WINNING

EN DGAM E STRATEGY
‘9'

I.
-. 214%
ALEXANDER BELIAVSKYAN
' 1' ADRIAN MIKHALCHISHIN '
ll l gs: _
Winning Endgame Strategy

Alexander Beliavsky,
Adrian Mikhalchishin

B.T. Batsford Ltd, London

First published in 2000
© Alexander Beliavsky, Adrian Mikhalchishin 2000

ISBN 0 7134 8446 2

British Library Cataloguing-in-Publication Data.

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

All rights reserved. No part of this book may be

reproduced, by any means, without prior permission
of the publisher.

Printed in Great Britain by

Creative Print and Design (Wales), Ebbw Vale
for the publishers,
B.T. Batsford Ltd,
9 Blenheim Court,
Brewery Road,
London N7 9NT

A member of the Chrjgigalis Group plc

A BATSFORD CHESS BOOK

 Contents

Page

Introduction 5

Pawn Endings 7

Knight Endings 34

Rook Endings 56

Shouldering: the struggle of the kings 109

Complex Endings 112

Defence in the Ending 131

The isolated pawn in the Ending 136

Rook and bishop against rock and knight 141

Rook and two pawns against rook and knight 150

10. Two minor pieces against a rook 153

11. Rook against knight and pawn 172

12. Queen Endings 180

Solutions to Exercises 191

 Introduction
For the authors the most import-
ant thing is—will readers study their
book? It was very pleasant for us
when Winning Endgame Technique
(in fact the first part of the present
book) was deeply studied by grand-
masters Boris Gelfand and Ognjen
Cvitan. Then grandmasters Suat
Atalik and Alex Yermolinsky also
studied the book for a month
(looking for mistakes!) and found
much of interest for themselves.
Particular thanks to grandmaster
Alexei Kuzmin, trainer of the Qatar
team, who discovered many import-
ant and interesting corrections in
pawn endings. We still cannot guar-
antee there are no mistakes—but
then again, generally speaking, any-
one who considers themselves free
from error makes more mistakes
than others.
The present book is a direct con-
tinuation of Winning Endgame
Technique and to some extent the
second part in the sense that many
themes, such as queen, knight and
complex endings, were not dealt
with in the previous book. However
in some cases we have devoted a
little more attention to methods of
play in definite types of position,
rather than concrete cases, which is
a fault, on the whole, of all books on
the endgame. For example the
Encyclopaedia of Chess Endings
gives replies only to some ques-
tions. We have tried to find more
examples in which play conforms
with general principles so that
readers can begin to apply these
methods in concrete practical situ-
ations. Of course, correlation of
methods of play in typical positions
and techniques is not always uni-
form, but the authors present their
vision of practical endgame prob-
lems and are a long way from ex-
hausting this theme.
It remains to thank our friend,
master Oleg Stetsko, for help with
the selection of practical examples.
 1 Pawn Endings

Guru Grigoriev was right! Shirov-Timman

Wijk aan Zee, 1996
Readers might know that the
greatest connoisseur of pawn studies
was the Soviet master Nikolai
Grigoriev, who achieved the im- e;
/7 /’/// ////

possible in his complicated pawn

7&7 7:7
studies. In fact he once said:
“Grandmasters do not like pawn 7/ 74/57
endings because they simply don’t
// a: //
///% / 7
understand them”. The statement
sounds paradoxical but the thought
is topical even today. It seems that
pawn endings are simple but the
/ 7/ 7e 7/

number of mistakes grandmasters

have made over the years has in no
way diminished. Since the publica-
But White does not win after
tion of our first book quite a few in-
l...@d6 2 M @xc6 3 f5! Qd6 4 f6,
teresting endgames have been
when there are two ways to draw:
played and quite a few instructive
mistakes made.
(a) 4...@d7 5 QB @e8 6 @e4
The most characteristic and sur~
prising mistakes are
$07 7 @d5 @e8! 8 $06 (8 @e6
@fS!) 8...@d8, and he does not win
the c7 pawn;
Premature resignation
(b) 4...c6 5 @B @d7 6 @e4
of a game
@dSI, and the king holds the
Yes, even today, this happens at squares corresponding to a5, c5,
grandmaster level. eS-d7; a4, b4, c4, d4, e4-d8, e8;
a6-e6.
In the following position Timman
resigned and both players thought Another surprising resignation
this to be in order, and appropriate followed in a game played two
comments appeared in ChessBase... years later...
8 Pawn Endings

Macieja-Grabarczyk Krumpachnik—Polak
Poland, 1998 Slovenia, 1 985

/// // / /
//@2//&//
xi;
%//;//w
’Il”’//////
’/

/?/ / %
ltnrr

n %/7/%
Although after l...ba+ 2 9x215
c5 3 @b5 @d6 it is asimple draw.
It is also possible to resign be-
cause of an incorrect calculation of
an arising pawn ending.
n a n/
%7/
Here Black resigned, reckoning
that he would lose both of his
doubled pawns, but after 1...Qd7 2
@xes @e7 3 g6 @fB! 4 @xe6 @g7
5 @fs @h6! 6 @f6 it all ends in a
well known stalemate

Svidler-Lobron Natapov—Schuravlov
Erevan, 1996 Moscow, 1 994

//
7/ ///// ///
//;%,f//////// / ///’/
//////
n g n /
Wx/,%/%

Here Black resigned, reckoning Black resigned, forgetting that

on the variation 1...g? 2 2 96
h1=fl 3 Exhl @xhl 4 @d5 @g2 5
Qes, but he did not calculate to the
end. 5...913 6 @f <$e3 and Black
captures the c3 pawn.
there was a last chance of saving the
game by stalemate after l...@c6 2
axb5+ $b 3 @d5 $34!
 Pawn Endings 9

Yudasin-Osnos After an arduous defence and

Leningrad, 1987 transposition to a pawn ending
White had the impression that an in-
teresting chance had cropped up,

WWWWW and he took it.

1 @f5??
/WW//W But he did not reckon on the

W// W;/ ,
simple
l...b4! 2 axb4+
// / Losing simply is 2 a4 @d4 3 §g6
W / W W @c3 and the black pawn is faster

WWW/WW; than the white one.

After playing 1 @fZ Yudasin of-
fered a draw, saying that this well-
known position is in the books and
is drawn! His experienced oppo-
nent, and international master, was
shocked by his own ‘ignorance’ and
accepted it there and then. However
it ought to be well-known to every-
one that after 1...?“ 2 @e2 f4 3
$12 13 4 @fl the triangulation
e5-f5-e4 is winning. It just shows—
don’t believe everybody! !
2...®xb4 3 ®g6 @xbfi 4 ®xh6 a5
5 ($n 34 6 @h6 a3 7 g5 a2 8 g6
al=W 9 g7 yf6+ and he had to
resign.
Correct was the more logical 1
$65 a5 2 @e4 a4 3 bxa4 bxa4 4
@65 $04 5 @e4 @b3 6 @d3 @xa3
7 @c3 @a2 8 @c2 a3 and here the
extra move 9 h3! makes a draw.
Another way to the draw is 5 @d3
QdS 6 h3!.
A lack of understanding of basic
principles of play in pawn endings

Erroneous play in pawn endings

Markovié-Ivanovié
Another type of very common Vrnjacka Banja, I 998
mistake—even strong grandmasters
have many technical shortcomings. W/
Hector-Speelman /”W,
Roskilde, 1998 W
M W W8W£
WWWWWWW;
W
W / WWW
W W8 W/
W: W W;W
/////W/
W W
W
W/ // / //:W
///// Because of his weakened structure
on the queen’s flank (doubled
////W
W /W.£/ pawns) White has the inferior posi-
tion, and the right way now was
10 Pawn Endings

l...hxg4 2 @xg4 @e5 3 @gS d5! 4 Schandorff-Speelman

cxd5 b5! 5 axb5 a4 6 @g6 a3 7 d6! Roskilde, 1998
@xd6 8 @xg7 32 etc. The only
chance of a draw would be the diffi-
Z Z
cult 4 d4+! @xd4 5 @136 @es 6 /Z%
/Z
fiZ/Z// Z
cxd5 @f4 (or 6.. .b5 7 d6! cxd6 8
—) 7 04 @e4 8 @xg7
axb5 a4 9 b6=
@f 9 $f7 QeS 10 @e7, and a
Z
draw. But Black played
l...g5+? 2 fxg6 h4?? // / ZZZ
It was still a draw after 2...hxg4 3
@xg4 9x6
ZEBZ
&Z Z
Z Z/Z
3 $13 xg6 4 @gZ @f6 5 @h3
@gS 6 d4 c6 7 c3 %Z%Z Z
And Black resigned since after
7...d5 winning is 8 c5 b5 9 axb5 a4 1 b4?? @f6 2 @114 g6! and White
10 bxc6, and in the resulting queen resigned since after 3 Q 3 g5 4 f4
ending White has too many pawns. exf3 5 <$xf3 @e5 6 $12 e4 7 @e2
f5! 8 gf @715 thanks to the dis-
tant passed pawn Black wins easily.
Ehlvest-Shirov Correct was 1 @g3! b4 (or 1...aS 2
Vienna, 1996 a4 bxa4 3 bxa4 g6 4 @h3 @f6 5
@h4 $66 6 Q 5 $65 7 @h6 @f6 8
g5+! @f5 9 g7 with a draw) 2
@113 f5 3 gxf5 $f 4 @114! 5+ 5
ZZZ
//// Z Z @g3 $16 6 @h3 @g7 7 @g3 h6 8
/.,// e /// ////,
@h3 @h5 9 @g3 a6 10 @h3 g4+ 11
@g3 @g5 12 @132 @114 13 @112 3+
Z /x 14 fxg3+ @g4 15 @g2 a5 16 f2
/ /Z”Z
&////" ’Z/’//
/‘/
@h3 17 g4!, and a draw. Black
lacks a single tempo move with the

Z Z /aZ a-pawn.
Plaskett-Rowson
Scotland, I 998
A simple position. White only
needs to meet HQdS with @e3 and Z Z / Z
there is no problem. However
Ehlvest played carelessly.
)Z Z
ZZZ:
///
1 @f4?? and after 1...Qe6 2 @e3
@ds 3 <§d3 f4! 4 gxf4 M 5 @e3 h3
6 gxh3 gxh3 7 $13 @xd4 8 @g3
VZ////, , , p
Z M/
@e4 he had to resign.

In the following position we have

//Z ZZ/
another example of an incorrect
pawn move.
 Pawn Endings 11

Black evaluated the position poor- Ostenstad-Kuzmin

ly and sacrificed a pawn. Biel, 1990
1...g4?? 2 fxg4 QgS 3 @f3 e5 4
c4 e4+ 5 @xe4 @xg4 6 b4 $t 7
@f5! @h4 8 CS 216 9 a4 h5 10 b5
21b 11 ab Qg3 12 c6! bxc6 13 /
b6!, and Black resigned. After the %//”/
correct ‘passive’ defence 1.. .Qg7! 2
c4 @h6 3 b4 @t 4 c5 @g6 5 @e5
W
’//% /
h5 6 b5 @f7 7 @d6 g4 he would
easily achieve a draw. But it looked 2/e//%%
all so simple—going directly for the
h5 pawn. / //////
Greenfeld-Golod 1Only one winning move is left:
Israel, I 998 ..Qe3! 2 904 @d2 3 @b5 @c3 4
@xb6 @b4, with a typical finish, but
there followed l...@d5?? and after
/ /
;/ g/V 2 @d3! draw.
i‘e

% %%7/% Ignorance of typical methods of

struggle. Typical breakthroughs.
//’</ 7/
%////%/ Adams-Lutz
Wijk aan Zee, 1995
///%é§//
Correct play was the simple 1.. f.5! / 7/7 4‘
/4/412
2 @xa7 (after 2 gf exf5 3 Qxa7,
simply 3. .g4 4 a4 f4 5 a5 g3) 2.. .e5
3 a4 e4 4 @b6 fxg4 (4. .f4!?—+) 5
a5 g3! 6 fxg3 e3 7 a6 e2 8 a7 el=§
9 a8=y %4+ 10 @c7 gd6+l with
/ // // ////
a transposition to a single, but easily
winning, pawn ending.
In the game Black decided to
25/ ///

"press the white king.

1...Qd6? 2 Qxa7 $c6 but here
followed 3 @bSEI f5 4 a4 fxg4 (or
4...f4 5 a5 @b5 6 QM! @xa5 7
<$06!) 5 a5 @b5 6 @b7! (Réti’ s
typical manoeuvre) 6..$a 7 @c6
@b4 8 @d6 @c4 9 @xe6 @dB with
a draw.
1 h5??
Correct was 1 bxa4 bxa4 2 g4! g6
3 gf gf 4 @d3! h5 5 c3! @05 6
c4 e4+ 7 fxe4 fxe4 8 @xe4 @xc4 9
@e3 @b3 10 @d3 @xa3 11 $03
with a draw.
1...b4! White resigned.
12 Pawn Endings

Overlooking stalemating defences l M?

