 Bd7 leads to a theoretically drawn position.
11 Nd3 Ke7
12 Nc5 Be8
13 Na6 Kd8
14 Bg4 Bd7
15 Nc5 Ke7
16 Kd3

“ Now White places his king on the

f l —a6 diagonal, in order to lull Black’s
vigilance and divert his attention from
the defence of the b5 pawn" (Bot-
vinnik). The camouflaging of the plan
resembles the previous example.

White cannot approach the g6 pawn,

16 . . . Be8 and hence the position is a draw.
17 Bh3 Ke8 Confident of a draw, Black makes the
first inaccurate move:
The decisive mistake. Black overlooks
White’s threats. He should have prepared 1 ... Bg4+?!
. . . g5 by 17 . . . Nh4. The immediate
17 . . . g5 would not have worked due He should not have voluntarily allow­
to 18 f4! ed the white king into the centre. To be

47
 Endgame Strategy

considered was I __ Be5, attacking the fore White continues making harmless
g3 pawn and the f4 square. moves, masking his intentions.

2 Kf4 Bc3 16 ... Bc3

3 Bd5 Bd2+ 17 Rb3 Bd4
4 Ke4 Bc3 18 Bf7 Kg7
5 Ra7+ Kf8 19 Rb7
6 Rc7 Bf5+
7 Kf4 Bd4 With the threat of a discovered check.

With the idea that on 8 Kg5 there is 19 . . . Kh8

a perpetual check by 8 . . . Be3+ and
9 . . . Bd4+. But not 19 . . . Kh6 20 Bg8.

8 Bf3 Bg7 20 Kd5 Bc3

9 Rb7 Bh6+? 21 Rb3 Bal
22 Kd6 Kg7
There was no necessity to allow the 23 Rb7
white king forward. Had Black guessed
at his opponent’s plan, he would possi­ Again threatening a discovered check.
bly have gained a draw. But Black’s
entire misfortune was precisely the fact 23 . . . Kh8
that he did not imagine that he could
lose this position. Note that White would Of course, not 23 . . . Kh6. I recall
not have achieved anything by 9 Bxh5 that at this point I jokingly regretted
Bh6+ 10 Ke5 Bg7+, but since the that it wasn’t possible to claim a draw in
pseudo-threat of capturing on h5 was view of the three-fold repetition of one
maintained, Black wanted to get rid of it and the same trap.
as soon as possible.
24 Bc4 Bd4
10 Ke5 Bg7+
11 Kd6 Bd4 Let us compare this position with the
12 Bd5 Bf2 one after Black’s 15th move. We see that
13 Rb3 Bd4 they are identical, except that the white
14 Rb8+ Kg? bishop is at c4 instead of d5, which is
15 Rb7+ Kh8 of no significance.

On 15 . . . Kh6 White has the un­ 25 Ke7!

pleasant 16 Bg8 g5 17 Rb5, while
15 . . . Kf8 did not appeal to Black The opponent has been influenced
due to 16 Rf7+ Ke8 17 Be6, when psychologically, and White can com­
his king is in a dangerous position. mence positive action.

16 Rb4! 25 . . . Bc3?

White’s plan includes gaining control The decisive mistake. Essential was
of f7 for his king. But on 16 Ke7 25 . . . Kg7 28 Ke8+ Kf6 followed by
Black can reply 16 . . . Kg7! 17 Ke8+ . . . g5. But Black reckoned that passive
Kf6, retaining drawing chances. There­ play would suffice for a draw, and did

48
 Do not Hurry

not want to bring his king out of its 35 Re7! Bg7

comfortable comer, since he did not
attach any significance to the difference On 35 . . . Be2 Black did not like
in the position of the white king at e7 36 Bd3!?
and f7. Evidently if White had played
his king to e7 on the 16th move, Black 36 Bd5 Be2
would have replied 16 . . . Kg7!, but at 37 Be4! Bb3+
the present time it was psychologically 38 Ke8 Ba4+
much more difficult for him to play this. 39 Kd8 Bb3
40 Rb7 Be6
26 Kf7!
With a trap: 41 Rb6?! Bf7 42 Ke7
With the white king at f7 the ending Bd4!, and after 43 . . . Kg7 the game
after the exchange of white-squared ends in a draw.
bishops is lost, and Black had not taken
this into account. 41 Ke7!

26 . . . Bd4 Black resigns. Against 42 Rb6 there

27 Rb5 Kh7 is no defence.
28 Rb3

The pursuit of the black bishops be­ Donchenko-Shereshevsky

gins. White has to hurry, since there are
not so many moves to go before Black Kaliningrad, 1 9 7 )
will be able to claim a draw on the 50-
move rule.

28 . . . Be4

Parrying the threat of 29 Bd3.

29 Be6

With the threat of 30 Rb4.

29 . . . Bc2
30 Ra3 Be5
31 Bc4 Bd4
In this position the game was adjourn­
So that if 32 Bd3? Bb2!, with a ed. Analysis showed that, in spite of his
draw. extra pawn, Black was unable to win,
but that to achieve a draw White would
32 Rf3 Be5 have to defend accurately.
33 Re3 Bd4
34 Re2! Ba4! 1 ... Bel
2 d6
The only move. Bad is 34 . . . Bf5
35 Be6 Bbl 36 R el Bc2 37 R cl, Forced, since 2 Bxh5 is bad due to
winning. 2 . . . Kxd5, while 2 Bg2 is met by

49
 Endgame Strategy

2 . . . Ke3 followed by the advance of 16 . . . Kf6

the f-pawn.
It is important for Black not to fright­
2 ... exd6 en off his opponent.
3 Bxh5 Bd2
17 Be8
Black achieves nothing by the direct
3 . .. Ke3 4 Bf7 f4 5 h5 f3 6 h6 f2 17 Kel is stronger. White fails to
7 Kg2. guess his opponent’s intentions.

4 Bf7 d5 17 . . . Ke5
5 h5 Ke4 18 Bg6 Kf4
6 Bg6 Bh6
The critical point of the ending has
After 6 . . . d4 White can hold the been reached. Had White played 19 Kel
position by 7 Kg2 d3 8 Kf2 Be3+ 9 he could still have drawn, since after 19
Kel Ke5 (9 . . . K f4 10 h6) 10 Bf7 . . . Kg4 20 Kdl d2 21 Be8 Be3 he
Kf6 (10 . . . f4 11 b6) 11 Bc4d2+ has 22 Bd7! Kg5 23 h6! But White
12 Kdl Kg5 13 Bd3 f4 14 Be2 Kh4 did not appreciate the fact that Black
15 h6 Kg3 16 h7 Bd4 17 Kxd2 f3 had embarked on active play.
18 Bxf3 Kxf3 19 Kd3, with a draw.
Therefore Black does not force events, 19 Be8? Bg5
but begins manoeuvring with his bishop, 20 Bf7?
so as to begin positive action in the most
favourable situation (in accordance with The losing move. White was obviously
the principle of “do not hurry”). expecting 20 . . .Kg4 21 Be6,when
he has everything in order. The best
7 Kg2 d4 drawing chance was probably 20 Bd7.
8 Kf2 Bg5 If instead 20 Bg6, then 20 . . . Bh4+!
9 Ke2 Be3 21 Kfl Kg5 22 Be8d2 23 Ke2 Bel,
10 Kel Ke5 and by the advance o f the f-pawn Black
11 Ke2 Ke4 wins the h5 pawn, diverting the bishop
12 Kel Bg5 from its defence. Note that it was al­
13 Ke2 Bh6 ready too late for 20 Kel due to 20
14 Kf2 . . . Ke3.

A move which indicates that White is 20 . . . Bh4+!

confident of a draw.
This White had n o t foreseen. The rest
14 . . . d3 is obvious.

Since the white king is in a dangerous

position, Black begins playing actively. 21 K fl Ke3
22 h6 Kd2
15 Kel Ke5 23 Bb3 Kel
16 Kf2<?) 24 h7 Bf6
25 Kf2 d2
The king moves away from the d3 26 Ke2 f4
pawn. White resigns.

SO
 Do not Hurry

In conclusion we will analyze two Rd8 6 Ra2 Rd3 7 Rc2 Rxh3 8 f 3,

examples in which it was imperative for and for a long time the black rook is
the stronger side not to delay taking shut out o f the game.
positive action.
2 Nel Be8

Kupreichik—Didishko Now the black bishop does no t head

for the defence of the a-pawn via the
Minsk, 1980 best route. To be considered was 2 . . .
Be6!P 3 Nc2 Bc8, so that if 4 Nb4
Bb7 or 4 Ne3 Be6, with drawing
chances. It should be mentioned that at
this point Didishko was in serious time
trouble.

3 Nc2 Bc6
4 Ne3!

The point: 5 Nf5 is threatened.

4, . . . g6
5 Nc4

The white pawns constitute a compact Loss of material for Black is inevit­
mass, restricting the black bishop. The able.
white rook is also much more active, and
Black’s outside passed pawn is securely 5 ... Rd8!
blockaded.
The best chance. After 5 . . . Ke7 or
1 ... Kf8?! 5 . . . Bb5 White wins a pawn in a more
favourable situation by 6 Nb6.
Black has to take measures against
N el—c2—b4. He can ensure the defence 6 Rxa6 Bb$
of his a-pawn using his bishop, while his 7 Rxd6 Rxd6
king heads for the centre to cover the 8 Nxd6 Bxd3
vulnerable points there. Black’s plan is 9 f3 Ba6
correct, but he implements it not al­ 10 c4
together exactly. The king can reach e6
in two moves via f7, while the bishop . Ke7 was threatened.
has to go to c8 via e6. Therefore by
1 . . . Be6 Black could have saved one 10 . . . Ke7
move, e.g. 2 Nel Kf7 3 Nc2 Bc8 4 11 c$ Kd7
Nb4 Bb7.
Interesting, but probably inadequate, If 11 . . . g5 12 Nf5+, and the
was the attempt to solve all the problems manoeuvre Ne3—d5 must win for
by tactical means: 1 . . . Rc8P! 2 Rxa6 White.
(if 2 c4, then 2 . . . Re6 followed by
the transfer of the bishop to c8) 2 . . . 12 h4 Kc6
d5 3 Ra3 dxe4 4 dxe4 Bg6 5 Nd2 13 Ne8 gS!

51
 Endgame Strategy

14 Nxf6? Kxgl 24 g7 Bb3 25 Kg5! (but not

25 Kg3 K fl 26 f4 Ke2 27 fxeS Ke3,
White has gained a won position. But with a draw) 25 . . . Kf2 26 f4 Ke3
Kupreichik hastens to increase his (no better is 26 . . . exf4 27 Kxf4,
material advantage, and makes the win followed by e4—e5) 27 fxe5 Kxe4 28
much more difficult. Following the rule Kf6, and wins. For a draw Black is short
“do not hurry!”, he should have played of one move, . . . Bg8. Were his bishop
14 h5!, winning. on that square, 28 . . . Kd5 would lead
to a draw.
14 ... gxh4
15 Nxh7 Kxc5 22 . . . Bb3!
16 Kg2 Be2 23 g5 BÍ7
17 Ng5 Kd6! 24 Kg4 Bg6
18 Nh3!
The bishop blocks the g-pawn and pre­
18 Kh3 Ke7 19 Kxh4 Kf6 leads tovents the advance of the white king. We
an amazing position, which, in spite of have a positional draw.
White’s two extra pawns, is drawn.
25 Kg3 Bf7
18 . . . Kc5! 26 Kg4 Bg6
27 Nh3 Be8
The black king reacts instantly to the 28 Kg3
movement of the white knight, and pre­
pares to go onto the attack. If 28 Kf5, then not 28 . . . Bd7+?
29 Kxe5 Bxh3 30 g6 and wins, but
19 Ngl Bdl 28 . . . Kxf3! 29 Kxe5 Bg6 with a
20 Kh3 Kd4 draw.
21 Kxh4 Ke3
28 . . . Bh5
(See next diagram) 29 Kh4 Bf7

22 Kg3?
Black sticks to the proven defensive
But this is procrastination. It was method, and avoids being diverted into
imperative for White to hurry! Correct calculating the variations after 29 . . .
was 22 g5!, e.g. 22 . . . Kf2 23 g6 Bxf3.

52
 D o not Hurry

30 Kg4 Bc8 defence.

31 Kg3 Bh5
32 Nf2 I ... Rb6!

White forces the opponent to capture The aim o f this move is to worsen the
on f3. position o f the opposing rook.

32 ... Bxf3 2 Rd3

33 g6 Be2
34 g7 Bc4 Forced. 2 Rb2 or 2 N cl is met by
35 Kg4 Kxf2 2 __ Rb4, winning a pawn.
36 Kf5 Ke3
37 Kxe5 Bg8 2 ... Ra6!
Drawn.
Creating a new threat of 3 . . . Ral+,
for a long time shutting th e white king
Lasker—Capablanca out o f play. Had Black not made the pre­
vious rook move to b6. White would
World Championship Match have had the strong reply 3 Rb2, with
Havana, 1921 counter-play.

3 g4 hxg3
4 fxg3
■ft
H wipB 1 Wm.
Totally bad was 4 Nxg3 R al+ 5 Kg2

||| ■
Nd6, when White has as many as four

WM ■i « P weak pawns.
m s ¡§ g 4 ... Ra2
ft m §¡ if ft 5 Nc3 Rc2!
J( 9 ô B
mm.
ft 1
m, ÉI IfH 1®sm
-- m With the threat of 5 . . . Nxd4.

6 N dl Ne7!
Black’s pawn formation constitutes a
compact mass, whereas White has two The black knight transfers to c6 to
weaknesses — at b3 and d4. But the attack the opponent’s weaknesses. With
distance between these weaknesses is every move White’s position deteriorates.
minimal, so that it is not easy for the
stronger side to exploit them. All White’s 7 Ne3
pieces are passively placed, and his king
is a long way from the main battle sector 7 b4 is b a d d u e to . . . R c l—b l.
- the b3 and d4 pawns. If in the dia­
gram position the kings were at d3 and 7 ... Rcl+!
d6 respectively, the game would end in
a draw. Therefore Black must all the The white king can now be allowed
time maintain the initiative, attacking into freedom, since the main role will be
White’s weaknesses and trying not to played by concrete variations. On the
allow the white king to come to their natural 7 . . . Rb2 there could have

S3
 Endgame Strategy

followed either 8 N dl, with a continu­ 13 . . . Nf5+

ation similar to the game, or 8 Rc3!? 14 Kf2 g5
Rd2 9 Nc2 Nf5 10 b4 Nxd4 11 15 g4
Nxd4 Rxd4 12 Rb3, with drawing
chances. There is no point in Black Passive defence would not have
allowing the opponent additional possi­ changed anything.
bilities.
15 ... Nd6
8 Kf2 Nc6 16 Ngl Ne4+
9 N dl Rbl! 17 Kfl R bl+
10 Ke2? 18 Kg2 Rb2+
19 K fl Rf2+!?
An oversight. White was bound to lose 20 Kel Ra2
his b-pawn, but he should have aimed for 21 Kfl
the exchange of rooks, since in the
knight ending Black would have en­ Forced. On 21 Nf3 Black wins by
countered certain technical difficulties, 21 . . . Nf2, or 21 Ne2 Ral+, 22 . . .
in view of the limited number of pawns. R xdl+, 23 . . . Nf2+ and 24 . . . Nxh3.
Correct was 10 K el!, defending the
knight (10 Ke3? Nb4), and if 1 0 __ 21 . . . Kg7
Na5 11 Kd2 Rxb3 12 Rxb3, with
drawing chances. Black’s plan includes playing his king
to d6 followed by . . . f6 and . . . c5.
10 . . . Rxb3! It is interesting to note that during the
11 Ke3 Rb4 twenty moves of this endgame Capa-
blanca has not made a single move with
his king — an exceptionally unusual
occurrence in his games.

22 Re3 Kg6
23 Rd3 f6
24 Re3 Kf7
25 Rd3 Ke7
26 Re3 Kd6
27 Rd3 Rf2+
28 Kel Rg2
29 K fl Ra2
30 Rc3 c5
31 Rd3 cxd4
Black has won a pawn. Now the 32 Rxd4
rhythm of the play changes sharply, and
the principle of "do not hurry" comes Bad is 32 Ne2 Rd2 33 Rxd4 Ng3+!
into force. Sec how calm Capablanca’s
actions become. 32 . . . Kc5
33 R dl d4
12 Nc3 Ne7 34 Rcl+ Kd5
13 Ne2
White resigns, since on 35 R d l there
On 13 g4 there follows 13 . . . g5. follows 35 . . . Ng3+ 36 Kel Rg2.

54
 CHAPTER 6

SCHEMATIC THINKING

Chess history knows of a number of on the K-side, and play against these
examples where, in a highly complex weaknesses promised more certain suc­
position, within literally a few minutes a cess. All this is very characteristic of
player has taken a decision, the correct­ modem chess. Of course, it is possible to
ness o f which has subsequently been have positions which allow the outlining
confirmed by lengthy analyses, although of an overall strategic plan, which the
to carry them out in actual play would opponent is unable to oppose. More
be totally unrealistic. Capablanca’s intu­ often plans have to be changed in ac­
ition was legendary, while Smyslov, cordance with changes in the situation
Petrosian, Karpov and many other on the board, caused by the actions of
players are renowned for their except­ the opponent.
ionally rapid and exact analysis of all the But thinking in schemes, in small com­
details o f a position. ponents of a plan, is necessary all the
In the endgame, schematic thinking time, except in highly tactical positions,
gives an experienced player the advant­ where general considerations fade into
age over an opponent who may be the background and give way to specific
superior to him in rapidity and depth of calculation. In the Znosko—Borovsky v.
calculation, but who relies mainly on Alekhine ending (p. 59) Black out­
this calculation. lined a highly complex plan which was
Schematic thinking should not be con­ brilliantly justified. However, it seems to
fused with the forming of a main stra­ us that it is much easier to describe such
tegic plan, although they have much in a plan after the completion o f a game
common. Both schematic thinking and than to form it during play. After all,
a general plan follow from a concrete had White on his second move played
evaluation of the position. For example, his pawn to f4, that would have been
in the Capablanca—Ragozin game (cf. the end o f Alekhine’s plan and he would
p. 1) White’s basic idea was to realize have had to form a completely different
his extra pawn on the Q-side. The one. It is more probable that during the
creation of a propitious moment for the game Alekhine was thinking in small
implementation o f this plan was pre­ schemes: exchange one pair o f rooks,
ceded by a great deal of preparatory retain the other, transfer the king to
work on improving the positioning o f e6, create a weakness fo r White on
the forces and on suppressing possible the K-side, and so o a By thus improv­
counter-play by the opponent. In doing ing his position, and not encountering
so Capablanca used logical set-ups o f his any resistance by the opponent at the
pieces, based on an evaluation o f the point when Znosko—Borovsky was
position, and designed to solve specific completely deprived o f counter-play,
and not very complicated problems. Alekhine was able to draw up his plan
During the course o f play one scheme in all its details and implement it in
was replaced by another, and at some full.
point Capablanca gave up altogether the We give the following game in full,
advance o f his extra pawn on the Q-side, since soon after the opening it went into
since Ragozin had acquired weaknesses an endgame.

CS-C
55
 Endgame Strategy

Janowski—Capablanca 11 ... e6
12 0 -0 Bd6
New York. 1916 13 Rfcl Ke7!
14 Bc3 Rhc8
15 a3?
1 d4 Nf6
2 Nf3 d5 A quite unprovoked weakening o f the
3 c4 c6 position.
4 Nc3 Bf5
5 Qb3 Qb6 15 . . . Na5
6 Qxb6 axb6 16 Nd2 f5!
7 cxd5 Nxd5
8 Nxd5 cxd5 Suppressing possible counter-play
9 e3 Nc6 with e3 -e4 .
10 Bd2
17 g3 b5
18 f3 Nc4

“ Black’s first plan is completed. White

now will have to take the knight, and
Black’s only weakness, the doubled b-
pawn, will become a source o f great
strength at c4. Now for two or three
moves Black will devote his time to im­
proving the general strategic position of
his pieces before evolving a new plan,
this time a plan o f attack against White’s
position” (Capablanca).

19 Bxc4 bxc4
10 . . . Bd7! 20 e4 Kf7
21 e5?
No prejudices. Capablanca thinks
schematically. Black’s plan includes A positional mistake, after which it is
transferring his knight to c4 after the unlikely that White’s game can be saved.
preparatory . . . b5. To support the With die centre closed. Black’s spatial
advance o f the b-pawn the bishop re­ advantage enables him without difficulty
treats to d7, whereas at the seemingly to prepare operations on the wings.
active position f5 it was out o f play. Correct was 21 exd5 exd5 22 f4!
followed by Nf3—e5.
11 Be2
21 . . . Bc7
In contrast to his opponent, Janowski 22 f4 b5
develops his pieces without any definite
plan. He should have considered playing It is difficult for Black to achieve
his bishop to b5, preventing the advance success by playing only an one wing,
o f the black pawn and preparing the de­ where White is able to hold the of­
velopment o f his king at e2 instead of fensive. Therefore, after preparing a
castling. break-through on one o f the wings and

56
 Schematic Thinking

tying down the opponent’s forces, a 34 Ral

blow must be struck on the other,
operating according to the principle of Or 34 R cl Rxf4+!
two weaknesses. Black first makes an
attempt to break through on the Q-side. 34 . . . Bc2
35 Bg3 Be4+
23 Kf2 Ra4 36 Kf2 h5
24 Ke3 Rca8
Loss of material is inevitable.
Threatening . . . b4.
37 Ra7 Bxg2
25 Rabl h6 38 Rxg2 h4
26 Nf3 g5 39 Bxh4 Rxg2+
27 Nel Rg8 40 Kf3 Rxh2
28 Kf3 gxf4 41 Bxe7 Rh3+
29 gxf4 Raa8 42 Kf2 Rb3
43 Bg5+ Kg6
Black readily switches his rooks from 44 Re7 Rxb2+
wing to wing. White’s lack of space, or, 45 Kf3 Ra8
as Nimzowitsch put it, his inferior ‘lines 46 Rxe6+ Kh7
of communication’, prevents him from White resigns.
keeping pace with his opponent.

30 Ng2 Rg4
31 Rgl Rag8 Bogoljubov—Lasker
32 Bel
Moscow, 1925
White has prepared for the defence of
his K-side. After the transfer of his
bishop to f2 followed by Ne3 he will
gradually neutralize Black’s pressure on
that part of the board. But just at this
point, when the co-ordination of the
white rooks is destroyed, the break­
through comes on the opposite wing!

32 . . . b4!

The inclusion in the game of the

white-squared bishop quickly decides
matters.
White has an undisputed advantage —
33 axb4 a rook and two connected pawns against
knight and bishop. It is very difficult for
In the event of the exchange of Black to find a reasonable plan of de­
bishops, the advance of the black h- fence, but he hit upon a brilliant defens­
pawn is decisive. ive formation and managed to save the
game. Not without reason have FIDE
33 . . . Ba4 awarded a Lasker medal for the best

57
 Endgame Strategy

game of the year in which defence

triumphs.

1 ... Ne5
2 Rd8

It is advantageous for White to ex-

change rooks.

2 ... Rc2
3 R8d2 Rc7
White’s subsequent plan could be as
An interesting point. 3 . . . Rc6 was follows: drive the black rook from c$,
objectively stronger. It would seem that transfer his own rook to c5, and by the
Lasker was masking his plan, hoping for threat of exchanging wrest control o f the
a greater effect if White should play fifth rank, and only then begin advanc­
routinely. Ke had obviously made a good ing the pawns over the entire front. Of
study of Bogoljubov’s at times over- course, it is easy to give such advice after
temperamental character, and was the analysis o f the position, but in a
assuming that White would try to ad­ practical game it would be extremely
vance his pawns in the centre without difficult to perceive the difference be­
sufficient preparation. There is also an­ tween the positions o f the pawns at a7
other possibility which cannot be ruled or a6.
out: Lasker had not yet formed the plan
in all its details. 4 ... Ke7
5 h3 Rc6!
6 f4 Nf7
4 Kf2?
7 Rd5?

An inaccuracy. This centralization of An incomprehensible move. True,

the king is untimely. Bogoljubov plans after 7 Ke3 Ra6 8 b3 Re6 Black
f3 -f4 , without considering any counter­ probably should not lose, since White
play by the opponent. He should first has advanced his pawns too early.
have broken up Black’s position by
4 Rd5 Nf7 (4 . . . Nc4 5 R d ) 5 Kf2! 7 ... Ra6!
(if immediately 5 Ra5 Rc2, while on 8 a3 Bc6
5 R ld 2 Black gains counter-play by 9 R5d4
5 . . . R c l+ 6 K f2 Bc6! 7 Ra5 a6,
when it is not easy for the white rook at Now Black’s plan o f defence takes
a5 to return to play) 5 . . . Ke7 6 R ld2 shape. The bishop has put the central e4
R cl 7 Rh5 h6 8 Ra5 a6 9 Rad5 pawn under fire, and it can advance only
Bc6 10 R d l Rc2+ 11 R5d2 Rc5 12 to a square attacked by the knight. The
Ke3. White has provoked the advance of black knight is also ready at any point to
the black pawn to h6, which is useful switch to f5 (with the white pawns at
for him, and also the very important (as f4 and e5) and can assist the undermin­
will be seen from the further course of ing o f White’s centre at an appropriate
the game) advance o f the pawn to a6. moment by . . . g5. While Black’s minor

58
 Schematic Thinking

pieces and king are holding the defence 24 Rc3! Nd6

in the centre, his rook breaks out via the 25 Rxc6! bxc6
Q-sidc and begins a counter-attack. 26 a5

9 • • • Rb6! The position has changed sharply.

10 b4 a5! Now Black is required to display a
11 bxa5 Ra6 certain accuracy, to avoid ending up in
12 R ld3 Rxa5 an inferior position.
13 Kf3 Rc5
26 • ♦ » c 5!
Each black piece is working to maxi­ 27 a6 Nb5
mum effect. 28 Ke3 c4

14 h4 The actions of the knight and the c-

pawn are co-ordinated to the maximum
White would like to play g2—g4, but extent. The white king’s passage to the
this is met by . . .g5! Q-side is blocked, and we have a posi­
tional draw.
14 . . . h5!
15 g4 hxg4+ 29 Kd2 Kd6
16 Kxg4 Nh6+! 30 Ke3 Ke6
Drawn.
Before beginning a counter-attack
with his rook from the rear, the enemy
king must be driven back.
Znosko—Borovsky v. Alekhine
17 Kg3 Rcl
18 a4 Rgl+ Paris. 1933
19 Kf3 Ral

19 . . . Nf5 also looks good.

20 R dl Ra3+
21 Rld3 Rxd3+!

Lasker avoids the repetition o f moves,

rightly assuming that after the exchange
of rooks it is White who will have to
fight for a draw. Unfavourable for Black
was 21 . . . Rxa4?! 22 Rxa4 Bxa4
23 f5! followed by 24 Kf4.

22 Rxd3 Nf5 This ending is highly instructive. The

23 h5 Ke6! position appears to be a ’dead' draw. It is
hard to imagine that, without the oppo­
Black threatens 24 . . . Nd6 25 Re3 nent blundering, one o f the sides can
Nxe4! with serious winning chances, hope for success. But that is precisely
but White has a possibility of gaining a what happened. Let us hand the word
draw. over to Alekhine himself:

59
 Endgame Strategy

“The play in this ending is by no 2 Kfl?

means so simple as it appears - especi­
ally for White. Black’s plan, which will After the correct 2 f4! White’s
prove completely successful, consists of chances would have been in no way
the following parts: (1) exchange one worse.
pair of rooks; (2) transfer the king to e6
where, being defended by the e-pawn, it 2 ... f5
can prevent the invasion at d7 by the 3 Rxd8+ Rxd8
remaining white rook; (3) operating with 4 g3
the rook on the open g-file and advanc­
ing the h-pawn, force the opening of the Defending against a possible . . . f4.
h-file; (4) after this White’s king, and
possibly his bishop, will be tied to the 4 ... Kf7
defence of h i and h2 against invasion 5 Be3 h5
by the rook; (5) Black meanwhile, by 6 Ke2 Ke6
advancing his a- and b-pawns, will sooner 7 R dl Rg8!
or later also open one of the files on the
Q-side; (6) since at this point his king Black confidently carries out his plan.
will still be on the opposite wing, White Three stages are already complete. It is
will be unable to prevent the invasion unfavourable for White to prevent the
of the first or second rank by the black advance of the rook’s pawn by h2—h4,
rook. It must be admitted that, had due to . . . Rg4.
White from the very beginning realized
that there was a real danger of turn losing 8 f3 h4
this ending, by careful defence he might 9 Bf2 hxg3
have been able to save the game. But 10 hxg3 Rh8
what happened was that Black played 11 Bgl Bd6
according to a definite plan, whereas 12 K fl
White played only with the conviction
that the game was bound to end in a The first weakness has been created,
draw. The result was an instructive scries and White’s king and bishop are tied to
o f typical patterns and strategems, much defending against the threats of the
more useful to students o f the game than black rook. Now the derisive stage o f the
the so-called ‘brilliancies’ of short one­ game commences. Black embarks on his
sided games.’* Q-side pawn offensive.
T o Alekhine’s words we can add that
this deeply conceived active plan is based 12 . . . Rg8
on the principle of two weaknesses. The 13 Bf2 bS!
first weakness of White’s position will be
the occupation by the black rook o f the Starting the attack. If White plays
h-file, the invasion squares along which passively there will follow . . . c 5 ,. . . c4,
White succeeds in covering with his king. . . . a5, . . . b4 etc. But this would have
The second and decisive weakness be­ been a lesser evil than that which occurs
comes the open file on the Q-side, where in the game.
the invasion cannot be prevented. It
should also be mentioned that a part of
any plan is the centralization o f the king. 14 b3? a5
15 Kg2 a4
1 Bh6 Rfd8 16 Rd2

60
 Schematic Thinking

On 16 b4 Black would have trans­ Black might just play 37 . . . Rcl+?,

ferred his rook to c6 via a8 and a6. when 38 Kd2 bl=Q 39 R xbl Rxbl
40 c7 follows.
16 . . • axb3
17 axb3 Ra8 37 . . . Kg3
38 c7 f3
The triumph of Black’s strategy! His 39 Kdl Rxc7
plan has been carried out. White has 40 Rxb2 f2
acquired a second weakness: the a-file White resigns.
occupied by the black rook. But the
game is not yet over.
Gligoric—Smyslov
18 c4 Ra3!
Candidates Tournament
Very strong. Zurich, 1953

19 c5 Be7
20 Rb2 b4
21 g4! ?

Realizing that his game is lost, White

seeks counter-chances.

21 ... f4
22 Kfl Ral+
23 Ke2 Rcl
24 Ra2

Otherwise after . . . Rc3 the white

pieces would be stalemated. In his book on the tournament,
Bronstein makes the following comment
24 . . . Rc3 on this position: “There exists a wide­
25 Ra7 Kd7 spread and therefore dangerous delusion
26 Rb7 Rxb3 that with an extra pawn the win is
27 Rb8 Rb2+ achieved automatically. Meanwhile, in
28 Kfl b3 the given position Black’s main advant­
29 Kgl Kc6 age lies not so much in his extra pawn,
30 Kfl Kd5 which cannot be realized for some time,
31 Rb7 e4! as in his control over many squares in
the central region of the board: d4, d5,
This energetic realization of his ad­ c5, f4 and f5.
vantage is typical of Alekhine. White has his counter-chances: a Q-
side pawn majority and the d-file. How
32 fxe4+ Kxe4 many such games have ended in a draw
33 Rxc7 Kf3 after inaccurate play! But Smyslov con­
34 Rxe7 Rxf2+ ducts such endings with an iron hand.
35 Kel b2 His plan divides into the following parts:
36 Rb7 Rc2 (1) Immediately exchange one rook,
37 c6! but retain the other for a possible battle

61
 Endgame Strategy

against White’s Q-side pawns and an But no t 19 . . . e5?? 20 Nd6 mate!

attack on the c4 and e4 pawns.
(2) By the threat of creating an out­ 20 Kd2 Kd4
side passed pawn, divert the white rook 21 c5 bxc5
onto the h-file, when his own rook can 22 Nd6 Ne5
occupy the d-file. White resigns.
(3) By the advance of the g-pawn to
g4. undermine the support of the e4
pawn —the white f-pawn. Polugayevsky—Vasyukov
(4) Tie down the white pieces by
attacking the e4 pawn. 34th USSR Championship
(5) Advance the king to win the Tbilisi, 1967
opponent’s weak pawns.
As we see, the winning plan is simple
—for Smyslov, of course.”

Rfd8
2 R adi Rxd2
3 Rxd2 Kf8
4 f3 Ke7
5 Kf2 h5!
6 Ke3 g5
7 Rh2 Rd8
8 Rhl g*

Black successfully advances towards

his goal. In this book there is hardly any
analysis of pure rook endings. In our
9 fxg4 Nxg4+ opinion, rook endings occupy a special
10 Ke2 Nf6 place in the classification of chess end­
11 Ke3 Rd4 ings. They are extremely complex and
12 R fl Ng4+ are subject to their own special rules and
13 Ke2 Kf8 principles, which often have nothing in
common with the principles of handling
Smyslov embarks on the final stage complicated endings. Frequently in rook
of his plan - he directs his king along endings, due to the abundance of possi­
the route e 7 -f8 -g 7 -g 6 -g 5 -g 4 . bilities, the actions of the two sides are
difficult to describe, the resulting posi­
14 Rf3 Kg7 tions are of an irrational nature, and the
15 Rd3 Kf6! choice of move has to be made mainly
16 Rxd4 on the basis of concrete variations. But
often we observe completely the oppo­
A desperate attempt to obtain counter- site picture. Concrete variations nave
chances. practically no significance, and the think­
ing is exclusively schematic.
16 ♦ ♦ « exd4 It is to this second type that the
17 Nb5 Ke5 Polugayevsky—Vasyukov ending belongs.
18 Nxa7 Kxe4 The weaker side’s method o f defence
19 Nc8 d3+! in such endings has long been known.

62
 Schematic Thinking

Black’s rook stands behind White’s pass­

ed pawn, and while the latter is advanc­
ing to b6, Black waits. When the white
king goes to the aid of the passed pawn,
Black wins one of the K-side pawns and
sets up a passed pawn on that part of the
board. After this he is ready to give up
his rook for White’s passed pawn, and
the ending normally reduces to king and
rook against king and pawn. The result
depends on the specific features of the
resulting position, of course, but the
game is more likely to be a draw than
for the side with the extra pawn to win. way immediately, by the dagger-blow
It would appear that in the given posi­ 1 hS! If Black captures on h5 or
tion the play should proceed according allows White to take on g6. White’s
to the scheme described. idea of creating a second passed pawn
is achieved in pure form, and a theo­
1 b5? retically won ending is reached.
During the game I was intending to
White’s sealed move is a serious mis­ play h4—h5 on my next move, but in
take, after which he should no longer my analysis I became aware that such a
have been able to win. Let us hand over hope was not feasible. After all, it was
th: commentary to grandmaster Poluga- now Black’s turn to move, and before
yevsky: posting his rook behind the white b-
“Only when I began my analysis did pawn, he could radically prevent all his
I discover a nuance in this position, and opponent’s aggressive intentions on the
a highly important one. The point is K-side, by first playing 1 . . . h5!
that, by advancing his pawn to b7, White I f in this case the white king were to
ties down the opposing king and rook, head for the b-pawn, play would proceed
and then, by an encircling manoeuvre as described at the very beginning, and
with his king, utilizing once again the (I have to ask you to take my word for
'triangulation' method, he wins the e5 this) White would at best be one tempo
pawn. But even after this, victory can be away from a win. However much I
achieved only if he creates a passed pawn racked my brains, I couldn’t find a win
on the f-file. for White.
If instead White wins the e5 pawn by
(See next diagram) ‘triangulation’ — which is possible —
then he succeeds in creating a passed
By playing f5—f6+, White prevents pawn only on the g- or h-file, which is
the black king from moving between the not good enough to win.”
squares g7 and h7, and after . . . Kf7
he wins by Rh8, while in the event of 1 ... Rb4?
■. . Kxf6 he has the opportunity for a
deadly check: Rf8+ and b8=Q. Obviously Black had failed to discover
In the adjourned position the white f- the essence of the position, and with his
pawn has no opposite number, but the very first move after the adjournment he
black g6 pawn stands in its path. This commits a decisive mistake. As already
pawn could have been cleared out of the mentioned, 1 . . . h 5! was correct.

63
 Endgame Strategy

2 h5! gxh5 nothing.

3 b6?! Therefore, in reply to 3 . . . Rb3+
White would have had to try 4 Kh4.
It is easy to understand Poluga- But after 4 . . . e4! (4 . . . Rb2, how­
yevsky’s joy on seeing the first move ever, is also possible) the tempting
after the resumption. But emotions, even 5 Kxh5 leads only to a draw after the
positive ones, are not always a good help quiet 5 . . . Rb4!!, when White is doom­
in chess. Pleasurably anticipating the ed to carrying on the fight ’a king down’,
implementation of his plan. White plays since he dare not step onto the ‘mined’
too hastily and allows his opponent a 4th rank. The thematic 6 f4 is just one
latent possibility of counter-play, which tempo too slow: 6 . . . e3 7 f5 e2 8
Vasyukov fails to exploit. 3 Kf3! was Re8 Rxb6 9 Rxe2 R bl. Also, 6 g4
correct. does not change anything: the further
advance g4—g5 is all the same impossi­
3 ... h4+? ble, in view of the reply . . . Rb5!”

Let us again hand over to grandmaster 4 Kf3 Kh7

Polugayevsky:
“Saving chances were offered by 3 No better is 4 . . . Kg6 5 b7 Kh5
. . . Rb3+!, when an amazing, study-like 6 g4+! hxg3 7 fxg3, when Black is un­
draw results after 4 f3 e4 5 b7 (or able to defend against the break-through
5 Kf2 Rb2+ 6 Ke3 Rxg2 7 b7 Rb2 8 g4+ and 9 gS!
8 fxe4 h4, and the black h-pawn is no
weaker than either of its white oppo­ 5 b7 Kg7
nents) 5 . . . h4+! (but not 5 . . . e3 6 Ke3 e4
6 f4 e2 7 Kf2), and after 6 Kf2
Black is saved by the straightforward Waiting tactics would also have been
6 . . . h3, and after 6 Kxh4 e3 7 Kg3 unsuccessful If 6 . . . Kh7, then 7 Kd3
by the highly subtle 7 __ Rb4!!, when Kg7 8 Kc3 R bl 9 Kc4Rb2 10 Kd5
White is in zugzvxmg. and the e-pawn is lost, since 10.. .Rb5+?
loses immediately to 11 Kc6!, when
White’s king approaches the b7 pawn
with gain of tempo, after which he wins
by moving his rook along the eighth rank.

7 Kf4 Kh7
8 Ke5 Kg7
9 Kd5 Rb2

The e-pawn was doomed. If 9 . . .

Kh7, then 10 Kc5 Rb2 11 Kc6Rc2+
12 Kd5 Rb2 13 Kxe4.

10 Kxe4 Rb4+
He has no other move than 8 f4 (the 11 Kd3 Rb3+
exchange of the b7 pawn for the e3 12 Kc4 R bl
pawn leads to a theoretically drawn end­ 13 f4
ing), but then 8 . . . e2 9 Kf2 Rxf4+
10 Kxe2 Rb4 once again gives White This was the sort of position White

64
 Schematic Thinking

was aiming for at the start of die ad- passed pawn.

2 Kb3 b5
13 ... Rcl+ 3 Kc2 a4
14 Kd3 R bl 4 Rh6 Kc5
15 f5 Rb6 5 Rxg6 Rxh4
16 f6+ Resigns. 6 Rg8 Rg4

We will now analyze another rook Both sides follow the pre-planned
ending which in many respects resembles scenario.
the previous one.
7 g6 Kb6

Smirnov—Shereshevsky But not 7 . . . Kc4? 8 g7 Rg2+ 9

Kbl Kb3 10 Rc8!, when on 10 . . .
Minsk, 1979 Rg l+ there follows 11 R cl Rxg7 12
Rc3 mate!

8 g7!

The advance of the king towards the

pawn at g6 was totally unpromising.

8 ... Kb7
9 Kd3

White’s plan begins to take shape.

After the arrival of the white king at f3
the black rook will be forced to allow it
across the fourth rank. The white king
On the K-side an exchange of pawns then intends to break through at a5.
is bound to take place, after which White
will remain with an extra passed pawn
on the g-file. The black rook will stand
behind it, and, if the white king should
head for the g6 pawn, Black will pick
up the b-pawn and set up a passed pawn
on the Q-side. The game will end in a
draw. This would be the normal pro­
cedure in this ending.

1 ... a5

It is useful for Black to remove his

pawns from the seventh rank. Besides,
the further the a-pawn is advanced, the Here White sacrifices his g-pawn by
more quickly it will be possible to create Rf8, and after . . . Rxg7 plays Rf5,
a passed pawn, if White should give up picking up Black’s Q-side pawns and win­
his b-pawn to take his king across to his ning. But Black finds a defence.

65
 Endgame Strategy

9 ... Ka7 is no win!

10 Ke3 Kb7
11 Kf3 Rgl 16 axb4 Rg6+
12 Kf4 Rg2 17 Kc5 Rg5+
13 Ke5 Rg3 18 Kc4 Ka7
14 Kd5 Rg5+ 19 b5 Rg4+
15 Kd6 20 Kc5 Kb7
21 b6 Rg5+
22 Kb4 Rg4+
23 Kb5

Black is again in zugzwang. He is

forced to give up his a-pawn, but this is
of no significance.

23 ... Rg5+
24 Kxa4 Rg4+
25 b4 Rgl
26 Ka5 Rg5+
27 b5 Rxb5+!
Drawn.
Up till now Black has stuck to waiting
tactics, but now he is in zugzwang. His
rook must not allow the white king to
reach c5, while his king must guard c6.
By the method of elimination it is easy
to find the only move, but one which
proves sufficient for a draw.

15 . . . b4!

Black sacrifices his Q-side pawns, but

breaks up his opponent’s pawn form­
ation. An amazing ending is reached, in
which White is three pawns up, but there

66
 CHAPTER 7

THE PRINCIPLE OF TWO WEAKNESSES

Black having two obvious weaknesses

(pawns at b4 and e6). White chose to
go into the ending, since Black had no
compensation for the two pawn weak­
nesses on opposite wings.

1 d4 Nf6
2 Nf3 b6
3 e3 Bb7
4 Bd3 e6
5 0 -0 Be7
6 Nbd2 d5
7 b3 0 -0
8 Bb2 c5
9 Ne5 Nc6
king across to the K-side and eliminate 10 a3 a6
the black pawns. 11 f4 b5
The first weakness in Black’s position 12 dxc5 BxcS
is the white passed a-pawn; the second is 13 Qf3 Nxe5
his K-side. If we slightly change the 14 Bxe5 a5
position by adding a white pawn at c4
and a black one at c5, White will be un­ 14 . . . Ne4 was seriously to be con­
able to win, since one weakness — the sidered.
extra a-pawn — is insufficient, and
Black's K-side is impregnable. As Nim- 15 Qg3 g6
zowitsch expressed it, in this second 16 f5 Nh5
example White lacks a ‘manoeuvring 17 fxg6! fxg6
pivot' - the square d4 for his king. The
principle of two weaknesses frequently A sad necessity. Of course, 17 . . .
determines the plan in an endgame, as Nxg3 failed to 18 gxh7 mate, while
we will see in numerous examples. on 17 . . . hxg6 there would have
followed 18 Rxf7!!, and now:
(a) 18 . . . Rxf7 19 Qxg6+ Kf8
Shereshevsky—Belyavsky (19 . . . Ng7 20 Qb7+ K f8 21 Bxg7*
Ke7 22 Bf6+) 20 Qh6+Ke7 21 Bg6.
Lvov, 1977 (b) 18 . . . Nxg3 19 Rg7+ Kh8 20
Rxg6+ Kh7, when White has at least a
After a rather uninteresting opening draw by perpetual check, but can also
and a complex but transient middle- continue his attack with 21 Rxg3+ Rf5
game, in which White gained an advant­ 22 R fl, with very dangerous threats:
age by tactical means, he was faced with (b l) 22 . . . Bc8 23 Rff3 (23 Rb3+
a choice: to win the exchange with a Kg6 24 g4 Qg5 25 Rg3 is also possi­
sharp and unclear position, or to go into ble) 23 . . . Bf8 (otherwise Rb3+ wins)
an ending with material level, but with 24 Rg4 Bh6 25 Rh3 with a decisive

67
 F.ndgame Strategy

attack. 25 axb4 axb4

(b2) 22 . . . Bxe3+ 23 Rxe3 Qb6 26 Nf3
24 Rff3 Kg8 25 Bxf5 exf5 26 c3,
with a big advantage. White’s plan is to transfer his knight
to e5 and his king to e3, and only then
18 Qh3 ?! to lay siege to the weaknesses. Black is
powerless to avert loss of material.
After 18 Rxf8+ Qxf8 19 Qh3
White would have won the b-pawn.
26 . . . Ne8
18 . . . Qb6 27 Ne5 Nd6
19 Qg4 Rf5 28 Kf2

The threat of Bxg6 had to be parried,

but Black should have included the inter­ Allowing the exchange of bishops, but
mediate check 19 . . . Bxe3+ 20 Khl this is no longer of any importance.
Rf5, with some counter-play for the ex­
change, although after 21 Bxf5 exf5
22 Rxf5 Bxd2 (22 . . . Bc8 23 Qf3) 28 ... Ba6
23 Rxh5 White has an obvious advant­ 29 Ke3 Bxd3
age. Now White has a choice. 30 Nxd3 Rc8
31 Nxb4 Rc3+
32 Kf4 Kf8
33 Rf3 Rc7
34 Ke3+ Ke7
35 Kd2 Ne4+
36 Kcl

White has won a pawn while main­

taining a positional advantage.

36 . . . Nc3
37 Re3 Kd7?
20 Bd41? 38 Na6 Resigns.

Forcing an ending in which Black has

two pawn weaknesses on opposite wings, In the above ending White exploited
and his minor pieces are unable to de­ two black pawn weaknesses on opposite
fend them. In addition Black is deprived wings. But the concept of a ’weakness’
of counter-play. is much wider than that of a lone (iso­
lated) pawn which can be subjected to
attack. A weakness may be the occupa­
20 ... Bxd4 tion of an open file by an enemy major
21 Qxd4 Qxd4 piece, an enemy outside passed pawn, an
22 exd4 Rxfl+ immobile piece, a king which is cut off,
23 Rxfl b4 and so on. In short, a weakness is pri­
24 g4 Ng7 marily a positional defect.

68
 ' ' The Principle o f Two Weaknesses

Alekhine—Samisch which was possible after 7 Qe4, and

with the aim o f fixing the weak pawn at
Baden Baden, 1925 h7. O f course, not 7 Bxh7 Qxf3 8
Qxb6? Q dl+ with a draw.

7 ... Qb5
8 h6 Qb3
9 Bc2! Qb5
10 Qd3 Qxd3
11 Bxd3

White has achieved the exchange of

queens in the most favourable circum­
stances: the black king is tied to the de­
fence of the h-pawn, and the knight is
quite unable to cope with the passed
b-pawn supported by king and bishop.
White h is an extra passed b-pawn
(Biack’s first weakness). However, the 11 . . . Nc8
immediate advance of this pawn would 12 Bxh7 Resigns.
expose the white king and give Black
serious drawing chances. Therefore
Alekhine sets about creating a second Tartakovcr—Boleslavsky
weakness in Black’s position.
Groningen, 1944
1 Qd4

"By this and his following move White

selects the correct winning plan, which is
the advance of his K-side pawns. The
passed b-pawn must be advanced only
later, when with the exchange of queens
the danger of perpetual check will be
eliminated" (Alekhine).

1 ... Qc7
2 Bd3! Qc7
3 g4 Kf7
4 h4 Nb6
5 h5 gxh5 Black has an undisputed positional ad­
6 gxhS vantage, and he carries out a manoeuvre
which wins a pawn by force.
The second weakness — the pawn at
h7 —has been created. 1 . . • Rcxc4
2 Nxe7+ Kf8
6 ... Qc6 3 NdS Rxcl!
7 Be4! 4 Rxcl Bb2!
5 Rc8+ Kg?
Avoiding the exchange of queens 6 K fl Rxa4

69
 Endgame Strategy

Thus White's first weakness is Black’s worsened the position of the white king
passed pawn on the Q-side. Guided by by 20 . . . Rf4 21 Ke2, and only then
the principle of two weaknesses, Black played 21 . . . Kd5, when 22 Ne3+
directs his efforts towards giving White can be met by 22 . . . Bxe3 22 Kxe3
a second weakness —on the K-side. a5, advancing the pawn to a4.

7 Ne3 Bd4 21 Ne3+ Bxe3

8 Nc2 Bb6
9 f3 Kf6 "After 21 . . . Kc6 22 Nf5 Black
is unable to win, due to the weakness of
The creating of weaknesses in the his f-pawn. For example: 22 . . . Kd7
opponent’s position is normally pre­ 23 Re7+ Kc8 24 Rf7 Bd8 25 Nd6+
ceded by improving the placing of the Kb8 26 Rb7+ Ka8 27 Rf7 Ra6 28
pieces: the black bishop has taken up Nb7 Bb6 29 Nd6 Bc7 30 Rf8+ Bb8
an excellent post at b6. Now Black 31 Ne4 Ra3+ 32 Ke2 Ra2+ 33 Kd3
activates his king. Rh2 34 Rxf6 Bf4 35 Rf5 Rxh3 36
Nxg5, with a draw" (Boleslavsky).
10 Ke2 Ke6
11 Kd3 Kd7 22 Kxe3 a5
12 Rf8 Ke7 23 Rd8+?
13 Rb8 h5
Missing a chance to save the game.
Black at last sets about creating a Had White anticipated his opponent’s
second weakness in White’s position. plan, by 23 Ra8 he would have pre­
vented the regrouping . . . Rf4 and
14 g3 . . . a4, followed by the advance o f the
black king on the Q-side.
This move could have been avoided,
but after . . . h4 and . . . Bgl White 23 . . . Kc6!
would have been left with a weak square
at g3. The game is decided. The remainder
does not require any explanation.
14 ... Bgl
15 h3 Bb6 24 Rc8+ Kb7
16 g4 h4 25 Rc8 Rf4
17 Rh8 g5 26 Kc2 Kb6
18 Rg8 f6 27 Rc3 Kb5
28 Kd2 a4
Black has carried out his plan.- White 29 Kc3 Kc5
has been given weak pawns at f3 and h3. 30 Kc2 Rb4
Now Black decides the game by advanc­ 31 Rc6 Rb6
ing his king. Since he too has acquired a 32 Rc4 Rb3!
weak pawn at f6, care and accuracy 33 Rxa4 Rxf3
are essential. 34 Ra6 Kd4
35 Kd2 Ke5
19 Rc8 Kd6 36 Kc2 Rxh3
20 Rc8 Kd57! 37 Kf2 Rd3
38 Ra5+ Rd5
Too hasty. Black should first have 39 Ra3 Rd4

70
 The Principle o f Two Weaknesses

40 Ra5+ Ke6 Preventing . . . Kd6.

41 Ra6+ Kf7
White resigns. 5 ... Rb6
6 Rb2 h6
7 Kf2 Ke6
Alekhine—Vidmar 8 Kf3 Nd5

Hastings, 1936 Black offers to go into a rook ending,

but White declines and continues break­
ing up the opponent's position on the K-
side.

9 h4! Ne7
10 Be4 Nd5
11 Rb3 Kd6

After a move by his rook Black would

have had to reckon with the exchange of
minor pieces and the advance of White’s
passed pawn.

12 g5!
White has an extra pawn at b4. Alek­
hine himself assessed this position as Alekhine consistently carries out his
follows: “White’s winning plan is easy to plan.
explain, but rather difficult to cany out.
White exploits the fact that the black 12 . . . hxgS
pieces are occupied on the Q-side to 13 hxg5
create, by the gradual advance o f his
pawns and their exchange, vulnerable The preparatory work is complete.
points in Black’s position in the centre The black pawns at e$ and g7 are
and on the K-side. Only after this pre­ separated and weak.
paratory work can the decisive offensive
be begun”. In other words, Alekhine in­ 13 . . . Ke6
dicates the need to give the opponent a 14 Bd3 Kd6
second weakness. 15 Ra3 Nc7
16 Ra7! Rb8
1 g4! Ke7
2 b5 e5 16 . . . Nxb5 17 Rxg7 would not
have left Black any chance o f saving the
Depriving th e white rook o f d4. game.

3 f4 f6<?) 17 Ke4 g6
It was better to exchange on f4, since
now Black’s e-pawn is isolated. 18 Kf5 was threatened.

4 fxe5 fxe5 18 Ra3! Rb6

5 Ra2 19 Bc4 Rb8
 Endgame Strategy

19 . . . Kc5 is well met by either

20 Ra7 or 20 Be2, and the knight
cannot take on b5 due to Ra5.

20 Rd3+ Kc5
21 Rd7!

The outcome of the game is decided.

21 ... Ne8
22 Bf7 Nd6+
23 Kxe5 Rb6
24 e4 Nxb5
2 Rg2+! Kf5
Black has succeeded in eliminating
White’s passed pawn on the Q-side, but
on the K-side White has become total Equally cheerless for Black is 2 . . .
master of the position. Kxh4 3 Nf3+! Kh3 (3 . . . KbS 4
Rg5+ Kh6 5 NxeS) 4 Rg6 Re8 (or
25 Rd5+ Kb4 4 . . . Bg3 5 Ng5+ and 6 Ne4)
26 Rd8 Na7 5 Nxe5 Rxe5+ 6 Kd3 Rd5+ 7 Kc2,
27 Rd6 Nc6+ when the rook ending is easily won for
28 Kf6 Kc5 White, since the black king is cut off,
29 Rd5+ Kb4 e.g. 7 . . . Kh4 8 a4 Kh5 9 Rg7 a5
30 e5! Kc4 10 R g2c5 11 Rd2 Rxd2+ 12 Kxd2
31 R dl+ Kc5 c4 (otherwise b2—b4) 13 Kc3 Kg6
32 Rcl+ Kd4 14 Kd4, and wins.
33 e6 Ke3
34 Bxg6 Nd4 3 Nf3 Bf4+
35 Bf7 Ne2
36 R el Kf2 On 3 . . . Bf6 White has the un­
37 Rxe2+ Resigns. pleasant 4 Rd2 Re8+ 5 Kf2, with
numerous threats.

Kovalyev—Azos 4 Kd3 c5

Tashkent, 1978 The first part o f White’s plan is com­

plete: Black’s Q-side has been weakened
(See n ext diagram)
5 Kc4 Kc4
White has an extra passed pawn at h4 6 Ng5+ Kf5
— Black’s first weakness. White’s plan is
to give Blade a weakness on the Q-side 6 . . . Kc3 can be met by 7 Rg4 Re8
(provoke . . . c5 or . . . d5), and then 8 Kd5 Re5+ 9 Kxd6 Rxg5 10 Rxf4.
use the passed h-pawn to divert Black’s
forces, giving it up when necessary to be 7 Kd5
able to attack Black’s weakened Q-side.
The h-pawn can now be given up: the
1 Ke3+ Kg4 invasion of the white king is decisive.

72
 The Principle o f Two Weaknesses

7 ... Rxh4 (a) fix Black’s K-side pawns, since other­

8 Nf7 Rh5 wise the bishop may escape from its
9 Nxd6+ Bxd6 ‘prison’; (b) wrest the d-file from the
10 Kxd6 opponent (in doing so he must weigh up
whether or not the minor piece ending
The outcome is decided, although in is won or whether a pair of rooks must
the rook ending White has to overcome be retained); (c) give Black a second
certain technical difficulties. weakness on the Q-side, since he already
has one, and a very serious one at that —
10 ... Ke4 the bishop at g7; (d) if possible, invade
11 Rg7 a6 the opponent’s position with decisive
12 Rc7 Rh2 effect.
13 Rb7 c4
14 Kc5 Kd3 1 Rfdl
15 a4 Rh5+
16 Kb6 Rh6+ White begins the battle for the d-file,
17 Ka7 a5 exploiting the fact that 1 . . . c4 loses a
pawn to 2 Nd6.
18 a5 and 19 Rb6 was threatened.
1 ... Kf8
18 Rb6 Rh8 2 Kfl
19 Ka6 Ra8+
20 Kb5 Ra7 The two players centralize their kings.
21 Rb8! White is not tempted into winning a
pawn by 2 Ng5 Ke7 3 Nxh7 Bh6 4
Within a few moves Black resigned. h4, after which his knight is exchanged.

2 ... Ke7
Alekhine—Euwe 3 c4!

London, 1922 3 . . . c4 was now threatened.

3 ... h6
4 Ke2 R xdl
5 R xdl Rb8

We give here Alekhine’s commentary

on this position: “ Black cannot ex­
change rooks, since after 5 . . . Rd8 6
Rxd8 Kxd8 White wins as follows:
1st phase: 7 h4 followed by g2—g4
and g4—g5, against which Black has
nothing better than . . . h5, since the
exchange o f pawns will give the white
knight the square h4.
White has the advantage. The differ­ 2nd phase: b 2 -b 3 followed by Kd3,
ence in strength between the white Nc3 and Ke4.
knight and the black bishop is obvious. 3rd phase: the transfer o f the knight
But to realize his advantage he must: to d3, which ties the black king to d6,

73
 Endgame Strategy

in order to hold the twice-attacked e5 22 Ke2 Rf7

pawn. 23 Nc3!
4th phase: finally f2—f4, forcing the
win o f the g- or e-pawn, after which The start of the concluding attack.
White wins easily.
By avoiding the exchange of rooks 23 ... Re7
Black makes his opponent’s task more 24 g5 hxg5
difficult”. 25 hxg5 Kc6
26 Kd3 Rd7+
6 Rd3 Bh8 27 Ke4 Rc7
7 a4! 28 Nb5 Re7
29 m Kd7
White sets about creating weaknesses
on the Q-side. Since 7 . . . a5 is not 29 . . . Kb7 1< to 30 Nd6+ and
possible due to the loss of a pawn after 31 Nc8, while if __ Rf7 30 Rc8+.
8 Rb3, the opening of the a-file is in­
evitable. This file will be occupied by the 30 Rb8 Kc6
white rook, and this will become Black’s 31 Rc8+ Kd7
second weakness. 32 Rc7+ Kd8
33 Rc6 Rb7
7 ... Rc8 34 Rxe6 Resigns.
8 Rb3 Kd7
9 a5! Kc6
10 axb6 axb6
11 Ra3 Bg7 Karpov—Parma
12 Ra7 Rc7
Caracas, 1970
Now after the exchange of rooks
White could carry out his winning plan,
but for this there is no necessity. At the
moment his rook occupies an ideal posi­
tion, and he has a quicker way to win.

13 Ra8! Re7
14 Rc8+ Kd7
15 Rg8! Kc6
16 h4

Do not hurTy! Black is deprived of the

slightest counter-play, and so before the
decisive offensive White ‘packs in’ the
bishop at g7 by h2—h4 and g2—g4—g5. Black has a weak d-pawn, which can
be finally nailed down by 1 Qd3 follow­
16 ... Kc7 ed by e3—e4. This is probably what the
17 g4 Kc6 majority of players would have done.
18 Kd3 Rd7+ But how then can Black be given a
19 Kc3 Rf7 second weakness? After making a deep
20 b3 Kc7 assessment of the position, Karpov takes
21 Kd3 Rd7+ what is at first sight a quite paradoxical

74
 The Principle o f Two Weaknesses

decision: he relieves Black of his weak­ 19 Rcl f6

ness at d6.
Parma succumbs to the opponent’s
1 f4 Re8 positional pressure and weakens the
2 fxe5! seventh rank. Karpov skilfully exploits
this error.
“2 f5 was also possible, but with the
opponent having only one weakness (the 20 Qd2 Kf7
pawn at d6) the win would have been 21 Kh2 Ke8
more than difficult. But now, in addition 22 Rd6 Rd7
to the weak e5 pawn, there are other 23 Rdl Rcxd6
advantages in my position, as for ex­ 24 cxd6 Qe6
ample control of the open file by the 25 Qd3 Qa2
rooks and the possibility of combined 26 Qxa6
play" (Karpov).
White has won a pawn. The game is
2 ... dxeS decided.
3 c5 Re6
4 Qd3 bxcS 26 ... Qc2
5 bxcS Qc6 27 Qa8+ Kf7
6 R bl Qc7 28 Qd5+ Kg7
7 Rfl Rf8 29 Rd2 Qc3
8 Khl Qc6 30 Ra2?! h5!
9 R bl 31 Rd2!

“White does not have, and there can­ “There was the threat o f a perpetual
not be, any clear-cut plan for realizing check after . . . h4. Yes; for many
his spatial advantage, since his specific players ‘taking a move back* is the most
goals vary depending on the opponent’s difficult thing of alL Now the rest is
replies. The ideal for White would be to easier" (Karpov).
force the advance . . . f6, when, with
the seventh rank weakened, he would 31 • • • h4
gain the opportunity of establishing a 32 Rdl Qc2
rook at b6 and beginning a combined 33 a6 Qa4
heavy-piece attack along the open files. 34 Qd3 g5
However, Black's position is most pro­ 35 Rbl f5
bably still defensible, although his task 36 Rb7 g4
is not an easy one” (Karpov). 37 hxg4 fxg4
38 Qe2 Resigns.
9 ... Qc7
10 e4 Rb8
11 R fl Rb7 Benko—Parma
12 Qc3 Rb5
13 a4 Rb8 Belgrade, 1964
14 Rcl Rc8
15 Rbl Kg8 (See next diagram)
16 R bdl Qe7
17 R fl Rc7 White’s bishop is much stronger than
18 a5 Rec6 the enemy knight, which does not have

75
 Endgame Strategy

draw.

4 ... Nxe8
5 Kf3 Kf8
6 Ke4 Nc7

Black has managed to check the oppo­

nent’s first onslaught. The white king
cannot break through on the Q-side.
Now White’s task is to create weaknesses
in Black’s K-side.

7 h4 Ke7
any strong points. Also actively placed 8 f4 h6
are the white rooks, whereas the black
rooks are merely carrying out defensive After 8 . . . h5 White has the highly
functions. unpleasant 9 Ke5 followed by f4—f5.

1 Bb3 9 Bc4!

Threatening to exchange on e8 A typical manoeuvre, which is repeat­

followed by Rd7, which did not work edly encountered in this ending. The
immediately due to . . . Ne5. black knight comes under the domina­
tion of the white bishop.
1 . .. Nf6
2 Kg2l 9 ... Ne6
10 g4 Nc7
White begins the centralization of his 11 Ke5
king, not fearing the exchange of rooks
on the e-file. Black is in zugzwang. Further posi­
tional concessions are inevitable.
2 ... Rxel
11 . . . f6+
Black could hardly have avoided this 12 Ke4 Kf8
exchange for long. 13 h5!

3 R xel Re8 Capablanca’s advice on the placing of

4 Rxe8+! pawns in endings with bishops (cf.p.110)
is in the majority of cases correct, but
White has accurately worked out his it pays not to be dogmatic. In the given
possibilities in the ensuing minor piece position it is more important to gain
ending. Contrary to the generally- control of the key e5 square for the
accepted rule that a rook and bishop are king — the ‘manoeuvring pivot’ for an
stronger than a rook and knight, White attack both on Black’s Q-side, and his
exchanges rooks so as to centralize his K-side. It should be noted that the
king as quickly as possible. After 4 routine 13 g5 would have allowed
Rcl?! Re7! Black would have not 13 . . . hxg5 14 hxg5 Ke7 with the
allowed the white king into the centre, threat of 15 . . . fxg5 and 16 . . . Ne6,
and would have had every chance of a when Black has excellent chances of a

76
 I The Principle o f Two Weaknesses

successful defence. 29 Ka6 Nf6

30 Bc6! Nxg4
13 . . . g5 31 a5 bxa5
32 b6 Ne5
No better is 13 . . . gxh5 14 gxh5 33 Be8 Resigns.
followed by 15 Kf5, when Black is in
w giw m g.
Faibisovich—Westerinen
14 fxg5 fxg5
15 Kf5 Kg7 Vilnius, 1969
16 Ke5

The break-through of the white king

to the Q-side is now inevitable.

16 . . . Ne8
17 Be6 Nf6
18 b4!

Do not hurry! Before the decisive in­

vasion of the king it is useful to advance
the Q-side pawns as far as possible.

18 . . . Ne8
19 b5! At first sight Black’s pawn weaknesses
— g6 and a5 —seem easily defended. But
Taking a concrete approach to the they can be subjected in turn to a com­
position, White places all his pawns on bined attack, because it is easier for
squares of the colour of his bishop. White to transfer his pieces, especially
from wing to wing.
19 • ■ • Nf6
20 a4 Ne8 1 Rd5!
21 Bf5 Nf6
22 Ke6 Nc8 The white rook occupies the staging
23 Be4! post from where it heads for g5, to tie
down the opponent’s forces to the de­
The black knight again comes under fence of the weak g-pawn.
the domination of the bishop.
1 ... Rg7
23 . . . Nf6 2 Rg5 Bd2
24 Bf3 Ng8 3 Ke2
25 Kd6
The king makes for Black’s second
Now to the Q-side! weakness —his a5 pawn.

25 . •• Kf6 3 ... Bc3

26 Kc6 Ke7 4 Kd3 Bel
27 Kb7 Kd6 5 Kc4 Kf7
28 Kxa7 Kc5 6 Kb5 Bc3

77
 Endgame Strategy

Again reminding Black of the weak­

ness of his g6 paw a

22 ... Re3
23 Kc7 Re7+
24 Bd7 Bf2?
25 f5!

The finish was:

25 ... Re3
26 Rxg6+ Kf7
27 Bc6+ Ke7
7 Rd5! 28 Rg7+ Kf6
29 Rf7+ Ke5
The white rook again occupies d5. 30 a7 Resigns.
White now threatens Rd3 and c2-c3, a
plan which Black is powerless to oppose. This ending is a classic example of
manoeuvring when there is an attack on
7 • • • Bel two weaknesses.
8 Rd3 Kf6
9 c3 Re7
10 Bb7 Re2 Holzhausen—Nimzowitsch
11 Kxa5 Rc2
12 Kb5 Bxc3 Hannover, 1926
13 a5

Black has won back his pawn, but

now White’s passed pawn advances
irresistibly. The game is decided.

13 . . . Bel
14 a6 c4

The best practical chance.

15 bxc4 Bf2
16 Ra3 c6+

If 16 . . . Ba7, then 17 Bd5 follow­ Even before Nimzowitsch many

ed by Kc6. masters understood the principle of two
weaknesses and were guided by it. But
for a clear-cut and easily understood ex­
17 Kxc6 Rxc4+ planation of this principle we are un­
18 Kxd6 Ba7 doubtedly indebted to Nimzowitsch.
19 Bc6 Rb4
20 Kd7 R bl 1 ... Rh6
21 Ra5 Rb3
22 Rg5 “ A move which arises from an exact

78
 ' The Principle o f Two Weaknesses

knowledge of the laws of alternation. 14 . . . Rxa4+

The point is that sooner or later Black 15 Kb3 Rb4+
will have to resort to . . . a6 followed 16 Kc3 Rb7
by • ■■b5, which, with the white pieces
fully committed to watching the e-pawn,
will give Black chances of invading (by “ In this ending, as soon becomes
opening the a- or b-file). This would dear, the K-side (the h3 and g4 pawns)
create the pivot necessary for the success cisive weakness” (Nimzowitsch).
of the alternating operation; what would
be lacking, however, would be the
essential ‘two weaknesses’, since the 17 Rf5 Ra7
necessity for White to ‘watch’ the e4 18 Kc4 Ra4+
pawn constitutes, of course, only one 19 Kb3 Rd4
weakness. The manoeuvres in the game 20 Re5 Kd6
(J . . . Rh6 etc.) have the aim of creat­ 21 Re8 Rd3+
ing a ‘second weakness’, the presence of 22 Kc4 Rxh3
which will acquire decisive significance 23 Rxe4 Ra3
in the rook ending which later ensues" 24 Re2 Ra4+
(Nimzowitsch). 25 Kb5 Rxg4
26 a4 Rb4+
2 h3 Rg6 27 Ka5 h5
3 Re2 a6 28 Rh2 Kc6
4 Rf4 b5 29 Re2 Rg4
5 b3 Rg5 30 Rc6+ Kd5
6 g4 Rge5 31 Re8 h4
7 Kc3 a5 32 Rd8+ Kc4
8 Ref2 a4 33 Kb6 h3
9 bxa4 bxc4 34 R dl Kb4
10 Rf8 R5e7 35 R bl+ Kxa4
11 Rxe8 Rxc8 36 Kxc5 g5
12 Nxc4 Nxc4 37 R hl Rg3
13 Kxc4 Ra8 38 Kd4 g*
14 Rf7 39 Ke4 Rg2
40 Kf4 h2
If 14 Kb3, then 14 . . . Kd5. White resigns.

79
 CHAPTER 8

THE STRUGGLE FOR THE INITIATIVE

Every chess player knows how import­ from the d-file, i.e. advance his Q-side
ant it is to seize the initiative during a pawns. But it is completely unclear
game. Nowadays one cannot hope for which would be the more important: the
success without being prepared to battle advantages White would gain by advanc­
for the initiative in a roughly equal ing his pawns, or the dangers and weak­
position. It is no accident that World nesses which would arise as a result of
Champion Anatoly Karpov and his pre­ this.
decessor Bobby Fischer have demon­
strated in many games their readiness at 1 Qb4
any moment to seize the initiative, not
Schlechter does not wish to take the
being afraid to take risks. For the sake of
the initiative one sometimes has to dis­ initiative.
regard a possible deterioration in one’s 1 ... c6
position. Courage, strength of will and
judicious audacity are needed for a Of course, not 1 . . b6 2 Qa4 a5,
player to decide on a step which is fre­ when Black’s Q-dide pawns are markedly
quently into the unknown. Just how weakened and lose their flexibility.
difficult and dangerous this can be,
especially in the endgame, will be shown 2 Qa3 a6
by the following examples. 3 Qb3 Rd8
4 c4

Schlechter—Lasker So as not to allow . . . d5. But the

pawn at d6 is by no means a weakness.
Vienna, 1910
4 ... Rd7
5 Q dl Qc5
6 Qg4 Ke8!
7 Qe2 Kd8

“ Lasker gradually assumes the initia­

tive and by extremely subtle play tries to
gain minute advantages. First the king
frees the rook from the defence of the
Q-side pawns” (Tarrasch in his Die
Modeme Schachpartie).

8 Qd2 Kc7

The position is level. The only active White has a pawn majority on the K-
plan which might be suggested for White side, while Black has an extra pawn on
would be to try and exploit his pawn the Q-side. The first signs of activity on
majority on the K-side. To do this he Black’s part have appeared. His king has
would have to remove the black pawn transferred to the part of the board

80
 The Struggle for the Initiative

where he has every right to develop the 14 Q dll?

initiative. But the simple advance of
Black’s pawns will merely lead to their Preparing a3 -a4 or Qh5.
exchange. It is necessary that White
should weaken his position on the Q- 14 . . . f6
side. It is this aim that is pursued by the
appearance there of the black king. The On 14 . . . Ra8 White has the un­
king acts as a nrovocator, inducing White pleasant 15 Qh5.
to throw his pawns forward. But to seize
the initiative in a level position, without 15 Qb3?!
sacrificing anything in return, is impossi-
ble. The slightest attempt to play active­ Schlechter is confused. He wants to
ly is often fraught with great danger. play actively, but does not wish to
weaken his position. 15 Qh5 would
9 a3 Re7 not have achieved anything due to 15
10 b4 . . . Qe6, but he should have played
15 a4!? After missing this opportunity
White has to switch to passive defence.

15 . . . Qe6
16 Qdl

With exact play White should perhaps

not lose the rook ending, but Black has
an undisputed advantage.

16 . . . Rh8!

Before attacking on the Q-side, Lasker

reduces the value of White's extra pawn
on the K-side.

“Very bold and energetic play, as one 17 g4 Qc4

might expect from the great master. Al­ 18 a4?!
though this move exposes the black king,
at the same time the white a-pawn is Inconsistent. White has opted for pass­
made backward. Black thus hopes to ive defence, which is of course undesir­
acate a passed pawn on the c-file” able, but in principle is possible. Since
(Tarrasch). earlier he avoided playing a3—a4 in a
much more favourable situation, he
11 cxb5 axb5 should certainly not have done it now.
12 g3 g5 Passive defence can be successful if the
opponent’s advantage is only very slight.
Lasker prevents f2—f4, and thus In such play one must retain complete
assures his queen of an excellent post in composure and ensure that the advant­
the centre. At the same time, however, age does not grow to considerable pro­
the h5 square is exposed. portions. To be considered was 18 Qf3
with the threat of 19 Rc3 and e4—e5.
13 Kg2 Rc8 Here is Tarrasch’s comment on this move:

81
 Endgame Strategy

"The opponent’s steady strengthening after 25 . . . Ra5! with the threat of

of his position begins to frighten White, 26 . . . Ra3 White would have been
and he stakes everything on this move, forced into a queen ending: 26 Rb3
sacrificing a pawn for an attack. But in Qxb3 27 Qxa5+ Kb7. Black would
fact, because of this advance he should have had every chance of winning.
have lost the game, even if it did give
him certain chances. There was as yet no 26 Ra4 c4
cause for desperation: all his weaknesses 27 Qal Qxe4+
(a3, e4 and h3) were sufficiently de­ 28 Kh2 Rb5
fended by his rook, and if he had avoid­
ed the exchange of queens, he could With the threat of 29 . . . Qe5+. De­
have readily continued play.” spite his two extra pawns, Black can
hardly have any serious hopes of win­
18 ... Qxb4 ning the game. The position of his king
19 axb5 Qxb5 is too dangerous.
20 Rb3 Qa6
21 Qd4 29 Qa2! Qe5+
30 Kgl Qel+
Black is a pawn up with a sound posi­ 31 Kh2 d5
tion.
By the queen checks Black has taken
21 . . . Re8 control of a5, and is now ready to meet
22 R bl Re5! 32 Ra7+ with 32 . . . Rb7

Defending against 23 R al. 32 Ra8!

23 Qb4 Qb5 Threatening 33 Qa7+ and 34 Qc5+.

24 Qel Qd3
25 Rb4 32 . . . Qb4

So as to answer 33 Qa6 with a queen

check at d6.

33 Kg2!

This cool king move renews White’s

threat.

33 . . . Qc5

Black should have reconciled himself

to the fact that after 33 . . . Rb8 34
Ra7+ Rb7 35 Ra8 the game could not
25 . . . c5? be won.

Fatigued by the hard struggle, Lasker 34 Qa6!

plays carelessly and throws away the
win. Black wins a pawn, but exposes his White’s attack is now irresistible. On
king too much. As shown by Tarrasch, 34 . . . Rb7 there follows 35 Qe6.

82
 The Struggle for the Initiative

34 ... Rb8 strong pressure. By the move in the game

35 Ra7+ Kd8 Bronstein sets his opponent a difficult
36 Rxg7 Qb6 problem: whether to allow the advance
37 Qa3 Kc8 of the black king to the centre when a
draw will be the most likely outcome, or
and, without waiting for the obvious 38 whether to continue the struggle for the
Qf8+, Black resigned. The game shows initiative by sacrificing a knight at f5.
just how risky the struggle for the initia­
tive can be in a level position. 2 Nexf5!?
By his uncompromising play Lasker
almost won, but he lacked the strength Suetin accepts the challenge.
to take the game to its logical conclusion.
2 ... exf5
3 Nxf5 Bg7
Suetln—Bronstein 4 Rg2!P

Moscow, 1968 4 Rxh8 Bxh8 5 Ne7 would have

won White a third pawn, but hardly the
game.

4 ... Bf8
5 Ne3

White has a spatial advantage, and a

powerful knight at d4, attacking the
weak e6 pawn. He has possibilities of
active play involving c3—c4 or the
sacrifice of a knight at f5. Black is re­
quired to defend accurately. A crucial position. White is threaten­
ing to capture on d5, obtaining three
1 ... b5! pawns for his sacrificed piece. The pair
of connected passed pawns at f4 and e5
Preventing c3—c4. Black could have are ready to sweep away everything in
approached the d5 pawn with his king their path. In this difficult situation
by 1 . . . Kc7, after which the knight Bronstein finds a brilliant defensive plan,
sacrifice at f5 would have promised one of the chief links in which is the
White little, but then White could have centralization of his king.
fixed the b6 pawn by 2 a4 and
followed up with c3—c4. For example: 5 ... Kb7!
1 . . . Kc7 2 a4! Bg7 3 Rxh8 Bxh8 6 Nxd5 Kc6
4 c4 dxc4 5 Nxc4 Bg7 6 Be3, with 7 Ne3 Ne6!

83
 Endgame Strategy

Following his king, Black’s entire ization of the black king: 16 . . . Kd5?
army has thrown itself into the battle 17 Rd8+, and 17 . . . Kxe5 fails to
with the opponent's main forces. Now 18 Ng4+.
White achieves nothing by 8 f5 Nf4
9 Rg4 Rh2+. 16 . . . Rh4
17 e6?
8 Ndl!
In his eagerness to win, White over­
Suetin continues to battle for the reaches himself. He should have forced a
initiative. The white knight vacates the draw by 17 Rc8+ Kd5 18 Rd8+ KxeS
diagonal for the bishop, while defending 19 Ng4+ Kxf5 20 Rd5+. Now the
the c-pawn. black king approaches the enemy passed
pawns.
8 ... Rh3!
17 ... Rf4+
Black consistently activates his forces. 18 Kg3 Kd6
19 Rf8 Be3
9 Rg6 20 Ng4

9 f5 can be met by 9 . . . Ng7 10 20 Nh3 o r 20 Nd3 is very strongly

Ne3 Ng3, blockading the passed pawns. met by 20 . . . Rc4.

9 ... Rh6 20 ... Ne4+

10 Rg8 Bg7! 21 Kh3 Rf3+
22 Kh4 Bg5+
Cutting off White’s rook from his re­ 23 Kh5 Rh3+
maining pieces. 24 Kg6 Rh4!

11 f5 Rh2+ White loses his knight, and his king,

which has gone to the aid of his passed
Only now is Black’s plan revealed. pawns, comes under a mating attack by
After 12 Kbl Rh8 13 Rxh8 Bxh8 14 the concerted action of the black pieces.
fxe6 Bxe5 his king eliminates the white
e-pawn, and the game ends in a draw. 25 Nf2 Rh6+
Suetin chooses a different way, which 26 Kg7 Nxf2
meets with energetic resistance by Bron- 27 Ra8 Ng4
stein. 28 Rxa6+ Ke5
29 Rb6 Bf6+
12 Kd3 N4c5+! 30 Kf7 Kxf5
31 Rxb5+ Ne5+
Obviously White had not expected White resigns.
this move.

13 bxc5 Nxc5+ He can defend against the mate only

14 Ke3 Bh6+ by giving up his rook.
15 Kf3 Bxcl Bronstein was awarded the prize for
16 Nf2! ‘The most interesting ending’. Also com­
mendable is Suetin’s persistent striving
Suetin prevents the further central- for victory, right to the end.

84
 The Struggle for the Initiative

Suetin—Gufeld 2 cxbS!?

Tbilisi. 1969 Highly original and bold! White ex­

ploits a tactical possibility and carries
out a combination, the consequences of
which were n o t easy to assess. O f course,
he could have played 2 b3 and then
prepared g3-g4, when it is difficult to
say how real his chances would have
been. By the move in the game White
risks losing his advantage, but he sharp­
ly complicates the play and sets his
opponent difficult problems.

2 ... axb5

2 . . . cxb5 loses to 3 BxbS! axbS

Not one pawn has yet been exchang­ 4 c6+.
ed. White holds the initiative and a spa­
tial advantage. He has the possibility of 3 Bxb5 Rxa3
a pawn break both on the K-side (g3—g4 4 Rxa3 cxbS
after appropriate preparation), and on 5 c6+ Kxc6
the Q-side (b2—b3 and c3-c4). But 6 Ra6+ Kb7
Suetin immediately begins play on the 7 Ra7+ Kc6
Q-side. 8 Rxf7

1 c4!? The position has changed sharply, but

the initiative is still with White. Here the
Unexpected and very interesting. game was adjourned, and Black sealed
White sets his opponent a difficult ex­ the strongest move.
changing problem: he can capture on c4
with the b-pawn or d-pawn, or else not 8 ... Bxb4!
at all. Gufeld does not find the best
reply. Not 8 . ' . Ra8? 9 Rf6 Kd7 10
Bc5!, with an obvious advantage.
1 . .. Bf8
9 Rf6 Kd7
1 . . . bxc4 would have failed to 2 10 Rxg6 Rc8!
b3!, but better, as shown by Suetin in
Shakhmaty v SSSR 1970 No. 4, was Forcing White to exchange rooks,
1 . . . dxc4!, when it is extremely diffi­ since he cannot permit the counter­
cult to pierce Black’s position. There­ attack with 11 . . . Rc2.
fore the correct plan for White would
have been 1 b3 followed by c3—c4, 11 Rg7+ Ke8
advantageously maintaining the tension 12 Rg8+ Kd7
on the Q-side. Of course, in a practical 13 Rg7+ Ke8
game it is often difficult to find the 14 Rg8+ Kd7
correct solution, especially in time
trouble. It is dangerous for Black to defend

85
 Endgame Strategy

against the checks by retreating his bis­ 25 . . . Kd7 loses to 26 Kg6 Bh8 27
hop, and anyway he has no reason to; he Kf7 followed by 28 f5.
is perfectly happy with a draw.
26 Kg6 Bh8
15 Rxc8 Kxc8 27 f5 d3
28 fxc6 d2
29 e7 dl=Q
30 e8=Q+ Qd8

“Bad is 30 . . . Kc7 31 Qxh8 Qg4+

32 Kf6! Qh4+ 33 Kf7! Qh5+ 34KÍ8,
when the white king escapes from per­
petual check” (Suetin).

31 Qc6+ Kb8

The position has greatly simplified,

and one gains the impression that Black
has achieved his desired draw. But even
here Suetin finds a way to pour fuel on
the fire.

16 g41!

This move demanded exact calcula­

tion.
What amazing changes occur in this
16 . . . hxg4 game! Who would have thought that a
17 Kf2 Bd2 tedious blocked position with rooks and
18 Kg3 e3?! bishops would have led to something like
this? Earlier White had merely an initia­
White’s persistent attempts to extract tive, which for a long time he skilfully
a win bear fruit, as Black chooses a maintained, whereas now he has an ob­
tempting but incorrect move. After 18 vious advantage, and the only question
. . . Kd7 19 Kxg4 Ke7 20 h5 Kf7 the is how to realize it. It is well known that
game would have ended in a draw. The in the endgame the value of pawns in­
subsequent play is forced. creases. White now has two pawns for
the bishop, and he can capture a third,
Black’s last, with check. But the essence
19 h5 e2 of the position is that, with the dis­
20 Bf2 el=Q appearance of the b5 pawn the scope
21 Bxel Bxel+ of Black’s queen is widened, and his
22 Kxg4 Bb4 chances of perpetual check are increased.
23 h6 Bf8 This means that White must play as if it
24 h7 Bg7 were a pure queen ending, where the
25 Kg5! d4! outcome is decided by king manoeuvres,

86
 The Straggle for the Initiative

and enemy pawns often serve as a screen The position has been repeated. But
against checks. Suetin nevertheless finds a manoeuvre
which wins. The position of the queen
32 Kf7! must be improved, and then the king
sent into the attack.
“After 32 Qxb5+ Ka8 33 b4Qd2!
Black’s queen breaks out into the open, 42 Qd7+ Kb6
which gives him a draw” (Suetin). 43 Qc6+ Kb7
44 Qd5+ Kb6
32 . . . Ka7 45 Kd7! Qh7
46 Qc6+ Ka7
33 Qd6+ was threatened. 47 Kc8! Qc2+
48 Kd8! Kb7
33 Qd6! 49 Qd7+

“Again the only way. After 33 Qc5+ Of course, not 49 e8=Q?? Qc7 mate.
Kb7! 34 Qxb5+ Ka7 the black queen
gives perpetual check" (Suetin). The 49 Kb6
variation can be continued with 35 Qc5+ 50 Qd6+! Kb7
Kb7 36 Qd6 Qg5, when Black has 51 e8=Q
everything in order.
The hour o f the white e-pawn has
33 . . . Qc8 finally arrived.
34 e6 Qc2!
51 . . . Bf6+
Not 34 . .. Qc4 35 b3!
With a last trap: 52 Qxfó?? Qc7 mate.
35 e7! Qxh7+
36 Ke8 Bxb2 52 Kd7 Qh7+
53 Ke6 Qe4+
Black has managed to eliminate two 54 Kf7 Resigns.
of the enemy pawns, but with the re­
maining pawn at e7 he is unable to Suetin was awarded a special prize for
cope. the best endgame of the tournament.

37 Kd8 Qh4!
38 Kc8 Qe4! Marshall—Lasker

Gufeld defends resourcefully. Now New York, 1907

39 Qd7+ Ka6 40 e8=Q allows Black
perpetual check by 40 . . . Qa8+ 41 (See next diagram)
Kc7 Qa7+ 42 Kd6Qd4+.
In his book Chess Fundamentals Capa-
39 Qc5+ Ka6 blanca comments as follows on this
40 Qd6+ Ka7 position:
“In this position it is Black’s move.
40 . Ka5 is stronger. To a beginner the position may look like
a draw, but the advanced player will
41 Kd8 Qh4 realise immediately that there are great

87
 Endgame Strategy

3 ... Rh5
4 Kgl c5

Black’s advantage assumes real pro­

portions. The d-pawn has become a
formidable force, the scope of the bis­
hop has expanded, and the white knight
is restricted by its own pawns.

5 Nd2 Kf7
6 Rfl+?

Possibly the decisive mistake. At this

possibilities for Black to win, not only point, in developing his initiative, Black
because he has the initiative, but because was forced to allow his opponent a re­
of White's undeveloped Q-side and the spite. Lasker would of course have been
fact that a bishop in such a position is happy to play 5 . . . Rh6, had he not
better than a knight. It will take some been left with his king cut off after
time for White to bring his rook and 6 R fl. Marshall should have exploited
knight into the fray, and Black can the situation to create immediate
utilise it to obtain an advantage. There counter-play. 6 a3! was correct. Now
are two courses open to him. The most on 6 . . . a5 White plays 7 Rbl
evident, and the one that most players followed by b3—b4. In reply to 6 a3
would take, is to advance the pawn to c5 Black can try to transfer his king to the
and c4 immediately in conjunction with centre by 6 . . . Ke6, e.g. 7 b4 Ke5
the bishop check at a6 and any other 8 bxc5 d3, or 7 Nf3 Kd6 8 b4Bg4
move that might be necessary with the with advantage to Black. But on 6 . . .
black rook. The other, more subtle, Ke6 quite in order is 7 R fl!, with a
course was taken by Black.” stubborn battle in prospect.
Capablanca goes on to explain that
with his rook Black must all the time 6 ... Ke7!
force White to defend something, when 7 a3 Rh6!
the áctivity of the white rook and knight
is restricted, whereas the black rook and After operating very effectively along
bishop retain complete freedom of the fifth rank, the black rook switches to
action. the sixth rank to attack White’s Q-side.