It is terribly dangerous to move
Hellers-Eingorn the‘wrong’ pawn—necessary was 1
Debrecen, I992 $03 $d5 2 $d3 h5 3 b4! axb4
(3. .a4 4 64+ $d6 5 M 65 6 $63
exf4+ 7 $xf4 $66 8 $63 g5 9 f4=)
/ 4 axb4 65 5 64+ $66 6 $e3 exf4+ 7
/% % % $xf4 5+ (if 7. .h4, then 8 $e3 A
/ 9
ra%az% a a f4) 8 63 $65 (8. .f5 9 f4 g4 10 65)
9 M! $66 10 f4 with a draw.
l...$d5 2 b4?
/%¢ Another incorrect advance—it
was necessary to play 2 64+ $d6 3
/ /% $63! (3 $d4? e5+ 4 fxeS fxe5+ 5
fl %% $d3 $66 6 $63 h5 A g5 —-+) 3.. .65
/ 4 f5! gxf5 5 exf5 $d5 6 $d3 a4 7
bxa4 bxa4 8 h5 h6 9 $c3 e4 10
Here easily winning is l...§xa4 2 fxe4+ $xe4 ll $b4, and, though
flxf6 (2 c6 fxeS 3 c7 §d4+! 4 $c5 Black has an extra pawn in the
§d1)2...§a2 3 c6 En 4 c7 Ec2. queen ending, White has chances of
However Hellers saw that in the a draw.
pawn ending he could force the win 2...axb4 3 axb4
of a pawn and decided that this was
sufficient.
1...§xc5+? 2 $1165 fxe5 3 $d5 /
$g6 4 $xe5 $g5 5 $e4 h5 6 $65 / Z %
$h4! 7 $xf4 Drawn. % %x
/ %‘/
The Black king is stalemated!
/%% (a /
Erroneous pawn advances 9 /
Indeed, this is a great mistake to
/ flgfifi/
make in pawn endings. // % / /
/7/2/;,
Karpov-Kasparov
LaS Palmas, 1996 3...h6?
Now comes a mistake from the
other side—after 3...eS! 4 64+ $66
// 5 6123 6am (5.617? 6 f5!=) 6 f5
2 a a 7/ (if 6 fxeS+, then 6...fxe$ 7 $f2 $66
M9 6 a 8 $g2 $f6 9 $g3 h6, and zug-
zwang) 6...gxf5 7 exf5 $d5 8 $d3
/ // ////// //// h5 9 $63 $c4 10 $64 $xb4 11

a///// (A
/2%///%/
//
/////_$
//// 7///
$d5 $c3 12 $66 b4 13 $xf6 b3 14
$g7 b2 15 f6 b1=w 16 f7 6157 17
$g8 Wxfl, winning.
4 e4+ $d6 5 $63 e5 6 fxe5+ fxe5
76n§w86g1
 Pawn Endings 13

Weak was 8 @g3 @f6 9 §g4 h5+ 6... @d5??

10 @g3 g5, and Black wins, while Correct was to force eZ-e3 by
now on 8...@f6 follows 9 @g3 h5 6...@e4! 7 e3 @d5 8 @e2 @CS! 9
10 f4, and a draw @d2 @d6!, so that on 10 @d3 @d5

Matlak-Tseshkovsky
Lubniewice, I 995 %//////
//////
/ ////////
////// ///”///
///7///£ / %é% ,
% / g 7
V//////// /% / %%/
y/V/y/M/
%/%V// ...it will be White’s move—after
Black moves his position is lost,
since he is forced to allow the White
king to e5, and ...@c5 loses after e4.
1 f4?? 7 @e3 @c4 8 @d2 @d4 9 e3+
Any other move wins—simplest @c4 10 $c2!,
was 1 @h4 @f4 2 @hS @e3 3 @g6 and Black resigned since White
@e 4 f4 @e3 5 f5. achieves the above-mentioned posi-
1...@e4 2 @g4 <m4: tion with Black to move.
White had reckoned only on
2...?63 3 @f5 with a win, but now Drasko-Vratonjié
on 3 @f5 follows 3...?e3!, and after Ulcinj, I 997
White moves this position is drawn!
3 @115 QM!
Again Black will not
3...@e3? because of 4 f5 winning
‘buy’
M y a;
31/9
:///-'/4,
a. A;
for White.
4. /

4 <é’g4 @d4 5 $13 f5 6 $12

/ % y /
as e/% M
/ //
%;7 fl /&
753%39%%3a%3,
// / 7% %
/323/3/3 'If/éfiy
l...f5!
f

////%// A colossal defensive resource,

%///,/§ / though it was possible to hold the

position even by simple defence,
e.g. l...@e6 2 @e4 (the assessment
is not changed by 2 f4 f5 3 g5 h5=)
14 Pawn Endings
.f5+ (the only move) 3 gxf5+
Qd6 4 f6 Q05 5 Qe5 a5 6 a4
Q06!=; or 2 c5 bxc5+ (losing is
2...b5 3 f4 a5 4 h4, 5 h5, 6 5!+—) 3
Qxc5 f5 4 gxf5+ Qxf5 5 b5 Qf4
6 Qa6 Q13 7 Qxa7 Qxf2=.
2 gf f6! 3 214?
White loses the thread of the
game, starting to play on his weak
flank. He should reconcile himself
to a drawn result after 3 h4 Q06 4
h5 Qd6 5 Qd3 Q05 6 Q03 a6 7 a4
35 8 f3=.
3...Qc6 4 f4??
A very serious mistake. Without
need White himself gives up the op-
portunity of exploiting the tempo
and hands back the move to his op-
ponent in a position of mutual zug-
zwang. It was still not too late to
lead the game to a draw by 4 h4 h5
5 Qd3 Q05 6 Q03 a6 6 f3» a5 7 f4.
4...Qd6 5 Qd3 Q05 6 Q03 h5 7
Qb3 h4—+ 8 Q03 h3 9 Qb3 a6 10
35
The assessment of the position is
not changed by 10 Q03 a5— +.
10...ba 11 Q34 Qxc4 12 Qa
Qd4 13 Qxa6 Qe4 14 QbS Qf
White resigned.
Sulipa-Gricak
Lvov, 1995
///// ///,,/
% 7
7 7 ////s//
%%/§%
///"7//
%///a
7 / /7/
A pawn ending which is quite
difficult to evaluate. To gain victory
Black needs to solve the problem on
the queenside, where his opponent
has the chance to create a passed
pawn. This is achieved by 1...Qd7!
2 h4 Q07 3 QB Q08 4 Qe4 Qb7 5
QB a6 -+. However in the game
followed
l...g5? 2 Q13 QdS 3 c6 Qd6 4
Qe4??
The decisive mistake. The signifi-
cance of a tempo in pawn endings is
far higher than in any other. White
should immediately exploit his op-
ponent’s mistake and himself break
up Black’s pawn chain. This is done
by 4 h4! gxh4 5 Qg4 a6 6 bxa6
Qxc6 7 Qxh4 Qb6 8 Qg4 Qxa6 9
M +-.
4. .a6 5 bxa6 Qxc6 6 Q13 Qb6 7
M gxh4 8 Qg4 Qxa6 9 Qxh4 Qb6
10 Qg4 Q06 11 M Qd6 White
resigned.
As a matter of fact, in their first
book, the authors made several inac-
curacies, which were corrected by
grandmaster Alexei Kuzmin.
Kupreichik-Mikhalchishin
Lvov, I 988
//, 7%
x%//%///// 7/// ////
77% // ////
g/ 7”?
/ 7 7 7
fl / fl 7%”
 Pawn Endings 15

1...Qd5
In their first book the authors
placed a question mark against this // ///y //
//
move, pointing out a ‘direct’ path to

\‘p

Ti}
a draw: 1.. Qf6 2 Qf2 Qg6 3 Qf3
h5 4 gxh5+ Qt 5 Qe4 Qg4 6 f5 ////////

\
Qg5 7 Qxd4 Qf 8 Qc5 Qe5 9 %,/ / §/%,
Qb6 Qd6 etc. However they did not

reckon on one finesse to which
A.Kuzmin drew attention: 3 f5+! (3
Qe2 h5 4 f5+ Qh6=) 3...Qg5? (3
Qf6, returning to the basic vari-
ation) 4 Qe2 h5 5 f6! Qxf6 6 gxh5
Qg5 7 Qd3 Qxh5 8 Qxd4 Qg5 9
Qc5 Qf6 10 Qb6 Qe7 11 Qxa6
Qd7 12 Qb7+—. Thus Black’s at-
tempt to simplify the position at
once is mistaken.
2 Qf2 Qe4
A dubious move. Better, missed
by the authors, is 2. a.5! 3 QB (or 3
Qe2 Qe4 4 f5 Q65) 3. .Qc4 4 Qe2
Qc5 (but not 4.. .Qd5? 5 Qd3 Qc5 6
g5 hxg5 7 fxg5 Qd5 8 g6+-) 5 Qd3
(5 Qd2 Qd6 6 Qc2 Qe6 7 Qb3
Qd5=) 5.. .Qd5 6 f5 Qe5 7 Qc4
Qe48f6d3 9Qc3 Qe310f7d211
f8=y d1=w 12 t6+ QB, and in
the resulting queen ending Black
achieves a draw without trouble.
3 f5 Qe5 4 Qe2?
Much stronger is 4 a5, but the
authors mistakenly assumed that
this was easily winning, giving the
variation 4.. Qd5 (4.. .h5? 5 gxh5
Qf 6 Qe2 Qg5 7 Qd3 A
\\\\\\\
/"'/¢//// //%
In this queen ending White has
some chances of a win but upon
correct defence it should probably
be drawn. For example: 13 g5 wel+
l4 Qc4 Qg4, and it is not easy for
White to improve his position.
4...h5?
Better 1S 4. a.5! (A 5 Qd3 Qd5 6
f6 Qe6) 5 QB Qd5 6 Qf4 QC4 7 f6
(13 8 f7 d2 9 f8=y dl=fi=.
5 gxh5 Qf 6 Qd3 Qg5 7 Qxd4
Qt 8 Qc5 Black resigned.
Now it is clear that best was an
immediate l...a5! 2 QB Qd5 3 QB
Qc4 4 Qe2 Qc3 5 l Qc4 6 Qd2
Qc5 7 Qc2 Qc4 8 f5 Qd5 with a
draw, since on 1...Qd5 could follow
2 a5! (the same as on 1...Qf6 2 Qf2!
Qg6 3 15+ Qf6!—4 a5!).
Vaganian-Portisch
Tilburg, 1992

/// /t/,’”
Qxd4--c5---b6xa6-b7+-) 5 Qe2
Qe5(?) 6 Qd3 Qd5 7 f6 Qe6 8
Qxd4 Qxf6 9 Q05 Qg5 10 Qb6
n4 11 Qxa6 115 12 Qb6 114 13 a6 8/ at; a /:
etc. Far more tenacious in the opin- //////////
ion of A.Kuzmin was 5.. Qd6 6 ///”V
Qd3 6 Qd2 Qd5 7 Qc2 Qd6 8
//////2
Qb3 d5=) 6...Qe5 7 Qc4 Qe4 8
f6 d3 9 Qc3 Qe3 10 f7 d211f8=§
dl=fl l2 yxh6+ Qf3.
///////
14 Pawn Endings
..f5+ (the only move) 3 gxf5+
Qd6 4 f6 Q05 5 Qe5 a5 6 a4
Qc6!=; or 2 c5 bxc5+ (losing is
2. .,b53f4a54h4 5h5, 6 5!+—)3
Qxc5 f5 4 gxf5+ Qxf5 5 b5 Qf4
6 Qa6 QB 7 Qxa7 Qxf2=.
2 gf f6! 3 a4?
White loses the thread of the
game, starting to play on his weak
flank. He should reconcile himself
to a drawn result after 3 M Q06 4
h5 Qd6 5 Qd3 Q05 6 Q03 a6 7 a4
35 8 B=.
3...Qc6 4 f4??
A very serious mistake. Without
need White himself gives up the op-
portunity of exploiting the tempo
and hands back the move to his op-
ponent in a position of mutual zug-
zwang. It was still not too late to
lead the game to a draw by 4 h4 h5
5 Qd3 Q05 6 Q03 a6 6 B a5 7 f4.
4...Qd6 5 Qd3 Q05 6 Qc3 h5 7
Qb3 h4—+ 8 Qc3 h3 9 Qb3 a6 10
35
The assessment of the position is
not changed by 10 Q03 a5— -l-.
10...ba 11 Q34 Qxc4 12 Qa
Qd4 l3 Qxa6 Qe4 14 Qb5 Qxf5
White resigned.
Sulipa—Gricak
Lvov, 1995
/ //%%// 7
%//Q//
at
4?
72%?
77/
A pawn ending which is quite
difficult to evaluate. To gain victory
Black needs to solve the problem on
the queenside, where his opponent
has the chance to create a passed
pawn. This is achieved by 1...Qd7!
2 h4 Q07 3 QB Q08 4 Qe4 Qb7 5
QB a6 -+. However in the game
followed
l...g5? 2 QB QdS 3 c6 Qd6 4
Qe4??
The decisive mistake. The signifi-
cance of a tempo in pawn endings is
far higher than in any other. White
should immediately exploit his op-
ponent’s mistake and himself break
up Black’s pawn chain. This 1s done
by 4 M! gxh4 5 Qg4 a6 6 bxa6
Qxc6 7 Qxh4 Qb6 8 Qg4 Qxa6 9
M +-.
4. .a6 5 bxa6 Qxc6 6 QB Qb6 7
M gxh4 8 Qg4 Qxa6 9 Qxh4 Qb6
10 Qg4 Q06 11 M Qd6 White
resigned.
As a matter of fact, in their first
book, the authors made several inac-
curacies, which were corrected by
grandmaster Alexei Kuzmin.
Kupreichik-Mikhalchishin
Lvov, I 988
%//
7/2/22y
//%/////
M a
/7 /
fl///%//.