1 ... Rb8 8 h4
2 b3 Rb5!
8
Along the fifth rank the black rook ** Bc2 12 R d d 3 .
can attack White’s pawns both on the Q-
side, and on the K-side. 8 ... Rao
9 R al Bg4!
10 Kf2 Ke6

On 3 Nd2 Black has the unpleasant 11 Nf3 does not help: 11 . . . Bxf3
3 — Rc5. 12 Kxf3 KeS followed by . . . Rf6+.

88
 The Struggle for the Initiative

11 . . . Ke5 Material is IcvcL The pawn formation

12 Kg2 Rf6 is symmetric, neither side has any weak*
nesses, and neither player has gained a
White is completely helpless. spatial advantage. It would seem that at
13 R cl d3 any moment the players will begin peace
14 R fl Kd4 negotiations. But the splendid Russian
player used to agree to a draw only when
The finish was: all possibilities had been exhausted.
Chigorin begins a persistent attempt to
15 Rxf6 gxf6 take the initiative.
16 Kf2 c6
17 a5 a6 1 d4!
Zugzwang.
To seize the initiative in a level posi­
18 Nfl Kxe4 tion, without giving the opponent any­
19 Kel Be2 thing in return, is not possible. White
20 Nd2+ Ke3 creates in his formation an isolated e-
21 Nbl f5 pawn, obtaining in compensation lively
22 Nd2 h5 piece play and an attack on the d6 pawn
23 Nbl Kf3 along the d-file.
24 Nc3 Kxg3
25 Na4 f4 1 ... exd4
26 Nxc5 f3 2 Nxd4 Nc7
27 Ne4+ Kf4 3 Radi R xfl
28 Nd6 c5 4 Nxfl Re8
29 b4 cxb4 5 Nf3 Rxe4
30 c5 b3
31 Nc4 Kg3 Practically forced, since 5 . . . Re6
32 Ne3 b2 6 Ng5 Rf6 7 e5 is obviously bad,
White resigns. while after 5 . . . d5 6 exd5 Nxd5 7
c4 Nb6 Black’s pieces are badly placed.

Chigorin —Schlechter 6 Rxd6 Re7

Monte Carlo, 1902 As before, the position is level. But

not without reason was Chigorin re­
nowned for his ability to play with
knights. The white cavalry embarks on
an open attack, attempting to change the
course of the battle.

7 Ne3! c5

Vacating c6 for the bishop.

8 Ne5! Be8
9 Nd5 Nxd5?

White’s persistence bears fruit: Black

89
 Endgame Strategy

chooses the inconect solution to an ex­ Whitc must aim to exchange rooks,
changing problem. After 9 . . . Rxe5! since the exchange of minor pieces is un­
10 Nxc7 Bc6+ 11 Kf2 Kf7 (not 12 favourable for him.
Nxa6 Ke 7) 12 Na8 Ke7 13 R dl a
roughly level position is again reached. 19 . . . Bb7
20 Ra5! Kd6
10 Rxd5 21 Rb5!

White has a spatial advantage. In three moves the white rook has
switched from b3 to b5, and Black’s
10 . . . Kf6?! position has immediately become hope­
less.
It would have been preferable to go
into the rook ending by 10 . . . b6 and 21 . . . Bg2
11 . . . Bf7, although even then White 22 Rb6+ Kc7
would have retained some advantage 23 Rf6!
thanks to the activity of his rook and the
weakness of Black’s Q-side pawns. The final finesse. 23 . . . Bxh3 is met
by 24 Rc6+, and otherwise 24 Rf7 is
11 g5+! Ke6 decisive.
12 c4
23 . . . Bb7
Not 12 Rxc5?b6. 24 h4

12 . . . b6 Do not hurry!
13 Kf3 a5
24 ... Ba8
Schlechter tries to activate his bishop 25 Rf7 Kd6
by advancing this pawn to a4. It would 26 Rxc7 Kxe7
have been better to transfer it to the 27 Nd3
long diagonal by . . . Bd7—c8—b7.
After the transfer of the knight to c4,
14 Kf4 a4 the white king breaks through to the K-
sidevia e5.
14 . . . Bd7 no longer works, since
the pawn ending is lost for Black. 27 . . . Kd6
' 28 Nf2 Bg2
15 Rd8! 29 Ne4+ Kc6
30 Ke5
A brilliant refutation of the oppo­
nent’s plan. Against the threat of Rb8, The game concluded:
winning material, Black has no satis-
factory defence.
30 . . . Bfl
15 . . . axb3 31 Nd2 Bd3
16 Rb8 Kd6 32 Kf6 Kd6
17 Rxb6+ Kc7 33 Kg7 Ke5
18 Rxb3 Bc6 34 Kxh7 Kf4
19 Ra3! 35 Nb3 Resigns.
 The Struggle for the Initiative

Stein—Averbakh 3 ... h6
4 Nf6+ Kh8
Riga. 1970
Perhaps 4 . . . Kf8 would have been
preferable, bringing the king closer to
the centre.

5 Nxd7 Rxd7
6 Be3

White had another interesting possi-

bility: 6 Bxe7!P Nxe7 (6 . . . Rxe7
7 Bxc6) 7 Bxb7 Rb8 8 Bf3 Rxb2
9 Na4 and 10 R fcl, w ith a positional
advantage.

6 ... Rb8
7 R fcl Nd4
Only the queens and one pair of 8 Kfl c5
pawns have been exchanged, so the dia­ 9 R abl b6
gram position can hardly be called an 10 b4!
ending. After possible exchanges the
game will inevitably pass into an ending. The position gradually becomes more
Now much depends on who can seize the and more open, which of course favours
initiative. If Black had time to play White.
. . . h6, his chances would not be worse.
But it is White to move, and he rapidly 10 • ♦♦ cxb4
builds up an initiative. 11 Rxb4 Rc8
12 R bbl Rdc7
1 Ng5! Bd7 13 Bd2 Nef5
2 Nge4! 14 c3 Ne6
15 Nb5!
Threatening the unpleasant cavalry
raids 3 Nd5 and 3 Nc5. Provoking a weakening o f Black’s Q-
side.
2 ... Nfe7
15 ... Rxcl+
Black tries to consolidate his position 16 R xcl Rxcl+
on the h i —a8 diagonal, but unpleasant­ 17 Bxcl a6
ness awaits him from another side. 18 Nc3 Nc5
19 Ke2
3 Bg5!
The exchange of rooks has further in­
3 Nc5 would have been simply met creased White’s advantage. It would
by . . . Bc8 and then . . . b6. White’s appear that Black’s position is already
pieces attack a new weakness — at f6. difficult to hold, and on top of every­
it is curious that for the consolidation of thing he was in serious time trouble.
his position Black is always short of one
move. 19 . . . Kg8

91
 Endgame Strategy

20 g4 Nd6 2 dxc5 Bxc5

21 Bc6 Bf8 3 Bxc5 Nxc5
22 Nd5 f5 4 Kd2 e4!

22 . . . b5 is no better due to 23 Ba3 Only four moves have been made, and
and 24 Nc7. Black’s advantage is clearly apparent.
The ability to begin active play at the
23 gxf5 gxf5 right time, using the minimum of pre­
24 Nxb6 e4 conditions, is an important endgame skill.
25 d4 Nd3
26 Bd2 Nb5 5 Nd4 Nd3
27 Bb7 Nb4 6 Rxc8 Rxc8
28 a4 Nd6 7 Ke2 g6!
29 Ba8 a5
Black makes a useful consolidating
Here Black overstepped the time limit. move, and at the same time opens an
White has an easy win after 30 Bxb4 escape square for his king. It would have
axb4 31 a5 Nb5 32 Bc6 Na7 33 Bd7. been wrong to go for the win of a pawn
by 7 . . . R cl 8 R xcl Nxcl+ 9 Kd2
Nxa2 10 Nc6!, or 7 . . . Nxb2 8
R bl Nd3 9 Rb7, when White has good
Eingom—Dolmatov compensation.

Tashkent, 1980 8 fxe4 dxe4

9 Rbl?!

As shown by Dolmatov, White should

have sought salvation in the knight end­
ing after 9 f3!? R cl 10 R xcl Nxcl+
11 Kd2exf3! 12 Nxf3 Nxa2 13 Ne5.
In spite o f being a pawn down, White has
every chance o f drawing, thanks to the
active placing o f his king and knight. But
it was psychologically difficult for White
to take such a decision.

9 — f5
10 h4?
In this complex endgame Black has a
slight advantage due to his superior pawn White loses the thread o f the game
formation. 1 . . . Bf8 suggests itself, and makes a second mistake. His pre­
but White replies 2 dxc5 Nxc5 3 Kd2, vious move could to some extent have
with approximate equality. Dolmatov been justified by 10 a3 with the idea
avoids the routine bishop move, and o f f2—f3. But in fact White later [days
finds a way to seize the initiative. f2—f4 and Kd2, when h2—h4 proves
not only to be a loss o f time, but also
1 ... e5! gives Black the possibility of taking his
king to h$. It should be noted that 10
Very timely and strong. b4? was bad due to 10 . . . Rc4, and

92
 The Struggle for the Initiative

if 11 a3 R c3. 27 Ke5 Nc5

28 Kxf5 Kc3
10 • • • Kg7 29 Kf6 a5
11 f4 Kh6! 30 Ke5 a4
12 Kd2 Kh5 White resigns.
13 b4 Kxh4!
On 31 Kd5 a possible variation is
A concrete approach to the position. 31 . . . Nd7 32 Kc6 Nf6 33 Kb5 a3.
13 . . . Rc4 was also good, answering
14 a3 with 14 . . . Nxf4, but the
move played is stronger.
Kinderman—Speelman
14 R hl+ Kg3
15 Rxh7 Nxb4 Dortmund, 1981
16 Rg7

16 Rxa7 is met by 16 . . . Nc6!

16 . . . Kf2
17 Rxg6 Rd8!

The point o f Black's play. It transpires

that White is helpless against the ex­
change sacrifice at d4.

18 Rh6 Rxd4+!
19 exd4 e3+
20 Kc3 Nd5+
21 Kc4 Nxf4 Although the queens have disappear­
22 d5 e2 ed, the position is more of a middlegame
23 R hl one. It is as yet early to give it a definite
assessment.
White loses quickly after 23 Rh2+
Ng2 24 R hl Ne3+ 25 Kc5 N fl. 1 ... Bxf2
2 Nhxf2
23 . . . el=Q
24 R xel Kxel The natural move, but a serious mis­
25 d6 Kd2! take. As shown by Speelman, annotating
this game in Volume 31 of Chess In-
The most exact. Black retains control format or, 2 Nexf2! was correct. How­
over d5. ever paradoxical it may seem, the
centralized knight should have been re­
26 Kd4 treated, and the one on the edge of the
board left in its place. This move never­
26 d7 loses to 26 . . . Ne6 27 Kd5theless has a logical basis, since it makes
Nd8 28 Ke5 Kc3 29 Kxf5 Kb2 30 it difficult for Black to develop an initia­
Ke5Kxa2 31 Kd6Kb3 32 Kc7 Nf7. tive, f4 being defended by the knight
and 2 . . . e5 3 R el leading to a very
26 . . . Ne6+ complicated game.

93
 Endgame Strategy

2 ... eS!
3 fxeS Nxe5

It has become clear that Black has

seized the initiative. But the pawn form­
ation is symmetric, and if White should
be able to co-ordinate his forces the
game will be completely level. The value
of each move begins to grow immeasur­
ably.

4 R el Nf5! White’s king has come out to c2, but

now his knight is pinned. How is Black
to maintain his initiative? White hopes
An excellent move. The difficulty of to free himself by 11 Rcl followed by
playing such positions is that there is an moving his knight off the back rank. In
abundance o f promising continuations, addition Black's knight is attacked.
but normally only one of them is
correct. 4 . . . Rhf8 looks tempting, 10 . . . Nd6!
but, as shown by Speelman, after 5 Kc2 11 R cl Rc8!
White succeeds in co-ordinating his
forces. E.g. 5 . . . Nc4 6 R adi NdS Here is the English grandmaster’s solu­
(6 . . . R xd l 7 N x d l) 7 Rd3. After tion. It becomes clear that 10 . . . Nf6?
the move in the game Black meets S Kc2 11 R cl Re8 would have led to equality
with 5 . . . Nc4!, while on S Bh3 he after 12 Ne3 Rxcl+ 13 Kxcl Ng4
has the unpleasant 5 . . . Nf3. 14 Kd2, but now die analogous vari­
ation is not possible due to 13. . .Nc4.
5 Nd6+
12 Rxe8 Nxc8
Probably the least evil 13 Kd2 R hl!

5 ... Nxd6 Accuracy to the end. After 13 . . .

6 Bxb7+ Kxb7 Nd6?! 14 Rc2! White would have had
7 RxeS saving chances, since 14 . . . R hl can
be met by 15 K cl.
The position has simplified. If it were
now White’s move, after Kc2 a draw 14 Ke2?!
could be agreed. But by constant threats
Speelman prevents White from making White intends to continue the struggle
this single move separating him from a pawn down after 14 . . . Rxh2+ 15
equality. Nf2. The idea is correct, but incorrectly
implemented. It was essential to play
14 Nf2! Rxh2 15 Ke3.
7 ... Rhf8!
8 Re2 Nc4! 14 . . . Nd6!
9 N dl Rfl
10 Kc2 Black wins a pawn in a much more

94
 I The Struggle for the Initiative

favourable situation. The outcome of the 13 Kf2 was better.

game is decided.
18 . . . R el +
15 h4 Nf5 19 Kf3 Nh4+
16 g4 20 Kg3 g5
21 Kh2 Kc6
On 16 Kf3 Black wins by 16 . . . 22 Nhl Re3
Rfl+, and if 16 Kf2 h5! (Speelman). 23 Rf2 Nf3+
24 Kg2 Ne5
16 . . . Nxh4 25 Rf6+ Kd5
17 Rc2 Ng2 26 Nf2 Re2
18 Nf2 White resigns.

95
 CHAPTER 9

SUPPRESSING THE OPPONENT’S COUNTER-PLAY

In his book M y System Aron Nimzo-

witsch put forward a new demand of
positional play — the necessity for pro­
phylaxis. He wrote: "We are in fact now
concerned with the warding off of an
evil, which has really never been under­
stood as one, yet which can, and in
general does, have a most disturbing
effect on our game. The evil consists in
this, that our pieces are out of, or in
insufficient contact with their own
strategically important points." In his
other book Chess Praxis Nimzowitsch
analyzes various forms of prophylaxis, White is a pawn up and has the ad­
and uses numerous examples to show vantage of the two bishops. But his
how important it can be to make a black-squared bishop has to be exchang­
timely move which suppresses the ed for the knight at d4, after which
opponent's play. Black acquires a stron^passed pawn sup­
In the endgame the contact of pieces ported by his rook! An interesting
with strategically important points has struggle is in prospect.
enormous importance. In this chapter
we will be mainly considering the most 1 Bxd4 cxd4
obvious form of prophylaxis, involv­ 2 b4!
ing the suppressing of counter-play by
the opponent which is directly threaten­ Ivkov allows the invasion o f the
ed, rather than in the process of genera­ enemy rook, since White docs not
tion. achieve anything by 2 Be4 Rb8 3 Raj.
The demands o f prophylaxis have
much in common with the principle “do 2 ... Rc3
not hurry” . Nimzowitsch expressed the 3 b5 Rxf3+
idea that waiting moves are the start of 4 Kg4!
any form o f prophylaxis. The timely
suppressing of the opponent’s counter- A move which demanded deep calcu­
play and the ability to employ waiting lation and a precise evaluation of the
tactics are characteristic of all great position. After 4 Kg2 White would
masters of the endgame. have won a piece, but Black would have
been able to set up an unusual fortress.
Here is a possible variation: 4 . . . RcJ
5 Ba4 Rxc4 6 b6 Rb4 7 Bc6 d3 8
Kf3 Kg7 9 b7 Nd8 10 Ra4! Rb3
Ivkov—Hartoch 11 Ra8 Nxb7 12 Ra7 d2+ 13 Ke2
Kf6 14 Bxb7 Re3+! 15 Kxd2 Re7,
Raach, 1969 and the advance of the h-pawn will cost

96
 Suppressing the Opponent's Counter-play

White his bishop. 14 Kf3 h5

15 Ke4
4 ... Rc3
5 b6 Zugzwmg!
With his king at g2 White would 15 ... h4
have had this move. 16 Rh7 g5
17 fxg5+ Kxg5
5 ... Rxc2 18 Rf7 Ng3+
6 b7 Rb2 19 Kxd4 h3
7 Ra8+ Kg7 20 Rf2 Kg4
8 b8=Q Rxb8 21 c5 Nf5+
9 Rxb8 Nh6+ 22 Rxf5!!

White is on the alert. He again pre­

vents the positional draw which was
possible after 22 Ke4?! Ng3+ 23 Kd3
Nf5 24 c6 Ne7 25 c7 Kg3 26 R f7h2.

22 ... Kxf5
23 c6! h2
24 c7 H1=Q
25 c8=Q+

Twenty-five moves ago it was impossi­

ble to foresee that the game would go
Thus White has managed to win the virtually by force into a queen ending
exchange, but the game is not yet over. w ith White a pawn up.
By the knight check Hartoch plans to set 25 . . . Kf4
up an interesting fortress. 26 Qf8+ Kg4
10 KgS!! 27 Qxd6

For the Second time the Yugoslav White has a theoretically won ending.
grandmaster suppresses the opponent’s 27 ... Qal+
counter-play by an exact king move. The 28 Kc5 Kf3
natural 10 Kf3 leads to a draw: 10 . . . 29 Qe6 Qa5+
Nf5 11 K e4h5 12 Rb7+ Kg8! But 30 Kc6 Qa6+
not 12 . . . Kf6, when White plays his 31 Kc7 Qa7+
rook to h7, and Black is forced to ‘tear’ 32 Kd8 Qb8+
his h-pawn away from his g-pawn due to 33 Ke7 Qc7+
zugzwang. 34 Kf6 Kg2
10 ... Nf7+ 35 d6 Qc3+
11 Kh4! Nh6 36 Kf7 Qf3+
12 Rb7+ Kf6 37 Kg7 Qc3+
13 Kg3! 38 Kh7 Qc6
39 Qg4+ Kfl
Of course, not 13 Rxh7?d3! 40 Qf5+ Kel
41 d7 Qd6
13 . . . Nf5+ 42 Qa5+ Resigns.
97
 Endgame Strategy

Timoshchenko—Makarichev 5 Bxc4 Bc5

6 Rgg2 Rh4!
Moscow, 1979
“The last precise move. The rook
occupies an ideal position and avoids the
possible attack Be2” (Makarichev).

7 b3 a5
8 Bb5 Ra8
9 Kc2 Rah8
10 a3 Rh3

With the threat of 11 . . . Rc3+.

11 Bd3 Rxh2
Black has the better chances in view
of the weakness of the white h-pawn and 11 . . . Rc8+ 12 Bc4 (12 Kd2 Bc3+)
the excellent strong point at e5 for the 12 . .. a4 would also have won.
black bishop.

1 Rg4? 12 b4 axb4
13 axb4 Rxg2
In Shakhmaty v SSSR 1979 No. 10 14 Rxg2 Rh2!
grandmaster Makarichev writes: “Black’s
plan is clear — pressure on the weak h2
pawn. Therefore White should first have The most convincing. The bishop end-
played the prophylactic 1 h3! Rh8 ing is hopeless for White.
2 B fl, and only then Rg4 with the
possibility of a2-a4. But White fails to
sense the danger.” IS Rxh2 Bxh2
16 Kb3 Kf6
1 ... Rh8 17 Ka4 KeS
2 R fl Ra7! 18 KbS g5
19 Kc6 g*
Black prudently defends the seventh 20 bS Bgl
rank with his rook. He now threatens 21 Be2
. . . Rxh2, which would not have work­
ed earlier due to Rgf4.
On 21 b6 there would have followed
3 Rf2 Rh5! 21 . . . Bxb6 22 Kxb6 KxdS with
an easy win.
Black consistently engages in prophyl­
axis. By the threat to the d-pawn he pro­
vokes c2—c4, depriving White of 21 . . . f5
counter-play with a2—a4. 22 Bdl Bc5
23 Be2 Ke4
4 c4 bxc4 White resigns.

98
 Suppressing the Opponent’s Counter-play

T aimano v—Aronin 1 ... Rxdl

2 R xdl Ra2
Moscow, 1949 3 Rd2

White has managed to parry the main

threats and to keep his extra pawn.

3 ... e4!

Black again sets White difficult

problems. He now threatens 4 . . . Ne5
followed by . . . Nc4.

4 f3!

By the use of tactics White ex­

tinguishes his opponent’s burst of
White is a pawn up, but the black activity.
rooks occupy open files, and his Q-side
pawns may come under attack. Much 4 ... exf3
depends on the next few moves, as to
whether White can stabilize the position Nothing is achieved by 4 . . . Ne5
and retain his material advantage. 5 fxe4N c4 6 Rf2+ followed by Bd4.

1 Rdd 1!! 5 Rf2!

A strong but by no means obvious Not 5 gxf3? Ne5.

move, 1 f4 and 1 Rfdl being the
continuations which suggest themselves. 5 ... Nh4
This is what grandmaster Taimanov had 6 g3!
to say regarding this in Sbakbmaty v
SSSR 1950 No. 1: Correctly solving the exchanging prob­
“First and foremost White must de­ lem. After 6 gxf3P! g5!all White’s
fend the first rank and not concede the pieces are awkwardly placed.
d-file. The tempting 1 f4 would have
been wrong due to 1 . . . exf4 2 exf4 6 ... Nf5
Ne7! 3 f5 Nd5, when 4 f6 fails to 7 e4 Ne3
4 . , . Rxd6. 8 Rxf3+ Ke6
1 R fdl Rxd6 2 Rxd6 Ra4! would 9 Bd4!
have led to great complications, e.g.
3 Rd7+ Ke6 4 Rxb7 Ra2 5 Bc3 Rc2 Again suppressing the opponent’s
6 Bel e4!, and in view of the threat of counter-play. “9 Bxg7 would have
. . . Ne5—d3 White must go in for the been a mistake due to 9 . . . Rg2+ 10
variation 7 f4 exf3 8 gxf3 Ne5 9 Khl Re2, when in spite o f White’s two
Bg3 Nxf3+ 10 Kfl Nxh2+ 11 Bxh2 extra pawns he cannot win, e.g.:
Rxh2 12 Rxg7 Kd5 with a draw. If (a) 11 h4 (or 11 B c3 N d l!) 11 . . .
after 2 . . . Ra4 White plays 3 Bc3, Rel+ 12 Kh2Ng4+ 13 Kh3 h5.
Black has the very unpleasant reply (b) 11 h3 h5, and White can­
3 . . . Ra3." not maintain his advantage: 12 Rf6+

99
 Endgame Strategy

(12 Bd3 Nc2 13 Bc3 Rxe4, or 12 white king" (Taimanov).

Bc3 N d l) 12 . . . Kc7 13 e5 R el+
14 Kh2 Rc2+ 15 Kgl Rg2+ 16 K hl 18 Bgl N fl
Rxg3 17 Bf8+ Kc8 18 Bd6 Rxh3+
Otherwise 19 Bd4+ wins immediately.
19 Kgl Nd5 20 Rf8+ Kd7 with a
draw" (Taimanov). 19 Bd4+ Kf7

9 ... Ng4 “If 19 . . . Ke6, then 20 Bc3 Kf7

10 Bxg7 Nxh2 (20 . . . Nb2 21 KgS N fl 22 Bel
11 Rf6+ KeS 23 KxbS Kxe4 24 g4 is hope­
less for Black) 21 Bel Kf6 22 eSt
Before playing Rf2, White worsens Ke6 (on 22 . . . Kg6 there would have
the position of the enemy king. followed 23 Bf2, and then as in the
game) 23 Kg5 Kxc5 24 KxhS Ne3
11 . . . Ke7 25 Kg5! Nd5 26 g4 Ke6 (26 . . . b6
12 Rf2 Rxf2 fails to 27 cxb6 Nxb6 28 Kb6, when
13 Kxf2 h5 the g-pawn advances irresistibly) 27 Kg6
Ne7+ 28 Kg7, and Black is defenecías
The position has clarified, and White against the following plan: White places
is a pawn up in a minor piece ending. his bishop at d2 and his pawn at g5.
Now, operating according to the princi­ and then takes his king across to the Q-
ple of two weaknesses, he must combine side" (Taimanov).
the advance of his extra e-pawn with an
attack on the h5 pawn with his king, 20 Bf2 Kg6
and in some cases on the b7/c6 pawn
pair. If 20 . . . Kc6 21 Kg5, or 20 . . .
Kf6 21 Bel followed by Bc3+.
14 Ke3 Ke6
15 Kf4 Ng4 21 e5 Nh2
16 Bd4! 22 Bd4! Kf7
23 Kf5 Kc7
Note how the black knight constantly 24 Be3 Ng4
comes under the domination of the 25 Bf4 Nf2
white bishop. If immediately 16 Kg5, 26 e6 Nd3
after 16 . . . Nf2 Black gains some 27 Bd6+ Ke8
counter-play. 28 Kg6

16 . . . Nh2 Black could have resigned

17 Bf2! Kf6 point. The finish was:

“After 17 . . . Ng4 18 Bgl! Black 28 ... Nxb4

ends up in an unusual zugzwang position: 29 Kxh5 Nd5
(a) 18 . . . Ne5 (18 . . . Kf6 19 30 g4 b5
Bd4+ Kg6 2 0 eS) 19 Bd4 Nf7 (19 31 cxb6 Nxb6
. . . Nd3+ 20 Kg5 Nxb4 21 KxhS is 32 g5 Nd5
hopeless for Black) 20 Bg7! 33 g6 c5
(b) 18 . . . Nf6 19 Bd4 Nh7 20 34 g7 Nf6+
Bg7 Kf7 21 Be5 Ke6 22 Bd4, and 35 Kg6 Ng8
Black has to allow the advance of the 36 Bxc5 Resigns.

100
 Suppressing the Opponent’s Counter-play

Botvinnik—Alekhine Preparing to bring the king out via f7

to e6.
AVRO-Tourniment, 1938
6 h4!

Timely and very strong. White attacks

Black’s pawns before his king can come
to their help.

6 ... Nd7

“ Incidentally, Alekhine avoided 6 . . .

Kf7 since he was afraid of 7 hxgS fxgS
8 Nf3 g4 9 Ne5+. To me this variation
did not seem so convincing, and there­
fore on 6 . . . Kf7 I was intending to
continue 7 Nf3! g4 8 N el Ke6 9 Nd3
The pawn formation is symmetric, but Kf5 10 g3 (10 f3 is also good) 10 . . .
Black’s Q-side pawns are weak, and Ke4 11 Nf4 when the position is
White’s pieces are much better placed. In hopeless for Black” (Botvinnik).
such a position, as in many others with a
spatial advantage, the main thing is not 7 Rc7 Rf7
to allow the opponent to free himself, 8 Nf3!
and to deprive him of the slightest
counter-play. Botvinnik copes brilliantly By continuing the attack on g5 White
with this task. provokes a further advance o f die black
pawns.
1 ... f6
2 Kfl Rf7 8 ... g4
3 Rc8+ Rf8 9 Nel
4 Rc3! g5
These knight manoeuvres have essenti­
There is nothing better. On any piece ally decided the game.
move, except 4 . . . Rf7, there follows
5 Rc7! 9 ... f5
10 Nd3 f4
5 Nel! 11 f3 gxf3
12 gxf3 a5
The routine 5 Kc2 would have 13 a4 Kf8
allowed Black to breathe more easily 14 Rc6 Ke7
after 5 . . . Rc8+ and . . . Kf7. By the IS Kf2 RfS
knight manoeuvre White forces his
opponent to undertake something, Black has more than enough weak­
otherwise there follows Nc2—e3. But nesses, but White is not in a hurry to
now any activity by Black is bound to attack them.
create weaknesses.
16 b3! Kd8
5 ... bS 17 Ke2 Nb8

101
 Endgame Strategy

Black is tired of waiting, and he sets a

little trap: 18 Rxb6? Kc7 19 R~Nc6.
Had Black continued moving his king be­
tween e7 and d8, White would simply
have strengthened his position by play­
ing his king to c3 followed by b3—b4.

18 Rg6 Kc7
19 Ne5 Na6
20 Rg7+ Kc8
21 Nc6 Rf6
22 Ne7+ Kb8
23 Nxd5
only to exercise a certain caution, and
The first gain of material. not allow the sacrifice o f the black
knight at b3 or a2. Black has no other
23 ... Rd6 active possibility (the pawn breaks
24 Rg5 Nb4 . . . dS and ¿5 arc too risky). For
25 Nxb4 axb4 safety's sake White decided to place his
26 Rxh5 Rc6 rook at d5, nipping in the bud any
pawn break in the centre, while after the
No better is 26 . Rxd4 27 exchange of knight for rode Black has
Now the rest is simple. nowhere to break through. In defending
against his opponent’s non-existent
27 Rb5 Kc7 threats. White overdid it, and created the
28 Rxb4 Rh6 possibility far Black of real play.
29 Rb5 Rxh4
30 Kd3 Resigns.
1 Rd5 Nc8
“One of those games where there arc 2 R ld2 Nc7
no brilliant moves; all the moves seem 3 Bdl Na6
very simple, but on the other hand not 4 Bc2 Nb4
one of them can be excluded, since they 5 Bbl Ra6!
are all closely connected. The difficulty
of playing such games lies not in the Black’s plan begins to take shape. Had
complexity of the calculation, of course, the white rooks been at d2 and d l, the
but in correctly evaluating positions draw would have been obvious.
reached during calculation" (Botvinnik).
6 R dl Nxd5+
7 RxdS Rxb3+1!
Brzozka-Bronstein
A bolt from the blue.
Miskolcz, 1963
8 Kxb3 Rb6+
(See next diagram) 9 Kc2 Rb2+
10 Kcl Re2
“A dead draw" the reader will say, on
glancing at this position. Indeed, such a In spite of his extra piece, White's
result seems most probable. White has position is difficult.

102
 Suppressing the Opponent’s Counter-play

11 R dl Rxc3
12 Rgl Rc3+
13 Bc2 Rxc4

The storm has died down. The white

bishop is unable to oppose the three con­
nected passed pawns in the centre. From
his calm and placid drawing fortress
White has been literally dragged into a
tactical game, in which he has few saving
chances.

14 Kd2 d5
15 R bl d4 2 Rc4! followed by 3 Rb4 b5 4 a4
16 Bdl Rc3 Black would probably have lost. After
17 Rb3 e3+ 1 . . . Rd3+ White has the problem o f
18 Ke2 R cl where to move his king. The centralizing
19 Rxa3 c4 2 Ke4, as played in the game, suggests
20 Ra7+ Kd6 itself, but correct was 2 Ke2! followed
by R cl, depriving Black of counter-
The black king comes to the aid o f the play. But it was very difficult to resist
pawns, and this decides matters. the temptation to obtain two connected
passed pawns.
21 Ba4 R hl
22 Rd7+ Kc5 3 Ke4 Rd2!
23 Rc7+ Kb4 4 Rcl?!
24 a3+ Kc3
25 Bb5 Rh2+ It was not yet too late to go back with
26 Kfl d3 4 Kf3!, but White follows the intended
27 Kd2 path.
+
o
*
X

28 Kgl e2
29 Kxh2 el=Q 3 ... Rxf2
White resigns. 4 Rc7+ Ke6
5 Rxb7 Re2+!

Balashov—Szabo A very important moment. It is

esseintial for Black to decentralize the
Sochi, 197) white king and to centralize his own.

(See next diagram) 6 Kf3 Rxh2

7 Rb6+ KdS!
White is a pawn up and his bishop is 8 Rxa6 Rh3+
stronger than the knight. In order to
realize his advantage, he must create While continuing to push back the
weaknesses in Black’s position. white king, Black prepares the invasion
of his own.
1 BeS Rd3+!
9 Kf2 Rh2+
In the event o f the natural 1. . .Rc8 10 Kgl Rc2!

U-M 103
 Endgame Strategy

11 Rxg6

Black loses another pawn, and the

white king is able to come into play after
Rg2, but White can no longer win. With
his limited forces the Hungarian grand­
master builds up an attack which obliges
White to settle for a draw. All Black's
pieces and his sole pawn take part in the
attack, his army being led by the king
itself.

11 ... Ke4!
12 Rg2 Rcl+ One is immediately struck by the un­
13 Kh2 Nh4 enviable position of the white bishop,
14 Rg7 shut in by its own pawns and forced to
guard the enemy passed pawn. Black’s
On 14 Re2+ th e king goes via f3 Q-side pawns have seized a great amount
to g4. of space, but in advancing so far have
themselves become a target for attack by
14 . . . Nf3 + the opponent’s pieces. The white rook
15 Kg2 has broken into the enemy rear, and
the knight at h5 may be able to go via
As shown by Szabo in Sbakbmatny f4 and e6 to c5. But Black is able to
Bulletin 1974 No. 2, Black also has a neutralize the opponent’s main threats,
draw after 15 Kh3 h5l, not allowing by exploiting the chronic defects in
the white king to escape, e.g. 16 Rg2 White’s position.
R hl+ 17 Kg3 h4+ 1 8 K g 4 N e l! 19
Rd2 (also after 19 Re2+ Kd3 20 1 ... Re7!
R xel R xel 21 Kxb4 R gl! Black
should not lose) 19 . . . Rgl+ 20 Kh3 Black forces the exchange o f rooks,
Nf3! 21 Rg2 (21 Re2? Kf5!) 2 1 . . . since otherwise he is threatening to win
R hl+ 22 K g4N el!, creating a drawing the bishop.
mechanism. After the move chosen by
Balashov a drawn rook ending is reached. 2 Rxe7 Ngxe7
3 Kf2
15 . • . Nxe5
16 fxe5 Rc2+ It was essential to parry the threat of
17 Kh3 Rxb2 3 . . . Ne3, but now the black king is
18 c6 Kf5 able to reach e6.
19 e7 Re2
Drawn. 3 ... Kg8
4 g4?

An interesting point. Euwe continues

Euwe—Averbakh the plan of transferring his knight to e6,
but fails to take any precautionaty
Candidates Tournament measures against the opponent’s counter­
Zurich, 1953 play. He should have gone promptly

104
 Suppressing the Opponent’s Counter-play

onto the defensive by Nf4, g2-g3, Ng2,

Ne3, Nfl and Nd2. In this case White
would have had chances of saving the
game, whereas now events develop al­
most by force.

4 ... Nd6
5 Ke3 Nb5
6 f4 Nc8
7 f5 Ncd6
8 Nf4

White has implemented his planned

set-up, but his game is now lost. The normal plan in such positions is
the minority attack by a2—a3 and
b2—b4—b5, but that is with the white
pawn at e3. Who is favoured by the
advanced position of the pawn at e5?
Most probably White, who, with his
advantage in space, can combine play
both on the Q-side and the K-side. True,
Black may also acquire counter-play by
the undermining . . . f6 o r . . . b6 and
. . . c5. A complicated battle now com­
mences.

1 Rc3

8 ... Nxa3! Vacating c5 for the knight.

"By positional play the maestro 1 ... Rf8

endeavours to secure and exploit true
values; by means of a combination he Black prepares . . . {6.
aims to refute false values” (Lasker).
The words of the great player are fully 2 Rel!
confirmed by the present example.
Suppressing the opponent's counter­
9 Bxa3 NbS play. On 2 . . . f6 White has the highly
10 Bel Nxc3 unpleasant 3 c6!
11 Ne2 Nbl!
White resigns. 2 . .. g6

With the aim of erecting a defensive

barrier on the K-side.

Boleslavsky—Averbakh 3 Bd3

Candidates Tournament White takes control of fS, so as to

Zurich, 1953 exchange off the black knight as soon as

105
 Endgame Strategy

it appears there. In this closed position it 16 Rc3 Rac8

is advantageous for each player to ex­ 17 Bxf5! Bxf5
change bishop for knight. 18 Reel Bd7
19 Nel
3 ... Bf5
4 Bfll In the tournament book Bronstein
indicates the possibility of 19 e6!
The side with a spatial advantage followed by Ne5, but Boleslavsky is not
should agree only to very favourable ex­ in a hurry to force matters. It only needs
changes. the white knight to penetrate to c5, and
Black’s position will immediately be­
4 . .. a4 come hopeless.

A committing decision. If now White 19 . . . Rb8

should succeed in provoking . . . b5, the
c6 pawn will be very weak, while Black
has eliminated only temporarily the
threat of the white knight penetrating
to cS.

5 h3 Bd7
6 f4 h5

All Black’s pawns are on squares of

the same colour as his bishop, and it is
well known that this is not the best
arrangement in the endgame.

7 Nf3 Kg7 20 Nd3?

8 Kf2 Rh8
9 gJ Kf8 This natural move allows Black to save
10 Kg2 Nf5 die game in amazing fashion. After Bron-
11 Bd3 Ng7 stein’s suggestion o f 20 Rc5! White
12 Ng5 Be6 was bound to win, e.g. 20 . . . b4 21
Ra5!, o r 20 . . . Rhc8 21 Nd3 BÍ5
Defending against a possible e5—e6. 22 Nb4. The game once again shows
how carefully one must watch for
13 Bc2! Ke7 counter-play by the opponent, even in
14 Ra3 Nf5 won positions.

Black is prepared to allow the ex­ 20 ... b4!

change of his knight for the bishop,
21 Rc5 Bf5
merely to avoid playing . . . b5. But
22 Nf2 b3!
after White’s next move he is forced to
23 a3
drain the cup completely.

15 Nf3! b5 Nothing is achieved by 23 axb3

Rxb3, while after 23 Rxc6? a3!! it is
Forced. Black who wins.

106
 Suppressing the Opponent's Counter-play

23 . . . Bc2! White would still have had some win­

ning chances after 24 N dl.
It transpires that on 24 Rxc6 there
24 Kd7
follows 24 . . . Rhc8 25 Rxc8 Rxc8,
25 Ke3 Ra8
when White cannot defend his d-pawn,
26 h4 Ra6
since 26 Kf3 fails to 26 . . . Be4+.
27 Nh3 Rb8
28 Ng5 Ke7
24 Kf3 Drawn.

107
 CHAPTER 10

POSITIONS WITH AN ISOLATED d-PAWN

Positions with an isolated pawn on the Forestalling__ d4.

d-file arise in many openings. In the
middlegame the side with the isolated 3 ... Ke7
pawn obtains compensation in the form 4 Ke2 Ke6
of active piece play, but in the endgame 5 Kd3 Ke5
the weakness o f the isolated pawn is 6 f4+ Ke6
more noticeable. It is the methods of 7 Kd4 Kd6
play in such positions that will be cover­ 8 b4 Ke6
ed in this chapter.
Black assumes that White will be un­
able to create any serious threats, and
Byelavyenets—Rauzer demonstratively moves only his king.
Such tactics are acceptable, but only up
Moscow, 1937 to a certain point. It would not have
done any harm to advance die f-pawn to
f6, so as not to ‘forget* to do it later. But
how can White win? Black already has
one weakness — the isolated pawn in the
centre. What is needed is a second weak­
ness, which can be created only on the
K-side. Byelavyenets first strengthens
his position to the maximum on the Q-
side, by advancing his pawn to b5 and
placing his knight at b4.

9 b5 Bb7
10 Na2 Kd6
11 Nb4 Ke6?
White has a strong knight against a
bad bishop, with the opponent’s d-pawn Here 11 . . . f6 was essential. Black
isolated. Is this ending won? Most pro­ calculated that after 12 Nc6 Bxc6 13
bably the position is drawn (cf. the bxeó Kd6 14 c7 Kxc7 15 Kxd5 f5!
following example). What happened in he would gain a draw, but he failed to
this game? White’s knight is attacked and take account of his opponent’s reply.
must retreat to c3 or d4 (taking the a7
pawn is unfavourable). The knight stands 12 g4!
very well at d4 in the middlegame, but
in the endgame this is the square for the Threatening by g4—g5 to fix the
king. Therefore: pawns at f7 and h7, while on 12 . . .
h6 there follows 13 Ncó! Bxc6 14
1 Nc3 Bc6 bxeó Kd6 15 c7 Kxc7 16 Kxd5,
2 Kfl Kf8 and the ending is won for White, since
3 f3 Black does not have . . . f5. Here is the

108
 Positions with an Isolated d-Pawn

variation given by Byelavyenets: 16 . . . 23 Ne2

Kd7 17 e4 Kc7 18 e5 Kd7 19 f5
gxfS 20 gxf5 Kc7 21 e6 f6 22 h3
hS 23 h 4 a 6 24 a4a5 25 Kc4 Kc6
26 e7 Kd7 27 Kb5, and wins.

12 . . . f5
13 g5

The h7 pawn is fixed. Now White

must advance his h-pawn to h6, when
Black will be faced with the threat of a
knight sacrifice at h6 or f 5, and hence
a second weakness will be created.

13 . . . Kd6 “Here the game was adjourned, and

14 h 3! adjudicated a win for White. He has
succeeded in implementing his plan com­
A concrete approach to the position. pletely, and Black, with three weak
“The more natural 14 h4 would have pawns at a7, d5 and h7, is helpless,
been weaker due to 14 . . . Ke6 15 for example:
Na2 Kd6 16 Nc3 Ke6 17 Ne2 Bc8 (a) 23 . . . Bd7 24 Nc3 Be6 25
18 Ng3 Bd7 19 a4 Be8, when Black Na2 Bf7 (Black cannot defend his d-
succeeds in preventing h 4 -h 5 " (Byela­ pawn with his bishop from b7, since
vyenets). after, say, 2 5 . . . Bc8 26 Nb4 Bb7
White wins by 27 Nd3 Bc8 2 8 NeS,
14 . . . Kc6 when there is no defence against Nxg6)
26 Nb4 Be6 27 Nc6 a5 28 bxa6!
Black moves his king to and fro as if Kxc6 29 a7! (essential, since after
nothing has happened. 29 Ke5 Bc8! 30 «7 Bb7 31 K f6 K d 6
32 K g 7 K e7 33 K xb7 K f7 White has
15 Na2 Kd6 no win) 29 . . . Kb7 30 Ke5 Bd7 31
16 Nc3 Ke6 Kf6 Bxa4 32 Kg7 b5 33 Kxh7 b4
17 h4 Kd6 34 Kxg6 b3 35 h7 Be8+ 36 Kf6 b2
18 Ne2 Bc8? 37 h8=Q bl=Q 38 a8=Q+ Kxa8 39
Qxe8-i-, and the queen ending is easily
“The last saving chance was 18 . . . won for White.
a6” (Byelavyenets). (b) 23 . . . Bf7 (an unsuccessful
attem pt to shut the knight in at a7) 24
Nc3 Be6 25 Na2 Bg8 26 Nb4 Bf7
19 Ng3 Bd7 27 Nc6 Be8 28 Nxa7 Bd7 29 Kd3
20 a4 Ke6 (Black is in zugzwang) 2 9 __ Kc7 30
Kc3 Kb7 (if 30 . . . Kd6 31 Kd4,
and Black is forced to free the knight)
There was no longer any defence 31 Kd4! Kxa7 32 Ke5, and White
against h4—h5. wins.
The entire ending is a good example
21 h5 Be8 of the battle between bishop and knight’’
22 h6 Kd6 (Byelavyenets).

109
 Endgame Strategy

Flohr—Capablanca a mistake.

Moscow, 1936 11 . . . h5?!

With his last move White vacated g2.

The Cuban grandmaster obviously
thought that White was intending to play
his knight to h4, advance f4 -f5 and
g3-g4, then transfer the knight to f t
and with the black king at c6 play Ne6,
forcing a won pawn ending.
But the black bishop does not have to
allow the knight to reach h4: with the
knight at e l it is sufficient to transfer
the bishop to e4, to say nothing of the
possible defence . . . g5 with the knight
at h4, to answer fxg5 with . . . hxg5!?
The position is virtually identical to and Ng2 with . . . g4! After the move
the previous one, but the play develops in the game Black gets into serious
rather differently. difficulties.

1 ... Ke7 12 b4 axb4

2 Kd2 Kd6 13 axb4 Kd6
3 Kc3 b6 14 b5
4 f4 Bd7
5 Nf3 f6 Black has no good moves. 1 4 __ Bf7
6 Kd4 a5 is unpleasantly met by 15 f5.

In contrast to Rauzer, Capablanca 14 ... g*

thoughtfully places all his pawns on dark 15 Na4 Kc7
squares. In his book Last Chess Lectures 16 Nc3 Kd6
Capablanca formulated the following 17 f5!
rule: "When the opponent has a bishop,
keep your pawns on squares of the same Breaking up Black’s pawn formatioa
colour as this bishop. But if you have a Bad now is 17 . . . Bxf5 18 Nxd5 Bd7
bishop, then, irrespective of whether or 19 Nxf6 Bxb5 20 Nd5, and 20 . . .
not the opponent has a bishop, keep Kc6 is not possible due to 21 Ne7+.
your pawns on squares of the opposite
colour to that of your bishop". 17 . . . gxf5
18 Ne2 Bd7
7 Nd2 Bc8
8 Nbl Be6 An inaccuracy. In anticipation of the
9 Nc3 Kc6 pawn ending Black should have played
10 a3 h6 18 . . . Bg8 19 Nf4 Bf7, provoking
11 g3 h2—h3, and only then attacked the b*
pawn by . . . Be8.
Black sticks to waiting tactics, and
White also does not huny. This manner 19 Nf4 Be8
of play led to Capablanca himself making 20 Nxd5 Bxb5

no
 Positions with an Isolated d-Pawn

21 Nxb6 Sc6

On no account must the white knight

be allowed to reach f4. Imagine that the
knight stands at f4. In reply to . . . Be8
there follows Nd5 winning a pawn,
since the pawn ending after . . . Ke6,
Nc7+ and Nxe8 is won for White.

22 Nc4+ Ke6
23 Nb2 BbS
24 Ndl Be2
25 Nf2 Bfl
Were Black’s Q-side pawns at a7 and
Now Flohr transposes into a pawn b6, his position could be considered
ending. quite satisfactory. But in the given posi­
tion, apart from the isolated d-pawn.
26 Nd3 Bxd3 Black has a complex of weaknesses on
27 Kxd3 Ke5! the Q-side, and in particular at c5.

"27 . . . Kd5 would have lost to 28 1 f3

Kd2 Ke5 (or 28 . . . Ke4 29 Ke2
Kd5 30 K f3 Ke5 31 b3 Kd5 32 Kf4 Karpov prepares to bring his king out
Ke6 33 b4\ this is where the two re­ at f2 followed by the invasion o f his
serve tempi on the K-side become im­ rook at c7.
portant!) 29 K ell Kd5 30 Kf2! The
king moves to el and f2 are given 1 . •• Rc8
exclamation marks, since e2 cannot be 2 Rxc8+ Bxc8
occupied, e.g. 29 Ke2 Ke4 30 Kf2 h4 3 Kf2 Kf7
31 gxh4 f4 32 h5, and the e3 pawn 4 Ke3 Ke7
is captured with check" (Bondarevsky). 5 b4!
The game continued:
With the exchange of rooks the weak­
28 Ke2 Ke4 ness of Black’s Q-side has become even
29 h3 more appreciable. By this last move
White consolidates his advantage.
Or 29 Kf2h4!
5 ... g6
29 . . . Kd5!
30 Kf3 Ke5 Black is forced to reckon with the
threat of g3—g4—g5. On 5 . . . h6
In this position the players agreed a White has the highly unpleasant 6 Bg6,
draw. while after 5 . . . g5 6 Nf5+! Bxf5
7 Bxf5 Black cannot avoid loss of
material.

Karpov-Hort 6 g4!

Budapest, 1973 White intends to create a passed pawn

ill
 Endgame Strategy

on the K-side, which will divert Black’s

forcesaway from the Q-side.

6 ... Nd7
7 f4 Nf8
« g5 Kd6
9 Kf3 Ne6?!

Inviting a bishop ending, which White

is happy to agree to. All Black’s pawns
are on squares o f the colour o f his
bishop, and Karpov elegantly realizes his
advantage.
1 ... b6?!
10 Nxe6! Bxe6
11 Ke3 Bg4 There was no necessity for this weak­
12 Bd3 Be6 ening. 1 . . . Kd8 was correct.
13 Kd4 Bg4
14 Bc2 Be6 2 Rxc8 Bxc8
15 Bb3 Bf7 3 Ke2 Kd6
16 Bdl Be6 4 Kd2 g6?!
17 Bf3 Bf7
18 Bg4 From previous examples we know
that this move is undesirable. 4 . . . h6
Here Black resigned, since the only was better.
defence against 19 Bc8 is 18 . . . Be6,
but the pawn ending after 19 Bxe6 5 Kc3 Nd7
Kxe6 20 g4Kd6 21 a3 is easily won 6 f4
for White.
Here we saw that the transition into White begins playing actively on die
the bishop ending proved fatal for Black, K-side.
although in any case his position was
lost. This ending shows how difficult it is 6 ... Nf8
for the weaker side to defend if, apart 7 Be2 Kc7
from the isolated d-pawn, there are also 8 g4 h6
other weaknesses. Equally difficult is the 9 h3 Kd6
problem of what to exchange in this 10 Bd3 Bb7?!
type of ending.
A further inaccuracy. The bishop lifts
its control over the g4 pawn and retires
Szabo—Korensky to a passive position. Preferable was 10
. . . f6 or the more energetic 10.. ,g5!?,
Sochi, 1973 not allowing the K-side pawns to be
fixed on white squares.
(See next diagram)
11 h4! Bc8
White has doubled b-pawns on the left
side of the board, and it might be ex­ Now 11 . . . f6 no longer works due
pected that the game will end in a draw. to 12 hSgxhS 13 Nf5+.

112
 Positions with an Isolated d-Pawn

12 g5 hxg5
13 hxg5

White has achieved a great deal: he has

firm control over the highly important
central squares cS and e5, he has the
d4 square at his disposal, Black’s K-side
pawns are fixed on white squares, and
the a6 pawn is weakened. But for a win
these advantages would appear to be in­
sufficient. Black is spved by the doubled
b-pawns. Imagine that the white pawns
were at b4 and a4. After a4-a5 it is
unlikely that Black would be able to save Black is in zugzwang.
the game: . . . bS is obviously bad,
while after the exchange of pawns on aS 21 . . . f6
the white king acquires an additional 22 gxfó Bf7
square for manoeuvring —b4. 23 Bc8

13 . . . Ne6?! White’s plan is to transfer his bishop

to b3 and then play e3—e4.
Black takes play into a bishop ending,
in which the white king gains free access 23 ... Bg8
to d4. A highly committing decision. It 24 Bg4 Bf7
would have been safer to continue the 2$ Bc8 Bg8
battle with two minor pieces. 26 Bh3 Bf7
27 Bg4 Be8
14 Nxe6 Kxe6 28 Bf3 Bf7
15 Kd4 Kd6 29 Bdl Ke6
16 Be2 a5
17 bxa5 bxa5 29 . . . Be6 would also have failed to
18 Bb5 Be6? save Black.

Black’s last and decisive mistake. The 30 Bb3 Kxf6

bishop ending could have been saved 31 BxdS
only by exceptionally accurate play. In
order to understand Korensky’s mistake, White has won a pawn, and he now
we will first analyze the game continu­ confidently converts his advantage into
ation. a win.

(See next diagram) 31 ... Be8

32 e4 g5
19 Be8 Ke7 33 e5+ Kf5
34 fxg5 Kxg5
Otherwise White penetrates with his 35 Kc5 Kf5
king to e5, with an easy win. 36 Bc6! Bf7
37 Kd6 Bb3
20 Bc6! Kd6 38 Bd7+ Ke4
21 Bb7 39 e6 Kd3

113
 Endgame Strategy

40 c7 Bf7 1 ... Kf8?!

41 Ba4 Kc4
42 Kd7 As was mentioned earlier, centraliza­
tion of the king is hardly ever incorrect,
and within a few moves Black resigned. it can only be inopportune. Black should
In this game White first put his oppo­ have preferred 1 . . . Ne6, when if die
nent in zugzwang, then transferred his knight moves from d4 he has 2__ NcS
bishop to b3 and by e3—e4 retained with counter-play, while after 2 Nxe6
his extra pawn. In Sbakbmatny Bulletin fxeó 3 f4 h6 he has good drawing
1974 No. 2 Szabo showed that Black chances.
could have put up a successful defence.
Correct was 18 . . . Bf5 or 18 . . . Bg4, 2 Kfl Ke7
not occupying e6 until White plays
Be8. After 19 Be8 Be6 20 b3 (other­ 2 . . .Ne6 is unpleasantly met by 3
wise a zugzwang position cannot be ob­ Nxe6, when 3 . . . fxeó is not possible
tained) 20 . . . Ke7 21 Bc6 Kd6 22 due to the loss of the h-pawn. But 2 . . .
Bb7 f6 23 gxfó Bf7 White wins a h6 followed by 3 . . . Ne6 came into
pawn, but he is no longer able to transfer consideration.
his bishop to b3, and the game must
end in a draw. 3 Ke2 g6?!

This mistake is so 'popular' that there

is no point in drawing attention to it
Polugayevsky—Mecking each time.

Mar del Plata. 1971 4 Kd2 Nc6

5 Nxeó

Polugayevsky takes play into a bishop

ending, in which all Black’s pawns are on
squares of the colour of his bishop. As
we have seen in previous examples, bis­
hop endings are very difficult for the
weaker side, if, apart from his isolated
pawn, he has even just one more weak­
ness. Nevertheless the decision of the
Soviet player is highly committing, since
Black gains the opportunity to unite his
central pawns. White could have con­
tinued 5 Kc3, retaining all the advant­
It is White to move, and Polygayevsky ages of his position, Annotating this
immediately fixes Black’s Q-side pawns game, grandmaster Timman gives the
by following interesting variations: 5 Kc3
Nc5 6 Be2! (6 f3 Nxd3 7 Kxi3
1 a 5! Kd6 8 b4 gS!, with drawing chances).
By 6 Be2! Timman offers a pawn
and we again encounter a typical ending, sacrifice: 6 . . . Ne4+ 7 Kb4 Nxf2
the principles of which have been ana­ 8 Kc5 Ne4+ (8 . . . Kd8 9 KxdS Kcl
lyzed in detail in previous examples. 10 Ke5) 9 Kb6 Nd6 10 Kc7 g5

114
 Positions with an Isolated d-Pawn

(if 10 . . . f5 , then 11 Bf3 Be6 12 Ne2)

11 Sf3 Be6 12 g4 followed by 13
Nf5+.

5 ... fxc6

Of course, capturing on e6 with a

piece would be a blunder.

6 f4 c5!
7 g3

pawn, is unable to win; (c) 17 Kc3!

bxa5 18 Kxd2 Bb5 19 Bdl Kd6
20 Kc3 Kd5! 21 Bc2 Bd7 intending
22 . . . Bf5, when Black is not in danger
of losing.
To be fair, it should be said that it was
hardly possible for Mecking to foresee all
these lengthy and complicated variations
at the board.

8 KcJ Be6

It transpires that here it is not possible

to go into the pawn ending by 8 . . .
This natural move, centralizing the Bb5, since after 9 Bxb5 axb5 10 Kb4
king, turns out to be a serious mistake. d4 White has the capture 11 fxe5 with
Timman showed that Black could have check.
drawn by 7 . . . Bb5!, when White is
forced to retreat 8 Bc2, since in the 9 Kb4 exf4
pawn ending Black gains a draw after
8 Bxb5 axb5 9 Kc3 Ke6, when on In view of the threat of 10 fxe5+
10 Kb4? there follows 10 . . . d4! Kxe5 11 Kc5, Black is forced to con­
After the retreat o f the white bishop cede the centre. If 9 . . . d4, then 10
Timman gives the variation 8 . . . Kd6 exd4 followed by h2—h4—h5.
9 Kc3 Be2 10 Bbl! Bfl 11 Kb4d4!
12 exd4 exd4 13 Be4 Kc7 14 Kc5 10 gxf4 Bg4
d3 15 K d4d2 16 Bf3 b6! 11 Kc3 Bf3
12 Kd4 Bg2
(See next diagram)

The Dutch grandmaster gives a detail­ Black keeps his bishop on the long
ed analysis of this position: (a) 17 b4 diagonal, in an attem pt to prevent
bxa5 18 bxa5 Kd6 19 Kc3 Kc5, e3—e4.
regaining the pawn after 20 Kxd2 Kb4;
(b) 17 axb6+ Kxb6 18 Kc3 Kc5! 19 13 h4 Bf3
Kxd2 Kd4, and White, despite his extra 14 b4 Bhl

US
 Endgame Strategy

If the black bishop moves off the long

diagonal, White wins by transferring his
bishop to g2 and then advancing e3—e4
when the black bishop is at f 7 or g8.

15 Be2 Bg2
16 Bg4 Be4
17 Bc8 Kc7
18 Be6 Kd6
19 Bg8 h6
20 Bf7 h5

White stands better. Although the

Black is in zugzwang, but it is Whitemost unpleasant pieces for Blade —the
to move. white-squared bishops — have been ex­
changed, the advantage of the bishop
21 3e8 Bc2 over the knight is fairly appreciable. The
22 Bf7 Be4 black knight is well placed, but it cannot
23 f5! take part in the defence of the d-pawn,
since the e7 square is easily controlled
The decisive breakthrough. by the white bishop from a3. Thus Black
is obliged to keep his rooks tied to the
23 . . . BxfS defence of his isolated pawn. With his
next few moves Black takes his king to
After 23 . . . gxf5 a possible vari­ the aid o f his rooks, and White attempts
ation is 24 BxhS Ke6 25 Be2 Kd6 to prevent this.
26 h5 Ke6 27 h6 Kf6 28 Bxaó! bxa6
29 b 5, and one of the white pawns 1 g4! Kf7
queens. 2 Ke2 Ke6
3 f4 f5?!
24 Bxd5 Bc8
25 e4 Black has prevented f4—f5+, but at
a high price. The long diagonal has been
Zugzwang. opened for the white bishop, and targets
have appeared for the white rooks on the
25 . . . Ke7 g-file. 3 . . . g6 was more circumspect.
26 Ke5 g5
27 hxg5 h4 4 gxf5+! Kxf5
28 g6 h3 5 Kf3 Ke6
29 g7 h2 6 h4! Rfg
30 g8=Q hl=Q
31 Qf7+ Kd8 6 . . . g6 was bad because of 7 h5!
32 Qf8+ Resigns.
7 Kg4 g6?

Ribli--Pinter Pinter decides to exploit the fact that

the white king has moved to g4 and
Bailie Hcrclanc, 1982 that 8 h5 is no longer very dangerous,

116
 Positions with sn Isolated d-Pawn

since 8 . . . gxh5 follows with check.

But he has forgotten that the white king
has moved away from the check that the
d-pawn will announce when capturing
on e4. To be fair, it must be said that
playing such a position with Black is
very difficult.

8 e4!

White’s advantage becomes decisive.

8 ... h5+
1 ... f6
8 . . . Rfd8 does not help due to 9
exd5+ RxdS 10 Rxd5 Rxd5 11 Rxd5 Black’s initial task is to bring his king
Kxd5 12 f5. to the centre and securely blockade the
d-pawn.
9 Kg5!
2 K fl Kf7
This bold raid by the white king wins 3 Ba5 b6
material. 4 Bc3 Rd8
5 Bb2 Rd6
9 . .. d4 6 g4!?
10 Bxd4 Nd8
11 Bg7! It is to White’s advantage to simplify
the position and reduce the number of
Pretty, and very strong. pawns.

11 ... Nf7+ 6 ... hxg4

12 Kxg6 Nh8+ 7 hxg4 Re6?!
13 Kh6 Nf7+
14 Kxh5 Resigns. After the exchange o f rooks this type
of ending is drawn, if the side with the
isolated pawn does not have any addi­
Averbakh-Keres tional weaknesses. In his book of select­
ed games, Keres makes the following
18th USSR Championship comment on this m ove
Moscow, 1950 “Not the best plan. Black wanted to
facilitate the access of his king to the
(See n ex t diagram) cherished d5 square, but, firstly, the
exchange on e6 cannot be forced, and,
Black has chances o f obtaining an out­ secondly, even if it could be forced the
side passed pawn on the Q*$ide (with his simplification would only favour White.
pawn at aS and the white pawn at a4, The ending after 8 Rxe6 Kxe6 9 Ke2
he can play . . . bS in a favourable situ­ Kd5 10 Ke3 is advantageous to Black,
ation). Black can also attem pt to give his but White has good chances o f resist­
opponent an additional weakness on this ing. Therefore, if account is taken of
part o f the board. the fact that the game was shortly to be

117
 Endgame Strategy

adjourned, better was 7. . .Nf8 follow­ Black, on the other hand, has sufficient
ed by 8 . . . Ne6 and 9 . . . Rd5, which time to prepare a break-through by
in the end is what in fact happened . . . b5. But this can be done only after
later.” thorough preparation. Thus, for ex­
ample, after 15 Ke2 it would be pre­
mature to advance 15 . . . b5 due to
16 a5 b4 (16 . . . «5 is slightly better)
17 a6!, when White unexpectedly ob­
tains counter-play (17 . . . Nc7 18
Kd3 Nxa6 19 Kc4, or 17 . . . Ra5 18
dS cxdS 19 Rxb4). Therefore Black
tries to strengthen his position to the
maximum, by transferring his king to d7,
when he will prepare . . . g6 a n d __ a6,
and only at the appropriate moment
play . . . b5” (Keres).

15 Kg3 Ke7
8 f3 16 g5!P

Here, and over the next few moves, Averbakh tries to catch his opponent
White wrongly avoids the exchange of in a trap, but in doing so bums his boats
rooks. behind him. Objectively stronger was
16 Kf2 with chances of a draw, but it is
8 ... Ne7 difficult to condemn White for taking
9 Bel Nd5 this risk. Had it succeeded he would have
10 Bd2 Rd6! immediately gained a draw, whereas
whether passive defence would have
"Black decided henceforth to avoid saved the game is unclear.
the exchange of rooks” (Keres).
16 ... f5
11 Ke2 Rd8 17 Re5 Kd6
12 Kf2 Nc7! 18 Rxd5+ Kxd5
19 g6!
A strong move, provoking a weaken­
ing of White’s Q-side and preparing to
bring the black king to the centre.
13 a4
The threat was 13 . . . Nb5, and if
14 Be3 Nd6, winning the exchange.
13 . . . Ne6
14 Be3 Rd5
“White’s pieces are tied to the defence
of his d-pawn, and he is obliged to re­
strict himself to passive defence, practi­
cally only king moves being possible. 19 . . . a5!

118
 Poiitions with an Isolated d-Pawn

Do not huny! "The subtle trap set by in terms of forcing a draw” (Keres).
White was that the pawn ending after We have given in full the analysis by
19 .. . Nxd4 20 Bxd4! Kxd4 21 Kf4, the celebrated grandmaster, in order to
which at first sight seems completely demonstrate once again the care with
hopeless, is in fact drawn! The best con­ which one must weigh up the con­
tinuation for Black is 21 . . . b5! 22 sequences of an exchange leading to a
axbS cxb5, but even then White un­ pawn ending. Of course, the practical
expectedly saves the game after 23 b4! player would not be obliged to waste
Kc4 24 Kxf5 Kxb4. Let us examine time on the calculation of such com­
this position in some detail: plicated variations, but on general
(a) 25 Ke6 is the first move which grounds would make a more useful
comes to mind, so as to answer king move, since in any case the d4 pawn is
moves with Kf7xg7—f6 and the advance doomed.
of the g-pawn. But Black replies 25 . . .
a5!, when White’s position becomes 20 Kh4 Nxd4?!
critical, e.g.:
(al) 26 f4 a4 27 f5 a3 28 fógxfó! Here too Black should not have
29 g7 a? 30 g8=Q al=Q 31Qf8+Kb3, hurried over the capture of the pawn.
with an easy win. He should have advanced his b-pawn to
(a2) 26 K f7a4 27 Kxg? ¿3 28 Kf7 b4 and only then taken on d4. By the
a2 29 g7 al=Q 30 g8=QQa2+, and move played Keres allows White to
it seems time for White to resign, but activate his bishop.
here, in this seemingly hopeless position,
he is able to draw! 31 Kf8 Qxg8+ 32 21 Bh6! Ne6
Kxg8 Kc4 33 f4 Kd5 34 Kf7! If 22 Be3 c5
now 34 . . . Ke4 35 Ke6 Kxf4 36 23 Kh5 Ke$?
Kd5 with a draw, or 34 . . . b4 35 f5
b3 36 f6 b2 37 Kg7 bl=Q 38 f7 By 23 . . . c4! Black could have won
with a theoretical draw. Amazing! easily (24 bxc4+ Kxc4 25 Bxb6 Kb4
Perhaps in this variation Black should and 26 . . . Kxa4), but he decided to
not exchange queens? However, after try and win without giving his opponent
30 . . . Qa8+ 31 Kg7 Qxf3 we reach the slightest chance. "But instead of
a queen ending in which it cannot be making a simple and dear move, Black
said with any confidence that the begins manoeuvring in the hope of gain­
stronger side has a win. ing an even easier win in the opponent’s
(b) 25 f4! is undoubtedly stronger: time trouble. Such tactics are complete­
(b l) 25 . . . Kc5 26 Ke6! b4 27 ly inappropriate, and lead to Black over­
f5 b3 28 f6, or looking an important defensive possibil­
(b2) 25 . . . Kc3 26 Kg5! b4 27 ity, which jeopardizes the win” (Keres).
f5 b3 28 f6 gxf6+ 29 Kxf6 b2 30 In this case Black should not have
g7 bl=Q 31 g8=Q, and in neither case abused the principle "do not hurry”.
does White have any difficulty. True, in a practical game it is difficult
(b3) 25 . . . a5!, and White’s posi­ for even a very strong player to find the
tion again seems critical, since 26 Ke6 golden mean.
a4 27 f5 a3 28 f6 gxf6 etc. leads to
a familiar won position for Black. But 24 Bel Nd4?!
even here White has the saving 26 Ke4!
a4 27 Kd3!, when his king arrives just As shown by Keres, he should have re­
in time, and it is Black who has to think turned his king to d5, and if 25 Bb2 c4

CS-1 119
 Endgame Strategy

26 bxc4+ Kxc4 27 Bxg7 Nxg7+ 28 has another possibility: 28 Kg5! with

Kg5 Nc8! 29 Kxf5 Kd5 followed by the threat of Bf6! The knight would
. . . b5. have had to return to d4, but after 28
. . . Nd4 29 Bxb6 Nxf3+ 30 Kf4 or
25 Bh6 Kf6 30 Kh5 White has good drawing chances.
26 Bg5+ Kc6 28 . . . f4 is no better due to the simple
29 Kxf4.
But even so Black has a way to win.
On 27 Bd8 he plays 27 . . . Kd7! 28
Bf6 (if now 28 Bxb6 Nxb3 29 KgS c4,
and wins) 28 . . . Ne6 29 Be5 Kc6
30 Bxg7 Nxg7+ 31 Kg5 Ne8 32 KxfS
Kd6! 33 f4 Kd5 34 Kg5 Ke6 35 f5+
Ke5 etc., or 33 Kf4 Ke6 34 Ke4
Nd6+ 35 Kf4 Kf6 36 g7 Kxg7 37
Ke5 c4! etc.
Thus 27 Bd8! would not have saved
White, but Black would still have had a
lot to do" (Keres).

27 Bh6? 27 . . . gxh6
28 Kxh6 Nc6!
"The decisive mistake in time trouble. 29 g7 Ne7
White misses the excellent chance of 27 30 Kh7 Kf7
Bd8! During the game Black thought 31 Kh6 Kg8
that this move was not possible due to 32 f4 Kf7
27 . . . Nxb3 28 Bxb6? c4, but White White resigns.

120
 CHAPTER 11

THE TWO BISHOPS

It is generally reckoned that in the White should have played 2 a4 follow­

majority of cases two bishops are stronger ed by the transfer o f one knight to c4,
than two other minor pieces. But while and the other via cl and b3 to d2.
in the middlegame the advantage of the White’s move could have been explained
two bishops is by no means always an by a desire to set up the above formation
important factor, in the endgame it is by Ng3—f l —d2. But the further course
often decisive. Thanks to their long of the game shows that he is not
range, the bishops are excellent for sup­ attempting to think schematically, but
porting pawn advances which seize space makes moves without any plan.
and create weaknesses in the enemy posi­
tion. In the end the stronger side fre­ 2 ... h5l?
quently uses the principle of two weak­
nesses, and, by combining threats, breaks In his notes to one of Steinitz’s games,
through on one of the wings. Euwe writes:
“What is the advantage of a bishop
over a knight? It is that it can influence
Richter-Tarrasch the battle from afar, whereas the knight
affects squares only in its immediate
Nurembcig, 1888 vicinity. How can the advantage of the
bishop best be exploited? To do this the
knight must be pushed back as much as
possible, and then all the time prevented
from coming into play. For this, pawns
are the most suitable.”
Later Euwe continues: “Of course, it
is by no means always that a pawn ad­
vance leads to such a favourable position
as in the given instance. But it can be
considered that, if it has the result of
weakening the opponent’s pawns, then
with the two bishops it must be con­
sidered favourable, since with the long-
range bishops the resulting position can
With two bishops against two knights be better exploited than with the knights.
Black has an undisputed advantage, in On the other hand, it should be mention­
spite of the symmetric pawn formation. ed that knights are better at exploiting
weak squares in the form of outposts
1 ... c5 from which they cannot be driven.
2 Ng3?! Therefore the pawns should be advanced
in such a way as to minimize the number
An àimless move. The only chance of of weak squares created. It is dear that
a successful defence is to create strong for this aim the most suitable are rooks’
points for the knights. In the given posi­ pawns, since each move of a rook’s pawn
tion c4 is a possible strong point, so weakens only one square (the one along-

121
 Endgame Strategy

side its new position), whereas a move

by any other pawn simultaneously
weakens two squares."
Euwe’s remarks are fully applicable
to the ending in question. After 1 . . . c5
White could have immediately trans­
formed c4 into an outpost.

3 f3?!

White quite unnecessarily weakens the

e3 square.

3 ... Bd7 White removes all the barriers preventing

4 Re2 his knight from occupying the outpost at
e4. It is not easy for Black to counter
It was not yet too late for 4 a4. this, e.g. 13 . . . Rf6 (hardly good is
13 . . . hxg4 14 fxg4, when White ob­
4 . .. bS! tains the additional square f3 for a
knight) 14 gxf5 Bc6 15 Rxdó Rxd6
Black also starts an offensive on the 16 Nxc5, and for the exchange White
Q-side. has fair compensation. 16 . . . Bxf3?
fails to 17 Nxf3 Rxc5 18 Nxg5+ Kg8
5 Rael Bf8 19 Ne4 Re5 20 Nf6+.
6 Nge4
12 . . . Ra8
Having missed the opportunity to
establish a knight at c4, White is unable Black prepares the advance of his
to set up a systematic defence. rook’s pawn on the Q-$ide. To be con­
sidered was the exchange of one pair of
6 ... Rg8 rooks, so that the bishops should not be
7 Nb3 Rc8 diverted by having to defend c6 and e7.
8 Ned2 Bd6
9 Ne4 Bf8 13 Na$ Rab8
10 Ned2 fS?!
Defending against 14 Nb7.
Tarrasch plays purposefully, but too
directly. He should have made one 14 Nab3 h4
further prophylactic move, 10 . . . Rg6, 15 Khl Rg6
and only then set his pawns in motion. 16 Kgl Be6
17 Rf2 Ra8
11 Re5!? Bd6 18 Rfe2?
12 R5e2?
White should not have allowed the ad­
vance of the a-pawn, although it is un­
The decisive mistake. Meanwhile, likely that 18 Na5 Bd5 would have
White had a chance to obtain counter­ affected the result.
play after 12 Rd5! Rg6 (12 . . . Rc6?
13 Nxc5!) 13 g4! 18 . . . a5!

122
 The Two Bishops

19 Nbl a4 have played 3 Kd2!, and if 3 . . . f6

20 N3d2 c4 4 Nxc6 Be3+, then 5 Ke2!, and it is
21 Nfl Rc8 not clear that Black can gain any advant­
22 Khl c3 age from a discovered check. On 3 Kd2
23 bxc3 dxc3 Rubinstein would most likely have re­
24 Ne3 b4 plied 3 . . . Bb7.

and within a few moves White resigned. 3 K fl Bb7

4 c3 f6

Tarrasch—Rubinstein Otherwise White plays d3—d4 follow­

ed by Nd3.
San Sebastian, 1912
5 Ng4

No better is 5 Nf3 Be 3 6 g3 Bc8.

5 ... hS
6 Nf2 Be 3
7 Bdl h4
8 g3 a5!

Black carries out a logical offensive

with his rooks’ pawns. This move pre­
pares . . . b4, creating vulnerable points
on White’s Q-side. Black docs not fear
9 Ng4 in view o f 9 __ b4.
With his next move Black exchanges
knights, and we again see a battle be­ 9 Bf3 b4
tween two bishops and two other minor 10 Kg2 bxc3
pieces. 11 bxc3 Ba6

1 . .. Nxe2 Attacking the d-pawn. White is unable

to avoid the creation of new weaknesses,
Of course, not 1 . . . Nxb3+, since since Black is threatening to invade with
after 2 axb3 Bb7 3 d4 the white his rooks along the b-file. 12 Rhdl is
knights acquire an excellent strong point bad due to 12 . . . Rab8.
at c5.
12 c4
2 Kxe2 Re8!

Before defending his c-pawn, Rubin­ Now the d4 square is decisively

stein sets his opponent a difficult prob­ weakened. Once it is occupied by the
lem, with which Tarrasch fails to cope. black bishop, the white knight will be
Black threatens 3 . .. f6, and so White completely helpless (one weakness), and
is forced to move his king. The natural Black can calmly set about winning the
3 Kd2 seems bad due to 3 . . . f6 4 a2 pawn (the second weakness).
Nxc6 Be3+ 5 Kc3 Be6, with a very
strong attack. But even so White should 12 . . . R ad 8

123
 Endgame Strategy

With the threat of 13 . . . dxc4 14 Domination! The white knight has no

Bxc6 Re6. moves. 28 Nf2 is bad due to 28 ..
Bxf2 29 Kxf2 Kf4.
13 cxd5 cxd5
14 R hdl Re7! 28 Kf3 Kd4

Played according to the principle “do First d4 was occupied by Black’s

not hurry!”. Black transfers his bishop bishop, and now his king uses it as a
to d4 only when the white knight will spring-board for transferring to the Q-
come under its domination at g4. Be­ side.
fore this there is no point in him lifting
his control of c l. He now threatens 29 Bb3 Bb7?l
. . . Rc7 followed by . . . Rc2.
Here Black could have ignored the
15 Ng4 hxg3 principle “do not hurry”, and won by
16 hxg3 Bd4 29 . . . Bxd3 30 Bxd5 B fl 31 Nf2
17 Racl Rb7 Bxf2 32 Kxf2 Kxd5 33 K xfl Kc4.
18 Rc2 Kf7
30 Ke2 Ba6
Rubinstein does not forget about 31 Bc2 Bb5
centralizing his king.
Threatening the advance o f the a-
19 Nf2 Rb2 pawn to a3.

Black exchanges one pair of rooks, 32 a4 Bd7

to safeguard the advance of his king to 33 Kf3 Kc3!
the centre.
After a concrete evaluation of die
20 Rxb2 Bxb2 position, Rubinstein goes for the ex­
21 Rd2 change of bishops.

21 d4 fails to 21 . . . Bc4. 34 Kxe3 d4+!

21 ... Bd4 An important link in Black’s plan. If

22 Nh3 Ke6 immediately 34 . . . Kxc2 35 Kd4,
23 Rc2 Kd6 with counter-play.
24 f5
35 Ke2
White tries to activate his knight, but
Black finds a convincing rejoinder. 35 Ke4 is hopeless due to 35 . . .
Bc6+.
24 . . . Rc8!
25 Bdl 35 ... Kxc2
36 Nf4 Bxa4
25 Rd2 is decisively m et by 25.. .Rc3. 37 Ne6 Bb3!
38 Nxd4+ Kb2
25 . . . Rxc2+ 39 Nb5 a4
26 Bxc2 Ke5 40 Ke3 a3
27 g4 Be3! 41 Nxa3 Kxa3

124
 The Two Bishops

42 Kd4 Kb4 draw. Perhaps White wanted to obtain

White resigns. winning chances and so he provoked the
advance of the g-pawn, which weakens
Black’s control of f5? Although rather
risky, such tactics are perfectly possible.

3 ... g5!?
Michel-Tartakover
Black begins advancing his K-side
Mtrienbad, 1925 pawns.

4 Nf3 h5
5 Be 2?

An incomprehensible move. White as

though taunts his opponent: “Do what
you want, all the same the game will end
in a draw”. But such passive tactics are
very dangerous against two bishops.
After 5 Bf5 White would have had
everything in order.

5 ... Re4!
6 Bd3 Rf4!
Black has two bishops, but there are 7 Ke2 g*
no weaknesses in White’s position and he 8 hxg4 hxg4
has the more compact pawn formation. 9 Nh2 g3
This suggests a draw as the likely result,
but to achieve this White must play Forced, unfortunately. Of course,
systematically, and in particular create Black would have preferred not to free
a strong point for his knight. the knight.

1 Kgl Kg7 10 Nf3 d4

2 Kfl 11 R fl b4

Why not the natural 2 Kf2? After All White’s pawns are fixed on white
all, 2 . . . Bh4+ is not dangerous in squares, and in the event of the knight
view of the simple 3 Kgl, when White being exchanged for the black-squared
threatens both 4 Nd4 and 4 R al. bishop, Black will win the ending.

2 ... Bc6 12 Nd2!

3 Ngl
Against passive play by White, Black
The knight is transferred to f3, where could have strengthened his position by
it occupies an insecure position — it can transferring his king to the centre, his
be driven away by the g-pawn. A good rook to the h-file, and his bishop via e5
strong point for the knight would be d4, and f4 to e3.
and here it would have been quite
sensible to play 3 c3 with a probable 12 . . . Rh4

125
 Endgame Strategy

18 K dl Kf6
19 Kd2 Kg5
20 Kdl Kg4
White resigns.

Averbakh—Botvinnik

22nd USSR Championship

Moscow, 19S$

13 Nf3?

It transpires that White made his pre­

vious move without any intention of
playing actively, and yet he had an inter­
esting possibility to do so: 13 Rf3! The
exchange sacrifice has to be accepted,
since after 13 . . . Be5 14 Rf5 Bb8
15 Nf3 Rh2 16 Kfl White is out of
danger. Thus there could have followed
13 . . . Bxf3+ 14 Kxf3 Be5 15 Be4!
Rf4+ (if 15 . . . Kf6, then all the same Here the pawn formation is not sym­
16 Ke2, and if 16 . . . B f 4 17 N f3) metric, and Black's K-side pawns can
16 Ke2 Rf2+ 17 Kd3, when White easily advance, cramping White's posi­
sets up a strong and possibly impreg­ tion. White has no way of opposing
nable defence. Having missed this possi­ Black’s plan. Play on die Q-side will
bility, White quickly succumbs to an ex­ merely create weaknesses, so he can only
change sacrifice, but this time by Black. watch the unfolding of events.

13 . . . Rh8 1 ... e5
14 Kd2? 2 Nel e4
3 Nc2
This allows a decisive blow. True,
after 14 Kel Rh5 followed by the Averbakh defends logically. By the
transfer of the black king to the centre advance of his e-pawn Black has weaken­
White would have had a difficult posi­ ed his control of d5. It is this square
tion. that the white knight aims for, since in
the event of its exchange for Black's
14 • • • Rh2! white-squared bishop, even with the loss
15 Nxh2 gxh2 of a pawn an ending is reached with
16 R hl Be5 opposite-coloured bishops and their
17 Bfl Be4! characteristic drawing tendencies.

A picturesque position. White has 3 ... Qd6

only king moves left. 4 Ne3 Qd4

126
 The Two Bishops

The queen has taken up an ideal posi­ 20 Kg2 aS!

tion, and White will be forced to ex­
change it. It will then be possible for the Only now does Black reveal his idea.
black king to advance into the centre. White was ready to meet 20 . . . b5, on
which there could have followed 21
5 Nd5 BgS! axb5 axb5 22 Nb4 with the threat of
Nc6, and Black’s plan caught him un­
Of course, neither now, nor subse­ awares. There is no way o f preventing
quently is Black tempted into winning a . . . f 4.
pawn after the exchange on d5. It is
interesting to follow how Botvinnik 21 K fl f4!
manoeuvres with his black-squared 22 gxf4 gxf4
bishop. 23 Nxf4

6 g3 f5 Forced, in view of the threat of 23

7 h4 Bd8 . . . f3 24 Bdl Bxd5 25 cxd5 Kc3.
8 Bc2 Kf7
9 Qdl Qxdl+
10 Bxdl Ke6 23 . . . Kc3
11 Nf4+ Kf6 24 Bdl Kd2
12 K fl g5 25 Bg4 Kc2
13 hxg5 hxg5
14 Nd5+ Ke5
15 a4 Kd4 The blade king has broken through
to White’s Q-side pawns. It is true that
Black has achieved a great deal: his he now succeeds in exchanging bishops
king occupies a dominating position in and winning the e4 pawn, after which
the centre, and his K-side pawns cramp there is little material left on the board.
White. But how is he to win the game? Black’s rook’s pawn nevertheless deddes
White's position is held together by the the game. It is interesting to follow the
knight at d5. Undermining its strong battle between bishop and knight, in
point by . . . b5 is ineffective in view which the bishop is dearly dominant.
of the considerable reduction in the
number of pawns. To Black’s aid comes
a pawn sacrifice enabling his king to 26 Be6 Bxe6
break through on the Q-side, which first 27 Nxe6 Be7!
of all he blocks. 28 Nd4+ Kc3
29 Nf5 Bf8!
16 Be2 Bc8 30 Ng3 Kxb3
17 Kg2 31 Ke2 Kxa4
32 Nxe4 Kb3
White should also have played his king 33 Kd3 a4
across to the Q-side, but for this he 34 Nd2+ Kb2
would have had to anticipate Black’s 35 c5 b5!
plan. 36 c6 Bd6
37 Ne4 Bb8
17 . . . Bd7 38 Nc3 a3
18 K fl Beft 39 f4 Bxf4
19 Kgl Bf7 White resigns.

127
 Endgame Strategy

Englisch—Steinitz position on both wings.

London, 1883 8 Rxd8 Rxd8

9 Be 3 h6
10 Rel f5
11 f4 Bf6
12 g3
Having gained space on the K-side,
Black switches play to the Q-side.

12 . .. a5!
13 Ncl a4
14 a3 Bc4
15 Kf2 gxf4

A concrete solution to the problem.

Black’s bishops rake the board, and he 15 . . . Rd5 followed by . . . b5—b4
threatens 1 . . . Bxb2 2 Rabl Bd4, was also very strong.
winning a pawn. White is forced to
weaken his position. 16 Bxf4 Bg5l

1 c3 After the exchange of bishops the

helplessness of the white knight becomes
There is hardly anything better. apparent. The invasion o f the rook at d2
1 Rabl is very unpleasantly met by is now threatened.
1 . . . Bf5.
17 Bxg5 hxg5
1 ... Rfe8 18 Ke3 Kf6
2 Nb3 b6! 19 h4

Beginning a consistent plan to restrict This loses quickly, but the position
the mobility o f White’s minor pieces. was already untenable.

3 h3 Be6 19 ... gxh4

4 R fd l c5 20 gxh4 Re8+
21 Kf2 Rxel
Black’s Q-side pawns neutralize the 22 Kxel Ke5
white bishop and deprive the knight of 23 Ne2 Bxe2
d4. 24 Kxe2 Kf4
25 c4 Kg4
5 Bg5 f6 26 Ke3 f4+!
6 Bf4 Kf7
After 26 . . . Kxh4?? 27 Kf4 it is
It is useful to centralize the king. White who wins.

7 f3 g5 27 Ke4 f3
28 Ke3 Kg3
The black pawns squeeze White’s White resigns.

128
 The Two Bishops

Blackbume—Lasker 10 . . . Bxc3 and 11 . . . Rxe2.

London, 1892 9 Bfl bS!

10 Nd3 Bd4
11 Ne2 Bb6
12 b3 Kg7!

There is no reason to hurry. The

centralization of the king is always
useful.

13 c3 Kf6
14 Kc2 Ne7

The time has come to aim for e3.

15 Neel Nd5
The pawn formation is as in the pre­ 16 Kb2
vious example. White has a weakness at
e3, but for the moment it cannot be The white pieces are extremely cramp­
approached. Lasker begins a systematic ed, and there is nothing to prevent Black
restriction of the white pieces. from continuing to strengthen his posi­
tion. But Lasker chooses the shortest
1 ... a6 path —a combinational one.
2 Bd3 f5!
16 . . . b4!
Black does not exchange his knight 17 Nxb4
for the white bishop, but erects a line of

E awns in its path. When there are two Forced. After 17 cxb4 Black wins
ishops and a knight against two knights by 17 . . . Bd4+ 18 Ka3 Ne3.
and a bishop, the exchange of the lone
knight frequently eases the defender’s 17 . . . Ne3
task. 18 R el Nc4+
19 Bxc4 Rxel
3 N4c3 Be6
4 Kbl Rfd8 Black has won the exchange. The rest
5 Nf4 Bf7 is straightforward.
6 Be2 Nc6
20 Bxa6 Rgl
Black takes control of d4, which in 21 g3 Rg2+
case of necessity can be occupied by 22 Ka3 Rxh2
either knight or bishop. 23 Ne2 Rg2
24 Nc2 g5
7 Rxd8+ Rxd8 25 Bd3 h5
8 R dl Re8! 26 Kb4 Bf2
27 a4 c5+
In the given position the exchange of 28 Kb5 Bxb3
any pieces favours White, who is short of 29 a5 c4
space. Now Black threatens 9 . . . g5. 30 Bxc4 Bxc2

129
 Endgame Strategy

31 a6 Bdl 2 Nd6 Rad8

32 Nd4 Bxd4 3 NfS Rd7
33 cxd4 Bxf3 4 BgS NhS
34 dS Be2 5 R fdl h6
35 Bxe2 Rxc2 6 Be3 Rfd8
36 a7 Ra2
White resigns. White’s pieces have taken up excellent
posts. It is new time for him to turn his
attention to the knight at d3. How can
it be driven away? Kotov approaches this
Kotov—Florian question differently: let the knight
remain where it is, but dislodge its sup­
Moscow, 1949 port — the pawn controlling c5, where
the knight has an alternative and no less
comfortable post.

7 Ba4!

“Such moves are often more difficult

to find than lengthy, forcing combina­
tions. However, their effect is just as
strong as that of many sacrifices"
(Kotov).

7 ... b5

Practically forced, since on 7 . . . Rc7

The position is an open one. White has White has the highly unpleasant 8 Rd2
the two bishops, but there are no weak­ followed by R adi.
nesses in Black's position, and his ad­
vanced knight at d3 seriously restricts 8 Bb3 Nhf4
the opponent’s play. White must first 9 Rd2 Ned
deploy his pieces well, and then deal 10 R adi b4
with the annoying knight. Which white 11 Bd5
piece stands worst? The knight at d2!
Kotov finds an excellent post for i t The knight is forced to retreat, and it
at f5. has no strong point.