1...Qd5
In their first book the authors
placed a question mark against this
move, pointing out a ‘direct’ path to
adraw: 1.. .n62Qf2Q63Qf3
h5 4 gxh5+ Qt 5 Qe4 Qg4 6 f5
Qg5 7 Qxd4 Qf 8 Qc5 Qe5 9
Qb6 Qd6 etc. However they did not
reckon on one finesse to which
A.Kuzmin drew attention: 3 f5+! (3
Qe2 h5 4 f5+ Qh6=) 3...Qg5? (3
Qf6, returning to the basic vari-
ation) 4 Qe2 h5 5 f6! Qxf6 6 gxh5
Qg5 7 Qd3 Qxh5 8 Qxd4 Qg5 9
Q05 Qf6 10 Qb6 Qe7 11 Qxa6
Qd7 12 Qb7+—. Thus Black’s at-
tempt to simplify the position at
once is mistaken.
2 Qf2 Qe4
A dubious move. Better, missed
by the authors, is 2..a5! 3 QB (or 3
Qe2 Qe4 4 f5 Q65) 3.. .Qc4 4 Qe2
Qc5 (but not 4. .Qd5? 5 Qd3 Qc5 6
g5 hn 7 fn Qd5 8 g6+-) 5 Qd3
(5 Qd2 Qd6 6 Qc2 n6 7 Qb3
Qd5=) 5.. Qd5 6 f5 Qe5 7 Qc4
Qe48f6d39Qc3Qe310f7d211
f8=y d1=w 12 t6+ Qf3, and 1n
the resulting queen ending Black
achieves a draw without trouble.
3 f5 Qe5 4 Qe2?
Much stronger is 4 a5, but the
authors mistakenly assumed that
this was easily winning, giving the
variation 4. Qd5 (4.. .h5? 5 gxh5
Qf 6 Qe2 Qg5 7 Qd3 A
Qxd4--c5---b6xa6-b7+-) 5 Qe2
Qe5(?) 6 Qd3 Qd5 7 f6 Qe6 8
Qxd4 Qxf6 9 Q05 Qg5 10 Qb6
n4 11 Qxa6 115 12 Qb6 114 13 a6
etc Far more tenacious in the opin-
ion of A.Kuzmin was 5.. Qd6 6
Qd3 6 Qd2 Qd5 7 @122 Qd6 8
Qb3 d5==) 6...Qe5 7 Qc4 Q64 8
f6 d3 9 Qc3 Qe310 f7 d211f8=W
dl=fl l2 yxh6+ Qf3.
Pawn Endings 15
////fl
£2 / 77/
7/ 77$
//M7/ ////
/ //7
In this queen ending White has
some chances of a win but upon
correct defence it should probably
be drawn. For example: 13 g5 @eH
l4 Qc4 Qg4, and it is not easy for
White to improve his position.
4...h5?
Better Is 4. .a5! (A 5 Qd3 Qd5 6
f6 Qe6) 5 QB Qd5 6 Qf4 Qc4 7 f6
(13 8 f7 d2 9 f8=y dl=fi=.
5 gt Qf 6 Qd3 Qg5 7 Qxd4
Qt 8 Qc5 Black resigned.
Now it is clear that best was an
immediate l...a5! 2 Qt2 Qd5 3 QB
Qc4 4 Qe2 Qc3 5 l Qc4 6 Qd2
Qc5 7 Qc2 Qc4 8 f5 Qd5 with a
draw, since on 1...Qd5 could follow
2 a5! (the same as on 1...Qf6 2 Q3!
Qg6 3 f5+ Qf6!—4 a5!).
Vaganian-Portisch
Tilburg, 1 992
7/7,
V/i
///'/ 7
7%7;7 7
7/7 7%
/77
I 6 Pawn Endings
A.Kuzmin also drew attention to
this pawn ending. We present his
more accurate analysis.
“The authors assessed this ending
in White’s favour, giving 1 $d2
$05(?-A.K.) 2 $03 g5 (2...$b5 3
g3g64h3h55g4h46g5$057
a3 $b5 8 $d4 $a4 9 $xc4 $xa3
10 f5 35 11 fxg6 fxg6 12 e5+-) 3
g3 (P-A...K)3. .g44a3 h5 5 a4 a5 6
e5 $d5 7 f5! $xe5 8 $xc4 $f 9
$b5 +—~. But the variation leaves a
strange impression. Firstly let’s look
at the final position.
/¢
////
/ $g2 Black has the better queen
////.
// //
After 9...$e6 (9...$g6!?) 10
$a f5 it is not Black, but White
who needs to think about saving the
game: 11 $b6 (11 $b4 f4 12 a5
fxg3 13 hxg3 $d6—+; 11 $b5 f4 12
gxf4 M 13 a5 g3 14 hxg3 h3!-+)
11.. f.4 12 a5 fxg3 (12.. .f3!?) 13
hxg3 h4 14 a6 hxg3 15 a7 g2 16
a8=§ g1=w+ with a theoretically
won queen ending.
Secondly, after 2. .g5 White can
also win directly by 3 fxg5 hxg5 4
a3 f6 5 a4 a5 6 h3 $d6 7 $xc4 $e5
8 $b5 $xe4 9 $a.
Thirdly, on 1 $d2 correct is an
immediate 1...g5 A 2 g3 g4 3 $03
$054a3h55a4a56e5$d57
$d2 (of course White is not obliged
to give u the pawn by the move 7
f5?) 7... d4 8 $02, sailing into a
drawn harbour.”
It remains only to see how the
game ended (comments by the
authors of the book).
1 h4? $c5?
After 1.. .h5! White 18 in no posi-
tion to create a passed awn: 2 $d2
$05 3 $03 $b5 4 a3 05 5 a4 35 6
g3 g6 7 e5 $d5 8 $d2 $d4 9 $02
C3 10 $d1 $d3-+.
2 h5 f6?
Making it easy for his op onent to
create a passed pawn: 2... d4 3 e5
$03 4 f5 $b2 5 e6 fxe6 6 fxe6 c3 7
e7 02 8 e8=fl CIT-w.
3 $e3 35?
Once again after 3...c3 4 $d3 02
5 $x02 $d4 6 $b3 $xe4 7 $34
$xf4 8 $a5 $g3 9 $xa6 f5 10 a4
ending.
4 a3 a4 5 e5! fxe5 6 fxe5 $d5 7
e6 $xa6 8 $d4 Black resigned.
Poor technique in transposing
to a pawn ending
It is surprising but true that mod-
em grandmasters experience great
difficulties in transposing to a pawn
ending. There are two factors—a
poor knowledge of pawn endings
and inability to switch from the fac-
tors of evaluation and play of com-
plicated endings to the factors and
methods typical for pawn endings.
This is a very great problem even
for high-level grandmasters and
here the authors simply must also
name themselves.
 Pawn Endings I 7

Beliavsky—Sveshnikov Black should correctly transfer to

Novi Sad, 1979 a pawn ending and this is achieved
by l...§a4! 2 a3, and the position of
/ the pawn on a3 gives Black the

22/ 2/7 2/
possibility of more quickly creating
a passed pawn. But in the game
2/ 2//22 there followed...

2 2/ 2 l...§xd4+ 2 §d3 Exd3+ 3 Qxd3

f6 4 Qc4 a6 5 f5!
7 [y 2" . // ””1 It is necessary to clear a path for
the king, otherwise the march of his
h2 pawn will lead to nothing.
5...Qd6 6 fxe6 Qxe6 7 Qd4 Qf5
2 / 22 8 Qd5 Qg5 9 Qe6! f5 10 e3 Qg4
ll h3+! QgS 12 h4+! Qg6 13 h5+
QgS l4 h6 Qg6 15 M Qxh7 16
How is it possible to lose here? Qf Qg7 l7 Qe6 Black resigned.
very simply—by transposing to a
pawn ending.
1 @e2?? 1 QB, with a draw. Adams-Lautier
1...Qc3! Takin on e2 indeed leads Tilburg, I 996
to a draw. 2 f1 yxe2+ 3 Qe
Qc2, and he had to resign since the
king goes over to the white pawn on
g3 while White’s during this time
goes to g6 after which Black wins
2% / /2 2/
by the well-known zugzwang.... 72:2 /Q
2222/4
Qg4!. .72 //
8/8. ////Q//
Kramnik-Lautier 282/ //// //
2
/&

//
%

Belgrade, 1995 ///// /

2/
22 @2 2 White has an extra pawn and can
a 22/2 win the position as he pleases. As he
pleases? In the game followed...
1 h4? §e6+!, and the pawn end-
2 22
23/7 H/ ing with an extra pawn was drawn
2 fieS Qf6 3 Exe6+ Qxe6 4 Qd4
2 2E
,..,//// 1‘3
a/Z"&//
Qd6 5 Qc3 Qc7 6 b4 cxb4+ 7
Qxb4 Qc6, and a draw.
Correct was 1 CS! or 1 fibS.
18 Pawn Endings

Oll-Benjamin 2 @xc6 ficB 3 EcS?

New York, 1995 Both players “drift” commented
A.Alekhine. After 3 Qe2 Exc6 4
fixc6 bxc6 5 b4! Qe7 6 Qd3 Qd6 7
Qd4 White has a typically winning
//% position.
///// 3...bxc6?
But here Black does not exploit
//////// his chance—better was 3... fixc6 4
27/ Exc6 (after 4 fid §c2 and Black’s
a/ active rook compensates for the loss
of a pawn) 4...bxc6 5 b4 Qe7 6 Qe2
//6 Qd6 7 Qd3 c5 8 bxc5+ Qc 9
7 Qc3 a5 reaching an equal position.
4 Qe2 Qe7 5 Qd3 Qd6 6 fiaS
Neither side can lose the rook gas 7 Qd4 f5 8 b4 abs 9 a3 :38
ending. But White decides to try to
win the pawn ending.
//7%

fi/
1 Ed4? fixd4 2 cxd4 Qg6 3 QB
Qf5 4 Qe3 Qe6 5 Qe4 f5+ 6 Qf4 / / ///,8
g5+ 7 Qe3 QdS 8 f3 Qc4 9 b3+
QdS! 10 h3 h5, and he had to
1//x/”’
resign since on 11 Qd3 follows ,[ffffyig/////,7,14%
11...g4. //
However, we should not think that
/41?/
classical players from the past /// / ///23
/// /&?
V//
handled analogous situations better.

Flohr-Vidmar 10 e4!
Nottingham, 1 93 6 A seemingly illogical move but
Black has only one weakness on a6
and White exchanges his weak e3

r§ na
4/95
// a ,,
pawn, activates his king and rook
along the fifth rank and then sets
about creating weaknesses for the
opponent on the kingside.
10...fxe4 11 fxe4 dxe4 12 Qxe4
/ 337 13 Qf4 h6 14 M! Q96 15 Qg4
£218 16 h5 g5
Or l6...gxh5+ 17 Qt EgS 18
// g4+-. I
17 g3 337 18 Qf3
Now the king transfers to the
1...5Dc6? other flank.
After the natural l...Qe7 Black 18...§a8 19 Qe4 3217 20 Ee5+l
has every chance of holding the Here Black has a choice: to allow
slightly inferior isolated pawn. the rook to e8 or the king to f5.
 Pawn Endings 19

20...?d6 21 fies c5 22 §d8+ Again typical—only this wins.

c6 19 @b6 h2 20 c7 @d7 White
After 22. ”@c7 23 §h8 it is time resigned.
to resign.
23 §c8+ @b6 24 fic Black Anastasian-Romanishin
resigned. Moscow, 1994