1 Nc4! Ba6?l 11 . . . Ndc5

12 Nd4!
Black does nothing to oppose White’s
plan. To be considered was 1 . . . Rfd8. Forcing further concessions by Black,
Black probably did n o t like the fact that in view of the threats of 13 Nc6 and
after 2 Bg5 h6 3 Bxfó gxfó 4 Ne3 13 Nxeó fxe6 14 Bxe6+.
his pawns would be broken up. But at
the same time White would have been 12 ... Nxd4
deprived of his splendid bishop, which 13 Rxd4 Rb8
would have allowed Black to put up a 14 R bl Nd3
stubborn defence. 15 Rb3

no
 The Two Bishops

This leads to gain of material. 2 Nxc4 Nxe4

3 R hfl Bd7
15 . . . Rc7? 4 R bl b6

An oversight in a lost position. The first weakening.

16 Rbxd3 Bxd3 5 f3 Ng5

17 Bf4! 6 Kd2 f5

Black was hoping for 17 Rxd3, which Black would like to play 6 . . . Ne6
is also good enough to win, but the move followed by . . . c5, but on 6 . . . Ne6
in the game forces the win of a piece. there follows 7 e4!

Black resigns. 7 a4! Nf7

8 a5 b5

Kotov—Katyetov The aim is achieved: a complex of

black squares on the Q-side has been
Moscow, 1946 weakened. But how is White to exploit
this weakening, since he has no knight?
This mission is assigned to the white
king. So that its journey should be a safe
one, in accordance with the prindplc of
two weaknesses White must strike a blow
on the other side o f the board, to create
vulnerable points there and divert the
enemy forces.

9 Rf2 h5
10 h4 Re7
11 Bf4 g6>!
12 g4!

The white bishops are strong, especi­ Black’s last move assisted White’s plan
ally the blade-squared one which has no to a considerable extent. There is now
opponent. White’s superior centre en­ the threat o f an attack on the black king.
ables him to prepare f2—f3 and e3—c4
or c3—c4. But Black has no weaknesses, 12 ... Kh7
and the advance of the white pawns may 13 gxh5 gxh5
lead to the creation of strong points for 14 Rg2 Rg8
his knights. Kotov chooses a different 15 Rxg8 Kxg8
plan, in which the leading role is played 16 Kc2 Kh7
by the bishop at h2. White gives Black 17 Rgl
w e a k n e ss on the Q-side, exploiting the
half-open b-file, the black-squared bis­ White controls all the key points of
hop and the a-pawn, w ithout under­ the position. Now it is the turn o f his
taking any activity in the centre. king.

1 Kc2 Nge4 17 . . . Re8

131
 Endgame Strategy

18 Kb3 Rc8 across to the Q-side, Black attacks the

19 Kb4! pawns at f4 or h4 with his bishop. But
when analyzing the adjourned position,
The white king boldly advances. By a my trainer Isaak Boleslavsky found that
desperate pawn sacrifice Black tries to White had serious winning chances. A
halt its advance, but he comes under brilliant analyst and a grandmaster of
attack by the white bishops. world class, Boleslavsky appreciated very
well the potential of the two bishops. In
19 ... c5+ spite of the fact that the play was
20 dxc5 a6 essentially confined to one wing and that
21 e4! dxe4 there was very little material, he found a
22 fxe4 Rc6 latent plan of playing for a win. It can­
23 exf5 Rf6 not be ruled out that, had Black realized
24 Rg6 Resigns. the danger facing him, he could have
found a defence. But it was evident that
my opponent also considered the posi­
tion to be drawn.

Shereshevsky—Yuf erov 1 Bf2

Minsk, 1971 Defending against 1 . . . Bel+.

1 ... Kg7?!

Black sticks to waiting tactics, since it

is not apparent how White can refute
them. It would have been stronger to
transfer die knight via d8 and b7 to cS.

2 Bf3 Kh6

Defending against h4—h5.

3 Be2

Now it is not easy fo r the black knight

This game was played in the last to break out, but so what? After all.
round of the Sokolsky Memorial Tourna­ Black has something akin to a fortress.
ment. Only a win would earn the author But the two bishops seem able to break
of these lines the title of USSR Master up any fortress, even the most solid.
of Sport. Shortly before the time con­
trol, in a difficult position, Black was 3 ... Rd2
able to land a clever tactical blow, as a 4 Kf3 Kh7
result of which the game went into an
ending where a win for White seemed im­ Having improved his position, White
probable. I thought that the position was begins to cany out his plan.
a dear draw. Indeed, how can White im­
prove it? The threat of h4—h5 is easily 5 ,Bd4 Nh6
parried, and if the white king moves 6 BbS Bel

132
 The Two Bishops

21 Bb4 Be 7
22 Kc4 Kh5
23 Bc3 Bh4
24 Kb5 Bg3
25 Bd2 Kh4
26 Kc6 Resigns.

Gheorghiu—Olafsson

Athens, 1969

7 h5!

The point of Boleslavsky’s plan. By

returning his extra pawn, White breaks
up the opponent’s pawns, after which
the two bishops go toi work.

7 ,,, gxh5
8 Be8! h4
9 Bd7 Kg6
10 Be6 Nf7
11 Bb6
White has a mobile pawn centre and
Black is completely tied up. The only two strong bishops, but the closed
way to avoid loss of material is to move nature of the position and the existence
the king between f6 and g6, but then of a strong point at c4 for the knight
the white king makes a decisive attack allow Black to put up a stubborn de­
on the d6 pawn. The move in the game fence. White will naturally aim to open
merely hastens the end. up the position and create scope for his
bishops.
11 ... Bb4
12 Bf2 Nd8 1 f3 Nd6
13 Bc8! Bc3 2 Bc3
14 Bxh4
2 e4 would be a serious positional
The rest is less difficult. By the march mistake due to 2 . . . f5!, when Black
of his king to the Q-side White concludes gains control o f the key squares in the
the game. centre.

14 ... Bf6 2 ... Bb7

15 Bf2 Nf7 3 g4!
.16 Kc2 Nh6
17 Bc6 Kh5 White carefully prepares an advance in
18 Bel Be7 the centre.
19 Bc3 Kg6
20 Kd3 Bh4 3 ... f5?!

133
 Endgame Strategy

One of those o ses where the threat is 19 ... Kxf6

stronger than its execution. Black pre­ 20 Bc8 Bd5+
vents e 3 -e 4 by placing his K-side 21 e4 Bf7
pawns on squares of the colour o f his 22 e5+ Resigns.
bishop, but in doing so he weakens
catastrophically the dark squares in the
centre. Preferable was 3 . . . f6, and if Flohr—Botvinnik
4 e4, then 4 . . . Nc4 followed by
. . . e5. Moscow. 1933

4 h3 Kf7
5 Kf2 Bd5
6 Bd3 Bb7
7 Bel! Bd5
8 Kg2 Bc4?!

After this move the opening up o f the

position can be prevented only at the
cost of loss of material. By 8 . . . g6
Black would have weakened his position
still further, but he would have main­
tained material equality.

9 Bc2 Bd5 The pawn formation is symmetric and

10 gxfSl exf5 fairly blocked. A stubborn battle is in
11 Bg3 Nc4 prospect, and it is difficult to predict
12 Kf2 Nd2 whether or not White will succeed in
advantageously opening up the game.
12 . . . g6 would have been answered
by 13 e4, when the two connected 1 Kf2 Ke7
passed pawns in the centre quickly de­ 2 Be3 Kd8
cide matters. 3 Kel Kc7
4 Kd2 Nc5
13 Bxf5 Bxf3 5 b4
14 Bc8
The first clash takes place on the Q-
The Rumanian grandmaster wins a side. White evicts the knight from c5,
pawn, and with it the game. but in doing so weakens his control of a4.

14 ... BdS 5 ... Ncd7?!

IS Ke2 Ne4
16 Be5 g5 Seriously to be considered was 5 . . .
17 Bxa6 Bc4+ Na4 followed by . . . bS, creating a
18 Kf3 Nf6 strong point for the knights.
19 Bxf6!
6 g3?!
The exchange of the black-squared
bishop for the knight is the quickest way This move could have been delayed.
to win. 6 a4 looks stronger, preventing Black

134
 The Two Bishops

from establishing a knight at a4. Besides, . . . g5. But without 19 . . . h6 this

the advance of rooks’ pawns is always cannot be played, since on 19 . . . Ng7
unpleasant for the side battling against 20 f5 g5 21 hxg5 fxg5 there would
two bishops. follow 22 f6+, when White wins. There­
fore Black adds an extra defence to g5,
6 ... Nb6 so that on 21 hxg5 he can recapture
7 Kc2 Nbd7?! with the h-pawn” (Botvinnik).

Again the knight should have been 20 Bel Ng7

played to a4. 21 fxe5 dxe5

8 a4! Nb6 Bad is 21 . . . Nxe5 22 Bc8, or 21

9 a5 .. . fxe5 22 Kf3 h5 23 Bg5+.

On the Q-side the black knights are 22 Kf3 h5

now rather cramped. 23 Be3 Kd6
24 Bh6 Ne8
9 Nbd7 25 g4 hxg4+
10 Bel Kd8 26 Bxg4 Nc7
11 Bb2 Ne8 27 Be3 Nb5
12 Kd2 Nc7 28 Ke2 Nc7
13 Ke3 Ke7 29 Kd3!
14 Bfl Nb5
15 h4! Nc7 Zugzwang. 29 . . . Nb5 is not possi-
16 Bh3 Ne8 ble, since after 30 Be6 Ke7 31 Bc5+
Nxc5 32 bxc5 Black’s 32 . .. Nd4 is
White has regrouped his forces for an without check.
attack on the K-side, and Black has pre­
pared a defence. There is no point in 29 . . . f5
waiting any longer. 30 exf5 gxf5
31 Bxf5
17 f4
The passed h-pawn and the chronic
Already threatening 18 Bxd7 and weakness of Black’s Q-side promise
19 fxe5. White a quick win.

17 . . . f6 31 ... Nxd5
18 Bf5 32 Bd2 N7f6
33 Kc4 Kc6
It is essential to provoke weakenings. 34 Bg6 b5+
35 Kd3 Ne7
18 . . . g6 36 Be4+ Ned5
19 Bh3 h6!
No better is 36 .. . Nxe4 37 Kxe4
“The point of Black’s entire defensive Kd6 38 h5 Ke6 39 h6 Kf6 40 h7
set-up. White’s main threat is f4—f5 Kg7 41 Kxe5.
followed by fxg6 and g3—g4, creating
a passed pawn on the h-file. Therefore 37 Bg5 Nh5
on f4—f5 Black must be ready to reply 38 Bf3 Ng3

ES-J
135
 Endgame Strategy

39 Bd2! allows Black to block the position.

Suppressing the opponent’s counter­ 1 ... c4!

play. After 39 h5? Nxh5! a n d __ Nxb4+ 2 Bc2 f5!
Black has drawing chances.
In just two moves the character o f the
39 . . . Kd6 position has changed. The white bishops
4 0 Bg4 are deprived o f their mobility, and the
knights have acquired some convenient
Aiming for c8. posts, in particular at d5.

40 . . . Nf6 3 Bel Rf7

41 Bc8 Kc6 4 Ba3 Rc6
42 Bel! 5 Bc5 Ra6!

Do not hurry! White provokes the advance o f Black’s

Q-side pawns, hoping to open up the
42 ... e4+ position on that part of the board. In­
43 Kd4 Ngh5 deed, after 5 . . . a5 6 Ba4! b5 7 Bc2
44 Bf5 Kd6 followed by a2—a4 White activates his
45 Bd2 Resigns. pieces. Instead of this, Chigorin himself
provokes

6 a4
Lasker-Chigorin
which restricts still further the oppo­
H utingi, 1895 nent’s white-squared bishop. 6 Ral or
6 a3 is unfavourable, since then 6 .. .b6
7 Bb4 Nc6 leads to the exchange of the
black-squared bishop.

6 ... Nc6
7 R bl Rd7
8 Rggl Nge7
9 Rb2 Nd5
10 Kd2 Ra5
11 Rgbl

White does not fear either 11

Rxc5, since the knight at d5 has no
favourable retreat, or 11 . . . Nxf4, on
White's strong pawn centre and two which there follows 12 Rxb7 (12 Bb4
bishops give him the advantage. After is also quite good).
1 f5 the position is opened to his ad­
vantage, and the bishops obtain scope. 11 . . . t>6
But the move made by Lasker on general 12 Ba3 g6
grounds 13 Rb5 Ra6!

1 Ragl? The correct solution to an exchanging

136
 The Two Bishops

problem. After 13 . . . KxbS 14 axbS 21 exf5 was also possible, but after
the a-file is opened to Black’s disadvant­ 21 . . . e5! 22 dxe5 Nxe5 23 Bxf4
age. Chigorin realizes that the imprison­ Nd3 Black would acquire counter-play.
ment of his rook at a6 is only temporary,
since the white rook at b5 is badly 21 . . . Nh5
placed and will be forced to leave its 22 Be3?
post.
Now Black again blocks the position.
14 Bel Nd8 After 22 Rgl+ Kf8 23 Bel foUowed
15 R al Nf7 by Ba3 White would have developed a
16 R bbl Nd6?! very strong attack.

Black provokes f2 -f3 , missing the 22 . . . f4!

favourable opportunity of 16 . . . g5! 23 Bf2 Ra5?
17 fxg5 Nxg5 18 Bb2 Ne4+ 19 Bxe4
fxe4, leaving White with a bad bishop Black should have immediately block­
against a strong knight. ed the centre by 23 . . ,e5 !, since 24
Rgl-f Kf8 25 dxe5 Nxe5 26 Rg5 is
17 f3 Nf7?l refuted by 26 . . . Rd2+!

Black has obviously noticed his over­ 24 Rgl+ Kf8

sight, and attempts to repair the mistake, 25 Raal?
but White has time to regroup his forces
and open up the position. Black should White returns the compliment. After
have played 17 . . . Rf7, preventing a 25 e5! Black has a difficult game. In his
possible e3—e4. notes to this game Chigorin writes: “If
25 e5 b5 26 Bxh7, then 26 . . . Nxe5!
18 Ra3! 27 Rg8+ Kf7 28 dxe5 b4! 29 cxb4
Rxc5+ 30 K fl Nf6, and by returning
Lasker does not prevent . . . g5, the piece Black obtains a probably won
rightly assuming that in an open game position". These variations are interest­
only he will have winning chances. ing, but what is Black to do if instead of
26 Bxh7 White plays Raal or Rgal?
18 . . . g5?
25 . . . e5!
It was not yet too late to return the
knight to d6. Erecting a new defensive barrier.

19 Ke2! 26 Rabl Ng7

27 Rb4 Rc7
Much stronger than 19 fxg5 Nxg5 28 B bl Ne6
20 Ke2 Rg7, and if 21 e4 fxe4 22 29 R dl?!
fxc4 Nf6, when Black has everything in
order. Played according to the principle “do
not hurry". But, as we know, such
tactics should not be abused. After 29
19 . . . gxf4 Ba2! b5 (29 . . . Nd6 30 dxe$ RxeS
20 e4! Nf6 31 Bd4) 30 Rxb5 the position is open­
21 Bxf4! ed up on the Q-side, and White retains

137
 Endgame Strategy

the advantage. Lasker, evidently assum­ 31 . . . Rxa4

ing that he will have time for Ba2, gives 32 dxe5 Nfxe5
extra support to his centre, but under­
estimates his opponent’s latent countcr- Black is a pawn up with a won posi­
play. tion.
33 Bh4 Rg7
29 . . . Ncd8! 34 Kf2 Rg6
35 Rdd5 Ral
36 Bd8 Nd3+
37 Bxd3
Otherwise . . . Ncb4 is decisive.

37 . . . cxd3
38 Rxd3 Ragl
39 Rf5+ Kg8
40 Bg5 R6xg5
White resigns.

This game is a good illustration of the

strengths and weaknesses of two bishops
in a battle against knights. The frequent
30 Rd2? mistakes can be explained by the tension
o f a crucial encounter between two out­
White fails to sense the danger. He had standing players upholding their creative
the possibility of opening the centre by conceptions.
30 dxe5!, when, as shown by grand­ In recent times opening theory has
master Vasyukov in the book M ik h a il made great advances. To make an assess­
C bigorin, after 30 . . . Nc6 31 e6! ment of any opening without linking it
(but not 31 R b 5 R x b S 3 2 a x b S N c x e S to the middlcgame is unthinkable, but
33 R d S K e 7 !) 31 . . . Nfe5 32 Bc2 more and more often the assessment of a
Nd3 33 Rb5 the chances are with particular variation depends on the pro­
White. By the move played Lasker pre­ spects of the two sides in the endgame.
pares Ba2, erroneously assuming that At the end of the seventies the Aver­
bakh Variation of the King’s Indian
30 . . . Nc6! Defence began a period o f rapid develop­
ment. A considerable influence was
is impossible. This knight move would made by some games in which an end­
also have followed on 30 Ba2, but not game was tested where White had the
30 . . . b5?, which is what Lasker was advantage of the two bishops. Now die
afraid of when defending a2 with his opinion of the specialists is unanimous
rook. —White has a big advantage. Few players
with Black are willing to defend this end­
31 RbS ing, and King’s Indian players have had
to find new paths, which has assisted the
A sad necessity. It transpires that 31 progress of theory. The initial game, in
Rxc4 is met by 31 . . . Nd6 (not which Black encountered difficulties,
31 . . . N x d 4 + 3 2 R d x d 4 ! ) , when was Polugayevsky—Uhlmann, Amster­
White loses the exchange. dam, 1970.

138
 The Two Bishops

1 c4 Nf6
2 Nc3 g6
3 e4 d6
4 d4 Bg7
5 Be2 0 -0
6 Bg5 c5
7 d5 e6

This line has now been discarded,

Black preferring to try his luck with the
sharp pawn sacrifice 7 . . . h6 8 Bf4
e6 9 dxe6 Bxe6 10 Bxdó Re8.

8 Qd2! exd5 known. The way for the bishops must be

9 exd5 Re8 paved by the pawns.
10 Nf3 Bg4
24 a5!
Games in which Black did not develop
his bishop here also confirmed White’s As in previous positions, the envelop­
advantage. ment of the opponent’s position begins
from the wings. Were Black himself to
11 0 -0 Nbd7 succeed in playing . . . a5, White’s win­
12 h3 Bxf3 ning chances would be sharply reduced.
13 Bxf3 a6
14 a4 Qe7 24 . . . h5
25 Bd2 Ne8
Knowing beforehand the further 26 g3
course of the game, one could suggest
here the ugly 14 . . . a5, which despite White unhurriedly strengthens his
all its drawbacks blocks the Q-side, and position, preparing an offensive on both
makes the coming ending more accept­ wings.
able for Black.
26 ♦ • • Bd4
15 R ael Qf8 27 Kg2 Ng7
16 Bdl Rxel 28 f4 Nf5
17 R xel Re8 29 N dl Nh6
18 Rxe8 Qxe8 30 Kf3 f5
19 Bc2 Nb6
20 b3 Nbd7 Black’s desire to block the K-side is
21 Bf4 Qc7 perfectly understandable.
22 Qe2! Kf8
23 Qxe7+ Kxe7 31 Bd3 Kd8
32 Ne3 Ke7
(See n ext diagram) 33 Nc2 Bb2
34 Ke3 Nf6
We have an ending with a symmetric 35 N el
pawn formation and the advantage of
two bishops for White. The plan for Blade is deprived o f the slightest
exploiting this advantage is now well counter-play, and Polugayevsky skilfully

139
 i
Endgame Strategy

combines the strengthening of his posi*

tion with action according to the princi-
pie “ do not hurry”.

35 ... Bd4+
36 Kf3 Bb2
37 Ng2! Nd7
38 Nh4 Kf6
39 Ke3 Nf7
40 Bc2 Bal
41 Ke2 Bb2
42 Bel Bal

42 . . . Nh6 was preferable, tempo­ 4 b3 Ng7

rarily preventing the inevitable opening 5 Bc2 Ne8
of the position. 6 Ne2 Bb2
7 f3 Ng7
43 g4! 8 Kf2 Bf6
9 Nc3 Bd4»
The bishops break free, smashing all 10 Ke2 f5
obstacles in their path. 11 f4!

43 ... hxg4 Events develop along familiar lines.

44 hxg4 fxg4
45 Nxg6 Kg7 11 . . . Ne8
46 Nh4! Kf8 12 Bd3 Bxc3
47 Bf5 Nf6
48 Bc8 Nd8 In many instances tw o knights battle
49 Nf5 Nh5 against two bishops better than do a
50 Bd2 Bd4 bishop and knight, by occupying strong
51 Nxd4 Resigns. points created in the opening up of the
position. Therefore it is difficult to con*
In subsequent games Uhlmann sue- demn Gligoric for this exchange.
cesfully employed this variation, but
this time from the white side. 13 Bxc3 Nef6
14 Bel Kf7
15 Ke3 Ke7
Uhlmann—Gligoric 16 Bc2 Kf7
17 b4!
Hastings 1970/71
The start!
(See next diagram)
17 . . . cxb4
The position resembles the previous 18 Bxb4 Nc5
one, like two peas in a pod. 19 Kd4 Nfd7

1 a5! Nc8 The black knights have become

2 Bd2 h5 established at c5, so White sets about
3 K fl Bd4 opening up the K-side.

140
 The Two Bishops

20 Bdl! Ke7 30 b7 Kc7

21 g4! 31 Bc8 Nc5
32 f5 Ne4+
The continuation! 33 Kc2 Kb8
34 Kb3 Nd2+
21 . . . hxg4 35 Ka4 Nxc4
22 hxg4 Kf6 36 f6 Ne5
23 Ke3 b6 37 Kxa5 Resigns.

Waiting tactics would not have saved

Black. 23 . . . Kf7 loses to 24 gxf5 Uhlmann—Andersson
gxf5 25 Bh5+ Ke7 26 Bg6 Kf6 27
Bh7l Ne4 28 Ba3, with the threat of Skopje Olympiad, 1972
Bb2+ (Marie).

24 gxf5 gxf5

The black knights are badly placed, a

factor which Uhlmann exploits to begin
immediate play on the Q-side.
25 Bxc5!
1 b4! cxb4
The end! One advantage of the two 2 Na4 Ngf6
bishops is that it is not difficult to ex­ 3 Nb6!
change one for a knight in a favourable
situation. Preventing a black knight from reach*
ing the strong point c5 via d7.
25 . . . Nxc5
3 — Bd8?
The other captures were no better.
25 . . . bxc5 26 Ba4 Nb8 27 Kf3 A serious mistake. It was essential to
Kg6 28 Kg3 Kh5 29 Bc8+ Kh6 30 ensure the knight access to c5 by 3 . . .
Kh4, o r 25 . . . dxc5 26 Ba4 Ke7 b3! 4 Bxb3 Ne4.
Bxd7 Kxd7 28 axb6,
4 Bxb4 Bxb6
26 axb6 a5 5 axb6 Kf8
27 Bc2 Ke7
28 Kd2! Kd8 5 . . . Nd7 6 Ba4 Nef6 7 Bxd6
29 Bxf5 Na4 Nxb6 is more tenacious, although after

141
 Endgame Strategy I

8 Bb3 White is close to a win. c6! bxc6

b7! cxd5
6 c5!!
No better is 8 • • . c5 9 Ba4 Nb8
The white pawns finally “break the 10 Ba5 Ke7 11 Bxe8 Kxe8 12 Bc7
ground” for the bishops on the Q-side. (Uhlmann).

6 ... Nd7 9 Ba4 Nb8

As shown by Uhlmann, annotating Or 9 . . . Nef6 10 Bc3! Ke7 11
this game in Volume 14 of Chess In- Bxf6+.
formator, on 6 . . . Nxd5 White would
have won by 7 c6! Nxb6 8 cxb7 Nd7
9 Ba4, or 7 . . . Nxb4 8 c7!, when 10 Bxe8 Kxe8
the pawn queens. 11 Bxd6 Resigns.

142
 CHAPTER 12

THE 3 -2 QUEEN-SIDE PAWN MAJORITY

In many openings — the Caro—

Kann Defence, Nimzo—Indian Defence,
Sicilian Defence, French Defence,
Queen’s Pawn Opening and others — a
pawn formation of the type shown in
the following diagram can arise.

Here is what Alekhine had to say re­

garding this position:
’’The ending in this game is note­
worthy in the sense that White’s cele­
brated Q-side pawn majority proves to
be completely illusory. Regarding this I
must remark that one of the most
The two sides’ plans normally follow characteristic prejudices of modern
from the pawn formation: the four theory is the widely-held opinion that
pawns advance against the three on the such a pawn majority is important in it­
K-side, and the three against the two on self — without any evaluation o f the
the Q-side, although cases of a minority pawns which comprise this majority, or
attack are also possible. Formerly it was of the placing of the pieces. In the given
considered more favourable to have the position Black has the following com-
extra pawn on the Q-side, since it is pensation.-
easier to set up a passed pawn there. (1) Great freedom for his king in
Modern-day practice has not confirmed comparison with the white king.
this unshakeable principle of the Steinitz (2) Dominating position o f the rook
theory. on the only open file.
Everything depends on the specific Used correctly, these two advantages
features of the position. should provide the basis for a win.”
In the majority of cases control of
the only open d-file confers an advant­ 1 g3 Kf7
age, irrespective of the number of pawns 2 c5 Kf6
on the wings.
Were White able to exchange the bis­
hops, he would gain a draw. After Black
has played . . . f4, following a prepara­
Yates—Alekhine tory . . . g5, his bishop will play an im­
portant role in the attack on the white
The Hague, 1921 king.

143
 Endgame Strategy

3 Bc4 Bc8! Marshall—Capablanca

4 a4?!
USA, 1909
White should have hastened with his
king to e l, although even in this case
he has a difficult game.

4 ... g5
5 b5?! f4
6 K fl

Too late. By invading the second rank

with his rook, Black creates an attack on
the white king.

6 ... Rd2!
7 gxf4 gxf4
8 K el Rb2
9 Be2 The diagram position is not an end­
game one, but play soon goes into an
Defending against 9 . . . Bg4. In the ending where Black seizes control o f the
event of 9 R dl Bg4 10 Rd6+ Ke7 d-file, which in fact decides the out­
11 Rd4 Bf3 12 Bd5 R bl+ 13 Kd2 come.
e3+ 14 fxe3 R dl+ 15 Kc3 Rxd4 16
Kxd4 Bxd5 17 Kxd5 fxe3 the black 1 Rfcl?
pawn queens.
White should have begun a battle for
the d-file and advanced his K-side pawns.
9 ... Ke5! Therefore 1 R fdl or 1 e4 was quite
10 c6 bxc6 appropriate, with a roughly equal game.
11 Rxc6
1 ... Rab8
2 Qe4?
After 11 bxc6 f3 12 Bdl or 12
Bfl Black wins by 12 . . . e3. With the threat of exchanging bishops
after Bh3. Again White deviates from
the correct path.
11 . . . Be6
12 Bdl R bl 2 ... Qc7!
13 Rc5+ Kd4
14 Rc2 e3 A simple move, yet at the same time
15 fxe3+ fxe3 difficult to find. It is not so easy volun­
16 Rc6 Bg4 tarily to place the queen opposite an
17 Rd6+ Ke5 enemy rook. But this move is part of a
18 h3 Bh5! plan to provoke White into opening up
White resigns. the game on the Q-side, which can only
favour Black.

There is no defence against . . . e2. 3 Rc3?

144
 The 3—2 Queen-side Pawn Majority

Of course, it is difficult to admit one’s Black wins a piece, and the realization
mistake. A t the given moment there of his advantage does not present any
would not have been anything particular­ great difficulty.
ly terrible in store for White after the
correct 3 R dl! But after the move 20 Rxb2 R xdl
played his position begins to deteriorate, 21 Rc2 Bf5
the reason being that Black seizes con- 22 Rb2 R cl
trol of the d-file. 23 Rb3 Be4+
24 Kh3 Rc2
3 ... b5 25 f4 h5
4 a3 c4 26 g+ hxg4+
5 Bf3 Rfd8 27 Kxg4 Rxh2
6 R dl Rxdl+ 28 Rb4 f5+
7 Bxdl Rd8 29 Kg3 Re2
30 Rc4 Rxe3+
“Black now dominates the entire 31 Kh4 Kg7
board” (Lasker). 32 Rc7+ Kf6
33 Rd7 Bg2
8 Bf3 g6 34 Rd6+ Kg7
9 Qc6 Qe5 White resigns.

Capablanca does not object to an end­

game, but only in a different version. Didishko-Maryasin

10 Qe4 Qxe4 Minsk, 1980

Now the b5 pawn is not hanging.

11 Bxe4 Rdl-*-!

The white king must be decentralized.

12 Kg2 a5
13 Rc2 b4
14 axb4 axb4
15 Bf3 R bl

White’s game is hopeless.

16 Be2 b3 White’s pieces are exerting unpleasant

17 Rd2 pressure on the opponent’s position.
Black chooses a forcing variation, involv­
ing a transition into an endgame.
Not 17 Rc3 due to 17 . . . Rxb2
18 Bxc4.Rc2. 1 ... g5
2 Bg3 Nxf3+
17 . . . R cl! 3 gxf 3 Bxg3
18 B dl c3 4 fxg3 Qxd4+
19 bxc3 b2 5 Rxd4 Rfd8

145
 Endgame Strategy

6 R fdl Rxd4 26 Kd4 Nb2

7 Rxd4 Re8
and Black resigned, without waiting for
In spite of the numerous exchanges, the obvious 27 Rxf7+.
Black’s position remains difficult.

8 Kf2 Re6 Smyslov—Szabo

Played in the hope of obtaining Hastings, 1954

counter-play on the Q-side. Passive de­
fence by 8 . . . Kf8 9 Nb5 c6 10
Nd6 Rd8 was also uninviting for Black.

9 f4 Rb6
10 Rd2 gxf4
11 gxf4 Rb4
12 Ke3 Kf8
13 b3!

White is not in a hurry to take positive

action, and first strengthens his position.

13 . . . c6
14 Rd8+ White’s Q-side pawns have gained
space, and all his pieces are ready to sup­
This invasion by the white rook leads port their advance. In addition the d-fUe
to gain of material, whereas 14 e5 would is under White’s control.
have afforded Black additional counter­
chances after 14 . . . Nh5. 1 a3 g5?

14 ... Ke7 Black fails to battle for the only open

15 Rh8 a5 file, and ends up in a very difficult posi­
16 Rxh6 a4 tion. “Black does not yet sense the
17 bxa4 Nd7 danger, and chooses an ineffective
18 Rb5! method of defence. He should have tried
21 . . . Rad8, so as after 22 Bxc6 bxc6
Preventing 18 . . . Nc5. 23 Rxd8 Rxd8 24 Rxe7 g5 to gain
counter-play at the cost of a pawn”
18 . . . Nb6 (Smyslov).
19 a5 Na4
20 Ne2 Rb2 2 Nd5 Red8
21 Nd4 Rxa2 3 Redi Bg4
22 Nf5+ Kd7 4 f3 Be6
23 Rh7 Ke6 5 b4
24 Rh6+ Kd7 White advances his pawns further,
avoiding the positional trap 5 Nc7? Bb3!
If 24 . . . f6, then 25 e5!
5 ... h6
25 R f6 R a3+ 6 Kf2 Rd7

146
 The 3—2 Queen-side Pawn Majority

Black resourcefully exploits White’s Bronstein—Rantanen

last move (now 7 N f4 is not possible,
since the rook takes on <12 with check) Tallinn. 197$
and successfully contests the d-file, but
it is too late.

7 Nc3 Rxd2+
8 Rxd2 RdS
9 Rxd8 Nxd8
10 f4 gxf4
II gxf4 Bb3

In order to stop the white pawns,

Black tries to bring his king across and
vacates e6 for it, but he runs into an­
other misfortune.

12 Ke3 Kf6 White has a spatial advantage, better

13 b5 e5 placed pieces, a centralized king, and
control of the only open d-file.
The intended 13 . . . Ke6 would
not have worked, since, as shown by 1 g4!
Smyslov, White then has a pretty win:
14 c6 Kd6 15 b6!, while if 14 . . . The doubling o f rooks on the d-file
bxc6 15 bxc6 Kd6 16 Nb5+. suggests itself, but Bronstein, after a
subtle evaluation of the position, begins
14 Ne4+ Kc6 advancing his K-side pawns. White’s idea
15 c6 is not to carry out a minority attack, but
to push back the opponent’s firmly
Exploiting the tactical features of the established pieces on the K-side and gain
position. control of die invasion squares on the d-
file, thereby creating the precondidons
15 ... exf4+ for a pawn advance on the Q-side, where
16 Kxf4 bxc6 White has a quantitative majority.
17 Nc5+ Kd6
18 Nxb3 1 ... a6

White has won a piece, and the rest is Black vacates a7 for his rook, and
not difficult. hopes for possible counter-play b y __ b5.
18 . . . cxb5 2 g5 Ne8
19 h4 Nc6
20 Bxc6 White has chosen a very favourable
moment for the advance of his g-pawn.
The quickest way to win. Black’s K-side is now paralyzed, and for
the moment his extra pawn on this part
20 ... Kxc6 of the board is of no significance at all.
21 Kg4! b4
22 axb4 Kb5 3 a4 Ra7
23 Nd4+ Resigns. 4 h4

147
 Endgame Strategy

In order to free his bishop from 14 Rb8! Bc5

having to defend the g5 pawn. 15 Ng3

4 ... Rb7?! After 16 Ne4 loss o f material is in­

evitable, and so Black resigned.
Black persists in his desire to obtain
counter-play on the Q-side, where he is
clearly weaker. He should have gone
y fcWV^. - ' 0
Larsen—Spassky
totally onto the defensive. To be con­
1
sidered was 4 . . . Kf8, transferring the Lugano Olympiad, 15*68
king to e7 after . . . Bc5, and thus
covering White’s invasion squares on the
d-file.

5 Rd3

White sets about doubling rooks on

the d-file.

5 . .. Rc5

With the threat of 6 . . . e5.

6 R el Rd77!
Black’s position is preferable: his Q-
This allows White, by using tacticalside pawns have advanced, whereas
White’s on the K-side are not yet ready
motifs, to seize control of the d-file.
to move. The centralized black king has
7 R edil g6 greater scope than White’s. But White
controls the important d-file. With his
If 7 . . . e5 8 Nf5 Rxd3 9 Rxd3 next move Spassky begins a battle for
Kf8, then 10 Be 3 with an overwhelm­ the only open file.
ing positional advantage.
1 ... Rcd8
8 Ne2! 2 Rh5?l

The d-file is completely in White’s Larsen plans to attack with his pawn
hands. minority. After the correct 2 Rcdl the
position would have been approximately
8 ... Rxd3+ level.
9 Rxd3 b5
2 ... h6
This merely accelerates Black’s inevi­ : b4 c4
table defeat. 4 a4?

10 cxb5 axb5 (See next diagram)

11 Rd7 Kf8
12 a5 Rc6 It was this position that Larsen was
13 Rb7 b4 aiming for. He should have exchanged

148
 The 3—2 Queen-side Pawn Majority

10 . . . Re4!

Black again exploits the unfortunate

position of the rook at h6.

11 g3

Not 11 Rxb3 Kg7.

11 - - - Rxb4
12 Rh4

The rook has finally broken free, but

it is too late.
one pair o f rooks by 4 Re5+ and
blocked the d-file by Nd4. 12 . . . Rxh4
13 gxh4 Rd3
4 ... Kf6! 14 Kg2 Ke5
5 axbS g6! White resigns.

A brilliant manoeuvre. Black sacrifices

a K-side pawn, shutting the white rook Levenfish—Flohr
out of play for a long time. He will then
attack on the Q-side with superior Moscow, 1934
forces.

6 Rxh6 axb5
7 Nd4?

This loses. After 7 Rh4 Rd3 8 Rf4+

Kg7 9 Nd4 Nd2+ 10 Ke2 Nb3 11
Rdl Nxd4+ 12 Rxd4 Rb3 Black would
have obtained two connected passed
pawns, but the outcome of the game
would still have been unclear.

7 ... Nd2+
8 Kgl
After White plays a4—a5 Black’s Q-
The king cannot stand on the e-file side pawns will be paralyzed. The open
due to 8 . . . Rxd4. d-file is in White’s possession, and his
minor pieces are better placed than his
8 ... Nb3 opponent’s.
9 Nxb3 cxb3
10 R bl 1 Re3+ Kf7
2 Red3 Ke7
10 Rh4 is too late due to 10 . . . b2 3 f4!
11 R bl Rd2 12 Rd4 Rxd4 13 exd4
Re2. It is time to set about realizing the

149
 Endgame Strategy

K-ãde pawn majority. At the same time forcing him to weaken d6 for the de­
a way for the king to the centre is cisive invasion by the white knight.
opened. . . . g5 fails to 17 hxg$ hxg5 18 Ne4,
and after 18 . . . g4 the white king goes
3 ... Nc4 across to f4 and wins the pawn.
4 Kf2 Rxd3
5 Rxd3 Rd8 16 . . . Nc7
6 Rxd8 Kxd8 17 Nc4 Be6

With the exchange of rooks the posi­ Otherwise the white king breaks
tion has simplified, and White’s advant­ through at b6.
age has become obvious.
18 Nd6 Bxc4
7 Be4! 19 Kxc4 Ke6
20 Nxb7
The knight has to be driven from its
strong position at c4. To this aim the Black resigned in view of the obvious
bishop is transferred to d 3, at the same 20 . . . Kxc5 21 Kc5 Nd5 22 Nd8
time provoking a fresh weakening of the Ke4 23 Nxc6 Nc3 24 Nb8 Kf3 25
opponent's K-side. Nxa6 Kxg3 26 Nc7.
In all the endings examined in this
7 ... h6 chapter, control of the d-file was a de­
8 Bd3 Nb6 cisive factor. It is true that one comes
across endings (usually with just one pair
After 8 . . . Nb2 9 a5 the exchange of rooks) where this does not give any
on d3 is little consolation. real advantage, but these are merely ex­
ceptions which confirm the rule.
9 e4 In conclusion we will examine an
example in which the side with the
Black has no way of opposing the ad­ initiative achieves success by play on the
vance of the white pawns in the centre. wing where the opponent has a pawn
majority. Here, with rooks on the board,
9 ... Na8 he has a bishop against a knight, which is
10 Ke3 Nc7 in itself an advantage when there is play
11 a5 on both wings.

11 . . . b6 was threatened.
Kasparov—Vukic
11 ... Ke7
12 Bc4 Kd6 European Team Championship
13 Kd4 Ne8 Skara, 1980
14 e5+ fxe5
15 fxe5+ Ke7 (See next diagram)
16 h4!
An experienced player will immediate­
Good technique. According to the ly be aware that the position has arisen
rules of the endgame the king and knight from a Caro—Kann Defence. Black con­
should have changed places. But Leven- trols the d-file, but this is of no signific­
fish puts his opponent in zugzwang. ance, since there are not, and cannot be,

iso
 The 3 - 2 Queen-side Pawn Majority

with f3—f4 and g4—g5. At the same

time he suppresses Black’s only chance
of counter-play — the attempt to create
a strong point for his knight at d5.
Black was planning to carry out the
counter-blow . . . b5 at a convenient
moment, and to set up the defensive
formation . . . Ne8, . . . f6 and . . .Nc7.
Kasparov radically forestalls Black’s
plan of removing the c4 pawn’s conirol
of d5.

3 ... Ne8
any invasion squares on it. On the other 4 Rel!
hand, Black’s pawn structure on the K-
side, where he has a numerical superior­ White again denies his opponent
ity, is completely fixed by the three counter-play. It is not at all in his
white pawns, and the difference in interests to allow . . . f6.
strength between the white bishop and
black knight is so great that the position 4 ... Rd6
can be assessed as won for White. 5 f4! Nf6?

1 ... Kc6 The decisive mistake in a difficult

2 Kc2 Rd7 position. 5 . . . f6 was bad due to
3 aS! 6 Bb4, while on 5 . . . Kd7 there
would have followed 6 g5. Now White
In the plan of creating a passed pawn takes play into a pawn ending.
on the Q-side, this move seems illogical.
But Kasparov makes a concrete assess­ 6 Bxf6 gxfé
ment of the position, and finds a plan 7 R dl Resigns.
based on exploiting the difference in
strength of the bishop and knight. This After the exchange of rooks Black has
aim is best served by a minority attack no satisfactory defence against g4-g5.

151
t$-t
 CHAPTER 13

COMPLEX ENDINGS

In the previous chapters we have been to block the Q-side and to begin play on
analyzing examples where one specific the opposite wing. In the first instance
principle of endgame play was most Black needs his knight at g4, but »t can
clearly revealed. In the present chapter be transferred via h8—f7 —h6 only when
the reader will meet endings where the White does not have the possibility of
players make use of a variety of methods playing e4—e5.
and principles. Along with examples of
impeccable endgame technique, we will 1 ... a5
also be considering endings where there 2 a4
are highly instructive mistakes by both
sides. There was no necessity for this.

2 ... Kd7
Factor—Rubinstein 3 Kfl Bc6
4 b3 Kc8
Lodz, 1916
Black is in no hurry to take positive
action. By manoeuvring with his king he
dulls White’s vigilance, and begins gradu­
ally strengthening his position. He must
first safeguard the retreat o f his bishop
from d6 in the event o f e 4 -e 5 , and
this can only be achieved by . . . cd.
This means that the white-squared
bishop must make way for the pawn.
Blacks ideal set-up would be: knight
at g4 and bishop at g6 or h7, but for
the moment this is not a reality.

5 Re3 Bd7
The position has arisen from the ex­ 6 Kg2 c6
change variation of the Ruy Lopez. 7 N bl Bc7
White has an extra pawn on the K-side, 8 Nc3 Nh8!
while Black has the two bishops. In such
positions White usually plays on the K- The right moment! White cannot play
side and in the centre, while Black 9 e5, since after 9 . . . fxe5 10 fxe5
creates counter-chances on the Q-side. Nf7 he loses a pawn. Black has taken all
But in the given position the pawn at the precautionary measures well in ad­
e4 is not a.'strength, but a weakness, vance.
since it is securely fixed and under direct
attack by the black pieces. In addition, 9 Rd3 Nf7
White has a ‘hole’ at g4. On the other 10 Redi Bg4
hand, Black has no prospect of active 11 R ld2 Nh6
play on the Q-side, so Rubinstein decides 12 Bgl Bb8!

1S2
 Complex Endings

Suppressing the opponent’s counter­ 31 Nxf4 Rxf4+

play. Black prepares the transfer of his 32 Ke2 R xfl
bishop to g6 via e6 and f7, and fore­ 33 R xfl Rxfl
stalls the possibility of NdS, which could 34 Kxfl Nh6,
have followed on 12 . . . Be6.
and within a few moves White resigned.
13 Bf2 Be6
14 R dl Bf7
15 Bgl Bg6 Najdorf—Stahlberg
16 Nd2 Nf7
Candidates Tournament
Do not hurry! Black has planned to Zurich, 1953
play his knight to g4 and his bishop to
h7, followed by . . . g5, but he decides
to manoeuvre a little more, so as to
hinder his opponent’s orientation.

17 Rel Bh7
18 Kf3 Nh6
19 Bf2 Ng4
20 Sgl g5!

The start of the attack. The tempo of

the play changes sharply, as Black is
transformed and becomes very active.

21 Re2 gxf4 The position is roughly level. The ad­

22 gxf4 Rg8 vance of White’s K-sidc pawns is hinder­
23 R el Reg7 ed by his doubled f-pawns, while on the
24 Ne2 Q-side it is not easy fo r Black to advance
his pawns. White has a good bishop, but
Defending against 24 . . . Nh6 the black knights may prove stronger
followed by . . . Rg4. due to the presence o f the doubled
pawns. A prolonged struggle fo r the
24 . . . f5! accumulation o f small advantages is in
prospect.
A fresh blow, which decides the game.
White cannot play 25 e5 due to 25 . . . 1 NfS Rad8
Bxe5 26 fxe5 Nxe5+ 27 Ke3 Í4+. 2 R fdl Nc8>!