We also certainly come across

transitions which are backed up by
splendid calculation.
Cruz-Seirawan
//
/,//%
Moscow, 1 994
/®% Zléfl//
/// 222 / /s //
// 1% /////z§// /
/”/4V/%
////// /% l...¢3c3! 2 £xc3 dxc3 3 a4 QM
/
//
// 22 4 35 @1135 5 @xc3 @b5 6 @d3
@b4!
Black meets the white king’s
roundabout route with one of his
own!
7e3
l...§aS+! Or 7 @e4 @03 8 @xeS @d2,
The best solution, after l...§h5 2 catching up with the white pawns.
@b4 Exh2 3 c5 and White has 7...?b3 8 exf4 exf4 9 <§e4 @c2
strong compensation for the pawn. 10 @f5 @d3 11 @n $93 12 @h4
2 @b3 3x32 3 @1132 @d6 4 @b3 @xf3 l3 g5 @e2 Drawn.
@c5 5 ®c3 e5
At first sight it seems that White Kuzmin-Petrosian
has the advantage because of his Moscow, 1 979
passed pawn, but the most important
factor in this ending is the paralysed
white pawn chain on the kingside.
6 @d3 f5 7 @c3 94 8 @b3 M! 9
7/7///
//%s
91:3 h5 10 @b3 f4!
t///
Now follows a typical, well-
calculated breakthrough. ’/ // /
//%//
ll gxf4 e3! 12 fxe3 h4 13 f5
@d6!
//”//
It is this which had to be foreseen
--bad would have been 13.. .g3? 14 ///////
hxg3 hxg3 15 f6 @d6 16 c5+
l4 QM @235 15 c5 @f l6 c6
/7
@e6 17 @cs g3 18 hxg3 h3! 1 c5!
20 Pawn Endings
The transfer to a pawn endin is
forced, since after l...bxc5 2 d3
and 3 Qc4 Black stands badly.
1.":c 2 Ec bc 3 Qd3 e5 4
Qc4 f5 5 Qxc5 h5 6 b4 axb4 7
Qxb4 f4 8 a5 e4 9 Qc3 e3 10 Qd3!
Black resigned.
An analogous idea was not taken
into account by Black when trans-
posing to a pawn ending in the
following game...
Finkel-Mikhalchishin
Belgrade, 1998
%,/ / //
//////%/
///x/
//:22 2//
9/ //,’//
/// //////
”/
% ,,,,, %////
Winning easily is 1...Qe5! 2 §c3
§c7 and 3...Qd5, but Black wants to
take the bull by the horns at once by
l...Qd5??
White is frightened by the pawn
ending and after...
2 Qf4 §c7
...he could quietly resign.
Meanwhile after 2 §d1+ Q06 3
Exd7 Qxd7 4 Qd3! Qc6 (4...Qe6 5
Qd4 changes practically nothing) 5
Qd4 a6 (after 5...f4 6 Qe4 Qc 7
Qxf4 Qb4 8 Qe5 Qa3 9 Qf6 Qxa2
10 Qg7 l 11 h4! a5 12 h5 an
endgame with an extra pawn for
White is reached. Therefore Black
rightly plays 7...Qd5 8 Qe3 Qe5 9
h4! a6 10 a3, with a draw) 6 a3!
(bad is 6 Qc4 f4!, and the white
king does not get to g7, while after a
capture on a3 the king on c3 can do
nothing since Black will have a
tempo after ...a5; h3 a4; h4 Qa2;
Qc2 a3!) 6...a5 7 Qc4 f4 8 Qd4 f3 9
Qe3 Qxc5 10 Qxf3 Qc4.
Ivanchuk-Kasimdzhanov
Elista, 1998
/
fr
gun/V/V/‘a(tux ’/
/
72
Here 1 Ed7 is quite simply win-
ning. But Ivanchuk was reckoning
on the pawn ending.
1 Exe6 yxe6 2 yxe6 fxe6 3
Exf8+ Qxf8 4 Qg2 and all of a
sudden
4...aS!!
The only move—-—bad was 4...Qe7
5 QB Qd7 6 Qf4 Qc6 7 Qg5 Qd5
8 Qf6 a5 because of 9 b3! with a
win for White.
5 Q13 Qf7 6 Qe4
Nothing is gained by 6 Qf4:
6...a4! 7 g4 hxg4 8 n4 Qg8!!,
creating the distant opposition. And
if 6 b3, then 6...g5 7 Qe3 Qg6 8
Qe4 Qh6 9 Qd4 Qg6 10 Qc5 Qf5
11 hn nS, with a draw.
6...g5 7 Qd4 Qg6!
We must mention the erroneous-
ness of the exchange on h4, e. g.
7.. .gxh4? 8 gxh4 Qg6 9 Qe4 a4 10
Qf4 Qh6 11 Qe3 Qg6 12 Qe4!,
triangulating, and Black IS forced to
move away his king to h6, which
gives White the opportunity of
entering via f4.

8 @CS gxh4 9 gxh4 @f5 10 @b5
@xeS 11 @a @d4 and he had to
agree a draw.
Pawn endings in the creative work
of Robert Fischer
Each of the great champions has,
besides his own style, also his own
methods of play in the various
stages of the chess game.
Thus, upon his ‘taste’ depends the
arising various structures and ac-
cording to his ‘taste’ his treatment
of them. The transfer to a pawn end-
ing is one of the instruments for
realisation of an advantage or a
method of defence. In Fischer’s case
this ending is met more frequently
than any other champion, and the
authors were interested in how the
great Fischer played them at these
moments. The examples show quite
a broad spectrum of quality of play.
Fischer-Letelier
Mar del Plata, 1 959
W W W /
AW W WIW
Wt W / W
W/W9W
W&&W W W
W W% / W
White clearly has the better king
but the asymmetrical structure gives
Black the possibility to reply to the
creation of a passed pawn on the
king’s flank with the creation of a
passed pawn of his own on the op-
Posite side. Fischer continued to
play for a win.
Pawn Endings 21
1 a4?
In principle, a serious mistake,
though White’s plan is understan-
dable———he places a pawn on a5 and,
by sacrificing a pawn on the king-
side, breaks through with his king to
the pawn on a6. However it is not
possible to win this position.
E.Mednis in his book How to beat
Bobby Fischer assessed the position
as a draw and did not criticise the
move in the game. Nevertheless
White has a path to victory and it
consists of the move 1 g4l. In reply
Black has the following contin-
uations:
a) 1.. .@d6 2 f5 gxf5 3 @xf5 c4 4
bxc4 bxc4 5 @e4 c3 6 @d3 @e5 7
@1103 @f4 8 QM @xg4 9 @a5 @f5
10 @xa6 @e6 11 a4, and White is
victorious;
b) 1...@d6 2 f5 g5 3 a4! $06
(3...b4 4 a5 $06 5 f6+—) 4 axb5+
ab 5 @e5 +—;
c) l...a5 2 a4 b4 3 @d3! (Why
not 3 5+ gxf5 4 gxf5+ @d6 5 f6?
Because of 6...c4! 7 bxc4 %e6!, and
for the time being White must think
about saving himself) 3...@d5
W
W/////
W W
\
W
W , /:/
,ageWW
//”//
\\\§
a4 W WaW
W//%W%//
W W //W/
4 g5! @e6 5 @c4 @f5 6 @xc5
@xf4 7 @b5 @n 8 9x35 @f4 9
@xb4 g5 10 a5, and White reaches a
queen ending with a b-pawn—and
every chance of a win. This
22 Pawn Endings

variation was found by A.Bely- Here is another example.

avsky. But we return to the game:
l...@d6 2 f5?? Gligorié-Fischer
Mednis passed over this move in Candidates (t), Belgrade I 95 9
silence. Fischer, apparently, noticed
that as a result of the correct 2 g3
$66 3 a5 @d6 4 g4 @e6 5 f5+ gxf5
6 gf Qf6 7 @d5 c4 8 bxc4 bxc4 9 //////%

\\\\\§
@xc4 @f the Black king hurries
back to CS.
///M
2...gxf5+ 3 @f MM/M/

\r>»
/MM//
M MM//
M /%// / MMMM
\\\\\
ill/MM» M,%M,
MxM MMM Here, leading to a draw is 1...§h5!
a’Mf/'M/M, M 2 ECS Ec, and whichever way
// ///////
///8
White retakes there follows 3...@c8
//// MM/ with a draw. But Fischer mistakenly
M played...
l...§h8? 2 @b?
Now Gligoric met mistake with
3...?d5?
mistake, wrongly transferring to a
(!)Mednis. Why not 3...c4!, and
pawn ending. Winning was 2 §c7+l
White must resign at once.
4 g4 @d4 5 g5?
@d6 3 §c6+ @d7 4 $b §b8+ 5
5 axb5 leads at once to a draw.
§b6 EhS 6 Eb7+ <$c8 7 @a6 §h6+
8 @a7.
5...c4! 6 bxc4 b4 7 c5?? 2...§b8+ 3 @a4 Ea8+ 4 @b3
After 7 g6 White reaches a slight-
ECB! 5 EXCS @xc8 6 QM @bS!
ly inferior queen ending but with
Gligoric did not reckon in his
chances of a draw. Now however
previous calculations that White
it’s all simple—7...b3, and it is time
could not maintain the opposition,
for White to resign.
therefore it’s a draw.
The pawn ending, apparently so
simple, is difficult in that on the
transfer to it the players must calcu- With time Fischer managed to
late the changes that take place in correct his shortcomings and his
the game by comparison with other transfer to the endgame became
aspects of the ending. immaculate.
 Pawn Endings 23

We also come across ignorance in

Lombardy-Fischer transferring to a pawn ending by
USA (ch), 1960/61 Fischer’s rivals.
Fischer-Bisguier
/W USA (ch), 1959

a:
//e%% Q \\\\
x%//
/,/ /// /
////a¢//y Q
\
\\

/////.: /////
Q

§\\\\\\\

Q
aA/a
\\\‘\\‘\
e/j/a/s
\\

a/%E/%/
\§\\\\

Q
% Vga,a%%

Returning the exchange to win a
pawn was also a recipe of Capablan-
ca himself.
l...§xc3+! 2 bxc3 Exe5+ 3 Qd2
Exel 4 Qxel QdS 5 Qd2 Qc4 6 h5
b6 7 Qc2 g5! 8 M f4 9 g4 35 10
bxa5 ba 11 Qb2 a4 12 Q33
Qxc3 l3 Qxa4 Qd4 l4 Qb4 Qe3
and White resigned.
Fischer-Larsen
Candidates (m) Denver, I 971
% % /t/
W % %x%
’/
/ / fl
:.\\\
Q
//Q//
Black should suffer a little in the
rook ending by 1...§g5 2 §d4 b5 3
Qe2 Qc5 4 QB EgS 5 Qf4 §f8+
with the idea of breaking through on
the second rank with the rook. But
Bisguier decides to transfer to a
pawn ending in which it requires in-
credibly accurate play to achieve a
draw.
l...§d5?! 2 Qe2 Exd2+
No help is 2...b5 3 Ed Qd 4
Qe3 a5 5 g5 b4 6 g6 Qe6 7 Qd4
+-.
3 Qd Qd5 4 Qe3 QeS 5 QB!
Fischer ‘waits’ for the weakening
of Black’s pawn structure—there is

no win by 5 g5 Qf5 6 Qd4 n5 7

.s

Qxc4 Qf6 etc.

/ % % / 5...aS
If 5...Qf6 6 Qf4 Qg6 7 QeS Qg5
w/a
'///
White goes not for the c4 pawn but
for the pawn on b7——8 Qd6! n4
9 Qc7 QB 10 Qxb7 Qe2 ll Qxa7
Qd3 12 Qb6 Qc2 l3 Qc5, and
Here White won easily after wins.
1 a4 Q18 2 .933! Rxc3 3 Qxc3 6 Qe3 a4
Qe7 4 Qd4 Qd6 5 35 f6 6 a6 Qc6 On 6. .b5 winning is 7 g5 b4 8 g6
a77 Qb7 8 QdS M 9 Qe6 and Qf6 9 Qd4.
Black resigned.
24 Pawn Endings

7g5 Qf5 8 Qd4 nS 9 Qxc4 Off at last!

Qf4 10 Qb4 Qe3 ll Qxa4 Qd2 20...Qc2
20...Qd3 21 Qb5 +-.
21 b4 Black resigned.
/ / /
Rossolimo—Fischer
USA (ch), 1962/63
%
///////a /
////
W as”? , //// //
I’M/W
/
A/Hn/z

12 Qb3
/a a%, %
4"“ ///
The position appears very simple,
fiy/fl /%/
but in fact it is just the opposite as 4....x6
8""II/l

testified by the mistakes made in

past analysis by authors even in the
Encyclopaedia of Chess Endings.
Fischer avoided 12 c4 because of
12...Q02 13 b4 Q03 14 Qb5 Qd4
15 c5 Qc3 16 Qa5 Qc4 17 Qa4
Qd4 18 Qb3 Qd5 19 Qc3 b6! with
a draw.
12.. Qd3 l3 c4 Qd2 l4 Qa4
Also no help is 14 c5 Qd3 15
Qb4 Qd4 16 Qb5 Qd3 17 QaS
Qc4!.
14 Qc2 15 Qa3!
Again 15 b4 draw, as pointed out
above.
15...Qd3 l6 Qb3 b6?
It is incomprehensible why Bis-
guier did not give Fischer the
chance ‘to show’ the win after
16...Qd2.
17 Qb4 QcZ 18 Qa3
Clearly not 18 Qb5? because of
18...Qb3!, but in the Encyclopaedia
of Chess Endings is given the vari-
ation 18 b3? Qb2 19 Qa4 Q03 20
Qa3 Qd3 21 Qb2 Qd4 22 Qc2
b5!=.
18...Qd3 19 Qb3 Qd2
19...Qd4 20 Qb4 wins.
20 Qa4!
W//
Black clearly has the more active
king but the White g5 pawn paral-
yses Black’s flank and it is imposs-
ible to win.
l...a4 2 Qe2
The Encyclopaedia recommends
2 b3! a3 3 h3 b4 4 M, and now
Black must play 4...d5 5 ed Qd
6 Qd3 Q05 7 Qe4 Qd6 with a
draw.
2. .a3!?
The last try.
3 bxa3 Qc3 4 a4! bxa4 5 Qe3
Qb2 6 Qd2 Qxa2 7 Qc2 a3 8 h3
And a draw.
Some conclusions
In the earliest stage of his career
Robert Fischer had technical prob-
lems in playing pawn endings, and
particularly when transposing to
them. But by the beginning of the
603 he had already managed to
eradicate these shortcomings and no
longer had technical problems with
pawn endgames.
 Pawn Endings 25

Exercises:
Pawn Endings

W
/W W¢W

&|’\
/
\

W/ Q
WfW W W1 “WW,
, g}
WW/
/ /, ,,,,, 4 {W
W
W§
W;/%/
§\
§

/’/
’,-’// //

W// W& ;
/W
What plan must White adapt
to realise his advantage?