25 exf5 Bxf5 Black defends d6, voluntarily with­

26 Ne4 Bxe4+ drawing his knight to a passive position.
27 Kxc4 Re8+ A more natural plan would be 2 . . .
28 Kf3 Rf7 NbdS or 2 . . . NfdS followed by
. . . g6 and the advance o f the king to
The f4 pawn cannot be defended. c7. Had White wanted to drive the
knight from its centralized position by
29 R ddl Ref8 e3—e4, this would have left him with a
30 R fl Bxf4 weakness —at f4.

153
 Endgame Strategy

3 K fl Rfe8 A slip, which could have had serious

4 Ke2 Kf8 consequences. 16 fS was correct.
5 Rxd8 Rxd8
6 R gl Ne8? 16 . . . b6?

Playing with fire. All Black’s pieces A mistake in reply. The correct 16
are now on the back rank. Not wishing . . . f5! would have improved things con­
to weaken his position, Stahlberg cedes siderably for Black.
his opponent more and more space.
After the correct 6 __ g6 the position 17 f5! cS
would have remained level. 18 f4 Rc6

7 Rg4 A belated attem pt at activity on the

Q-side.
The rook advances for die decisive
offensive. 19 a4 b5
20 Bc2!
7 ... Ne7?
So as to answer 20 . . . c4 with 21
An incorrect approach to the ex­ Be4 Rb6 22 b3!
changing problem. It is easier to defend
with two knights than with one, which 20 ... Ne8
in addition is badly placed at e& Even 21 Be4 Rc7
here 7 . . . g6 was not yet too late. 22 Bd5 c4
23 e4 Nd6
8 Nxe7 Kxe7 24 axb5 axb5
9 Re4+ Kf8 25 Ke3 Ra7
10 Ra4 a6 26 R gl Kf8
11 Rf4 f6? 27 Kd4

This finally ruins Black's game. After White centralizes his king, not allow­
11 . . . Nf6 things would not yet have ing the opponent’s into the centre.
been hopeless, although after 12 Rf5
followed by the advance of the f- and 27 . . . Rc7
e-pawns White would have retained a 28 Rcl!
marked advantage. Now the white
squares on Black’s K-side are decisively Preventing the slightest attem pt at
weakened. counter-play. On the natural 28 Ral
there could have followed 28 . . . b4!
12 Rh4 h6 29 Ra8+ Ke7 30 Rg8 c3 31 Rxg7+
13 Rh5! {31 bxc3 Nb5+) 31 . . . Kd8 32 Rxc7
cxb2!
Paralyzing the opponent’s pawns
along the rank. 28 ••• Nb7
29 R al Nc5
13 ... Nc7 30 Ra8+ Ke7
14 f4 Ke7 31 e5!
15 Rc5 Rd6
16 Rcl?! White finds the quickest way to win,

154
 Complex Endings

having accurately worked out all the A pretty way o f realizing his advant­
variations. age, but evidently not the strongest.
Black’s defence would have been diffi­
31 . . . Nb3+ cult after the natural 1 Rg5!, e.g. 1 . . .
32 Kc3 Ncl Rae8 ( / . . . Rg7 2 Rx¿7 Kxg7 3
Bxh7) 2 Rh5 h6 3 Ne2 Kg7 4 Nf4
On 32 . . . Rc5 there could have Rf8 5 Rf3, with an overwhelming
followed 33 Ra7+ Kf8 34 Rf7+ Ke8 advantage.
35 Be6 fxe5 36 Rxg7 Nd4 37 Kb4
Nxe6 38 fxe6, with a won rook end­ 1 ... Rae8
ing. 2 Nd2 c5

33 Rg8 Ne2+ Practically forced. If Black allows the

34 Kd2 Nxf4 white knight to occupy e5, his position
35 Rxg7+ Kd8 becomes hopeless.
36 exf6! Rd7
37 Rxd7+ Kxd7 3 dxc5 Rxe5
38 Bc6+! Resigns. 4 Nf3 Rc3?

A splendid cieative achievement by This loses a pawn. After 4 . . . R5e7

the Argentinian grandmaster! 5 Ng5 Be6 White has the advantage, but
no forced win is apparent (6 R e l Bd7).

5 Ng5 Rxg3
6 hxg3 h6
Lasker—Pillsbury 7 Nf7+ Kg7
8 Nd6 Re7
Paris, 1900 9 Nxb7

White has won a pawn. The realization

of the advantage should not be too diffi­
cult due to Black’s numerous weaknesses
on the Q-side.

9 ... Nf6
10 Nc5 Bg4
11 Rf4!

White defends against 11 . . . Nh5,

and switches his rook for an attack on
the opponent’s Q-side.

Black has weak pawns at e6 and h7, 11 . . . Bc8

and his king is badly placed. Lasker 12 Ra4 Ng4
carries, out an interesting knight man­ 13 Ba6?!
oeuvre, to attack both Black’s weak­
nesses. The natural 13 Kd2! was stronger,
the move demanding a rather lengthy,
1 N bl although not difficult, calculation of

155
 Endgame Strategy

variations: 13 . . . Nc3 14 Rb4Nxg2 when Black would appear to have no­

15 Rb8 Bh3 {15 . . . Bg4 16 Rb7 Kf6 thing better than 1 7 __ Rc7. Then by
17 R xe7 K xe7 18 B fl Bf3 19 Be2) 18 a4 White could have forced Blade
16 Rb7 Kf6 17 Rxc7Kxc7 18 Ke2 onto the defensive after 18 . . . Nc4.
Kd6 19 Nb7+! {19 b4 d4!) 19 . . .
Kc7 20 Na5 c5 21 Kf2 Kb6 22 b4>. 1 7 __ Bxg2
and wins. The move in the game allows 18 Bd3+ Kg57!
Black to evict the white rook from the
Q-side and to obtain some counter-play. Actively played. The cool 1 8 . . . Kg7
was also perfectly possible, when it is
13 . . . Bf5 not altogether dear how White is to de­
14 Rf4 monstrate die correctness o f his plan.

Not 14 Bb7 Re2. 19 Rf8

14 . . . Ne3 Lasker begins an attack on the h6

15 c3 Kg6 pawn.
16 Rf2 Be4!
19 ... Kg4
20 Rg8+ Kf3
21 Rg6 Ng4
22 Bf5 h5
23 Rg5 Rel+

On 23 . . . Re2 White has the un­

pleasant 24 Bg6!

24 Kb2 R hl?

This defensive move is the decisive

mistake. 24 . . . Nf2! was correct,
when the threats of 25 . . . Ndl+ and
The black pieces have come markedly 25 . . . Ne4 give Black excellent counter­
to life. play, e.g. 25 Bg6 N dl+ 26 Ka3
(26 Kc2 Ne3+) 26 . . . Nxc3 27 Bxh5+
17 b3?! Ke3, and Black at least equalizes.

A difficult move to find, and one 25 Bg6! Kxg3

which was unanimously praised by all 26 Bxh5
the commentators. White plans to ex­
change the K-side pawns (two for one!), White has achieved his aim. With the
so as to then attack Black’s weakened exchange of bishop for knight Black is
Q-side with all his forces. 17 b3 is an deprived of any counter-play.
important step in White’s plan, preparing
the transfer of his king via b2 to the Q- 26 • • • Bh3
side, and to some extent follows the 27 Bxg4 Bxg4
principle of “do not hurry’’. But in the 28 Rg6 Rh2+
given position more energetic measures 29 Ka3 Rc2!
were required. 17 Bb7! was correct, 30 Nd3!

156
 Complex Endings

Lasker avoids a clever trap: 30 Rxc6? 61 Nb4 Bg2

Rxc3! 31 Ne4+dxe4 32 Rxc3+Kf2, 62 Na6 Bf3
with a draw. 63 Nc7+ Kb8
64 a6 Resigns.
30 . . . Kh4

There is nothing better. 31 Ne5 was

threatened.
Fischer—Reshevsky
31 Ne5 BfS
32 Rxc6 USA, 1963

White is a pawn up with a positional

advantage.

32 , . « Kg3
33 Rc5 Rd2
34 Nc6 Kf4
35 Nb4 d4
36 cxd4 Rxd4
37 Ra5 Rd7
38 Nc6

38 Nd5+ would have immediately

concluded the game.
White has a clearly superior bishop, a
38 ... Be4 spatial advantage, and a better pawn
39 Nxa7 Rd2 formation, while Black's d-pawn is very
40 Nb5 Rd5 weak. With his next move Fischer fixes
41 Kb4 Bd3 the h7 pawn, giving Black a further
42 Nc7 RxaS weakness.
43 KxaS Ke5
44 Kb4 Kd6 1 g5! Be7
45 Nb5+ Kc6 2 Ke2
46 a4 Kb6
47 Na3 Be2 White improves the positions of all his
48 Nc4+ Ka6 pieces.
49 Kc3 Bdl
50 Nb2 Bh5 2 ... Raf8
51 b4 Be8 3 Be3 Rc8
52 Kb3 Bc6 4 b4
53 Kc4 Bd7
54 Kc5 Bg4 White’s pressure grows with every
55 Nc4 Bdl move. He threatens both b4—b5, and
56 b5+ Ka7 also the simple strengthening o f his posi­
57 a5 Bf3 tion by 4 Kd3 followed by c3—c4.
58 Ne5 Bg2 Black therefore decides to upset his
59 Nc6+ Ka8 opponent's plans, even at a cost o f weak­
60 Kb6 Bh3 ening his own position.

157
 Endgame Strategy

4 ... b5 Forcing the win of a pawn by the

threat of 16 Rf3+. Black is not helped
At the sides of the board Black has by 15 ...K g 4 16 Rg3+Kh5 17 Rf 3
weak backward pawns, while ‘adorning’ Bg7 18 Rxh8 Bxh8 19 Rf8 Bg7 20
the centre is his d6 pawn. But three Rf7.
weaknesses are too many, and later
Fischer allows Black to rid himself of 15 . . . Bg7
one, while attacking in turn the weak­ 16 Rxh8 Bxh8
nesses which are far from each other — 17 Rxh7 Re8
the pawns at h7 and a6. 18 Rf7+

5 R ddl! Winning control of e4.

Instantly changing the target of attack. 18 . . . Kg4

19 f3+ Kg3
5 ,,, Ke6 20 Kd3
6 Ral Rc6
7 Rh3! Bf8 Consistent, but even stronger was
8 Rahl Rc7 20 Kfl with mating threats.
9 Rh4!
20 • ■ • e4+
Zugzwang! White provokes . . . d5. 21 fxe4 Rd8+
Rook moves along the seventh rank are 22 Bd4 Kg4
not possible due to R al, while after 23 R fl! Be5
9 . . . Rc4 19 f3 an analogous situ­ 24 Ke3 Bc7
ation arises a move later. 25 Rgl* Kh4
26 Kf3 Rd7
9 ... dS 27 e5 Rf7+
28 Ke4 Rf5
This leads to a decisive opening of 29 e6 Bd8
lines. 30 Bf6! Bxf6
31 gxf6 Rxf6
10 R al! 32 Kd5 Rf2
33 R el Resigns.
The final finesse. The rook must be
lured onto the sixth rank, diverting it
from the h7 pawn. Smyslov—Simagin

10 ... Rc6 19 th USSR Championship

11 exd5+ Kxd5 Moscow, 1951
12 Rdl+ Ke6
13 Rd8 (See next diagram)

If Black’s rook were at c7, he would The white pieces are more actively
have 13 . . . Bg7. placed, but the win is still far off. It is
interesting to follow how, by moves
13 . . . Kf5 which are simple but are often difficult
14 Ra8 Re6 to find, Smyslov consolidates and then
15 Rh3! increases his advantage.

158
 Complex Endings

the only open file, his king and knight

will be much more active, and, very im­
portant, Black will be deprived o f the
slightest chance o f counter-play.

5 ... Rf7

On 5 . . . Rfe8 White has the un­

pleasant 6 hS.

6 NxcS R xdl
7 R xdl Nxh4
8 Nd7!

1 Ng4! An excellent move. White does not

prevent the exchange o f knight for bis­
The knight takes control of the weak hop, since it is much more important to
e5 and f6 squares in Black’s position. forestall the possible counter . . . eS.
The threat is 2 Nf6+, winning the d5
pawn. 8 Nxf3

1 ... Nf5 Otherwise White will withdraw his

2 Kf2! bishop.

Very strong. White renews the threat 9 Kxf 3

of winning the central pawn. It tran­
spires that 2 . . . Nxh4 is bad due to The position has clarified, and White’s
3 Nf6+ followed by R hl. advantage has grown markedly.

2 ... dxc4 9 ... NdS!7

2 . . . d4 is no better due to 3 e4 Clever, but insufficient.

followed by h4—h5. White’s pressure
grows with every move. 10 Nc5!7

3 bxc4 a5 White ignores the opponent's offer to

go into a rook ending. After 10 cxd5
Black tries to create counter-play by Rxd7 11 d6 Kf7 followed by 12
advancing his pawns on the Q-side, . . . Rd8 and 13 . . . Rb8 Black would
where he has a numerical superiority. gain counter-play.

4 Nf6+ Kg7 10 . . . Nc7

5 Nd7! 11 Ke4!

Smyslov rejects the possibility of a K- Again preventing . . . e5.

side attack, prefering to exchange his h- 11 . . . h6
pawn for the black c-pawn. He reckons
that in this case he will gain a number of There is no other possibility of
other advantages: he will have control of counter-play.

159
 Endgame Strategy

12 Rd6 hxgS
13 fxg5 Kf8
14 Rc6!

Smyslov again avoids going into a

rook ending, but continues to play for a
squeeze on his opponent's position.

14 . . . Ke8
15 Ke5!

By his previous move White tied down

the enemy rook, and now comes the de*
cisive invasion of the white king.
the endgame a rook and bishop are
15 . . . Kd8 normally stronger than a rook and
16 Rb6! knight. By a veiled manoeuvre White
ensures the invasion of his rook at d7,
Each move by White deserves an ex­ after which his advantage assumes real
clamation mark, although they seem proportions.
simple. The immediate 16 Kd6 would
have been premature due to 16 . . . 1 R cl! Kd7
Ne8+ 17 Kxe6 Re7+, but now c6 is 2 Bc6+ Kc8
vacated for the white king.
After 2 . . . Kd6 or 2 . . . Ke7
16 . . . Kc8 White has the very unpleasant 3 Ba4.
17 Kd6 Rf2
18 Kc6! 3 Ral!

Do not hurry! 18 Rb7 would not Following the principle “do not
have worked due to 18 . . . Na6! hurry", White forces the black pawn to
move to a6, where it restricts the knight
18 ... Rc2 still further, and only then carries out
19 Rb7 Ne8 the invasion.
20 Ra7 Kb8
21 Re7 Resigns. 3 ... a6
4 R dl!

Smyslov—Matanovic The aim is achieved. Black is unable to

prevent 5 Rd7, since 4 . . . Rd8 loses
Monte Carlo, 1967 to 5 Bb7+.

(See next diagram) 4 »♦ ♦ Rf8

5 Rd7 Kb8
At first sight the position seems 6 e5 Ka7!
roughly level. Black is threatening
. . . Rb8, and 1 Bc6 is met by 1 . . . By this clever manoeuvre Black frees
Rd8 with a probable draw. But Smyslov his king. Now 7 Rxc7+ allows 7 . - -
again succeeds in demonstrating that in Kb6, equalizing.

160
 Complex Endings

7 Be4 Kb6 Geller—Shcherbakov

Now it is time for the white king to 22nd USSR Championship

come into play. Of course, not 8 Bxh7 Moscow, 195$
Nd5.

8 Kg2 h6
9 Kg3 (6

Otherwise the advance of the white

king to h5 will decide the game.

10 exf6 Rxf6
11 Kg4 Rf8
12 Rh7 Rd8
13 Rxh6 Kb5
14 Rh7 Nd5
15 Rb7+ Kc6
16 Rg7 As a rule, in the endgame a queen and
knight are stronger than a queen and
bishop. This is the case when the knight
Before exchanging his b-pawn for the occupies a strong point of the opposite
g-pawn, White, operating according to colour to the bishop and in die im­
the principle “do not hurry", improves mediate vicinity of the enemy king, and
the position of his rook. creates the threat of a mating attack.
The present ending is an exception. The
16 . . . Kb5 only square that can become a strong
17 Kxg5 Kxb4 point for the black knight is f5 after
18 h4 the preparatory . . . h5. But as soon as
the knight tries to go to f5 via e7. White
Both sides have obtained an outside begins an attack with his queen and
passed pawn, and in such cases a bishop bishop on the weak c-pawn. Therefore
is much stronger than a knight. Black is forced to stick to passive, wait­
ing tactics, whereas White has the
possibility of combining pressure on the
18 ... Nxe3 weak black a- and c-pawns with play for
19 h5 Rf8 an attack on the opponent’s king.
20 h6 Nf5
21 Rb7+ Kc5 1 ... Qc6
22 h7 Nd6 2 Qc2 Qb6
23 Rc7+ Kd4
24 Bg6! Black defends against the threat of
3 Qc4 followed by 4 Ba7.

Domination. 3 Be5 Qc6

4 Bb2 Qd6
24 . . . Nf5 5 Qc4 Kf8
25 Rf7 Resigns. 6 h4

161
 Endgame Strategy

A useful prophylactic move. White the K-side.

CTeates the potential threat of h4—hS,
and vacates h2 for his king. 17 ... Nd5
18 g4 hxg4
6 ... hS 19 fxg4 Nc3+
7 Qe4 Kf7 20 Kf4 Nb5
8 Ba3 Nb4
9 Qc4+ Kf8
10 Bel!

Geller transfers the bishop onto the

c l —h6 diagonal, where it has many
excellent posts.

10 . . . Qdl+?

Black fails to withstand the tension.

He should have returned his knight to
d5, controlling e3.

11 Kh2 Qd5

A sad necessity. If 11 . . . Qd6, The most energetic and convincing

then 12 Bf4 Qc6 13 Be3. way of realizing the advantage.

12 Qxd5! 21 ♦♦• Nd4

22 h6 Ne6+
The simplest way of realizing the ad­ 23 Ke4 Ng5+
vantage. In the resulting minor piece 24 Kd3 Ke5
ending the bishop is much stronger than 25 Bxc5
the knight. White's king is closer to the
centre than Black’s, and White’s K-side 25 Bel would also have won.
pawn majority allows him to create an
outside passed pawn. 25 . . . Nh7
26 Be3 g5
12 . . . Nxd5
13 Ba3! Black resigned, without waiting for
the obvious 27 Bd2.
The black knight must be diverted
away from the centre, to ensure the
approach of the white king. Spassky—Karpov

13 ,,, Nb4 Montreal, 1979

14 Kg2 Ke7
15 Kf3 a5 (See n e x t diagram)
16 Ke4 Kd6
17 f3 Black’s pieces are excellently placed,
and he has a more promising pawn form­
Geller sets about creating a passed ation than his opponent. He now has to

162
 Complex Endings

to the conclusion that it was necessary

to activate my rook straight away."
(Karpov).

4 ... Re3!
5 Ngl

As shown by the World Champion,

after 5 Rxd6 Rxd6 6 Rxd6 Bxf3 7
gxf3 Ne5 8 f4 Nf3+ 9 Kg2 Rxc3
10 Rd5 Nd4 11 Rxc5 Rb3 12 Rd5
Rb2+ Black wins easily.

solve an exchanging problem. 5 ... Kf7!

1 ... bxc5! This pawn sacrifice is part of Black’s

plan.
“The strongest of the three possible
continuations. After the plausible 1 . . . 6 Rxd6 Rxd6
dxc5 White develops pressure on the d- 7 Rxd6 Ke7
file. Then the exchange on c3 would 8 Rd3 R el
not be dangerous for him, and he could
even go after the pawn at bó. 1 . . . White’s game is lost.
NxcS looks sensible, since . . . Bxc3 is
then a strong threat, but after 2 Nd5 9 Ba2
Bxd5 3 cxd5 I thought it doubtful
that Black would be able to win." After 9 Bc2 Black wins by 9
(Karpov). Ne5 and 10 . . . Nxc4.

2 R dl

If 2 Nd5, then 2 . . . Bxd5 3 cxd5

Rb8 with a big and possibly decisive
positional advantage.

2 ... Bxc3!
3 bxc3 Rf6!
4 Rfd2

“How should Black proceed further?

After the apparently natural 4 . . . Bxf3
5 gxf3 Ree6 the white king moves
across to f2, and then White himself 9 ... Rcl!
can prepare an invasion on the b-file.
The immediate 4 . . . Ne5 is also possi­ “Perhaps one of the most exact moves
ble, but then 5 Ngl Rd7 6 Ne2, and in the game. So as to use the full power
all the time the capture on c4 is not of the black bishop, the rook heads for
possible because of the pin Ba2. After c2.” (Karpov). The World Champion’s
looking deeply into the position, I came entire play is imbued with a striving for

163
 Endgame Strategy

complete harmony, whereas the white bishop. Therefore White's main task is to

E scattered around the board are

g in co-ordination.
deprive Black of counter-play and to
blockade the d5 pawn.

10 Nf3 Bxf3! 1 Rfel

11 Rxf3 NeS
12 Re3 Kf6 It is to White’s advantage to exchange
13 Bb3 a5 rooks, since in the minor piece ending
his king will be much more active than
White is in complete zugzwang. the opponent’s.

14 Ba4 Nxc4 1 ... Rfe8

15 Re8 Rxc3
16 Rc8 Ne3 After 1 . . . R xel+ 2 R xel d4 3
17 Bb5 c4 Ne2 Black could have lost a pawn, but
gained some counter-play, after 3 . . .
17 . . . Rc2 18 Bc6 Nfl+ 19 KglRb8 4 b 4 d 3 5 Nc3 a5 6 bxa5 Rb3
Nxg3 and 20 . . . f4 was more exact. 7 N b l! If instead 4 Nxd4, then after
4 . . . Rxb2 5 Re7 Rb7 the advantage
18 Kgl Rc2 is with White, but nothing decisive is
19 Bc6 c3 apparent. After missing this possibility
20 Bf3 g5 which, though rather risky, was the most
21 g4 f4 unpleasant for White, Black gradually
White resigns. ends up in a bind.

2 Rxe7 Rxe7
Botvinnik—Konstantinopolsky 3 Kf2 Kf7
4 R dl!
Sverdlovsk, 1943
It is time to take the d-pawn under
control. 4 Rel? would have been a
mistake due to 4 . . . R xel 5 Kxel
d4! followed by the centralization of
the king. White must seize the e-file in
such a way that the d -pawn is unable to
advance.

4 ... Re8
5 Rd2! h6
6 Re2 Rb8
7 Ke3

The blockading square is best occu­

White’s knight is significantly stronger pied by the king, the knight being need­
than the enemy bishop, and his pawn ed for the preparation o f b4—b5.
formation is more compact. He plans
active play on the Q-side (the pawn 7 ... Rb3
break b4—b5 after suitable preparation), 8 Kd4 Kf6
whereas Black must try to activate his 9 Na2

164
 Complex Endings

White follows his plan. 30 Ng7 Ba4

31 f5 Kg5
9 ... Rb8 32 Ne6+ Resigns.
10 b4 g5
11 g3 gxf4
12 gxf4 a6
13 Nc3 Rg8
14 a4 Rg4 Goldin—Korzubov
15 Rr‘2
Dushanbe, 1980
Black has slightly activated his forces,
but it is too late —White is all ready for
the decisive break.

15 . . . Be6
16 b5

This same move would have followed

on 15 . . . Be8.

16 ... axb5
17 axb5 cxb5
18 Nxb5 Rgi
19 Nc3! Kf7
20 Rb2! R fl In spite of Black's material advantage,
21 Ne2! it is very difficult for him to break up
the opponent’s fortress. If he advances
Do not hurry! After 21 Ke5 R el+ his pawns by . . . a5 and . . . b5—b4,
22 Kd6 Black would have gained White’s defences hold if he exchanges on
counter-play by 22 . . . d4. b4 and plays Rc2. Only play on both
wings can give Black winning chances.
21 . . . R el
22 Ke5 d4 1 ... Qcl+
2 Kf2 fS!
This pawn sacrifice merely delays the
inevitable. A highly important move, which has
to be made immediately. If White were
23 Kxd4 Kg6 to succeed in playing g2—g3 and f3—f4,
24 Nc3 Kh5 the game would most probably end in a
25 Re2 Rxe2 draw.
26 Nxe2 Kg4
27 Ke5 Bc8 3 Kg3
28 Nd4 h5
29 Nxf5! 3 g3 would have been met not by
3 . . . Qhl 4 f4! with a probable draw,
In conclusion — a tactical blow. If but 3 . . . f4!, exposing the white king.
29 . . . BxfS 30 h3+, winning.
3 ... Qg5+!
29 ... Bd7 4 Kh2 f4

165
 Endgame Strategy

The first stage o f Black’s plan is com- After 24 Nxb4? KgS White is in zug-

Í ilete: White’s K-side pawns have been zwatig, since 25 Rf2 is bad due to 25
ixed. The next stage is a Q-side pawn . . . d4 26 cxd4Q el.
offensive.
24 . . . Qd3
5 Re8+ Kf7 25 R cl Qe3
6 Rc2 a6 26 Rc2 Q el
7 Nc6 Qh4 27 Rc2 Q bl
8 Nd4 b5 28 Rd2 Q al
9 Kgl Qf6 29 Rc2
10 Kh2 Qb6
11 Re5 The series o f manoeuvres has not
brought any tangible result, and Black
The pawn ending after 11 Re6 Qxe6 sets about transferring his king to the
12 Nxe6 Kxe6 13 g4 a5 followed by Q-side.
. . . d4 is hopeless for White.
29 . . . Ke7
11 . . . Qb7 30 Re2+ Kd7
12 K hl Qd7 31 Rc2?
13 Kgl aS
14 Kh2 Qd6 After this Black is able to cany out
15 Re2 b4 his plan. Interesting complications would
16 axb4 axb4 have arisen after 31 Re5!, e.g.:
(a) 31 . . . Qxb2 32 Rxd5+Kc7 33
This concludes the second part of Rc5+ (33 . . . Kb6 34 Rb5+ Ka6 35
Black’s plan. Now he must take his R xb4 Qxc3 36 Nc6 also promises Black
queen into White’s rearguard. little) 34 Rxc4 bxc3 (or 34 . . . b3
35 N xb3) 35 Rxc3, with a draw.
17 K hl Qa6 (b) 31 . . . bxc3 32 Rxd5+ Kc8 33
18 Kh2 bxc3 Qxc3 is Black’s best chance, when
much depends upon whether the posi­
White loses immediately after 18 tion after 34 Nb5 Qb4 35 Nd6+ Kc7
cxb4? Q al+ 19 K h2Q dl. 36 Nxc4 Qxc4 37 Re5 is a win or a
draw.
18 . . . Qal
19 Rc2 Qel

The king must now be brought to a4.

Before advancing his king. Black, operat­
ing according to the principle “do not
hurry” , manoeuvres with his queen, try­
ing to avoid revealing his plans to his
opponent.

20 Re2 Qdl
21 Rf2 Q al
22 Rc2 Kf6
23 Nc6 Qdl
24 Nd4

166
 Complex Endings

The winning plan can be as follows: Black has carried out his plan in full,
Black places his g-pawn at g5, his queen and White is in zugzwang. To any knight
at cl and his king at d5. Suppose that, move Black replies 36 . . . Kb3, and if
with his rook at e l, it is White to move: 37 Nd4+ Ka2 followed by 38 . . . b3,
(i) The rook moves to e8. Black ad­ and 39 . . . Qxb2, winning. If White
vances his king to d3, plays . . . Qe3!, plays 36 Nc6 Kb3 37 Nxb4, then
and then proceeds with his king to f2, 37 . . . d4 is decisive.
after which at some point the . . . g4
break is decisive. 36 g3
(ii) The rook keeps on the fourth
rank. Black achieves the following posi­ White is forced to expose his king.
tion with White to move: rook at e4,
black king at cS, queen at a l. White is 36 . . . gS!
forced to allow the black king to reach
d3. Of course, not 36 . . . fxg3+? 37
Thus by playing 31 Re5! White Kxg3, when the white king breaks out
would not have attained a theoretically to the centre.
drawn position, but would have retain­
ed drawing chances. He is not obliged to 37 g4 Q el
give up his knight for the pawn immedi­ 38 Kg2 Qg3+
ately. 39 K fl Qxh3+
But after the insipid move in the game 40 Kf2 Qh2+
White quickly ends up in a lost position.
The “do not hurry" tactics have played Black has captured the h-pawn, but it
their part: White has been uable to is still by no means easy to win the game.
switch from passive defence to active In order to breach White’s position he
play. has to sacrifice his queen for the rook.

31 ... Kc8 41 Kel Qhl+!

32 Ne6 Kb7 42 Kf2!
33 Nd4 Kb6
34 Ne6 Kb5 42 Ke2 loses to 42 . . . Qgl
35 Nd4+ Ka4 Kd2 Q fl.

42 . . . Q dl
43 Re2 Qd3?!
44 Rc2

(See next diagram)

44 ... Qxc2+?

Fatigued by the tense struggle, Korzu-

bov makes it much harder for himself to
win. The idea of the queen sacrifice at
c2 is correct, but it should be made with
the white king at g2. This can be
achieved by a study-like manoeuvre,
making repeated use of zugzwang:

U -L
167
 Endgame Strategy

54 Kdl Qxf3+
55 Kd2?

A new phase of the snuggle has begun

— a queen ending where Black is a pawn
up. Queen endings are mainly a battle of
the kings. Black’s plan includes taking
his king to f3, and therefore White
should have kept his king at e l , f l and
f2. Here he should have retreated to e l,
and the outcome would have been not
altogether dear. After the move played
Black cuts the white king off from the
44 . . . Q dl! 45 Rc2 Qcl! 46 Rc2 K-side, and the win becomes merely a
Q al 47 Kg2 (any knight move is question of time.
decisively met by 47 . . . Kb3, while on
47 Ke2 Black wins by 47 . . . Qgl) 55 . . . Qe3+!
47 . . . Qel! 48 Re2 Qcl! 49 Rc2
Qxc2+! 50 Nxc2 bxc3 51 bxcí Kb3 After 55 . . . Qxg4?! White would
with an easy win. After the move played have had every chance o f giving per­
the battle flares up anew. petual check.

45 Nxc2 bxc3 56 K dl Qe5

46 bxc3 Kb3 57 Kd2 Kc4
47 Nb4! 58 Qc8+ Qc5
59 Qa6+ Kd5
The only move, the alternatives being 60 Qb7+ Ke5
hopeless: 61 Qg7+ Ke4
(a) 47 Nd4+ Kxc3 48 Ne2+ Kd2 62 Qh7+ Kf3
49 Nxf4gxf4 50 g5c3 51 g6 c2 52 63 Qh3+ Kf2
g7 cl=Q 53 g8=Q Qel+ 54 Kg2 Qe2+ 64 Qh2+ K fl
and 55 . . . Qxf3. 65 Qh5 Qd5+
(b) 47 Nel Kxc3 48 Ke2 d4 49 66 Kcl Kg2
K d ld 3 50 Ng2Kd4 51 Kd2 c3+ 52 White resigns.
Kdl d2! 53 Kc2 Kc4 54 Nxf4gxf4
55 g5 Kd5. The f-pawn will queen.

47 . . . d4
48 cxd4 Kxb4
49 Ke2 Eliskascs—Flohr

Here is where it becomes significant Scmmering, 1937

that the king is at f2, and not g2.
(See next diagram)
49 • • ♦ c3
50 d5 Kb3 The black pieces occupy more com­
51 d6 c2 fortable positions, and the a2 pawn
52 d7 cl=Q may prove a weakness. White has to
53 d8=Q Qe3 + play accurately, so that Black’s initiative

168
 Complex Endings

5 Bxtt Rac8
6 Rd2 c5 !

The positional blows follow one after

another.

7 dS Nc4
8 Re2 Nd6

The blockade is complete. From being

roughly level the position has been trans­
formed into one which is won for Black.
But the win is still far off.
should not be transformed into a solid
positional advantage. Possibly he should 9 R bl Rc4
have castled long, so as to co-ordinate his 10 g3 Rdc8
rooks and bring his king closer to the de­ 11 Bg2
fence of his a-pawn.
For example: 1 0 -0 - 0 !? Be6 2 Kbl Now Black has many tempting possi­
Nc4 3 Nd2 Rac8 4 Nxc4 Bxc4 5 bilities: . . . b5, . . . Rc2 or . . . f6
Bxc4 Rxc4 6 R cl Rb4+ 7 Kal e5 followed by the transfer of the king to
8 dxe5 Rd2 9 R bl, or 2 ...R a c 8 c7. Flohr takes a different, and interest­
3 Nd2 Bf5+ 4 e4 Bd7 5 R cl, with ing decision: he exchanges rooks. One
chances of equalizing. White chooses K- cannot help being struck by the confi­
side castling, and takes his king away dence of the grandmaster in the correct­
from the centre. ness o f his evaluation o f the position.
After all, if the resulting minor piece
1 0 -0 Be6 ending should not prove to be won for
2 e4? Black, the exchange of his rooks, ex­
cellently deployed on the only open file,
And this pseudo-active move is a will have thrown away the win.
serious positional mistake. To be con­
sidered was 2 Nd2 Rac8 3 R fcl, re­ 11 ... Rcl+!?
straining Black’s offensive. 12 R xcl Rxcl+
13 R el Rxel+
2 ... Bg4! 14 Kxel f5!
3 Rfdl?
The base of White’s central pawn
There was no point in stopping half­ chain has to be weakened.
way. 3 d5 was more correct.
15 f3
3 ... e6!
After 15 exf5 gxf5 the d5 pawn
Flohr skilfully exploits his opponent’s would in time be lost, while on 15 f4
errors. First he restrains White’s pawn Black has a simple reply: 15 . . . Kf6.
centre, and then embarks on a complete
blockade of it. 15 . . . fxe4
16 fxe4 b5
4 Kfl Bxf3! 17 Kd2 a5

169
 Endgame Strategy

18 Kd3 Kf6 “The decisive mistake. The correct

19 Bf3 Ke7 continuation was 28 Bbl Nd2 29 Bd3.
20 h4? not allowing the knight to attack the K-
side pawns.” (Averbakh).
In the end this leads to a weakening
of White’s position on the K-side. The 28 . . . Nd2
immediate 20 Bdl followed by 21 a4 29 Bc2 Nfl
was better, as suggested by Euwe. 30 Kxa5

20 . . . h6! 30 g4 Ne3 would not have changed

anything.
N ot allowing White to fix the pawns
at h7 and g6 and obtain counter-play 30 . . . Nxg3
after g3—g4—g5. N ow on 21 g4 there 31 Ka4 Nh5
follows 21 . . . g5, restricting the white 32 Kb3 Kd4!
bishop still further.
Now precise calculation is required.
21 Bdl Kd8
22 a4 33 Kb4 Nf6
34 d6 g5
Following the change in the K-side 35 hxg5 hxg5
pawn structure, this move is not as good 36 Kb5 g4
as it was earlier. 37 Bdl g3
38 Bf3 Kc3
22 . . . bxa4?! 39 Bhl Kf2
40 Kc6 g2
22 . . . b4! is stronger. Euwe gives 41 Bxg2 Kxg2
the following variation: 23 Bb3 Kc7 42 d7 Nxd7
24 Bdl Kb6 25 Bc2 (25 Bb3 Nb7 43 Kxd7 Kf3
26 Kc4 Nc5 27 Bc2 hS) 25 . . . Nb7 White resigns.
26 Kc4 Nc5 27 g4g5 28 h5 b3 29
Bbl b2 30 Kc3 Nxa4+ 31 Kb3 Nc5+ We give the following game in full,
32 Kxb2 Nd7 and 33 . . . Nf6. The since the exchange o f queens takes place
difference in the position of the white fa- as early as the ninth move, and the sub­
pawn leads to a win for Black. sequent play follows the typical princi­
ples of endgame strategy.
23 Bxa4 Kc7
24 Bc2 Kb6
25 Kc3 Kb5 Shereshevsky—Veremeichik
26 Kb3 Kc5
27 Ka4 Minsk, 1978

White makes a desperate attem pt to 1 d4 Nf6

obtain counter-play. On 27 Kc3 there 2 Bg5 c5
would have followed 27 . . . a4! 3 Nc3!? Qa5
4 Bxf6 gxf6
5 c3 f5
27 . . . Nc4 6 Qh5 cxd4
28 Bb3? 7 exd4 Qb6

170
 Complex Endings

8 0 -0 -0 QH6+ lar prospects.

9 Qxh6 Bxh6+
10 Kbl d6 14 . . . Bg7
15 Nge2 h5
16 h4

White leaves himself with a backward

pawn at g3, but it is easily defended. At
the same time the black h5 pawn also
becomes vulnerable.

16 ... Na6
17 Rhel Nc7
18 Nd4 Kf8
19 Re3 Bf6
20 Bh3 Bxd4

A complex endgame position hzs been After this exchange Black is doomed
reached. Black’s K-side pawn formation to passive defence. After 20 . . . e6
is compromised, but he has the two bis­ White could have made the piece sacri­
hops. White must play thoughtfully and fice 21 dxeó fxe6 22 Nxf5!? Bxc3
consistently, otherwise he may gradually 23 Rxd6, although, of course, he would
be saddled with even the inferior game. not have been obliged to do so. Never­
First he must restrict the scope of the theless, this continuation would have
black bishops, and then think about given Black counter-chances in a tactical
creating weaknesses in the opponent's struggle, whereas, without his black-
position. squared bishop, his position will gradu­
ally deteriorate against correct play by
11 g3! White.
21 Rxd4 Re8
With the idea of developing the bishop
at g2 and of setting up the pawn White must again form a plan. Black
column h2/g3/f4, restricting the oppo­ has weak pawns at h5, f5 and e7 on
nent’s black-squared bishop. 11 . . . f4 the K-side. But these weaknesses are
fails to 12 Nd5. close to one another and are not difficult
to defend. Therefore, following the
11 ••• Bd7 principle of two weaknesses. White must
also create some vulnerable targets on
Black plans to exchange the white- the Q-side. But first he should improve
squaicd bishops, but he merely aids the the placing of his pieces and tie Black
development of White’s game. down on the K-side. To do this he
centralizes his king and transfers his
12 Bg2 Bc6 knight to e3, after whidi his rooks gain
13 d5! Bd7 the opportunity to attack the oppo­
14 f4 nent’s Q-side.

Now the possible targets for attack 22 Kcl! Kg7

take shape — the pawns at f5 and e7, 23 Kd2 Kf6
while the black bishops have no particu- 24 Kel Rc8

171
 Endgame Strategy

25 B fl Ne8 This rook will for a long time be shut

26 Rd2 out of play. White has a clear procedure
for realizing his advantage: (1) fix
Preparing the transfer of the knight Black's Q-side by a2—a4—a5; (2) ex­
to e3. change the black rook at c7 by Rc4;
(3) with his bishop at d3 and knight at
26 . . . Ng7 e3, tying Black’s pieces to the defence
27 N dl Rhe8 of the f5 pawn, transfer the king to d4;
28 Ra3 (4) advance c2—c4—c5, after which
Black’s position will begin to collapse.
It would have been better to avoid White mentally outlined this plan, but
this weakening and play 28 . . . Ra8. then, fearing the possibility of time
trouble, decided to act according to the
29 Rb3 Rc7 principle “do not hurry’’ and to repeat
the position once or twice.
After 29 . . . b5 White has the un-
I easant 30 a4. 37 Rd4?

30 Rb6 But this principle should be employed

only when the player is confident that
Now the pawn at b7 is fixed. such tactics will lead to a deterioration
in the opponent’s position and to a
30 • . . Rd8 favourable outcome. During the game
31 Be2 Kg6 White was convinced that there was al­
32 Ne3 Ne8 together nothing that Black could do,
33 Rd4 Nf6 and he overlooked a latent possibility of
34 Bd3 Rg8 counter-play.

Black is completely tied down, and 37 . . . Bb5!

can only watch as White improves his
position. Being short of time, White is An unpleasant surprise. It transpires
not in a hurry to take positive action. that the f5 pawn is immune, while after
38 Bxb5 axb5 39 a3 Rc5 the black
35 Ke2 Ra8?! pieces break free. White does not wish to
36 Rdb4 Ra7?! play 38 c4, since this square is ear­
marked for his rook, and so he decides
on an exchange sacrifice.

38 Rxb5!? axb5
39 Nxf5 Ra4
40 b4 Rxa2
41 Kf3?!

Again abusing the principle of “do not

hurry’’. The immediate 41 Nxe7++ was
simpler and better, but time trouble and
a sharp change of situation can put many
players out of their stride. Fortunately
for White, 41. . .Ng4 fails to 42 Nxd6+!

172
 Complex Endings

f5 (or 42 . . . Kb6 43 N xp+ Kg7 The immediate 47 . . . Rgl is more

44 Ne5) 43 Nxf5! (but not 43 BxfS+? logical.
Kf6 44 Ne8+ KxfS 43 Nxc 7 Rxc2),
but he should not have allowed Black a 48 Bd3 Ne8
wide choice of continuations by playing 49 Kc4!
41 Kf3. This was clearly irrational.
Now we see White's plan. His king
41 ... Rc3 goes onto the attack, shattering Black's
42 Nxe7++ Kg7 defences.
43 Nf5+ Kf8
44 Ke2 49 . . . Ng7
50 Kb5 Nxf5
White has no right to give up his c2 51 Bxf5 Ke7
pawn, since it and the bishop at d3 are
the base of his entire position. 51 . . . Re3 is obviously too slow.

44 . . . Rc8?l 52 Kb6 Kf6

Black again selects a comparatively With the idea of answering 53 Bd3

passive plan, transferring his rook to d8 with 5 3 __ Re7.
to defend the d-pawn. Much more un­
pleasant for White would have been an 53 Kc7!
attempt at counter-attack. A possible
variation would be: The king lands the decisive blow.
44 . . . Rc7 (so that Nxd6 should worthily crowning its lengthy journey.
not be with gain of tempo) 45 Kd2 R al
46 Bxb5 Rgl 47 Bd3 Ng4 48 bS! 53 . . . Rg8
Nf2 49 b6R c5 50 Nxd6Nxd3 51 54 Bd3 Re7+
Nxb7 Rc8 52 d6l, and White must 55 Kxd6 Rc8
win. But Black has other possibilities.
Instead of 47 . . . Ng4 he can play 47 With the threat of 56 — Rcc7.
. . . Rg2+ 48 Kel Ng4 49 b5 with a
very sharp game. 56 Rc4 Rd8+
In choosing passive tactics, Black had 57 Kc5 Rc7+
possibly not guessed at White's idea. At 58 Kb6 Rcd7
first sight, Black’s set-up does not seem 59 Rc7
at all bad: one rook is transferred to the
defence, the second will attack at gl, The simplest.
and the knight can go via e8 to g7.
59 . . . Rxd5
45 Kd2 Rd8 60 Kxb7 Rxd3

All the same the b5 pawn cannot be Despair.

saved. White was threatening 46 Nxd6
Rd8 47 Nxb5, when 47 . . . Nxd5 61 cxd3 Rxd3
is not possible due to 48 Bc4. 62 b5 Rxg3
63 b6 Rb3
46 Bxb5 R al 64 Rc5 Rb4
47 Kc3 R el 65 Kc7 Resigns.

173
 Endgame Strategy

Rubinstein—Tackacs point, better chances were offered by

the cunning 2 . . . Nxd4?!, and if 3
Budapest, 1926 exd4 Bd2!, when it is Black who wins.
But White, of course, is not obliged to
take on d4; he can simply play 3 Nxb7,
retaining all the advantages of his posi­
tion.

3 Kfl!

White does not hurry, and centralizes

his king. The immediate 3 b4 was also
possible, but Rubinstein prevents
. . . Nxd4 once and for all, while the b-
pawn can always be advanced later.

3 ... Bd8
The position has been reached from 4 b4 f5
the Carlsbad Variation of the Queen’s
Gambit. The pawn formation is charac­ Black’s K-side activity is easily neutral­
teristic for the carrying out of the so- ized by White, and leads merely to new
called minority attack, by which White weaknesses. Once Black had chosen pass­
advances his Q-side pawns to a4 and b5 ive tactics, he should have stuck to diem
to create weaknesses in his opponent’s as long as possible. A tenacious, planned
pawn formation. But Black has succeed­ defence would have been much more
ed in exchanging the white-squared bis­ appropriate than unprepared counter­
hops, which in this position is to his ad­ play. He should have played 4 . . . Kf8,
vantage, and White’s pawn attack can be but not 4 . . . Bc7 5 Bxc7 Nxc7 6
met by the direct . . . b5 (with the white Nc5.
pawn at b4) followed by the transfer of
the knight to c4. Therefore there is no 5 Nb2!
point in White forcing events. He must
attempt to break up the opponent's de­ White keeps a careful eye on his oppo­
fences and worsen the placing of the nent's counter-play. The transfer of the
black pieces, and only then seek a possi­ knight to d3 forces Black to abandon
ble breakthrough. his active play on the K-side.

1 ... Ne6 5 ... g5

2 Na5 Ra7? 6 Nd3 Kf7
7 Rc2 Bb6
Such moves are made only in ex­
tremis. Black’s position is inferior, and Stronger was 7 . . . Ke7 with the
he should have decided on 2 . . . b5 idea of . . . Bb6, . . . Kd8 and . . . Bc7,
3 Nc3 Rac8, aiming f o r __ c5. In this taking the king to the defence of the Q-
case, despite his obvious pawn weak­ side.
nesses, Black's pieces would have occu­
pied active positions, and White’s task 8 Bd6!
would have been markedly more diffi­
cult. From the purely practical view­ Again Rubinstein skilfully suppresses

174
 Complex Endings

the opponent’s counter-play, while con­

tinuing to cramp Black’s position. ■ □ ■ ■
Tackacs is unable to persist with his king
manoeuvre, since after 8 . . . Rh8 9
ansa bet,■
Reel Kc8 10 a4 Kd8 White has the
decisive sacrifice 11 Rxc6! bxc6 12
Nxc6+ Kc8 13 Nxa7 Bxa7 14 Rc8+
Nd8 15 Ra8 Bb6 16 a5.

8 ... Nd8

Removing one of the attacks on c5,

which White promptly exploits.
18 h4!!
9 Nc5! Nxc5
10 Bxc5 Bxc5 Rubinstein sacrifices a pawn, to gain
11 bxcS control o f the g-file. It is this that will be
the second weakness in Black’s position.
The position has simplified. Black’s
basic weakness — his b7 pawn — is 18 ... gxh4
fixed, the rook at a7 is still out of play, 19 gxf5 gxf5
but for the moment White does not 20 Rg7 Nd8
appear to have anything concrete. He 21 Rg8 f4
needs to create a second weakness on
Black’s other wing. This is assisted by This merely accelerates the end.
the position of the pawns at f5 and g5.
Were these pawns on the 6th rank, 22 Rh8 fxe3
White’s task would be more difficult. 23 fxe3 Kd7
24 Rg2! Re8
11 . . . Ke7 25 Rxh4 Re7
26 Rh8 Kc7
In an attempt to help out his rook at 27 Rgg8 Rd7
a7, Black marches his king to the Q-side,
but this merely helps White’s break­ Black resists to the last.
through on the K-side.
28 Nb3!

12 Rb2 Kd7 The transfer o f this knight i e5 con-

13 Rebl Kc8 eludes the game.
14 Ke2 Re7
15 Kf3 Re4 28 . . . a5
16 g4! 29 Ncl Ra8
30 Nd3 b5

The start of the attack. A last try: White might just play 31
Ne5 Re7 32 Rxd8? Rxd8 33 Rxd8
Rxe5! Passive defence would not have
16 . . . g6 saved Black, a possible continuation be­
17 Rgl Nf7 ing 30 . . . a4 31 Ne5 Re7 32 a3!

175
 Endgame Strategy

RcS 33 Kg4 Rb8 34 Kf5 Rc8 35 attack there, and so he transfers the king
R e8R xe8 36 Rxc8 followed by Re7+ to f5 by the shortest route via h7.
and the transfer of the king to b6.
3 ... h5!
31 cxb6+ Kxb6
32 Nc5 Rd6 In achieving the required set-up,
33 a4! Rc8 Fischer does not waste a single move. It
34 Kg4 Resigns. is advisable not to huny only when the
projected plan has been earned out, or
Black has no defence against Kf5, when it has to be masked from the
Rh7, Kc5 and Rxd8. opponent.

4 Be3
Saidy-Fischer
White sticks to waiting tactics. It
USA, 1964 would possibly have been better to
activate his bishop by a3—a4, b2—b3
and Ba3 with his king a t e3.

4 ... Kh7
5 f3 Kg6
6 a4 Kf5
7 Ke2 g5
8 Kf2

Black has achieved his planned set-up,

and now sharply changes the tempo of
the play. In his M y 60 Memorable
Games, commenting on a game from
Portoroz, 1958, Fischer writes: “Pet­
Were it not for the pawns at d4 and rosian likes to play cat-and-mouse,
d5, White’s chances would be in no way hoping that his opponents will go wrong
worse. But the mutually isolated pawns in the absence of a direct th reat The
create an impenetrable barrier to the amazing thing is — they usually do!” A
white bishop, without restricting the highly graphic exposition o f the princi­
mobility of the black knight. White is ple “do not hurry”. The further course
therefore faced with a difficult task in of the game shows how well Fischer had
trying to draw. learned this particular lesson. Although
Saidy does not make any obvious mis­
1 Kfl Nf8 takes, he also begins moving here and
2 Ke2 Ne6 there, not wishing to create any weak­
3 Kd3 nesses in his position. But at the critical
moment, when immediate activity is re­
What plan should Black choose? First quired, White proves to be psychologic­
he must activate his king. It would ally unprepared for it.
appear that it should be brought to the
centre to d6. But the American grand­ 8 ... Nd8
master intends to play on the K-side 9 Bd2 Kg6
with the aim of creating a target for 10 Ke3 Ne6

176
 Complex Endings

11 Kd3 Kf5 27 Kd3

12 Be3 f6
13 Ke2 Kg6 The pawn ending is obviously hope­
14 Kd3 f5 less, while after 27 Kf2 Black gains e4
for his king by 27 . . . Kf5 28 Kf3
Note how slowly, almost unwillingly. Ne4 followed by . . . Ng5+. If 29 Bh4,
the black f*pawn advances. then 29 . . . Nd2+ 30 Ke2 Nc4 31
b3 Na5 32 b4 Nc6, and White loses a
15 Ke2 f4 pawn.
16 Bf2 Ng7 27 . . . Nf5
17 h3 Nf5
18 Kd3 g4!? It transpires that against the threat of
. . . Nh4 White has no satisfactory de­
Unexpectedly Black makes a break, fence.
although he could have played his knight
to e6 and his king to f5. 28 Bf2 Nh4
29 a5
19 hxg4 hxg4
20 fxg4 Nh6 A belated attempt at activity.
21 Bel?
29 . . . Nxg2
The critical point. What is now re­ 30 Kc3 Kf3
quired of White is a concrete approach 31 Bgl Ke2
to the problems facing him. He should 32 Bh2 f3
have immediately activated his bishop 33 Bg3 Ne3
by 21 Bh4 Nxg4 22 Bd8, with the White resigns.
idea of a4—a5. True, after 22 . .. Ne3
White loses a pawn, but 23 h3 Nc4 There is no defence against. . . Nf5.
24 gxf4 Nxb2+ 25 Kc2 Nxa4 26
Kb3 b5 27 Kb4 a6 28 Ka5 Nc3 29
Be7 gives him every hope of a draw. In­ Lilienthal—Bondarevsky
stead, through inertia, White continues
his waiting tactics, and very soon finds Moscow, 1940
himself in a hopeless position.

21 . . • Nxg4
22 Bd2 Kf5
23 Bel Nf6
24 Bh4 Ne4
25 Bel

It is now too late for Bd8, since after

. ■. Kg4 the white king has to switch to
the defence of the g2 pawn, and there
can be no question of any activity on the
Q-side.

25 . . . Kg4 This game was played in the last

26 Ke2 Ng3+ round of the 12 th USSR Championship,

177
 Endgame Strategy

and had enormous significance regarding 9 Rxc8! Rxc8

the final placings. Bondarevsky was lead­ 10 h4!
ing, with Lilienthal a point behind. In
such a situation, where a draw is equi­ White opens and seizes control o f the
valent to defeat, the majority of players h-file, the invasion squares along which
begin complicating the position, burn­ are a weakness in Black’s position.
ing all their boats behind them. But in
this game White happily went for favour­ 10 . . . h6
able simplification, and as a result gained
a highly promising ending. The mutually 10 . . . gxh4 is bad: 11 R h l f5 12
isolated pawns in the centre give the ad­ g5-
vantage to the knight over the bishop.
In addition, Black’s doubled b-pawns are 11 hxgS hxgS
weak, so that it is dangerous for him to
go into the minor piece ending. Perhaps 11 . . . fxg5 was the lesser
evil, but it was difficult for Black to de­
1 Rfcl cide on giving himself a backward pawn
at h6 and weakening his control over e5.
Of course, not 1 Nd7?! Rfd8 2
Nxb6? Ra6. 12 R hl Re8
13 Kd2 Bd7
1 . .. Rf c8 14 Rh6 Rf8
2 a3 Bf5
3 g4 Be6 On 14 . . . Ke6 White has the un­
pleasant 15 f4.
The rook ending after 3 . . . f6 4
gxf5 fxeS 5 dxe5 gxf5 6 f4 is also 15 Nel Kc7
unfavourable for Black. 16 Nc2 Rf7
17 Ne3 Be6
4 h3 f6 18 Kc3
5 Nd3 gS
6 f3 Kf7 White has completely tied down his
7 Kf2 Ke7?! opponent’s forces on the K-side, and he
now transfers his king to the Q-side to
A routine move. Black follows the attack the weak b-pawns.
principle that the centralization o f the
king is always useful. This is correct, but 18 . . . Kd6
first he should have suppressed White’s 19 Kb4 Bd7
attempts to give him weaknesses on the 20 Nf5+
K-side. By 7 . . . h5! Black could have
gained counter-chances. In general the exchange of minor
pieces eases Black’s defence, but White
8 Ke3 Kd6? has calculated that the rook ending is
won after 20 . . . Bxf5 21 gxf5 Kc6
And here . . . h5 was simply essential. 22 a4 Rf8 23 Rh7 Rd8 24 Rf7 Rd6
The natural king move is possibly the 25 b3 b5 26 a5 b6 27 a6.
decisive mistake, allowing White to give
his opponent a weakness on the K-side, 20 . . . Kc7
in addition to those which he already has. 21 a4 Be6

178
 Complex Endings

22 Ng5 Bd7
23 Nh5! f5
24 Nf6!

By this knight manoeuvre White wins

a pawn, accurately judging that the
counter-play obtained by Black is in­
sufficient.

24 . . . fxg4
25 Nxd5+ Kb8
26 fxg4

White avoids the tempting, but errone­ The position is an open one, and the
ous attempt to play for mate: 26 Rh8+? bishop is d early superior to the knight.
Ka7 27 Nc7 b5 28 a5 b6 29 a6 It is true that realizing this advantage is
Bc6!, and Black defends against the mate very difficult, since the pawn formation
at a8, simultaneously attacking the is symmetric, and there is no possibility
knight. of setting up a passed pawn. First White
restricts the knight’s mobility and
26 , . . Bxg4 creates weaknesses in his opponent’s
27 Nxb6 Rf2 position.
28 b3 Bdl
29 d5 Kc7 1 Re5 b6?!

If 29 . . . Rf3, then 30 d6 Rxb3+ 1 . . . Kd6 was preferable, and if 2

31 Ka5, and either Black is mated or Bxb7?! Rb8.
the white pawn queens.
2 Bfl! a5
30 a5 Rd2 3 Bc4 Rf8
31 Rh7+ Kb8
32 d6! Just three moves have been made, and
how the position has changed! The black
White concludes the game energetic­ rook is tied to the f7 pawn, Black’s Q-
ally. 32 . . . Rxd6 fails to 33 Rh8+ side pawns have lost their mobility, and
Kc7 34 Rc8 mate. he has acquired a weak square at b5.
Now White centralizes his king.
32 . .. Rd4+
33 Kc5 Rh4 4 Kg2 Kd6
34 d7 Kc7 5 Kf3 Nd7
35 d8=Q++! Kxd8 6 Re3 Nb8
36 Rd7+ Resigns. 7 Rd3+ Kc7
8 c3 Nc6
9 Re3 Kd6
10 a4
Fischer—Taimanov
White's play has proved successful —
Candidates M-Final Match the black knight has no strong points,
Vancouver, 1971 the white king is centralized, and Black’s

179
 Endgime Strategy

pawn weaknesses on the Q-side are fixed. 19 Ke2 Kd8

But for a win this is insufficient. Black
must also be given a second weakness on Black does nothing to hinder White’s
the K-side. intention. He could of course have play­
ed 18 . . . Rf6, but after 19 Kd2 he
10 . . . Ne7 would have had to agree to the exchange,
11 h3! Nc6 since to allow the white rook to reach e8
12 h4 would be even less attractive.

Do not huriy! Remember the advice 20 Rd3 Kc7

of Byelavyenets: “ If there is a possibility 21 Rxd6 Kxd6
of advancing a pawn two squares or one, 22 Kd3 Ne7
advance it first one square, look care­ 23 Be8
fully around, and only then advance it a
further square.” The knight is tied to the g6 pawn,
and White’s problem is to break through
12 . . . h5?! with his king to a6. If he can attain the
following position. Black will be in com­
Black fails to withstand the tension. plete zugzwang.
His desire to clarify the position as soon
as possible is psychologically understand­
able. Unfortunately, he is now forced
to place all his K-side pawns on white
squares, and it is this that will constitute
Black’s second weakness. Up to a certain
time it would have been better to stick
to waiting tactics, leaving the K-side
pawns in place. After 12 . . . h5 White
realizes his advantage in highly instruct­
ive fashion.

13 Rd3+ Kc7
14 RdS fS
15 Rd2 Rf6 23 «• • Kd5
16 Re2 Kd7 24 Bf7+ Kd6
17 Re3! 25 Kc4 Kc6
26 Be8+ Kb7
Black is practically in zugzwang. 27 Kb5 Nc8

17 . . . g6 Clever, but inadequate. White cannot

18 BbS of course take the pawn: 28 Bxg6??
Nd6 mate, but he is able to push
Black’s K-side pawns present an ex­ back the black king by a subtle bishop
cellent target for the bishop, while on manoeuvre.
the Q-side the king can easily approach
via c4—bS—a6, so that White’s next 28 Bc6+ Kc7
problem is to exchange rooks. 29 BdS! Ne7
30 Bf7 Kb7
18 . . . Rd6 31 Bb3!

180
 Complex Endings

The bishop switches to the long dia­

gonal.

31 • • • Ka7
32 Bdl Kb7
33 Bf3+ Kc7

The a6 square has been won for the

king. Black cannot play 33 . . . Ka7,
since he is in xugiwang after any waiting
move by White.

34 Ka6 Ng8
35 Bd5 Ne7 more promising game, since White will
36 Bc4 Nc6 not have any invasion squares on the d-
37 Bf7 Ne7 file, while in the event of the exchange
38 Be8 of rooks the Q-side pawns at a3, b2 and
c3 may become an excellent target for
The required position has been attain­ the black bishop.
ed. Black is in zugzwang.
1 ... Re8
38 . . . Kd8 2 Rd3 Raa8
39 Bxg6 3 K fl Rab8!

The tireless bishop now sacrifices it­ By this move Portisch reduces still
self. further the value of a possible c3—c4,
on which there follows . . . b4!
39 .., Nxg6
40 Kxb6 Kd7 4 N el g6
41 Kxc5 Ne7 5 Nc2
42 b4 axb4
43 cxb4 Nc8 The white knights rush around the
44 a5 Nd6 board in search o f strong points, and are
45 b5 Ne4+ quite unable to find any.
46 Kb6 Kc8
47 Kc6 Kb8 5 ... h5
6 f3
and Black resigned.
An important moment. The f3 square
is now occupied by the pawn, which
Keres—Portisch means that there is no longer any threat
to the e5 pawn, and Black can ex­
M oscow, 1967 change rooks. To some extent the pawn
formation resembles that in the previous
(See next diagram) Fischer—Taimanov game, where on one
wing all the weaker side’s pawns are on
A complicated ending. White controls squares of the colour o f the bishop, and
the only open file, but a careful study of on the other wing — on squares of the
the position shows that Black has the opposite colour.

181
 Endgame Strategy

6 ... Red8! White’s Q-side pawns, while the king

7 Rcdl Rxd3 takes on the defence of the b5 pawn.
8 Rxd3 c5
9 Ne2 c4! 20 Ne3

A subtle understanding o f the posi­ Not 20 Na7? Kd6 21 Nxb5+ Kc6

tion. Black paralyzes White’s Q-side, 22 Na7+ Kb6 23 Nc8+ Kb7, when
after which he exchanges the second pair the white knight is trapped.
o f rooks. There was no sense in main­
taining the pawn tension on the Q-side, 20 ... Kd6
since the . . . b4 break was not in the 21 Nb4 Nc5
spirit of the position. 22 Kf3 Bh6
23 h4! Nd3
10 R dl Rb7
11 Nb4 Rd7 The end seems to be close. White can­
12 Kel Rxdl+ not take the knight: 24 Nxd3? cxd3
13 K xdl BcS 25 g3 Kc5, and the black king breaks
through to the Q-side pawns after the
First of all White must be deprived o f preliminary exchange on e3.
counter-play involving an attack on the
b5 pawn by Na6—c7. The c7 square 24 N dl Bel
will be guarded by the bishop, while the 25 Kc2!
black king prepares to go to gS via g7
and h6. A brilliant defence.

14 Nc6 Nd7
15 f4

Keres does not wish to await the

squeeze, and attempts to activate his
game on the K-side, which leads to the
creation of weaknesses for both players.

15 . . . f6
16 fxc5

It is unlikely that 16 f5 gxf5 17

exf5 h4 was any better.
It transpires th at 25 . . . Nxb2 26
16 . . . fxe5 Nxb2 Bxb2 27 Kd2 Bxa3 28 Kc2
leads to a very pretty positional draw.
Now White has a weak pawn at e4, White moves his king between c2 and
and Black —at e5. b l. As soon as the black king goes to a5,
he checks with his knight at c6, and if it
17 Ng3 Kf7 goes to c5, he checks at a6. There is
18 Ke2 Ke6 also no way o f breaking through on the
19 N fl Bf8 K-side.- . . . g5 is met by g2—g3. The
pawn ending after the exchange on b4
The bishop is switched to h6 to attack is also drawn, in spite o f Black’s two

182
 Complex Endings

extra pawns! Black is forced to retreat, 41 gxh4+ Kf4!

and the battle flares up anew.
Black has accurately calculated that
25 . . . Nc5 he will be able to stop the h-pawn, while
the loss of the e4 pawn will be fatal for
Attacking the e4 pawn. White.

26 Kf3 gS! 42 h5 Kxe4

43 h6 Nf4+
Forcing White to open up the K-side, 44 K fl Bh4
since 27 g3 is not possible due to 45 Nb4 Bf6
. . . g4+. 46 Kel Kf3!
47 h7 Bg7
27 hxg5 Bxg5 48 Nc2 Nd5
28 Na2
It is time to pick up the h-pawn.
Now after . . . Nd3 and . . . Bel
White is no longer able to set up a 49 Kd2 Nf6
fortress, and so Keres defends the cl $0 Nel+ Ke4
square. 51 Nf2+ Kf5
52 Ng2 Nxh7
28 ... Ke6
29 Nf2 Kf6 Black has finally won a pawn. White’s
30 Ndl Nd3 Q-side pawns are weak, and his second
31 g3 Kg6 weakness is Black’s passed e-pawn. Al­
32 Kg2 though the distance between these
pawns is not great, Black’s advantage is
32 Ke2 fails to 32 . . . Ncl+. sufficient to win.

32 ... Bd2 53 Ne3+ Ke6

33 Kf3 Kg5 54 Ne4 Bh6!
34 Ke2 Bel
35 Kf3 Bd2 It is essential to simplify the position.
The knight ending is won.
Do not hurry!
55 Ke2 Bxe3
36 Ke2 Bel 56 Kxe3 Nf6
37 Kf3 Kf6! 57 Ng5+ Kd5
58 Kf3 Nh5
Now Black pushes back the white king
and breaks through to the e4 pawn. The Aiming for the b2 pawn.
game enters its decisive phase.
59 Ne4 Nf4
60 Nf6+ Kc6
38 Kg2 Kg6!
61 Ke4 Nd3
39 Kf3 Kg5
62 Ng4 Kd6
40 Kg2 h4!
63 Nh6 Nxb2
64 Nf7+ Kc5
Obtaining control of f4. 65 Nxe5 Ndl

tS -N 183
 Endgame Strategy

6 6 Nd7+ Kd6 blow in the centre.

White resigns.
4 ... f5?I
A highly interesting battle between
two outstanding players, in which both
attack and defence were o f very high
quality.

Andersson—Franco

Buenos Aires, 1979

5 Bxb 6 !

The situation on the board has chang­

ed (the e5 pawn has been weakened),
and Andersson immediately changes his
plan, transforming a dynamic spatial
advantage into a static one: a good
knight against a bad bishop.

With the pawn formation symmetric. 5 ... axb 6

White has a slight advantage thanks to 6 Nc4 Bf6
his better placed pieces, and in particular
his bishop. But Black is threatening by After this move White’s idea is com­
. . . Bc5 to equalize completely, and so pletely vindicated. Correct was 7 . . .
Andersson prevents the exchange of fxe4! 8 Rael Rhf8 9 R hfl Bh4! 10
bishops. g3 Bg5 11 Rxe4 (11 a4 Rd3 12 Rxe4
R 8f3) 11 . .. b 5 ! , and White has only a
1 b4! Nb6 minimal advantage in the rook ending
2 Na5 0- 0- 0+ after 12 Ne3 Bxe3 13 Rxe3 RdS,
3 Kc2 Be7 since 12 Rxe5 bxc4 13 RxgS does
4 a3 not achieve anything due to 13 . . .
Rfe 8 . Andersson would obviously have
White has a persistent positional ad­ had to play 10 Rxe4 Rxf2+ 11 Rxf2
vantage, which he intends to increase by Bxf2 12 Rxe5, maintaining a minimal
the advance of his c-pawn. Painstaking advantage.
work is required o f Black to neutralize
White’s initiative. A set-up which de­ 7 a4!
serves consideration is the following:
. . . Kb8 , . . . Nc8 , . . . c6 , and possibly The f-file is blocked, and White can
. . . Rd7, . . . Bd8 and . . . Bc7. Instead delay defending his e-pawn. It is much
Black decides to follow the principle of more important to secure the ‘eternal’
answering a flank attack by a counter­ post for his knight a t c4.

184
 Complex Endings

7 ... Bg7 The co-ordinadon o f Black’s pieces

8 R hel Rne8 has been completely destroyed, ana he is
9 b5 f4 literally hanging on by his last legs. It
only needs one final effort by White to
Now White’s hands are freed in the break Black’s defences, and Andersson
centre, but it is very difficult to suggest easily finds a winning piece $et*up.
anything better for Black.
20 Rd5! Ke8
10 aS! 21 h3!

With the possibility of play on the Q- White intends to transfer his knight to
side and the prospect of an attack on the d3, but does not hurry over carrying this
black king, together with the weakness out, so as not to allow counter-play with
of the eS pawn and the advantage o f . . . c6 .
knight over bishop, White’s position can
be considered won. It is interesting to 21 . . . Ke7
follow the ease with which the Swedish 22 Nb2 Ke8
grandmaster realizes his advantage, in 23 Nd3 Bg7
which one senses his complete mastery
of endgame technique. The last chance was 23 . . . c6 , al­
though after 24 Rdd7 cxb5 25 Nb4
10 ... bxa5 the outcome of the game is not in doubt.
11 Rxa5 b6
12 Ra7 Bf6
24 c4 Bf6
12 Nxb 6 + was threatened. 25 c5! bxcS
26 NxcS Re7
13 R eal Re6 27 Ra6 Bh8
14 R la 6 28 Kc4!

Now threatening 15 Na5 and 16

Nc6 . Even here Andersson does not hurry,
but strengthens his position to the
14 . . . Rde8 maximum.
15 Kb3!

Do not hurry! The opponent must be 28 . . . Bg7

allowed a little play, since the slightest 29 f3! Rb 8
activity on Black’s part will merely
worsen his position.
Black tires o f moving his bishop, and
15 . . . Bd8 the game concludes even more quickly.

Now the eighth rank is blocked.

30 Ne6 Bf6
16 Ra8 + Kd7 31 Rc6 Resigns.
17 Ra2! Bf6
18 Rd2+ Ke7
19 Ra7 Rc8 On 31 . . . Rc8 there follows 32 b 6 .

18$
 Endgame Strategy

Miles—Byme 5 Kt2!

Reykjavik, 1980 Once White has brought his king to e4

and defended the e5 pawn, it will be
possible for him to invade with his rook
via b l. Byrne makes a desperate attempt
to escape from the vice, but in doing so
creates new weaknesses in his position.

5 -. • gS
6 Nh5
Pressure on the e 6 pawn is exchanged
for an attack on the weak squares on the
f-file, which are in the immediate vicini­
ty of the black king.

The advantage is w ith White, who has 6 ... Be7

a spatial superiority and better placed 7 Ke3 Kf7
pieces. Black must simplify the position, 8 R fl+ Kg6
and he has two possibilities: 1 . . . Qa8 9 g4!
and 1 . . . Rd 8 . The American grand­
master chose Now Black has not only to play the
endgame without his king, but he must
1 ... Qa8 ? also constantly watch that he doesn’t
get mated.
which was evidently the decisive mis­
take. In the endgame the white rook be­ 9 ... h6
comes much stronger than its black 10 R bl Bd8
colleague, with good prospects of invad­ 11 Ke4 Bc7
ing along the open b- and d-files, and 12 R fl Bd8
later along the f-file. The black knight 13 Rf3!
has no strong points, and in the absence
of the queens the white king can advance Black is in zugzwang.
into the centre without fear, whereas the
prospects for the black king are consider­ 13 . . . Be7
ably more modest. All these factors indi­
cate that White has a big advantage in 13 . . . Rh8 can be met by 14 Nc6 ,
the ending, whereas after the correct and if 14 . . . Nxc4 IS Rd3 BaS 16
1 . . . Rd 8 ! he would have merely had Rd7.
an insignificant advantage.
14 Rb3 Bd8
2 Qxa8 Rxa8 15 Nc6 Bc7
3 NaS Bf8 16 Rf3 Nd7

Black cannot contest the d-file, since The white rook, by switching from
3 . . . Rd 8 is met by 4 Rbl and 5 Nc6 . the f-file to the b-file and back, has
caused total confusion in the enemy
4 Nf4 Re8 ranks. With difficulty Black has parried

186
 Complex Endings

the immediate threat of an invasion, but Miles—Ljubojevic

his pieces occupy pitiful positions. Now
Miles embarks on a decisive strengthen­ Puerto Madryn, 1980
ing of his position — the advance of his
pawn to a5, after which it will be im­
possible to avert the invasion of the
white rook down the b-file.

17 a4! Rf8

This loses material, but there

thing better.

18 Ne7+ Kh7
19 Rxf8 Nxf8
20 Nf6 + Kg7
21 Ne8 + Kf7 White’s position is better. Black’s
22 Nxc7 Kxe7 extra doubled pawn is of no significance,
23 Nxa6 the active placing of all the white pieces
being more than sufficient compensa­
tion. Black has a difficult exchanging
In the knight ending White is an out­ problem to solve, and has to choose be­
side passed pawn to the good and has a tween 1 . . . exd5, 1 . . . RxdS and
spatial advantage. All that he is required 1 . . . Nd4. The Yugoslav grandmaster
to demonstrate now is elementary chooses the most aggressive and least
technique. successful continuation.

23 . . . Nd7 1 ... Nd4?

24 Nc7
He should have gone into the rook
It was still possible even to lose the ending. Best was 1 . . . exdS, answer­
game after 24 a5?? Kd 8 . ing 2 R lxcó with 2 . . . Rf8 , as
suggested by Miles in Informator No.
24 ... Nb6 30. In this case Black would have been
25 a5 Nxc4 able to cover his main weakness — the
26 a6 Kd7 seventh rank — by . . . Rf7, and White
27 a7 Nb6 could hardly have hoped for more than
28 a 8 >Q Nxa8 a rook ending with four pawns against
29 Nxa8 Kc6 three on the K-side.
30 Kd3 Kd5 One can understand the unwillingness
31 Nb6 + Kxe5 of the Yugoslav grandmaster to play for
32 Ke3 Kd6 a draw without the slightest chance of
33 Ke5 Kc6 anything more. A player normally finds
34 Nc4 Kb5 it difficult to take this kind o f decision.
35 Nd2 Kb4 To do this, apart from an exact appraisal
36 Ke5 c4 of all the details of the position, he
37 Nxc4 Kxc4 needs great confidence in his own
38 Kxe6 Resigns. powers, and, if you like, a certain degree

187
 Endgame Strategy

of courage. After all, if in such a cheer­ 13 . . . Ne7

less, passive struggle he fails to gain a 14 Kg3!
draw, he will be annoyed that he did not
go in for more complicated play, where Black is completely deprived of
there would have been fair practical counter-play, and in such situations, as
chances. But after Black fails to ex­ is well known, it is very useful not to
change the white bishop, a rook ending hurry but to strengthen the position to
with an extra pawn for White on one the maximum.
wing becomes a mere dream.
14 . . . Nd5
2 Bc4! Rd 6 15 f3!
3 Kg2 !
White guards against a check on the
Do not hurry! third rank in the event of the exchange
of the pawns at b3 and g7.
3 ... Kf 8
4 R dl Rb 8 15 . . . a5
5 b3!
After 15 . . . Ne7 Black, who was
It is important for White to retain his already in time trouble, would have had
b-pawn. to reckon seriously with 16 Ra7.

5 ... Rbd 8 16 Bg6 Rb 6

6 Rxa7 Nb5 17 Rf7+ Kg8
7 Rxd 6 Rxd 6 18 Ra7 Kf8
8 Rb7 19 e4!

The position has stabilized. White has To be considered was 19 Rxa5 Rxb3
regained his pawn and obtained a big 20 Ra 8 + Ke7 21 Rg 8 , but after 21
positional advantage. Black’s a-pawn is . . . Kd6 22 Rxg7 Ke5 the threat of
weak, his king is cut off on the back 23 . . . Nf4 would have given Black
rank due to his weakened seventh rank, some counter-play. The move chosen by
and his rook and knight occupy passive the English grandmaster is significantly
positions. White’s first problem is to stronger and more energetic.
create weaknesses in his opponent’s
position on the K-side. 19 . . . Nb4

8 ... Nc3 As shown by Miles, after 19 . . . Ne3

9 Kf3 h6 20 Rf7+ Kg8 21 Rc7 Kf 8 22 Kf2
10 g4! Nd5 N dl+ 23 Ke2 Nb2 24 Rf7+ Kg8 25
11 h4 Nc3 Re7! Rb 8 26 Re8 + Rxe 8 27 Bxe8
12 h5 Nd5 followed by K d2-c2 the knight is
13 Bd3! trapped.

White has fulfilled another task — he

has fixed the g7 pawn and given Black (See next diagram)
a whole complex of weak white squares
on the K-side. He already threatens
Bg6 . 20 e5!!

188
 Complex Endings

32 Be4 Nxb3
33 Rb7 Resigns.

Bogoljubov—Lasker

New York, 1924

Excellent play. White had many

tempting continuations, but analysis by
Miles shows that all the rest were much
weaker:
(a) 20 Rxa5 Nc6 21 Ra8+ Rb8.
(b) 20 Ra8+ Ke7 21 Rg8 Nc6
followed by . .. Ne5.
(c) 20 Rf7+ Kg8 21 Re7 Rb8 22
Rxe6 Nc2! (intending . . . Nd4) 23 Black’s extra pawn is a considerable
Re8+ Rxe8 24 Bxe8 Nd4! 25 Ba4 advantage. But the activity of the white
Kf7, with drawing chances. pieces, the small amount of material left
on the board, and the advantage of the
20 . . . Nd5 bishop over the knight when there is
play on both wings make Black’s task
On 20 . . . fxe5 Miles had prepared of realizing his advantage very difficult.
21 Ra8+ Ke7 22 Rg8 Kf6? 23 Rf8+ By his next move, preventing . . . f5,
Kg5 24 Be4!!, when Black has no de­ Bogoljubov makes it easier for his oppo­
fence against Rf7, Rxg7, Rg6+ and nent to draw up a plan.
Rxh6, since 24 . . . Nd5 is met by
25 Bxd5 and mate at f5 with the rook. 1 g4?!

21 Ra8+ Ke7 Black can hardly hope that the a-pawn

22 Rg8 fxe5 itself will queen. To win he must create a
23 g5! second weakness in White’s position.
After 1 g4 the pawn at h3 is signifi­
The decisive blow. cantly weakened. Black’s problem is now
to provoke a further advance of the
23 ... hxg5 white pawns, then fix and eliminate
24 Rxg7+ Kf8 them. As we will see, it is fairly easy to
25 h6 Ne7 describe Black’s plan, but its implement­
26 Rf7+ Ke8 ation requires Lasker’s supreme techni­
27 Bh5 Kd7 que.
28 h7 Rb8
29 Kg4 Kd6 1 ... Nd2
30 Kxg5 Nf5 2 Rc8?! Kh7
31 Bg6 Nd4 3 Ra8?!

189
 Endgame Strategy

In the endgame it is normally ad­ 10 . . . Nc6

vantageous to place a rook behind an 11 Bc7 Nd4+
enemy passed pawn. But for the time 12 Kg3
being Black has no intention o f advanc­
ing his a-pawn. His aim is to attack Compared with the position after
White’s K-side. Had Bogoljubov guessed White’s sixth move, the only change is
at Lasker’s plan, he would have hardly that his bishop stands at c7 rather than
taken his rook away from the defence of d 8 .
the K-side.
12 . . . Ra3+
3 ... Ra2 13 Kf2 Ra4!
4 Kg2 Nb3+
5 Kg3 Nd4! Now Black starts to take positive
action.
Of course, Black does not play 5 . . .
a5. The scope o f White’s rook would 14 Kg3 Ne6
immediately be widened, and he would 15 Bb6 ?!
gain counter-play, e.g. S . . . aS? 6
Ra7 Kg6 7 h4! f 6 8 h5+ Kh6 9 Bb6 , In the event o f 15 Bd6 White was
with th e threat of 10 Be3+ and 11 h 6 evidently afraid o f 15 . . . aS. But it
(indicated by Alekhine). By the move in would have been much better to permit
the game Black acates the strong posi­ the advance o f the a-pawn that to allow
tional threat of 6 . . . Ne6 followed by his king to be cut off from the g- and fa-
7 . . . g5!, in the event of the bishop pawns. Bogoljubov had obviously been
moving from d8 . The h3 pawn would lulled by Lasker’s preceding series of
then be doomed, so that White’s reply is manoeuvres, and he considered his
practically forced. defensive set-up to be perfectly correct

6 h4 15 . . . Ra3+
16 Kg2?
A serious achievement by Black. It
remains for him to provoke the advance Again a careless move, this time lead­
of the g-pawn, and the game will be ing to a forced loss. Right to the last
practically decided. To this aim Lasker moment Bogoljubov fails to guess at
plans a new regrouping of his pieces. Lasker’s idea. 16 Kf2! was correct, so
But, in order to mask his intentions and as to use the bishop to hinder the black
to achieve the maximum effect, for knight from attacking the g4 pawn, e.g.
a certain time he follows the principle 16 . .. N f 4 17 Bc7 Nd5 18 Bd8 .
“do not hurTy”, dulling his opponent’s
vigilance. 16 . . . Nf4+

The outcome of the game is decided.

6 ... Ra3+
7 Kf2 Nc6 17 Kf2
8 Bc7 Ne7
9 Bd6 Ra2+ White loses after 17 Kgl Rg3+, or
10 Kf3 17 Khl Rh3+. On 17 Kfl Black does
not play 17 . . . Rf3+ 18 Kel Ng2+
10 Kg3 is more exact. 19 Ke2 Nxh4 20 Bf2 Rh3 21 Rxa6 ,

190
 Complex Endings

but simply 17 . . . Ra2!, with a decisive 29 Kf3 f6

positional advantage. 30 Bd6 Rd4
31 Bc7 Rc4
17 . . . Nd3+! 32 Bd6 Rc 6
18 Kg2 33 Bb8 Kh6
White resigns.
Forced. On 18 Kg3 Black wins by
18 . . . Ne5+ 19 KF4 Ng6 + 20 Kg5 We will now analyze a more modem
R f3! 21 Bc7 f 6 + 22 Kh5 Nf4+ 23 variation on the same theme.
Bxf4g6 mate.

18 . . . Nc5! Spassky—Petrosian
19 g5
World Championship Match
Black has earned out his plan. Now Moscow, 1969
comes an energetic elimination of
K-side pawns.

19 • i ■ Ng6 l
2 0 Bf2 Nf4+
2 1 Kh2 Kg6
2 2 Ra7 a5

Only now does Lasker dedde to ad­

vance his a-pawn, to divert White’s forces
from the K-side.

23 Bg3 Ra2+
24 Khl Nh5!
The first game o f the match was ad­
Lasker’s knight manoeuvres in this journed in this position, with Black
ending are above all praise. having to seal his move. In comparison
with the Bogoljubov—Lasker game,
25 Be5 Black’s task looks more difficult. Firstly,
White’s rook is much more active than in
25 Bel loses to 25 . . . R al 26 the previous example, and secondly, the
Ra 6 + (26 Re7 K f5!) 26 . . . Kf5! pawn structure on the K-side is sym­
(stronger than 26 . . . K b7 27 g6+!) metric, which also favours the defender.
27 Rxa5+ Rxa5 28 Bxa5g6! 29 Kg2 The publishing house Fizkultura i Sport
Kg4 30 Bel Ng7 (Euwe). brought out an interesting book on the
match, written by the two players’
seconds, grandmasters Boleslavsky and
25 . . . Ra4! Bondarevsky. In spite o f the fact that
26 Kg2 Rxh4 the grandmasters worked separately on
27 Ra6 + Kxg5 the book, the majority o f the variations
28 Rxa5 Kg6 coincide almost exactly. The book
contains an exhaustive analysis of
The game could have concluded here. the given ending, and we will make use
The finish was: of it here.

191
 Endgame Strategy

1 ... Ne4! “What should White play? After 5

Bd5 b3 6 Bxb3 Nxb3 7 h4 Nd4! 8
The commentators unanimously con­ g5+ Kf5 9 h5 gxhS 10 RxhS Kg4
sider this stronger than the alternatives: 11 Rh8 Rc2+ 12 K fl Nf3 he loses.
(a) 1 . . . Kf6 2 Rd4, when Black We considered the most tenadous to be
achieves nothing by 2 . . . Ke5 3 Rxb4 5 Rf7+ Ke5 6 Rf3, but the rook end­
Ne4 4 Rc4 and 5 Bc2, while after ing after 6 . . . Nxb3 7 Rxb3 Rc4
2 . . . Rb 6 his forces are too passively gives Black good winning chances”,
placed. writes Boleslavsky. But Spassky had
(b) 1 . . . Rc3 2 Rxd 6 Rxb3 3 Kg2 managed to find a brilliant saving man­
leads to a rook ending which Bonda- oeuvre.
revsky judges to be drawn. Accurate play
is demanded of White, it is true. A possi­ 5 Bf7!! b3
ble variation is 3 . . . Rc3 4 Rb 6 b3 6 g5+!
5 g5 Rc2+ 6 Kf3 b2 7 Rb7+ Kf8 8
Ke4 Ke8 9 Ke5 Rf2 10 Kd6 , and “The point o f White’s plan. He gains
White does not allow the black king to time to exchange the K-side pawns, and
break through to the b 2 pawn. for the b-pawn he intends to give up his
bishop” (Bondarevsky).
2 Rd7+
6 ... Kxg5
On 2 Rd4? Black had prepared 2 . . . 7 h4+ Kf6
Rcl+ 3 Kg2 Nc5 4 Rxb4? R bl. 8 h5 Rc2+

2 ... Kf6 ! “On 8 . . . gxhS there follows 9

Bxb3 Kg6 10 Bg8 , while if 8 . . .g5
On 2 . . . Kh6 White would have 9 Bg6b2 10 Rf7+ Ke6 11 Rg7, and
continued 3 Rd4! R cl+ 4 Kg2 NcS there is no win for Black" (Boleslavsky).
5 g5+ Kg7 (5 . . . Kxg5 6 R x b 4 ,o r
5 . . .K b 5 6 BdS) 6 Bd5 b3 7 Rf4 9 Kf3 b2
(Bondarevsky). 10 Ba2 gxhS

3 Rxh7 Rcl+ “If 1 0 __ R cl, then 11 hxg 6 Kxg6

4 Kg2 Nc5 12 Rh2 Na4 13 Rh4 Nc3 ( U . . . ftal
14 Rxa4 Rxa2 15 Rb4 with a draw)
14 Rb4 Rc2 15 Bc4 Rd2 16 Ke3
etc." (Bondarevsky). On 10 . . . g5
there would have followed 11 Rf7+
KeS 12 h 6 .

11 RxhS R cl
12 R h 6 +?

“A blunder. Spassky knew that 12

Ke3 led to a draw, but th e quite under­
standable desire to win the dangerous
pawn as soon as possible suggested to
him a ‘second’ solution. But 12 Rh2
would no t have drawn: 12 . . . Na4

192
 Complex Endings

13 Rh4 Nc3 14 Rh2 (if 14 Rb4 Rc2)

14 . . . R fl+ 15 Kg3 (15 Ke3 R al)
15 . . . Na4 16 Rh4 R al 17 Rxa4
Rxa2 18 Rb4 Ke5" (Bondarevsky).
“The position resembles a study, and
the draw could have been achieved by
12 Ke3 Na4 13 Rh4! Nc3 14 Rb4
Ral 15 Kd3” (Boleslavsky).
These beautiful and instructive vari­
ations show just how complicated a
seemingly simple ending can be.

12 . . . Ke5
3 Kf2!
The king approaches one square
nearer to the b 2 pawn, and this factor 3 c5 looks good, but after 3 . . .
acquires decisive significance. bxc5 4 Rxc5 Rd 8 Black obtains
counter-play.
13 Rb 6 Na4
14 Re6 + 3 ... Nf6
4 Ne5
It transpires that after 14 Rb4 Ral
15 Rxa4 Rxa2 16 Rb4 Kd5 17 Ke3 Preparing Ke3 and preventing
Kc5 and 18 . .. Kc4 Black wins. . . . Ng4+.