W W WW /
”2 WW W //W /
/WW W W/ W W
W WéWsW
Wax
W / W W

How do you assess the position What is correct 1...ficg4 or I...hxg4?

with White to move?
2 6 Pawn Endings

// //

\\\\
/ W / % / y} /
%§% %/ / fl %
W/%%y@ %1%/%///
/é /
/ ///

\
/// 3/23

/ % ߤ%
//%/ n¢ o / /

§
///
//r//////S
How does Black save the endgame? What is correct: I...@e5 or @g5?

6 9
\§

%/ 7 % / // % % /gs
%// 7% //@///
i§

/n/7%4%/
//;A4
fi % /%///

fl
/ @8”a
//%
//
%n%%%%%%
Assess the outcome ofplaying What is correct: I...g4 or 1...@f6?
the typical 1 g4?

% W / /
/ % fl
////
/,, ’/ %//
Al/t/
/‘
WWI/i / /’//////
/m /:%:/
&/
/%%%/A
%'% ’n%% %%W
/g/
/
Should White resign? What is the simplest way
for White to win?
 Pawn Endings 2 7

ll 14

A/ A AAA A A A:
A AAx/AA
My A//
% % ,,,,, %7%
AAAA/AW A AA A
\

§§
A/AyA/A
A A A AA //
////A
How does White win?

12 15

AyAyAW/
\

///A /%
AyAyAMA
A A A
AA% AA
AyA/
A A/
AAA AA
A/A/A/ A
\

//// A / AAA
Can Black save himself?

13 l6
\\‘\\\\\\\:\

AA A/4/A
/AV////A7W A

A Ax/flA
/ M AAA
\\

/ A 8AA
A

A A A A A A
A A, A AA/
§

/ //////
What 13' right I...®e6 or 1...§c6 .7 Point out the right movefor White.
28 Pawn Endings

20

/ W WWf/
/ /WaWW
W/W/W
///
W W/
///// W / /&
W W WeeW
//W WWW/WgW/
//
/’"WW W
///WE/
How does White make a draw?

18 21

W W
W/x,"’/ W/W W:
WWW/W/
/W/ W& /%///M
WWW/W W/W W
// /./ / W W W//
W//////
/ W
What IS correct: I...e4, 1...g4 What is correct:
or 1...@g8? I...@g7 or I...@h5?

19 22

W/WW W/xW/ / /
/W/ / / W/
W&//// Wa W W W
WWWW / é/W /%W W
\

WW/WW
W/Vx // W////WW
///WW /W////
What zs the right movefor Black? What IS correct: I...a5 or I...axb5?
 Pawn Endings 29

23 25

///
%%W/
%,4flr
/

/ //
/Q%§% ///// /”/
/”
fl: //: fly‘

// /£s/
/7/
////// /// //
/////////
Find the right continuation How should White continue?
for Black.

24 26

//////
//f%£%
9456/ //////‘/ /////
fififiy/fi
/%'a%/24 e’% /1@
a u,/ atr‘wzn
uénir /u,/;r2r
What is correct: 1 c4 or 1 b4? What is correct:
I...@xb4 or I...@c4?
30 Pawn Endings

Exercises:
Transposition to a Pawn Ending

1 3

4/474 ,/,//,7

/ 4
4 / 47
4
4
4 4 W 4
W
Can Black transpose to Demonstrate a plan to realise
a pawn ending? the extra pawn.

2 4

4 , /
/7 4 4141
4 E4 4 4
74 /4 4
/ 8484 4
/ 48/ 4 47,
4 474 WZ4
Can White transpose to How do you assess the endgame
a pawn ending? after 1 c5?
 Pawn Endings 31

A
A”AAK A ”A ”A A

\
A
4/

A? A”;
AA%,&/ A/IA W

\§
1% ///A
/ A / Wot/s

\
Ag“ ”A /A/”A”MA////

\\\§
”A ”AA/”A ”A/
W ”W W ”W
In whosefavour is the pawn ending Which should Black prefer,
after I...b5 ? the rook or the pawn ending?
6' 9

A
\\,\\\
::§\

W” ”A
\ \§

////
% //&//

MWI/
W‘
// //A”A
A ”A / ”A
”A
Is it worth White avoiding the Can Black transpose to
transposition to a pawn ending? a pawn ending?
7 10

M4 /M//W/?%
// //A ////

AA AA AA / / 4
/A/ W/A/A/Aas

Can 1 @3613 be played? In what way can White transpose to

a pawn ending?
32 Pawn Endings

11

////Z////////
////Z
//Z//A
Z// /Z
Z/Z/Z:
Z Z
Is it possible to transpose to Find the right continuation
apawn ending by 1...g5? for Black.
12 15

//Z
ZZ

// /§/ , , ,
Find the right continuation
for Black
13 16

/Z % /
///¥x // // flw
Z // Z/ // g Z/fiZ IIIIIII

////
/<////£Z%/%
11111111

/ M /// Z”
Z/éZ/ /
’tllll/

Can Black transpose Find a winning plan for Black.

to a pawn ending?
 Pawn Endings 33

l7 20

nurse W ////1/1
% /5fl
, , , 7e/ //// AI“
9 ’1

/W/%nfn
\D33

/ ///”’ ,,,,,

Can White transpose to

a pawn ending?
18

//
/ / e n
fr n n/e
% n/ //
\\

ea/ % /
\

@%e//fie4
// /
Find the right continuation How does White transpose to
for Black. a winning pawn endgame?
19 22

/I r I I
I y y

r ,, /%/ /%l ’
// ’ 6 // /’//
// 4....3 %, am,
i, ,//z '41;/',.:

/ 23% n7/;%
////
How should Black realise Find the right continuation
his extra pawn? for Black.
 2 Knight Endings
Due to their small radius of action
knights and pawns have very defi-
nite features. M.Botvinnik described
these features well by the very com-
prehensive and deep statement:
“Knight endings are pawn endings”.
Indeed the method of playing knight
endings is very close to those with
pawns. Here also a space advantage,
active position of the king, pawn
structure are all significant. An
important role is played by the
presence of outside passed pawns.
Often the zugzwang motive is ex-
ploited in the game. However, apart
from the general principles char-
acteristic for playing pawn endings,
play in knight endings is compli-
cated by the possibility of pursuing
a knight with an enemy king. Apart
from this, knight endings with
limited pawn material have specific
features associated with the
possibility for the weaker side of
sacrificing the knight to liquidate
the opponent’s pawns and thereby
achieve a draw. This feature creates
significant difficulties in the
realisation of a positional and
material advantage.
Let us look at a few examples
where there is a compact pawn
mass.
Upon the central placement of
pawns the pieces can have quite a
broad range of activity. In the first
position White has a positional ad-
vantage due to the active position of
his king and the presence for Black
of a weak backward pawn on b6.
Fedotov-Arkhipov
Moscow, I 9 78
%/ /%e// ’///’
//
%’<
/ ///M€
However the fact that there are
only a few pawns on the board com-
plicates White’s task.
An insufficiently deep penetration
into the features of the position
leads to Black’s defeat, practically
without a struggle. However,
analysis shows that Black has very
considerable defensive resources,
and to achieve a win White has to
demonstrate very accurate play
throughout.
Let us look at the possible course
of events. The first part of White’s
plan consists of the creation of a
passed pawn and exploitation of the
weakness of the b6 pawn.
1 15 ef 2 @3f @d4 3 @do
White carefully controls the b5
square, preventing the move

"b6b5 e.g. 3 @g7? $06 4 e6
@xe6 5 @xe6 b5, and the white
pawns are exchanged. Also no good
is 3 @xd4 cxd4 4 @f7 d3 5 e6+
@c7 6 67 d2 7 68=g dl=¥, with a
draw. Black, for his part, is obliged
to control the 66 square. Now, for
example, 3...@c6 with the idea
...b6-b5, does not work because of 4
e6! t§xd6d5 67, and the pawn cannot
be sto pd
3. e6 4 @b5 @d4 5 @c3 55e6
'//4/9% //x
fie
/,a/a/
/////,
//a3/////
///%////
6 @d5
And so the first part of the plan is
completed: White wins the b6 pawn,
since the opponent’s pieces are
forced to control the assed e5
pawn. Also possible 1s 6 £a4 How-
ever it is interesting that, with the
knight on a4, tactical resources
suddenly appear for Black which
requ1re precise attention from his
Opponent. For example. 6. b5 7
cb @307. This trappy move was
not possible with the knight on d5.
Now, on the natural 8 5Dxc5+ fol-
lows 8.. @e8! with a draw after 9 b6
d5+. However after 8 5236 White
0311 return to the channels of the
main variation, looked at below.
ls0 possible of course is 8 b6
d5+ 9 @fl c4 10 e6+ @c6 11
hC3' 55x03 12 67, and White
S ould Win. But 7. .Qc7 1s not obli-
gatory and is given to illustrate
Knight Endings 35
Black’s possibilities. Now his best
counterchance lies in the formation
of a passed c-pawn, since giving up
the b6 pawn without a murmur leads
quickly to defeat. And so...
6...b5! 7 cb @d4 8 @c3
The realisation of the extra pawn
requires accuracy. Black’s hopes are
linked to the win of the b-pawn and
the subsequent sacrifice of the
knight for the e-pawn. Since his
king is already close to the b-pawn,
any advance of this pawn must be
backed up by accurate calculation.
The second part of White’s plan
lies in the win of the c-pawn while
holding on to his own pawns. Let us
see whether Black can hold the de-
fence by the active advance 8...c4 or
the passive move 8...5De6.
I. 8...c4
/.
/%e% / //'///,
6/ //
%// //
// /7 //
On c4 the pawn supported the ma-
noeuvres of the white knight but
now Black is left without a base on
d4 for his own knight which is also
deprived of its defence. This creates
a motive to carry out the threat of a
breakthrough with the king to the c4
pawn, in combination with threats to
advance the passed pawn.
But the direct approach of the
king to the c-pawn does not achieve
its objective. For example: 9 <g’gS
36 Knight Endings
@c7 10 @f4 @b6 11 @e4 @xb5
with a draw.
The solution lies in the creation of
a zugzwang position where Black
simultaneously controls the e6 and
f5 squares, ie. in this position it is
necessary to hand over the move to
Black. This is achieved by the
endgame king manoeuvre, ‘trian-
gulation’, characteristic of pawn and
indeed also of knight endgames. In
the present case, along the g6, f7
and f6 squares. This method is ac-
tively exploited also later on.
9 @fi @e6 10 @g6 @d4 11 @f6
The objective is achieved! Now in
the position shown in the previous
diagram it is already Black to move.
ll...5be6 12 @f5 @cs 13 @d5
@d3 14 e6+ @d6 15 e7 @d7 16
@e4
The c4 pawn is doomed.
Instead of 10...£Dd4 there is no
salvation in 10...@e7, on which
ossible are both 11 @135 @c5 12
d5+ @f7 13 b6 @e8 14 e6, and 11
®d5+ @eS (11...@d8 12 @fi @c7
13 @c3 @d7 14 b6) 12 @f6 @d4 13
b6.
And so 8...c4 loses. But what if
the pawn is left on c5?
II. 8...®e6
Also here the immediate break-
through with the king to the c-pawn
is not successfill. For example: 9
@f5 ®d4+ 10 @e4 @e6 11 b6 @306
12 b7 @d7 and ...@c7. Therefore
White first endeavours to provoke
the advance of the pawn to c4, aim-
ing to obtain the position in the
previous diagram. The direct 9 @e4
does not achieve its objective be-
cause of 9...®c7!, and he has to go
back, 10 @303, so as not to allow a
draw in the variations 10 @xc5+
@e8! 11 b6 58d5+ or 10 b6 @d5+.
It becomes clear that the manoeuvre
53e4 is good when the king is situ-
ated on f7 and cannot immediately
be attacked. The solution to the
position is again achieved by the
triangulating king manoeuvre
g6-f7-f6.
9 @g6 @e7
After 9...5Dd8, preventing the
move to f7, the king breaks through
to the c-pawn: 10 @f5 @e7 11 @e4
@e6 12 @d5 followed by ®f4+ and
s.
The continuation 9...c4 10 @f5!
leads to a position looked at under
the analysis of the first variation.
The continuation 9...®d4 10 @fl
@e6 11 58e4 is looked at below (see
moves 11-13).
10 58d5+ @d7
On 10...@e8 winning is 11 b6
@d8 12 @f6 @d7 13 e6+.
11 @f7 523d4
Losing at once is 11...®d8 12
@f6 @e6 13 ®b6+.
12 @c3 @e6 13 559,4 58d8+
The position arising after 13...c4
14 @303 @d4 15 @f6 is examined in
the commentary to the continuation
8....c4. Also losing is 13...®d4 be-
cause of 14 5Dxc5+ @c7 15 @e6+.
14 @f6 c4
Black already cannot hold on to
the pawn on c5, since in the event of
14...5De6 winning is 15 b6 03d8 16
@xc5+ @c6 17 <$e7.
 Knight Endings 3 7

15 4&3 @e6 16 @f5, and White @xg7 However he does not notice
wins by continuing as in the analy- this possibility.
sis to Variation I. 6...5bf4? 7 @es ®g6+ 8 @f5 gxf6
The difficulties also encountered 9 gxf6
by the defending side with a flank
pawn formation are well demon-
strated by the following example. 7/

///
/% //W
Kremenetsky-Razuvaev
Moscow, 1981 % %/
/4W /////V/%/
// ////W
/ .4 W, W W
//Y /6/7
//
04... ////
7W ///”
%//43 fa4//////
How can he fight against the
passed f6 pawn? Black cannot allow
its further advance to f7 since, with-
//, out support from the king, the
knight will not be able to cope with
it, e.g. 9. .5Df8 10 f7 @hS 119305
@h6 12 @f6 58h7+ 13 @e7 @g7 14
White has a material and posi- @d7. The transfer of the knight to
tional advantage, but his chances are this key point ends the struggle,
based on the formation of a passed since the f8 square is indefensible.
pawn which is inevitably linked to This is a typical knight manoeuvre
exchanges and gives Black defens- with a pawn one square from queen-
ive resources. ing. And so the knight is forced to
1 @e3 ®d5+ 2 QM @f4 3 g5+ move to a passive position on h8. In
@hS contrast to the variation given after
More natural looks 3.. .fn 4 Black’s 3rd move, Black’s pieces
hxg5 @hS 5 @e5 @g4, creating de- now have no room for manoeuvre.
fensive possibilities after 6 g6 523d3+ 9.!t 10 @d6 @115 11 @e6
7 @e6 5Df4+WAand also upon 6 f6 @g6 12 @e7
gxf6 7 gxf658 6+ 8 @e6 55f4+9
@d6 @f5 10 t7 £36.
/x. , .....
4 @xf6+ $xh4
//W%W-- l l l §//
\\\*

After 4... gxf6 5 gxf6 the white / 7/ ///” I

/ fl“% %W
pawns are very dangerous.
5 @e4 ®h3 6 f6?
On 6 g6 @hS 7 @e5 White main-
\

tains the advantage, but now Black

can momentarily exploit his own
//%/2//
drawing resource, linked to a knight
7%W/Wa/x
\

sacrifice to liquidate the pawns:

6...5n5! 7 fxg7 @e6+ and 8...
38 Knight Endings
The black king has two squares at
its disposal: h5 and h7. Black played
12...Qh5
Why did he reject 12.. .Qh7? Let
us look at a possible continuation of
the struggle. 12.. .Qh7 13 Q17
Qg6+ 14 Qe8Q 8 15 Qh6+ Qh7
16gQg4 Qh8 17 Qe5 Qg8 18 Qe7,
and there arises a well-known posi-
tion from Chéron, presented in
Y.Averbakh’s monograph Compre-
hensive Chess Endings (2nd edition,
1980). This position is won in the
following way: 18.. .Qh7 l9 Qf8
Qh6 20 Qg8 Qg5 21 Qg7 Qf5 22
Qd7 Qg6 23 17 Qg5 24 Qe5 Qf4
25 Qg8 Qe6 26 Qf3+ and 27 Qd4.
Instead of 14.. .Qg8, there is no
saving himself either by 14.. .,Qf4
or 14.. .Qh.4 For example: 14.. Q14
14 Qe5 Qd5 16 17 Qf6+ 17 Qe7
Qg8 18 Qe6 Qg7 19 Qd7; or
14.. .Qh4 15 Qg5+ Qh6 16 f7 Qg6
17 Q13 Qg7 18 Q65 Qf8 19 Qe7
Qh7 20 Qd7 Qh8 21 Q18 (21 Qf6
Qf8!) 21. .Qg5 22 Qg6+.
After 12...Qh5! the king at the
necessary moment can attack the
pawn from the rear. This method of
defence is presented in the same
book by Y.Averbakh in examples
470 and 471. For example:
(a) 13 Q17 Qg6+ 14 Qd6 (also
insufficient is 14 Qe8 because of
14.. .Qg4 15 Qd6 Qe5!=) 14. .Qg4
15 Qe5+ Qf5 16 17 Qf6=;
(b) 13 Qd7 Qg5 14 Qe6 Qg6.
In the game followed.
13 Qe6 Q 6 14 Qe7 Qh5 15
Qe6 Qg6 l6 Qf5 QgS
Great accuracy is required of
Black. Simpler is 16.. .Qh7, since
White does not manage to reach
Chéron’ $5530osition. For example:
@(a) 17 h4 Q 818 Qe7 Q1719
Qh6 20 e5 Qf5, defends
successfiilly;
(b) 17 Qe7 Qh6 18 Qc6 QgS 19
Qe5 Qf4 20 Q17 Qg6 21 Qd6
Qh8! 22 Qe7 Qg6+ 23 Q17 Qe5+
24 Qg7 Qg4! (it is important to
leave the file where the promotion
of the pawn will take place!) 25
Qc4 Qc6 26 Qg8 Qd8.
l7 Qd4!
The knight heads for the key e5
square!
17...Qh6 18 Q13 Qg6
On 18...Qh7 19 Qe5 Qg8 20 Qe7
arises the above-mentioned Chéron
position.
19 Qd7 Qh8 20 Qe5 QgS 21
Qe6
79/ 7 47%
//
\
//
///
\\
IIIII
7 / 7 7.
7% / %7 /
21...Qh5?
Black cannot stand the tension
and makes a mistake. With the only
move 21...Qf4! he holds the posi-
tion. A.Kremenetsky intended the
continuation 22 Q17 Qg6 23 Qd6
Qh8 (on 23...Qe5 winning is 24
Qd5) 24 Qe7 Qg6+ 25 Q17 Qe5+
26 Qg7 Qg4 27 Qc4, so as on
27...Qc6! to offer a draw, while on
the natural--looking 27.. .Q1‘3? there
is the beautiful win. 28 Qh6! Qg5
29 Qg6 Qf4 30 QdZ! Qg4 31 Q64
4313 32 233+ @114 33 an Qg5 34
Qe5 (zugzwang!) and then 35 17.
22 Qe7 Black resigned.
On 22.. .Qh4 winning is 23 Qf8
QgS 24 Qg7 $15 25 Qd7.

The following example shows the
possibilities of the weaker side,
when he has compensation in the
form of actively placed pieces.
Gufeld-Grigorian
Daugavpils, 1979
//
/7%/%
/ / //
//
/ %% %‘
////?/8”
2/%7 M 7
lll/
// /
4/7// /
White has an extra pawn but
Black has sufficient counterplay due
to the active positions of his pieces.
Now White should secure the elas-
ticity of his pawn chain by 1 h3,
retaining the possibility of the ex-
change gxf5 for an appropriate mo-
ment. However, there followed:
1 gf gf 2 5813 @f6
It is important to maintain the
possibility of play in the centre and
on the king’s flank. Passive is
2...¢Zd6 3 @d3 h6 4 @d4 with a
win.
3 5AM
On 3 @d3 possible is 3...5Z)d7 4
@d4 556+.
3...Qe4 4 @b5 @dS
Further activity by 4...®d5?? ends
in a sudden mate by 5 @d6!
5 $13 58g4!
The activity of the opponent’s
pieces does not permit White to
strengthen his position without ex-
changing pawns. But this allows
Black drawing chances linked to the
possibility of sacrificing the knight
at an appropriate moment.
Knight Endings 3 9
//l
//
// %//
// % 2 /%
/,/
///¢Y///
/7%7%m
2/97a327
/7/ //7/,
For exam le.
(a) 6 h3 I53Z1h6 7 @d4 @d6 8 @g3
@d5! 9 QM @e4 10 @gS 5bf7+ 11
@f6 @d6 12 h4 ®e8+ 13 @fi
@xe3! 14 5866 @c7!=;
(b) 6 e4+ QCS 7 exf5 @xh2+8
@g3 58H+ 9 @g4 @c6! 10 f6 @d7
11 @c7 @e3+ 12 @gS 5804 and
@d6=.
Probably the maximum that White
can extract from the position is to
organise a passed pawn by 6 h3
@h6 7 @d4 @d6 8 e4.
The attempt to obtain two con-
nected passed pawns, undertaken in
the game, is met surprisingly by an
energetic black counterattack,
linked to the activity of the king.
6 @d4 58t+ 7 @g3 931+ 8
$12 $94!!
An apotheosis of the activity of
the king! On 9 @3f @xe3! the
white pawns are liquidated.
9 ($1t @xe3 10 939.6 $13 11
l @g3 Draw!
The black king confidently ties
the knight down to the f4 pawn and
does not allow its own white col-
league out of the cage. On 12 @hl
follows 12.. .116! 13 @gl h5 and then
h4-h3-h2, stalemating the king.
Now let us look at an example of
playing knight endings with the
presence of pawns on Opposite
flanks. In this case a decisive role is
often played by the organisation and
40 Knight Endings

energetic exploitation of a distant 6...@f5? 7 58215 ads 8 b7 58b4+

passed pawn. 9 @e2! @a6 10 $13
And so White blockades the pawn
Vladimirov-Novopashin and with a fine knight manoeuvre
Volgodonsk, I 981 wins it.
10...5Z)b8 11 M @e5 12 @b3 @c6
13 @cs @f5 14 @d3!
An important finesse! If 14 @d7
\

Black holds on by 14...®d4+ 15

/ /§4{4 :12 @c6 16 b8=¥ 58b 17 @3b
4
4% 4?/4x4
W// 4
14...@d4 15 $12 @c6 16 @b4!
\ .33

@b8 17 $13 @e5 18 58d3+ @do 19

4.494
///
/ 4 @3114 @w 20 @gs $xb7 21 QM
@c6 22 $11M
/%/’/>,//
7/
,,,
White’s advantage lies in the 444 / /g
possibility of quickly creating a 44 / 4
passed pawn while retaining a good
coordination of pieces which allows
him to support the advance of this
4f4%//4
pawn and at the same time to con-
4/4%/ /4
trol Black’s activity on the kingside.
However the peculiarities of the 4 4' 4 /
position are such that great accuracy
is required from both sides for the
fulfilment of their plans. Now White This ending has theoretical sig-
can quickly organise the manoeuvre nificance. Will the black pieces
1 $d4 @ds 2 @c5 @3744 3 b5, but succeed in neutralising the h-pawn?
he prefers prophylaxis. 22.43127 23 h5 @c6 24 @g7 @115
1 @e3 g5?! 25 @f6 @g8+ 26 $17 ®h6+ 27
A hasty decision, facilitating @g7 5815+
White’s task. After 1...h6 2 M g5! On 27...@g4 winning is 28 5812!.
Black’s defence is far easier. 28 @f6
2 1‘n f4 3 @c4 @f5 4 b5 21b 5
ab $n 6 b6
4
Let us pay attention to the differ-
ences in the possibilities for the two 4% 4 4N4/
sides. The passed b-pawn distracts
/// 4
\

the black knight whereas both white

4 $4?/’
\

pieces can attack the passed pawn.

Nevertheless Black can put up stub-
born resistance by breaking through
;4;424 4
with his king to the g4 square— 44:4 / 4
6-- Qg4“ However he chooses // 4 /'4’
another route-march for his king.