14 . . . Kd4! 4 ----------- Rd8

15 Re4+ Kc5 5 Ke3
16 Rxa4 R al
White resigns. After the exchange o f one pair of
rooks the d-file is not o f any great im­
portance.
Smyslov—Karlsson
5 ... Ne4
Las Palmas Interzonal, 1982 6 a5! Nc5

(See next diagram) On 6 . . . Rd2 White has the un­

pleasant 7 c5!
White’s advantage is obvious. Black’s
K-side pawn majority has no significant 7 a6 Kg7
role to play, whereas White’s extra pawn 8 Rc2!
on the Q-side is a very serious factor.
The white knight is excellently deployed White takes control o f the second
in the centre, and Black’s only ‘trum p’ rank and threatens 9 Rd2.
is his possession of the d*file. Therefore
Smyslov first wrests the open file from 8 ... Ne4
his opponent. 9 g 4!

1 Rd3 Rxd3 Smyslov does not miss the chance to

2 Nxd3 Nh5 strengthen his position on the K-side too.

193
 Endgame Strategy

9 ... Kf 6 24 . . . Kg8
10 g5+ Kg7 25 Rf7 Rd3+
11 Nc6 Rd7 26 Nxd3 Resigns.
12 Rg2
On 26 . . . Kxf7 the manoeuvre of
White is not tempted bp the possibil­ the knight to c6 via e5 is decisive.
ity o f 12 cS, winning a piece. After 12 Such games appear simple and clear,
. . . Nxc5 13 Rxc5 bxcS 14 b 6 axb 6 and they are easy to annotate. The only
15 a7 Rxa7 16 Nxa7 h 6 ! 17 h4 difficulty is in finding and making
hxg5 18 hxg5 e5! 19 fxe5 Kf7 Black White’s moves. It is possible that White
has fair counter-play, whereas after the could have won the game differently: his
move in the game he has none at a ll advantage in the initial position was ob­
vious. However, it seems to us that the
12 ... Kf8 path chosen by Smyslov, by which the
13 h4 Ke8 opponent was not allowed the slightest
14 Rh2 Kf8 counter-play, is the most technically
15 h5 gxh5 correct, and corresponds best of all to
16 Rxh5 Kg8 die aims of this book.
17 R hl Kg7
18 Ne5 Rd 8
19 Rh 6 ! Rd 6 Petrosian—Ivkov

Black’s moves are very much forced, Bugojno, 1982

as he is almost in zugzwcmg. On 19 . . .
Nc5 White wins by 20 Rf 6 with the
threat of 21 Rf7+.

20 Ke2!

A move which is highly characteristic

of the Soviet grandmaster. Smyslov does
not trouble himself with calculating the
consequences of 20 Nc6 Nd2, where
Black threatens to capture on c4 with
check.

20 . . . Nc3+
The players have exchanged only the
After 20 . . . Rd2+ 21 Kel Rd 6 the queens, the white-squared bishops, and
move 22 Nc6 wins without any un­ one pair of pawns, but the game has al­
necessary complications. ready gone into an ending.
The pawn chains will immediately tell
21 Kel Rdl+ an experienced player that the opening
22 Kf2 Rd4 was a King’s Indian Defence, in which
23 Kf3 Rd 6 the pawn formation often determines
24 Rf 6 the plans of the two sides. White’s pawn
wedge in the centre creates the precondi­
Further resistance is pointless. The tions for play on the Q-side. By . . . f4
game concluded: Black can set up an analogous wedge on

194
 Complex Endings

the K-side, but it is fairiy obvious that With the disappearance o f the rooks
he will be unable to obtain any serious White's advantage has increased consider­
counter-play on this part of the board. ably, to a great extent due to the differ­
Summing up all that has been said, it is ence in the positions of the kings.
not difficult to decide that White has a Petrosian’s problem now is to weaken
positional advantage, but to convert it the opponent’s Q-side and create the pre­
into a win is no easy matter. Let us see conditions for the approach of his king,
how Petrosian solves this problem. exploiting the remoteness of the enemy
king from the Q-side.
1 Nc3!
7 ... h5
Threatening to capture on a7, which
was not possible immediately due to This attempt to exchange the black-
1 . . . Ra 8 . squared bishop is obviously too late.

1 ... a6 8 Nc4 Nc8

2 Kc2 Kh7 9 Bd2! Bf6

Ivkov plans to exchange off his bad 9 . . . Bh6 would of course have been
bishop by . . . h5 and . . . Bh6 . met by 10 Bb4.

3 R acl Nd7?! 10 NaS b6

11 Nc 6 ! Kg7
Black is inconsistent. He should have 12 3c3 Kf7
continued 3 . . . h5, with chances o f a 13 Nd3 Ke8
successful defence. In vindication o f the 14 a4! Bd8
Yugoslav grandmaster, it should be IS Ndb4 aS
stated that to foresee Petrosian’s sub­ 16 Na2!
sequent manoeuvre was very difficult.
Outwardly the position appears fairly The way for the white king is pre­
blocked, and it is hard to imagine that pared. Petrosian has carried out his plan
on the next move it would already be with precision and consistency, and
too late to play . . . h5. White now has a decisive advantage.

4 N b l!! 16 ... f4
17 Bf2 g5
It is all wonderfully simple. Once this 18 Kd3 Nf8
move has been made, everything be­ 19 h3 Ng6
comes clear. But the plan of transferring 20 Nc3 Bf6
the knight from c3 to c4, at the same 21 Kc4 Kd7
time exchanging both pairs of rooks, 2 2 KbS Kc7
could be found only by a player with a 23 Ka6
complete mastery of endgame technique,
of which Petrosian is undoubtedly one. This concludes the king’s march, and
23 Nb5+ is now threatened. Ivkov
4 ... Rxcl makes a desperate attem pt at counter-
5 R xcl Rc8 play.
6 Na3! Rxcl
7 Bxcl 23 . . . g4!?

195
 Endgame Strategy

24 hxg4 hxg4 white c-pawn advancing Black acquires

25 fxg4 Bh4 an excellent strong point at d4. Black's
26 Ndl! rook, king and bishop also occupy excel­
lent positions. This makes White's plan
This puts an end to Black’s counter* all the more interesting, a plan which he
play. 26 Nb5+ would have been techni­ carries out in full without encountering
cally less correct. any resistance on the part o f the oppo­
nent, on account of its very originality.
26 . . . Bxf2
27 Nxf2 Nce7 1 Kc2 Rd5
2 Be3 g5
On 27 . . . Nh4 White wins by 28 g5 3 h3!
followed by the transfer o f his knight
from f 2 to f3 via h3. White restrains his opponent's activity
on the K-side.
28 Nxe7 Nxe7
29 g5 Ng6 3 ... Rd 8
30 Ng4 Nh4 4 Bh5 Rd5!
31 Ka7 Ng6 5 Be2 Rd 8
32 Nh2 Nh4 6 Kb3 Rc8
33 Nf3 Ng6 7 Bh5
34 b3 Nf8
White’s manoeuvres appear rather ab­
and without waiting for 35 Nh4, Black surd, but, thanks to their very illogical­
resigned. ity, they provoke Black into trying to
seize the initiative.

Svyeshnikov—Browne 7 ... Nd3?

Wijk aan Zee, 1981 Browne fails to guess at his oppo­

nent's plan. 7 . . . Nc4 followed by
. . . Nd6 was correct, maintaining rough­
ly equal chances.

8 R dl Nc5+
9 Kb4! Ne4
10 Ka5! Nd6
11 Kb6 ! Rc6 +
1 2 Ka7
(See next diagram)

We are accustomed to the king being

an active fighting unit in the endgame.
White has th e advantage o f the two But when White's entire army is back in
bishops with a typical Q-sidc pawn camp, and the king alone goes into the
majority. But it is not possible to exploit attack, it must be agreed that this is not
this advantage, since the black knight often seen. Nevertheless, the Soviet
is very strong, and in the event o f the grandmaster’s unusual evaluation o f the

196
 Complex Endings

The correct solution to the exchang­

ing problem. The bishop ending is very
easily won for White.

22 . . . R xdl
23 Bxdl a5

White only has to play b2—b4, and it

will all be over. In the next few moves
Svyeshnikov does not hurry, but seeks
the most favourable moment for the
advance.

24 Kb6 Bc7+
position proves to be correct. The white 25 Kc6 Bf4
bishops now rush to the aid o f their 26 Kb7 Be5
king, and the entire black army is im­ 27 Be2 Bd6
prisoned in its own camp. 28 g3 f5
29 Kc6 Bb8
12 . . . Rc7 30 Bc4 e$
13 Bb6 ! 31 b4!

Panying the threat of 13 . . . Nc8 +, The finish was:

which is decisively met by 14 Kb8 !
31 ... Ba7
13 Rc6 32 Kb7 Bxf2
14 Ba5 Be5 33 bxa5 Bxg3
15 Bf3 Rc5 34 a6 Bf2
16 Bb4 Rc7 35 Be6 ! f4
17 Kb6 Rd7 36 Bd5 h5
18 Rel 37 Bf3 Resigns.

Black is completely tied up, and loss

of material is merely a question of time. Larsen—Maijanovic

18 ... Í6 Bled/Portoroz, 1979

19 a4 Kd8
20 Bxd6 ! Rxd 6 + (See next diagram)
21 Kxb7
It is difficult to imagine that the game
The bold white king was first into the will end in a win for White. If he gives up
attack, and is the first to win spoils for his two knights for a rook and pawn, this
its army. The ending with opposite leads to a drawn rook ending. But Larsen
coloured bishops and rooks is easily won won the game, and, as we will see, this
for White, thanks to his two-pawn major­ was no accident. The ending may seem
ity on the Q-side. tedious to the reader, but it is rather
instructive, and after a study of the
21 . . . Rd2 preceding chapters the turning points are
22 R dl! not so difficult to understand.

197
 Endgame Strategy

series of harmless moves. We beg the

reader, in making White’s moves, not to
fall asleep, since they were made with
the aim of lulling only Marjanovic.

4 ... R bl+
5 Kg2 Rb2
6 Nc4 Rbe2
7 Ncd2 Rd7
8 Nb3 Rde7
9 Ra4 Rb2
10 Nbd2 Reb7
11 h3
1 Ne3 Rbb7?!
This pawn has to be advanced to h5.
As shown by Larsen, annotating this But there are no sharp advances, so as
game in Informator No. 27, 1 not to frighten the opponent.
Rb5! was correct. Marjanovic obviously
reckoned that he would gain a draw as 11 ... R2b4
he pleased, and did not attempt to delve 12 Ra5 R7b5
into the subtleties of the position. In­ 13 Ra3 RdS
deed, there appears to be nothing 14 Ra7 RdbS
threatening Black. If the Yugoslav grand­ IS h4 Rb7
master had tried to find a winning plan 16 Ra 6 R7b6
for his opponent, he would possibly have 17 Ra3 R 6 bS
been able to forestall this plan. 18 Rd3 Kh7

2 Re4 Re7?

Black’s first move was not the best,

although quite reasonable, but his
second is a direct mistake. In order to
obtain winning chances, White has to
weaken his opponent’s position. There
is no way of approaching the f7 and h7
pawns, which only leaves g 6 . The black
pawn which is there is twice defended.
In order to attack it, White must first
provoke the advance o f the black f- and
h-pawns. By 2 . . . Rb5! Black could
have forestalled this plan. 19 Ngl!

3 Rh4! h6 The Danish grandmaster obviously de­

4 Nf3 cided that it was time to take positive
action. Maijanovic does not sense this
Things are going well for White. For moment, and continues carelessly mov­
the success o f his subsequent plan, ing his rooks around the board, while the
Larsen carefully masks it ana, following white knight begins stealing up on the h5
the principle “do not hurry”, makes a square.

198
 Complex Endings

19 • • . Kg7 31 Rd7 R6a7 32 Ne6 + Ke8 33 Nc7

2 0 Ne2 Ra5 Rc 8 34 Ned5, with a decisive ad-
2 1 Nf4 Rba4 vantage.
2 2 Nb3 Ra7
23 Rd5 28 Ne8 + Kf8
29 Nc7! Ra7
Larsen evidently has a good under* 30 Rd 8 + Ke7
standing of his opponent’s psychology, 31 Rg8 ! Rc6
and he decides to advance h4—h5 in die 32 Ncd5+ Kd6
most favourable circumstances. Now 33 hxg6 fxg6
Black could himself have played 23 . . . 34 Nb4 Rb 6
h5, and, in order to pierce Black’s de* 35 Nfd5 Rbb7
fence, White would have had to prepare 36 Rxg6 + Kc5
the advance of his f-pawn. The immedi­ 37 Rxh 6 Rf7
ate 23 h5 came into consideration, 38 Rc6 + Kb5
although then White would have had 39 Rc2 Rad7
certain difficulties to overcome after 40 g* Resigns.
23 . . . Ra3, with the threat of ex­
changing rooks by 24 . . . Rb7. The The deftness with which Larsen
best continuation for White would have directed his cavalry in this ending would
been 24 hxg6 fxg 6 (24 . . . Rb 7 25 have been the envy o f any horseman.
Nc5 Rxd3 26 N fxd3) 25 Ne6 + and
26 Nec5.
Andersson—Stean
23 . . . R4a6?
Sao Paulo, 1979
The decisive mistake. As already
mentioned, Black should have played This game began with a currently
23 . . . h5. popular variation of the English Open­
ing, and almost immediately went into
24 h5! an ending, by-passing the middlegame
stage:
Black’s game is lost. Against the
attack on his K-side he has no defence. 1 c4 Nf6
2 Nf3 c5
24 . . . Kh7 3 g3 d5
25 Nd4 Kg7 4 cxd5 Nxd5
26 Nb5 Rb7 5 Bg2 Nc6
6 d4 cxd4
Black had to reckon with the threat of 7 Nxd4 Ndb4
27 Rd 8 and Nd6 - e 8 +. 8 Nxc6 Qxdl+
9 Kxdl Nxc6
27 Nd6 Rbb 6 P!
(See n ext diagram)
This allows the concluding attack, but
even the better 27 . . . Re7 would not 10 Bxc6 +!?
have saved Black. Larsen gives the vari­
ation 28 Nc4 Rc7 29 Ne3 Rca7 30 In this age of rapid growth of opening
R d 8 R a 8 (Nc4—d6 was threatened) information, it is difficult to guarantee
E S -N
199
 Endgame Strategy

anything, but it would appear that the The exchange of one of the bishops
exchange of the bishop in this position has taken place, but is White’s advantage
was first employed in the game Miles— sufficient for a win? After all, there are
Tukmakov, Las Palmas, 1978, although opposite coloured bishops on the board,
the idea itself is by no means new. White which foreshadow a draw. It is interest­
breaks up his opponent’s Q-side pawns, ing to follow how Andersson realizes his
giving him in return the advantage of the advantage. First he must fix the oppo­
two bishops, taking account of the fact nent’s weaknesses on the Q-side, i.e. not
that the side with the two bishops does allow . . . a5, and this must be done
not have a knight, and that White’s posi­ energetically.
tion is therefore to be preferred. Sub­
sequently White must aim to fix the 19 b4! Rdb 8
opponent’s pawn weaknesses and to
rovoke the exchange of one of the 19 . . . a5 fails to 20 bxa5 and
lack bishops. Black’s task is less spe­ 21 Bb6 .
cific — to organize piece pressure on
White’s position. 20 Bd2! Ke6

10 ... bxc 6 20 . . . a5 again did not work, due to

11 Nc3 e5 21 bxa5 and 22 Re3.

11 . . . g 6 came into consideration. 21 Rc5!

12 Be3 h5 It is useful to provoke . . . Rb5,

13 h4 Bb4 which will allow White to gain a tempo
14 Kd2! by a3—a4.

White allows his king’s rook across to 21 ♦♦♦ Rb5

c l, and completes his development with 2 2 Reel f6
his forces well co-ordinated. 23 a4 Rb7
24 Rc5
14 • »« Ke7
15 R hcl Rd 8 + By energetic play White has achieved
16 K el Be6 a bind over his opponent’s Q-side. Now
17 a3 Bxc3+ he might try pressurizing the a7 pawn,
18 Rxc3 Bd5 say, by transferring his rooks to a5 and

200
 Complex Endings

a 6 , his king to c3, and his bishop to cS. absolutely no need for White to have the
But White’s advantage is not so great option of play on two wings. Therefore
that he can achieve success by such Andersson decides not to try and give his
straightforward play. Black would pro­ opponent a second weakness, and blocks
bably gain counter-play on the K-side. the K-side.
The next stage of White’s plan should be
to neutralize any possible initiative by 30 . . . Rd 6
the opponent on the K-side, and to 31 R bl Rdd7
create weaknesses there if the opportun­
ity should arise. Note how ’gentle’ and Black sticks to waiting tactics, since
‘ingratiating’ Andersson’s moves become. he has no basis at all for active counter-
play.
24 . . . Rg8
25 f3! f5?! 32 Bd4 Kf7
33 Rb2!
The first signs! Soon Black’s other K-
side pawns will also be forced to move The Swedish grandmaster does not
onto white squares. It would have been hurry with the decisive regrouping of his
better to defend against the thrtat of pieces, by which he deploys them in the
26 e4 by 26 . . . Kd6 . most favourable positions. This reduces
to a minimum die probability o f any
26 Kf2 Rd 8 ‘surprise’ on the part o f the opponent at
27 Bc3 e4 the time of the regrouping itself.

There is no other defence against 33 ... Ke6

27 e4. 34 Kd2 Kf7
35 Kc3 Ke6
28 R dl Rdd7 36 Ra5 Kf7
29 Ke3 g6 37 Ra6

One white rook has reached its

appointed place. With the arrival o f the
second rook at a5 the a7 pawn will falL

37 «• • Re7
38 Bc5 Red7
39 Rd2 Rbc7
40 Be 3 Rb7
41 Bc5 Rbc7
42 Rc2!

Right to the end Andersson masks his

intentions. The transit point for the rook
30 f4! is now occupied by the bishop, and the
white king is on the c-file.
The final stage of White’s plan com­
mences. Since, in the event of a con­ 42 . .. Rb7
centrated attack by White, it is unlikely 43 Kb2 Rdc7
that Black can save his a-pawn, there is 44 Ka3 Rd7

201
 Endgame Strategy

45 e3! the weak black d-pawn, and if in doing

so he can retain at least one pair of
White suppresses even the most in­ rooks, the opposite coloured bishops
significant attem pt by his opponent to will not guarantee Black a draw.
gain counter-play. The bishop obtains a
post at d4 and blocks the d-file. In 1 ... Rc7
addition. Black is obliged to forget about 2 b3!
...e 3 .
Yusupov sets a positional trap, into
45 . . . Rdc7 which Razuvayev falls.
46 Bd4 Rd7
47 Rc5 Rbc7 2 ... Rfc 8 ?

A fter this White is able to pick up the This natural move is a positional mis­
a7 pawn ‘at his leisure’. 47 . . . Rdc7 take. 2 . . . b5! was essential.
was slightly better, forcing White to win
the pawn with his king at a3. But 48 3 a4
Rea5 Bc4 49 Rxa7 Rxa7 50 Bxa7
would not have allowed Black any hope The white bishop is now assured o f an
of saving the game. excellent post at c4.

48 Kb2 Rb7 3 ... Kf8

49 Kc3 Rb6 4 Bd5 Ke7
50 Rca5 Rxa6 5 Bc4 Rc5
51 Rxa 6 Ba2 6 Rd3 d5!7
52 Rxc 6 Resigns.
Razuvayev does not wish to be tied to
the defence of his d-pawn, and he de­
Yusupo v—Razuvayev cides to sacrifice it for the sake o f ex­
changing one pair of rooks and activating
Kislovodsk, 1982 his remaining rook.

7 Bxd5 Rcl+
8 Rxcl Rxcl+
9 Kg2 Rc2
10 Bc4 Be5

The position has clarified. White is a

pawn up, but it is difficult to realize his
insignificant material advantage, in view
of the active placing of his opponent’s
pieces and the absence o f any defects in
the opponent’s pawn formation. Yusu­
pov embarks on a lengthy phase of
The previous move saw the conclusion manoeuvring, with the aim of improving
of a mass exchange of pieces, and the the placing of his own pieces and of
game went into an ending which is slight­ creating weaknesses in the opponent’s
ly better for White. He can hope to win position.

202
 Complex Endings

11 Bb5 Bd6 ! White has markedly improved his posi­

tion, and it is natural that Razuvayev is
On 11 . . . Rc7 there could have unhappy fo r the game to continue in the
followed 12 Bc4!, when the black rook same vein. He therefore seizes the oppor­
is cut o ff from the second rank. tunity to set his opponent a difficult ex­
changing problem. It should also be
12 Rd4 a5 taken into account that Yusupov had to
solve it on the last move before the time
All the same Black cannot get by control. (The initial position was reach­
without this move, but now his b6 pawn ed after White’s 21st move.)
is weakened, and in the future he will
have to reckon with an exchange sacri­
fice at b 6 or c5 when the white king is
at b5.

13 Re4+ Kf6
14 Bc4 Rd2
15 Re8 Bc5
16 Kf3

On the Q-side the pawn structure has

stabilized. Now changes can be expected
on the K-side. White is preparing to
attack the f7 pawn with his rook along
the seventh rank, so Black has to take
counter-measures.
In endings with opposite-coloured
16 . . . g5l? bishops plus rooks, exchanging problems
become o f primary importance. In this
Black vacates g6 for his king and example —and it is a fairly typical one —
moves his K-side pawns ortto squares of White had the possibility of winning a
the colour of his bishop. Passive defence, second pawn, exchanging rooks, but the
involving the transfer of his rook onto opposite coloured bishops would have
his second rank, was unpromising. In led to a draw. On the other hand, it
this case White could have prepared the often happens that transposing into a
advance of his king to b5 under the bishop ending with one extra pawn is a
cover of his bishop at d5, supported by sure way to win.
the pawn at e4. In endings with opposite-coloured
bishops without rooks, schematic think­
17 g4 h6 ing comes to the forefront. In the possi­
18 Rc 8 Kg6 ble variation 20 Rc6 + Rd 6 21 Rxd 6 +
19 h3 Bxd6 22 Bd3 Kf 6 White has two
captures on f5: with the bishop or the
pawn. In the event of the capture with
White overlooks his opponent’s the bishop Black’s problem is simplified,
counter-play. As indicated by Yusupov, and consists o f not allowing the oppo­
he should have first checked at g8 . nent to create two connected passed
pawns advancing on black squares. Let
19 . . . f5! us analyze the capture with the pawn,

ts-o 203
 i
Endgame Strategy

23 gxf5. Now it is dangerous to allow attem pt is 5 e5+! KxeS 6 KgS. Black

the white king to reach hS, in view again has to find a particular defensive
of the risk o f ending up in zugrwang. set-up, shown in the next diagram.
Imagine this position.

As soon as the white pawn advances

Black’s king is tied to the defence of to f7, the black king stands at e7, while
his h-pawn, while his bishop has to re­ if the white king approaches the h4
strain the advance of the white pawns in pawn, the black bishop defends it from
the centre. If it is Black to move he has e l. Draw.
to relinquish one of these tasks. Note Let us now return to the game con­
that is bad to take on h4 due to Kg4! tinuation.
Therefore the best reply to 23 gxf5 is
23 . . . h 5 . 20 ♦♦♦ f4
Suppose now that White succeeds in 2 1 Bf5+ Kg7
provoking the advance o f the black pawn 2 2 Rc7+ Kg8
to h4. The following is now a probable 23 Rc6 fxe3
position: 24 fxe3 Rd 6
25 Rc8 + Kg7

The pawn structure on the wings has

become stable, while in the centre White
has acquired a passed e-pawn. If he
should succeed in advancing it to eS,
without allowing the opponent any
serious counter-play, his advantage will
become sufficient for a win.

26 Rc7+ Kf6
27 Rh7 Ke5
28 Rh 8 Rc6
29 Bd3
White can try playing for a win as
follows: 1 f4gxf4 2 e4! Bg3 3 Be2, The white bishop has two excellent
to which Black does best to reply with posts at f5 and c4. Yusupov begins
the counter-sacrifice 3 . . . f3! 4 Bxf3 operating according to the principle “do
Bel. After this White’s only winning not hurry”. In doing so he keeps the

204
 Complex Endings

position in a constant state o f tension, Now 46 e4 is answered by 46 . . .

threatening to advance e3 -e4 —e5 both Ke5, with good counter-play.
with the bishop at f 5, and at c4.
46 Rf8 + Kg7
29 ... Rf6 +! 47 Rg8 + Kf6
30 Ke2 Bf8 48 Re 8 ! Bc3
31 Rh7 Bc5 49 e4!
32 Bc4 Kd6
At last. For twenty-five moves Yusu-
Black has successfully parried White’s pov has been preparing this advance, im­
first onslaught. proving the placing o f his pieces and
wearing down his opponent by constant­
33 Rh 8 Ke7 ly threatening it. White has chosen a very
34 Rh7+ Kd6 apt moment to commence positive
35 Rg7 Kc6 action. The black pieces have lost their
36 Ra7 Kd6 co-ordination, whereas all the white
37 Rh7 Kc6 pieces, including the e-pawn, co-operate
38 Bb5+ Kd5 splendidly with one another.
39 Bd3
49 ... Bd4
White again prepares to switch his 50 Bd5 Kg7
bishop to f5. 51 Kd3 Bc5
52 Rg8 + Kf6
39 . . . Ke5 53 Re8 Kg7
40 Bf5 Rd 6 54 Rc8
41 Rh 8 Rc6
Threatening a possible exchange sacri­
Black must all the time be on the fice on c5. After 54 . . . Kf6 there
alert: 41 . . . Rf 6 42 Re 8 + Kd6 44 e4 follows 55 Kc4, when the capture on
is bad for him. c5 is a real threat to Black.

42 Bd3! 54 . . . Rf6
55 Rc7+ Kf8
The indefatigable bishop again 56 e5
switches to c4.
The actions o f the black pieces have
42 . . . Bb4 finally lost all harmony. The outcome of
43 Re 8 + Re 6 the game is decided.
44 Rb 8 Rd 6
56 ... Rf4
44 . . . Bc5 came into consideration. 57 Be4 Rf7
58 Rc 8 + Ke7
45 Bc4 59 Bf5

The white bishop, like a pendulum,

Threatening the advance o f the e- oscillates between c4 and f5, the latter
pawn.
post being especially good.

45 . . . Kf6 59 . . . h5

205
 Endgame Strategy

This loses, but Black’s position is al­ Material is level and the pawn form­
ready on the way downhill. ation symmetric. The position is roughly
equal, but not drawn! We have already
60 Rc7+! Kf8 seen in numerous examples that main­
61 Rxf7+! taining the balance in a complex ending
against a strong opponent is an extreme­
The correct approach to the exchang­ ly difficult matter. The Swedish grand­
ing problem. In the given situation the master is renowned for his great skill in
transition into the bishop ending is the the playing o f this type of ending. He
quickest way to win. has frequently succeeded in demonstrat­
ing that the concepts o f equality and a
61 . . . Kxf7 draw are by no means synonymous.
62 gxhS Bf8
63 Ke4 Bg7 1 Bd2 0 -0
64 KdS Ke7
65 Bg6 Kd7 The choice o f move in such apparent­
6 6 e6 + Ke7 ly simple positions has to be approached
67 Bf7 with a great degree o f responsibility. As
shown by Kovacevic, annotating this
Black is in an unusual form o f zug- game in Informator No. 32, 1 . . . Bd7
zwang. He is forced to allow the advance was weaker due to 2 Ne4! Bc6 3 Nd6 +
of the h-pawn. Kd7 4 Bxc6 + Kxd 6 (4.. Kxc6 5 Nc4)
5 Bf3 Bxb2 6 R bl Bg7 7 Ke2 . with
67 . . . Bf6 advantage to Whkc. In this variation the
6 8 h6 Bc3 black king may well come under a strong
69 Kc 6 Bd4 attack. Miles correctly removes his king
70 h7 Resigns. from the centre, since 1 . . . Ke7 2
R cl Bd7 3 0—0 Bc6 4 b4 favours
After 70 . . . Kd8 71 Kb7 Ke7 72 White (Kovacevic).
Kc7 it is again zugzwang.
2 R cl Bd7
3 0 -0 Bc6
Andersson—Miles 4 Rc2!

Tilburg, 1981 A strong move. Andersson prepares

for play on the Q-side, involving the ex­
change on c6 and die transfer of his
knight to c5. In this case, in order to in­
crease the pressure on the opponent’s
Q-side, White needs the c-file.
The routine 4 R fdl would have
eased Black’s problems.

4 ... Bxg2

The knight must be moved from d 8 ,

to co-ordinate the rooks.

5 Kxg2 Nc6

206
 Complex Endings

6 Nc4 Rfd 8 11 . . . f5
12 Nc5 Rd 6
To be considered was Makarichev’s
suggestion of 6 . . . a5 with the idea 12 . . . Kf6 was to be considered.
of 7 . . . Nb4. Then 7 a3 could be
met by 7 . . . a4, and if 8 Nd6 Ra 6 13 Rb3 b5
9 Nxb7 Rb 6 . 14 Rd3 Rxd3
15 Nxd3 Rb 6
7 Bc3 Rac8 16 f4
8 Bxg7 Kxg7
9 R fcl Rb 8 White’s initiative has transformed into
10 a3 a stable positional advantage, which he
has consolidated with his last move. But
By threatening to cramp Black on the Black has few real weaknesses, and there
Q-side with b2—b4, Andersson pro­ is very little material left on the board,
vokes a weakening of the opponent’s so that the game is still closer to adraw
pawn formation on this part o f the than to a win for White.
board.
16 . . . Kf6
10 ... a5 17 Kf3 e5!

Not 10 . . . Ne5 11 Rc7 Nd3 12 Not 17 . . . b4 18 a4 followed by

Rl c3 Nxb2 due to 13 Ng5. Rc5, or 17 . . . g 5 18 e4. Miles goes
into a rook ending. It is interesting that
on the previous move Andersson could
have prevented this by 17 Rc5, but
he considered the rook ending to be
favourable. Objectively speaking, 17
Rc5 would seem to b e stronger than
17 Kf3, but the final result o f a game
is often influenced by subjective factors
no less than by objective ones. This is
confirmed once again by the present
game.

18 fxe5+ Nxe5+
19 Nxe5 Kxe5
20 Rc5+ Kd6 ?!
11 Rc3!
It can be assumed that Miles was short
There would not seem to be anything of time. 2 0 . . . Kf6 looks much more
difficult about this move. But this sim­ natural, depriving the white king of the
plicity is merely apparent, and we would possibility of approaching the K-side via
invite the reader to try finding such a f4 and g5.
move. The position has hardly changed,
but it is not at all easy for Black to de­ 21 b4 Rb7
cide what to play. The white rook has 22 h4 a4
gained access to b3, and an attack on 23 Kf4 Ke6
the b7 pawn may become a reality. 24 h5 Kf6

207
 Endgame Strategy 1

he did not like his position. It would

seem that Andersson also investigated
the consequences of 26 h 6 , but he con­
sidered them insufficiently clear and
decided to repeat the position, taking
account of the fact that the opponent
did not have anything better. The two
exclamation marks are attached for
White’s excellent understanding of
psychological subtleties.

26 . . . Rd7?

The decisive mistake. Remember

Byelavyenets: “The repetition of moves
Kovacevic shows that 25 . . . Kg7 in the endgame plays an important role.
was objectively stronger, not allowing Disregarding the fact that it gains time
h 5 -h 6 . As confirmation he gives the for thinking, it can be mentioned that,
following variation: 26 hxg 6 hxg 6 27 by repeating moves, the active side
Kg5 Re7 28 Rxg 6 + Kh7 29 Kxf5 acquires certain psychological gains. The
Rxe3 30 g4 Rxa3 31 Rb 6 Rf3+ 32 defender, whose position is inferior,
Kg5 Rf7! 33 Rxb5 Ra7 with a draw. often cannot stand it, and creates a
The variation is correct, but it should be further weakening which cases his oppo­
added that in time trouble it is unlikely nent’s task. In addition, repeating moves
that anyone would play 25 . . . Kg7. enables the position to be clarified to the
One could say that, when pressed for maximum extent.’*
time, an experienced player's hand
would itself make the king move to f7. 27 Rxb5! Rd3
28 Rb7+ Kf6
26 Rc5!!
28 . . . Kg8 is hopeless: 29 h 6 ! Rxa3
Kovacevic shows th at White does not 30 Rg7+ Kh 8 31 Re7 Kg8 32 Kg5,
win by 26 h 6 . Here are the variations with the threat o f Kf 6 .
given by the Yugoslav grandmaster:
26 h 6 (26 hxg6+ bxg/6 27 Kg5 Re7) 29 Rxh7 g5+
26 . . . Rd7! 27 Rb6 Rd3 28 Rxb5 30 Kf3 Rxa3
(28 Rb7+ Kf6 29 Rxh7g5+ 30 Kf3 31 Ra7 Ra2
Kg6) 28 . . . Rxa3 29 Ra5 Kf6 30 32 Ra6 + Ke5
Ra 6 + Kf7 31 b5 Rb3 32 Rxa4 Rxb5 33 g*
33 Kg5 f4+ 34 Kxf4 Rh5, with a
draw. All these variations demand The rest is not so difficult.
serious consideration, and it is possible
to find them in a calm situation, but not 33 . . . fxg4+
in time trouble. In our opinion, what 34 Kxg4 Rg2+
happened was that, after playing 25 . . . 35 Kh3 Rb2
Kf7, Miles saw the possibility o f 26 h 6 , 36 Rxa4 Ke4
and began feverishly seeking a way out. 37 Kg3 R bl
He obviously overestimated some possi­
bilities on the part o f his opponent, and 37 . . . Kxe3 White wins by

208
 Complex Endings

38 h 6 R b l 39 Ra3+ Ke4 40 Kg4 3 ... Kf8

R gl+ 41 Rg3 (Kovacevic). 4 Nf3 Nxf3+

38 h6 Rgl+ 4 . . . Nb3 was preferable, with a

39 Kf2 R hl complicated game.
40 Ra6 Kd5
41 Kg3 Rh4 5 Bxf3 Nd7
42 Rg6 Ke4 6 Be4 Ke8 ?
43 b5 Kf5
44 Rc6 Re4 This natural move proves to be a mis­
45 Kf2 Re7 take. Black aims to transfer his king to
46 b6 Rb7 c5. If he should succeed in this, he will
47 Kg3 Resigns. have a perfectly reasonable, and perhaps
even more promising, position. But the
plan proves to be impracticable. 6 . . . e5
was better, with chances for both sides.
Vaganian—Rashkovsky

Vaganian begins carrying out a far

In this complicated ending White has from obvious, but highly effective
the advantage. The Mack c4 pawn is counter-plan, involving the advance of
cut o ff from the main chain, and White his K-side pawns. In doing so White had
can quickly bring his king to the centre, to work out the consequences o f sacrific­
ing his passed pawn on the Q-side.
1 ... Rb8
7 ... Kd8
The unpleasant 2 e3 was threatened, 8 h4! Kc7
driving the black knight into the comer. 9 a5!

2 Rxb 8 + Nxb 8 The pawn is sacrificed, so as to gain

3 a4! several tempi and to squeeze the oppo­
nent’s position on the opposite wing.
Black must not be allowed to dislodge
the white knight from its excellent 9 ... Nc5
blockading position by . . . Nb5. 10 Kf2 Nb3

209
 Endgame Strategy i

11 g4 Kd8 Black is a pawn up, with a bishop

against a knight, two passed pawns on
The black king returns to the defence. the c- and g-files, and. . . nevertheless a
White’s threat o f advancing his pawn to lost position.
h 6 and then breaching Black’s defences
by f4—f5 is highly dangerous. 21 . . . g5

12 h5 Ke8 Passive tactics would not have chang­

ed anything. White would have placed
On 12 . . . gxh5 there would have his king at d4 and knight at c3, and by
followed 13 g5. Ne4 or Nb5 would have won the d6
pawn.
13 h 6 ! Nxa5
14 f5 22 Nc3 g4
23 Kf2 Ke8
It is not often in the endgame that 24 Nb5 Kd7
one sees such a furious pawn attack. The 25 e4
double capture on g6 is threatened, and
so Black’s king, which is the sole de­
fender of his K-side, is forced to take
one further step back.

14 ... Kf8
15 g5 Nb3
16 Ke3 Nc5
17 Bc2 Bc8

Black wishes to clarify the position on

the K-side. In the event of 17 . . . Nd7,
with th e aim of preventing f5—f 6 , White
would first have strengthened his posi­
tion by 18 Kd4, and then all the same A pretty zugzwang position. On his
played 19 f 6 , meeting 19 . . . gxf6 next move Black is forced to give up
with 20 Ba4. something.

18 f6 Bh3 25 . . . Kd8
19 Ba4!
Rashkovsky pins his last hopes on his
Vaganian forestalls Black’s threat to passed pawns.
play 19 . . . Nd7 20 Ne4 Bg2!
26 Nxd 6 c3
19 . . . exf6 27 Nxf7+ Kc7
20 gxf6 Nxa4 28 Ne5

Black could hardly have avoided this The black pawns are easily stopped,
exchange. White was threatening 21 Bc6 whereas White’s three passed pawns in
followed by 22 Nb5. the centre are impossible to stop.

21 Nxa4 28 . . . g3+
 Complex Endings

29 Kxg3 c2 Miles plans the set-up: Re2, N el,

f2 -f3 and Kf2, with the aim o f co­
29 . . . Bfl 30 f7 would have wonordinating his K-side pieces. By his con­
quickly for White. trol over c2 White intends to reduce
to the minimum the effect o f the black
30 Nd3 Bfl rooks on the c-file, while by his active
31 Nel Kd7 rook on the a-file he intends to worry
32 e5 Bc4 the opponent and force him to exchange.
33 d 6 Be6
34 Kf4 Kc6 1 ... Rec8
2 Nel Rb4
and Black resigned without waiting for 3 f3 g5
his opponent’s move. White brings up his
king and eliminates the c2 pawn, after Yusupov makes an active attem pt on
which further resistance is pointless. the K-side. The threat is 4 . . . g4.

4 g4! Rb3
5 Kf2 Kg7
Miles—Yusupov 6 h3

V rbu, 1980 Before switching to active play on the

Q-side, the English grandmaster makes
all the useful moves on the K-side.

6 __ hS

Yusupov creates a slight weakness in

his opponent’s position — the pawn at
h3.

7 Ra3! Rxa3

Black cannot avoid the exchange of

rooks. On 7 . . . Rb5 there could have
followed 8 Nd3 with the threat of
The only serious defect in Black’s 9 b4.
position is his compromised central
pawn formation. Were the d 6 pawn at 8 bxa3 Rc3
e 6 , the game would be absolutely level. 9 Ra2 h4
The drawback to Black’s position, which 10 Ke2
amounts to his having one extra pawn
island, is highly insignificant, but the Although White is still engaged in de­
presence on the board of knights gives fence, it is already apparent that Black’s
White the preconditions for exploiting activity has reached an impasse. Very
his advantage. First he has to exchange soon his pieces will be completely
one pair of rooks and suppress Black’s thrown back. Yusupov tries to latch
temporary activity. onto White’s weakness at h3.

1 Re2 10 ... R cl

211
 Endgame Strategy )

11 Nd3! R hl change anything.

12 Rb2!
14 Rxd 6 Ra2+
Miles demonstrates a concrete ap­ 15 Kfl Ral+
proach to the position. He does not try
to defend his h-pawn, since variations After 15 . . . Ra3t White holds onto
indicate that its capture favours White: his pawn by tactical means: 16 Nb4!,
12 . . . Rxh3 13 Nf2! Rg3 14 Rxb 6 and 16 . . . Rxe3 fails to 17 Rxf 6 !
h3 15 Rbl h2 16 a4 Rgl 17 R fl
Ne8 18 a5. 16 N cl Ra3
17 Nc2

It becomes clear that White has

managed to retain his extra pawn.
Black’s game is lost.

17 . . . Ra2
18 Nb4 Ra5

18 . . . Rh2 is no better.

19 Ke2 Rb5
20 Nd3 Ra5
21 R b6 Ra2+
If now Black tries to stop the a-pawn 22 Rb2 R at
with his knight, White gains a decisive 23 Ne5!
advantage on the K-side: 18 . . . Nc7
19 Nhl Rg2+ 20 Rf2 Rgl 21 Rxh2 Aiming at f7.
Ral 22 Rh5 and 23 Ng3.
Therefore 18 . . . f 6 , but then 19 23 ... Ra7
a 6 Nc7 20 a7 Kf7 (20 . . . Kg6 21 24 Rb5 Ra2+
N h l Rg2+ 22 R f2 R gl 23 Rxh2 25 Kd3 Ra7
R a l 24 R h8 Rxa7 25 Ng3, with a 26 Kc3 Rc7+
decisive advantage) 21 N hl Ke7 22 27 Rc5 Rb7
Kf2 R xfl+ 23 K xfl Kd7 24 Kg2 28 Ra5 Kf8
Kc6 25 Ng3 Kb7 26 Nf5 Ne8 27 29 Ra6 Ne8
Ne7 Nc7 28 Kxh2 Kxa7 29 Nc8 +,
and White wins. 29 . . . Kg7 was more tenacious, but
All these complicated variations are all the same Black would be unable to
given by Ugrinovic, annotating the game prevent e3—e4.
in lnformator No. 30.
30 e4! Nc7
12 . . . R al 31 Ra5 f6
32 Nc6 dxc4
A sad necessity. 33 fxe4 Ne6
34 Rf5 Kg7
13 Rxb 6 Rxa3 35 e5!

13 . . . Ra2+ 14 Nb2 does not The endgame play o f grandmaster

212
 Complex Endings

Miles is characterized by unhurried In contrast to the Miles—Yusupov

manoeuvring and the painstaking accu­ game, White, apart from his superior
mulation of small advantages, according pawn formation, also controls the c-file.
to all the demands of the principle “do Black is faced with a difficult defence.
not hurry". But when his advantage
attains decisive dimensions, the English 1 h4 Ke8
player is transformed, and he uses all his 2 Nf4 Kd8
tactical skill to reach his goal by the 3 Bf3?!
shortest path, although quieter, more
lengthy roads might be found. A player An inaccuracy. As shown b y Kasparov,
from the past who acted in this manner annotating this game in Informator No.
was the outstanding Russian Champion 32, he should have played 3 h5, secur­
Alexander Alekhine. ing the post for his knight at f4.

35 . . . fxe5 3 ... Rc7?

36 Rxe5 Nf4
A mistake in reply, possibly caused by
Not 36 . . . Kf7 37 Rxe 6 . time trouble. 3 . . . g5 was correct

37 Rxg5+ Kh8 4 Rxc7 Kxc7

38 d5 Nxh3 5 g4!
39 Rf5! Kg7
40 Nd8 Ra7 It is easier fo r White to exploit his ad­
41 Ne6 + Kg8 vantage in the minor piece ending than
42 g5 with the rooks on.

Black resigned, in view of the vari­ 5 ... g5

ation 42 . . . Rf7 43 Rxf7 Kxf7 44
g6 + Kxg6 45 d 6 . 5 . . . Ne4 does not solve Black’s
problems: 6 Bxe4! dxe4 7 Nh5 g6
8 Nf6 .
Kasparov—Ivanov
6 hxg5 hxg5
Moscow, 1981 7 Nh3! Nh7
8 e3 f6
9 Be2 Bc8 ?

The decisive mistake. As shown by

Kasparov, Black should have tried to
hold the position by 9 . . . Nf8 10 f4
Ne6 .

10 f4!

Now the black knight is tied to the

defence of the g5 pawn, and Black has
no way of preventing the transfer of the
white king to g3 followed by Bd3.

213
 Endgame Strategy

10 • • • Kd8 18 Bb7, winning the d5 pawn.

11 Kf2 Ke7
12 Kg3 Be6 15 Bxh7! Bxh7
13 Bd3 gxf4+ 16 f5! Resigns.
14 exf4 Bg8
In view of the possible variation: 16
On 14 . . . Nf8 White was intending . . . Bg8 17 Nf4 Kf8 18 Kh4 Kg7
15 f5 Bg8 16 Ba6Nd7 17 Nf4 Nb8 19 g5 (Kasparov).

214