An instructive position! The black
knight can control the h-pawn from
the g4 and f5 squares, therefore to
advance it White must deflect the
knight from these points. This task
can be fulfilled by the knight from
the e3 square: 28...¢3h6 29 9M2!
Qd6 30 Qg6 @g8 31 Qg7 @e7 32
@g4 Qe6 33 @e3.
Black chooses the best plan, en-
deavouring to break through with
the king to the h-pawn and at the
same time controlling the e3 square,
and this attempt is rewarded.
28...Qe4! 29 QgS 523d6 30 @f2+
QeS 31 Qg6 @f5 32 @d3+ Q94 33
53c]? @14 34 ae2+ Qg4 35 @g3
@e7+ 36 Qf6 @fS Drawn.
The king attacks the pawn and de-
fends the knight, observing from the
h6 square.
Is White’s advantage in the last
diagram really insufficient for a
win? No way! After the game
E.Vladimirov demonstrated a clear
way to realise the passed pawn. In-
stead of 29 QgS he gave 29 Qg6!!
as leading immediately to victory.
//.// /
2 ////%§
///
7//
7///////
//
A picture of a position, having an
instructive character! White’s pieces
are clearly fiilfilling their functions:
the king confidently squares up to
the enemy knight, while the white
knight remains triumphant on the d3
square.
Knight Endings 41
Now every move by Black loses.
(a) 29.. fle7+ 3O QgS @f5 31
9312+ QeS 32 Qg4+ Qe6 (32.. Q.e4
33 @e3!!+—) 33 Q6 @h4+ (or
33.. .Qe7+) 34 Qg7 f5+ 35 Qf8!
Zugzwang! On any king move, win-
ning is 36 @e3! Leading to the same
outcome is 30.. .QgS 31 933+ QeS
32 Qg4+ Qe6 33 Qg6 5867+ 34
Qg7 @f5+ 35 Qf8!.
(b) 29.. .Qh4+ 30 QgS $313+
(30.. @f5 after 31 @fZ leads to
Variation (a) above) 31 Qf6!, and
the h-pawn is unstOppable.
But to conclude White missed the
Win by playing 33 93c]? After
Black’s 32nd move there arises the
position shown in the last diagram,
but with White to move With the
tried and tested ‘trian ulation’ king
manoeuvre 33 Qf6! 531% 34 Qg7
933+ (34...5Ag4 35 @fZ!) 35 Qg6
he hands the move over to his oppo-
nent and wins.
Returning to the position shown
in the above diagram, it is useful to
bring to mind a fragment from the
game Botvinnik-Simagin (Moscow
1955), where, by exploiting his
opponent’s inaccurate play, M.Bot-
vinnik was able to save himself in a
similar ending.
Botvinnik-Simagin
Moscow, I 955
//////
///%///
$\\
x§§
// /
\\\‘
\\::\\\
// // / //
/////////%%
%// ////
42 Knight Endings
Leading to a win here 1s 1.. .QfS 2
5M3 @gS 3 58h4+ @g4 4 55g6 5238
5 @b4 @gS. However there
followed:
1 .uél? 2 Qb4 @fs 3 @c3!
With tempo the king draws closer
to the centre of events and now if
3. @xeS 4 Qd he succeeds in
containing the awn.
3. .5De4+ 4 d4 @gs 5 @d3 @g4
6 ®e5+ $15 7 @ds @g4 8 @e5+
@g3 9 @g6!
In contrast to the previous
example White has this defensive
resource, since the pawn finds itself
one move further from the queening
square.
9...®e6 10 ‘36 5818 11 53x18 h4
12 @e6 h3 l3 Z3Dg5 Drawn.
Let us look at some more
examples of a struggle with pawns
on different flanks.
Vasiukov-Timoschenko
Volgodonsk, 1 981
/////
/4%””/
6 7/
/ //¢
//”/my //
//%
/%// /////
///////
In this position White can quickly
convert his positional advantage to a
material one but playing this ending
is complicated due to the fact that
the pawns are situated close to one
another which might allow Black to
regroup his forces successfillly for
defence. With fine play White pre-
vents this possibility.
1 @e4 @e8 2 @d6!
After 2 @3n 5M2 White’s task is
more complicated.
2...@$ 3 @3c 58f2 4 58d7+ @fl
5 @e5+ @f6
And so White creates a distant
passed pawn. He combines its ad-
vance with fine manoeuvring of
pieces, creating on the way threats
to the g5 pawn.
6 @d5 @d1 7 c5 536+ 8 @c4
@e4 9 @d3! 99.6 10 @d4 @f6 11
Qt? @137 12 c6 $9.6
On 12...@d6 winning is 13 5De4+.
l3 c @e7 l4 9% @d6
%/
///'///m//
//'///
///”///
////////a//
// %,%
x //
//
7/,
How can he realise the extra
pawn? Insufficent is the natural 15
@b7 @e8 16 @e4+ c$e5 17 53n
@f4 18 @h7 55d6+! (for the present
the g4 pawn is untouchable because
of 18.. .n4 l9 @f6+ @xf6 20 c7,
and Black does not succeed 1n play-
ing ...@3e8 since the pawn queens
with check) 19 @c7 5&7 (also poss-
ible is l9...¢3b5+ 20 @b6 @d6) 20
g5 @fSL—z On the g5 square the
pawn becomes vulnerable.
However E.Vasyukov finds an el-
egant manoeuvre and wins the g5
pawn while the black knight oc-
cupies the d5 square———which is
rather poor for the struggle against
the c-pawn.
15 @h3! @d5+ 16 @b7 @e5 17
55n QM 18 @h7

In this lies the main point of the
refined manoeuvre begun with the
move 15 ®h3!. In contrast to the
position looked at in the previous
note, the pawn remains on g4 with
Black to move——-and he is in zug-
zwang. On 18.. .Qe4 or 18.. .Qe5
follows 19 g5 Qf5 20 ®f6 @e7 21
c7 n5 22 @d5 @f5 23 Qc6! and
24 c8=W. This same manoeuvre
wins on 18.. 5/3.e7—l9 c7 n4 20
®f6+ Qg5 21 @d5 On 18...n4
decisive is 19 @f6+ @xf6 20 c7
@e8 21 08=W+.
In the game followed...
l8...5./3c3 l9 g5 Qf5 20 c7 and
Black resigned.
A very instructive ending to the
game.
Kochiev-Lerner
Beltsy, 1981
% /
%’//
//za/ 7 ’
/
//
fl} % / W
% // 7/ /
/// M/
/// /
In contrast to the previous
example Black’s positional advan-
tage is felt less 1n view of the more
active position of the white pieces.
With energetic play, 1 58e5+ Q03 2
5M7 or 1 b4 Qc3 2 Qe3, White
could create defensive chances.
However if he deprives his pieces of
activity, Black achieves a decisive
advantage.
1 @217? b4 2 @c6 Qc3 3 £835
4M7
Knight Endings 43
The main thing is to organise a
distant passed pawn and support its
advance. The white knight is in no
position to struggle against the
whole of Black’s position.
4 Qe4 523c5+ 5 QfS @xb3 6 58M
Alas, on 6 @c6 follows 6. ®d4+.
6...¢3d4+ 7 ns 43e6+z
It is important to restrict the white
knight; now it is deprived of the
important c5 square.
8 Qf6 b3 9 @d6 b2 10 Qxe6
b1=y ll Qf6 Wb6 White
resigned.
Let us look at another interesting
ending, demonstrating the import-
ance of active defence.
Nikolaevsky-Gufeld
Kiev, 1 951
/
1///
\§‘
W7 / / \
% 1 /,fi/
/ / ///
n/@////
rrrrr
\
After 1...5Da5 2 b4 cxb4 3 axb4
@xc4 White gradually loses. How-
ever analysis shows that he is not
exploiting his defensive possibi-
lities. After 2 l93d g6 the first im-
pression is that White gets into
zugzwang since 3 Qe5 n5 4 QdS
loses because of 4.. .Qf4! 5 b4
Qe3!!, and the pawn cannot be
stopped. But nevertheless White
finds a savin resource:
3 @e2!! 5%)b 4 ®g3+ Qh4 5
435+!
44 Knight Endings
And Black is forced to reconcile
himself to a draw, 5...@h5 6 @g3+,
since he would risk defeat after both
5...@h3 6 @e7, and 5...gxf5 6 g6
@d4 7 @e5! ®c6+ 8 @d6 @d8 9
@d7.
With passed pawns on opposite
flanks, as in pawn endings, the
strongest side might win even with-
out the participation of the king. An
interesting example is from the
game...
Azmaiparashvili-Novopashin
Volgodonsk, 1981
Way / //y//a
W /%//& g?
/
a a/ aaa
,,,,,/
a aéa a
y/ ay/
1 a5 @f6 2 f4 ($27 3 f5 @gs 4
@d3 @216 5 @es 58M 6 f6 c4 7 h6,
and Black lost on time.
Van der Sterren-Douven
Netherlands, 1985
A position which, despite its ap-
parent simplicity, is quite insidious.
Black’s pawn weaknesses require
due attention from him. The cardi-
nal decision in the position is the
exchange of knights, which requires
accurate calculation. For example,
after 1.. .Qe6+! 2 @xe6 @xe6 3 g5
fn 4 hn @d6 5 @g3 %e5 6 14+
@f5 6 QB d4 the chances are even.
In the game, however, there
followed...
1...h6? 2 h5 gt?
The only chance of saving the
game remains with 2.. .@f7 3 hxg6+
@xg6 4 f3 @d7 5 QBfS h5 6 @e7+
$17, and White’s advantage is not
so tangible.
3 @f5+ @e6 4 gt @e4 5 13
5236 6 e3 ®e2+ 7 @g4 @e5 8
52)xh6 d4 9 5517+ 99.6 10 @d8+
<§d7 ll exd4! and White won.
Godena-Lalié
Portoroz, I 998
a aa
Q
Wig/fl a
.1
/gW/s
y/y/
OW
{gs/ a
After 1...5Dd3! 2 b3 @ds the
black king cuts off the enemy knight
from the centre.
3 @3215 99.4 4 58% e5 5 M
White presents his opponent with
good chances. Correct was 5 b4!
@d5 6 b5 e4 7 @a7! @c5 8 (2308
 Knight Endings 45

@b 9 58d6+ $05 10 @xe4+ @d4 Barlov—Abramovié

11 @gS with a clear advantage. Yugoslavia, 1989
5...<§d5
A critical position has arisen.

X EX
6 (9335 e4 7 g3
4 4/

\
Again better was 7 @c4 5801 8
523d; exchanging the b- for the 4 /4/44
e-pawn.
7...@d4 8 42)“ @cl 9 @a5 @d3 %/////4/
10 @e2 @c1+ ll QdZ @d3 12 @e2
And White reconciled himself to a ”/4 A4 44
,,, ,
draw.
////////
Polnareva-Akhsharumova
Moscow, 1984
l...Q)e7+ 2 @1'6 @c8 3 @gs h6+!
4 @h4 If 4 Qxh6 @g4. 4...?e4 5
®g7 @e5 6 @hS $f6 7 @xh6 58d6
4 4’/ 71/ Drawn.
White is not able to break the
/ 444/4/ blockade surrounding him.
4 / / 4
4 Vyzhmanavin-Chibu rdanidze
USSR (ch), 1 984
/ 4444 4
/ 444 \
4/4/ M4
Q

//// /
At first sight it seems that a com-
/4%/ 4
/ /
plicated and long struggle is in pros- /
llllll

pect, but Black, exploiting the bad

position of the white king, quickly
achieves victory.
1 @e5 15 2 @d7 @f7 3 @c5 @e7
4 5836 @d6 5 @b4 e5 6 @d3 @d5
7 ®b4+ m4 8 @3216 @d3 9 £81m
@e2 10 @c6 £813 11 @e7 5De1+ 12
@gl @d3 13 43f @xfz l4 @gZ
e4 15 58d4+ @d3! and White
resigned.
In the next game, White has a ma-
terial advantage but Black finds an
interesting possibility to
himself.
7 444 / /
4 4 4'4
////%/ ,,,,,
White has the advantage, but the
limited amount of material left com-
plicates the task. White’s plan is
linked to a breakthrough with the
king to the queen’s flank, since
there is nothing in 1 58b6 @a5 2
@d4 because of 2 .5Dc6+!
save 1 @e3 @g6 2 @112 @g7 3 @c2
$g6 4 @b2 523e7
46 Knight Endings
He cannot wait any longer, the c4
pawn cannot be held. To exploit the
remoteness of the white king, Black
strives to simplify the position.
5 @b6 f5 6 exf6 @xf6 7 @xc4
@g6 8 @c2 5AM
In the pawn ending——-8...¢3e5 9
@xeS @xe5——White gains victory
by 10 @d3 @f4 11 04 e512 c5 QB
13 c6 e4+ 14 @d4 e3 15 c7 e2 16
c8=fi e1=w 17 Wf5+ @g2 18
fie4+ fixe4 l9 @xe4.
9 @d3 @113 10 h3 @g1 11 QdZ
@e5 12 @e3 @xh3 l3 c4
With this move White rejected
Black’s offer of a draw. In fact, if
13...5Df4 White retains the advan-
tage by 14 5bf3+ or 14 @e4. Also
other defences are no help.
13...®gl 14 @e4 @h3 15 QCS
@d6 16 @d3 e5 17 @134 99.6 18
@xes 5732+ 19 @e3!
19 @d4 @d6 20 05+ 966 21 C6
@d6 leads to a draw.
l9...5Ddl+ 20 @d4 QBfZ 21 CS
65111 22 c6 @d6 23 @133 @g3 24
gm @xc6 25 523d2 @d5 26 $13
hl
//flfl
/////
///
/ //
/// . . . ,
///“
_In this apparently arid desert a
Wln can still be achieved due to the
poor osition of the black knight.
27 e4
But not 27 931‘] because of
27. .Qd4! 28 ®h2 §d3 29 @gZ
@e3 30 @xhl $12 with a draw.
27...?e5 28 @e3!
The last finesse: if 28 @n, then
28...¢3f2! 29 $f @f4.
28...?d5 29 ®f6+ @e5 30 @hS
Black resigned.
He loses the knight, without man-
aging to capture the g4 pawn in
return.
An instructive and complicated
knight ending, where the advantage
of one of the sides lies only in a
slightly better structure, is encoun-
tered in the game...
Timman-Ree
Netherlands, I 984
”/flfi/fi
yfl flx/ fl
7/// fl
fl/ ///’/7//g
yéfi fly
The more obvious 1...e5 looks
stronger. For example: 2 @e2 h6 3
@e4 b6 4 $d3 $17 5 $04 @e6 6
@b5 58b8, and it is difficult for
White to obtain real chances of a
win.
2 b4 @216 3 a3 @c7 4 @e2 h6
His defensive problems are not
solved by 4...5Db5 5 a4 58c3+ 6 @d3
@xa4 7 @xe6 and Black has
difficulty involving his knight in the
future play.
5 589,4 918
More circumspect was 5.. .b6 6
@d6 a6.
6 @d6 b6 7 @d3

Accurate calculation was re uired
for the initiative-seeking 7 (Qc8l.
For example: 7...Qb5 8 a4 Q03 9
<$d3 Qxa4 10 Qxa7 Qb2 11 @d4
e5+ 12 $64 l 13 f3 Qf2+ 14
@d5 Qd3 15 b5 Qf4+ 16 $64
n2 17 Qc8, and White’s position
is close to a win.
7...a6 8 Qc4 Qd5 9 @d4 @e7 10
g3 @d7 11 f4 @c6 12 <$9.5
@//
/ /
// /@
%%§/ 7/
11/ 2"?
aa/
7/ % /7:.‘
%%//
/ //I
I/
12...Qc7
Black chooses passive defence—
and wrongly so. In his comments to
this game Timman wrote that
12. .@b5 13 Qb2 Qxb4 14 axb4
@xb4 15 @xe6 a5 16 @17 a4 17
Qxa4 was losing. Considerably
stronger, however, is 16...@b3!,
after which 17 @xg7 @xb2 18 f5 a4
might lead to a queen ending with
an h--pawn, where White’s chances
of winning are problematical.
l3 Qd6 @d7 14 f5 ef 15 Qf
With simple and convincing
moves White has succeeded in in-
creasing his advantage and placing
Black in a zugzwang position.
15...Qe8 16 g4 Qf6 17 h3
17 Qxh6 $06 18 g5 Qd7 19 @e6
is more quickly decisive.
17...h5 18 g5 Qh7 19 M Q18 20
n7 Qg6 21 @f6
21 @d5 is also sufficient to win.
Knight Endings 4 7
21..Qxh4 22 Qt @c6 23 Qg3
@d5 24 34 b5 25 35 @c4 26 551.5
Qg2 27 @e5 Black resigned.
The difference between pawn and
knight endings is seen when there
are passed pawns on opposite
flanks. For example, the game...
Vukovié-Eingorn
Belgrade 1987
//
/ //7/
/1377/
/////
/§m/./}
7////
/////// ://77/
White should continue 1 @gZ! M
2 Qd4 @e4 3 Qe6 @e5 4 Qg5, and
the outcome of the struggle is still
not quite clear, since it is difficult
for Black to advance both pawns to
the third rank. But in the game
followed...
1 Qd4 @133! 2 Qf5+ @f2 3 Qg3
After 3 Qh4 g3+ 4 @hl the
pawns are frozen but the poor posi-
tion of the white king 1S decisive—4
.Qe3 5 b6 @fl 6 b7 Qg4 7 b8=fi
Qf2 mate.
3...h4 4 Qe4+ @e3 5 Qd6
On 5 Qg5 g3+, with the following
interesting variations:
(a) 6 @gZ (6 ®h3 Qf4+ 7 @xh4
2) 6. @1‘4 7 Qh3+ @g4 8 @gl
f4+ (8. .Qe3+ 9 @h1 h3? 10
Qxh3=) 9 <éhl h3 10 b6 h2 11 b7
Qh3, and mate in two moves.
(b) 6 @gl @f4 7 Qh3+ @g4 8
@gZ Qe3+ 9 @gl @xh3 10 b6 Qg4
48 Knight Endings
11 b7 Qe5 12 b8=¥ Qf3+ l3 Qfl
g2+ 14 @e2 g1=g 15 @x13 ¥g3+
5.. .g3+ 6 @gl
Forced. If 6gQgZ Qf4 7 @gl then
7. $13 13 decisive.
6...h3 7 Qf5+ QM 8 Qd4 ®g4 9
Qe2
The threat was 9...h2 10 @h1
@h3. Now, however, on this could
follow 11 n3 @xg3 12 b6 with a
draw. In order to achieve victory
Black must broaden the range of ac-
tivity for his kni ht.
9.. .Qb6 10 fl Qc4 11 @gl
Qd2
White resigned, without waiting
for Black to promote to a queen in
the variation 12 b6 Qf3+ 13 Qfl
g2+ 14 $12 g1=¥+ 15 n1h2.
The comparison between knight
and pawn endings is borne out with
the exploitation of zugzwang in the
two following endgames.
Sajtar-Benkii
Budapest, I 954
///////
7’7
// 72M
///////
////////,%
/////////
If in this position we exchange
knights for bishops or rooks then the
game would be drawn. It is interest-
ing that Reshevsky managed to win
a queen ending with such an align-
ment of forces against Geller, Inter-
zonal tournament, Sousse 1967,
and, with c~pawns, Mikhalchishin-
Kasparov, USSR (ch) 1978.
Of course, according to Botvin-
nik, playing a knight endgame, is
like playing a pawn ending. White’s
plan consists of advancing the first
pawn with help of the king, after
which Black gradually lands in
zugzwang.
$1 Qf4 d6 2 g4+ @gs 3 Qh3+
6
Igf 3...@h4, then 4 @f4 with the
irresistible threat of 5 Q12 and 6 g3
mate.
4 @g3 Qc4
Attempting to create counterplay
by means of an attack on the pawn.
5 @h4 Qe3 6 Qf4+ @h6 7 g5+
@g7 8 @hS Qf5 9 Qd5!
Inferior was 9 g6 in view of
9...Qg3+! 10 @gS Qe4+ 11 @f5
Qg3+, preventing White attacking
the coordinated forces.
9...Qg3+
On 9...Qd6 or 9...Qd4 would
have followed 10 Qe3!, and Black
is forced into a worse position.
10 @g4 Qfl
If 10...Qe4 11 @f4 Qd6 decisive
is 12 Qe71.
11 QM!
The main principle of the end-
game is not to rush! White restricts
the activity of the black knight,
which, in order to get into play, has
to go to d2 or h2, and then the white
king finds itself in a very favourable
position——safe from the
knight along the diagonal.
11...Qd2 12 Qe3!
black
Still more restriction of the black
knight.
12...? 6 l3 QfS Qfl 14 Qh4+
@g7 15 g4! QM
15...Qe3+ 16 @hS.
l6 g6+ @h6
16.. .Qg7 17 @g5.
17 Q“!
Zugzwang 1n action.
17...Qd2 18 $15 @g7
 KnightiEndings 49

18...@c4 l9 Qf6. 29...®c4

19 g4!
Zugzwang in turn. It should be
mentioned that White does
hurry with the advance of his re-
serve pawn, since this cuts off
squares from the king.
19.. .Qg8
On a move of the knight follows
20 Qg5, while on 19...Qh6——20
@g2 Qg7 21 9M4.
20 Qg5 Qe4+ 21 Qf4!
21 Qh6 Qf6.
21...5Ac5 22 @f5
Step by step White creates a very
strong position, and Black must
continually watch the threat of a
king infiltration to f6 or h6
22.. .Qd3+ 23 Qg5 @eS
Black continually attacks
pawn and accuracy is required from
White. Thus there is nothing in 24
g7 QM, after which the coordin~
ation of the white pieces is
destroyed.
24 @g3!
The knight crosses to h5, after
which the pawn pushes on to g7.
24...®c4 25 @hS 5863
25...@d6 26 Qh6 @e4 27 g5;
25...@e5 26 Qf5 and 27 g7.
26 g7 QM
26.. .QdS 27 Qh6; 26...Q17 27
In this case White wins in the fol-
lowin
0néinstructive way.
not f6 Qg8
30...®d6 31 Q67 and then 32
Q18.
31 Qg6 936+ 32 Qf5 @c4 33
Qf6
The familiar ‘triangulation’.
33...Qh7 34 @g3!
A new reconstruction. Now bad is
34.. .Qg8 because of 35 @f5 and 36
@367, and also 34. .5De3 because of
35 Q17
34...@d6 35 @f5 58e4+
35...5De8+ 36 Q17.
36 Qf7 @g5 37 Qe7! 939.4
37.. .Qg8 38 Qf6 58e4+ 39 Qg6.
38 Qf8 Qf6 39 Q17 @gS
the Black IS on the final frontier.
40 g5 @h6! 41 Qf8 @g8 42 g6+
And White wins.
It is interesting that precisely the
same endgame was met in the fol-
lowing game.
Matulovié-Uitumen
Palma de Mallorca, I 970
////
y /
\

Qh6 @xg4 28 QM.

27 Qf4 @d1 28 Qf5 58e3+ 29 // / //£”
Qg5
White hands over the move to his
.... % //
opponent and at once places him in / //,/ /Q/
\

zugzwang. Black, apparently tired

from a difficult defence, here made
a mistake by playing 29...Qg8?, and
// //////
/// //%///////
after 30 Qg6 resigned, since on
30...5n4 follows 31 @f6 @xf6 32
Qxf6 QM 33 Q17. This ending also ended in victory
There was a chance for him to put for the stronger side and we present
up more stubborn resistance by... it without commentary since here
50 Knight Endings
the same idea was utilised—only
White probably defended in weaker
fashion.
l...58b6 2 531:6 523d7 3 @e7 @f6+
4 Qh4 g5+ 5 Qg3 66 @c6 QhS 7
4M4 ae4+ s an d6 9 Qg2 @315
10 @e6 4 11 ®f4+ QgS 12 58e6+
@114 13 f4 ®e7 14 @e2 523d5 15
Qh2 g5 16 523g3 523e3 17 @e4 58f5
18 Qg2 g3 19 @f6 523e3 20 l
573g4 21 53M @e5 22 Qg2 @d3 23
l 55M 24 t g2+ 25 Qh2 g4
26 @f6 g3+ 27 l Qh3 White
resigned.
It is interesting that in both
examples the stronger side had
doubled pawns on the g-file. But
what will be the case if the pawns
are situated on another file? To us it
seems that the weaker side will
make a draw only with pawns on
the edge file, since then the king of
the stronger side has no exit.
Van Wely-Adams
Grom'ngen, 1997
/a:
/ transposition of moves gives Black
////
7 /Q
1 @d4?
Correct was the transfer to a
ending by 1 @d6! Qe6 2 @e4 £164
3 fxe4 QxeS 4 QB Qf6 5 h4! gxh4
g Qxf4, and then the king travels to
3. . following game where despite, the
l...®d7 2 e6 5565 3 @112 @116 4
Qc3 Qe7 5 Qb4 Qd6 6 e7
What else can he do? He must
give up a pawn—though it was still
possible to attempt a breakthrough
with the king to the d8 square, but
then, with the king on e7, the knight
from e5 goes to d3, and White is
again in a blind alley.
6...Qxe7 7 Qc5 Qf6 8 QdS @d3
9 @c6 @el 10 Qe4 @g2 11 @d4
@e3 12 @f5 58“ 13 Qd3 @e5+ l4
Qe2 Q 6 15 Qf2 @c4 16 Qg2 Qf6
l7 h3 e5 18 @d6 Qg6 19 @e4
Qh6 20 Qf2 Qg6 2l Qe2 Qh6 22
@c5 Qg6?
% /,
7/ / 7
/
7/ ,l/l,/W//
’2 //
//
’/// /£s 78
/ //Q/ /
Black must have the possibility,
on @d3, to reply.... @g6, defending
the f4 pawn.
23 M?
Correct was 23 @d3! @c4 24 h4,
obtaining a winning position. The
the possibility of saving himself.
23...gxh4 24 @e6 h3! 25 @xf4+
QgS 26 Qxh3+ Qh4 27 @312 Qg3!
The Black king breaks through to
the white pawns and a draw is
inevitable.
28 g5 55x13 29 g6 @d4+ 30 Qd3
@e6 31 Qe3 Drawn.
An interesting ending arose in the
approximate equality, the struggle
was still not over.
 Knight Endings 51

Ivanchuk-Eingorn exploitation of this advantage

USSR (ch), 1988 requires filigree technique.
1 f3 58d7
Neither now, nor later is there any
ssibility of 1.. .c5 since then 2
%% % %l% $05 a6 3 @d6 @e7 4 @b7 @d7,
{% / /x// and the white king goes to the

/fl//%1
'“/ 15% Z/
centre followed by a3 and b4.
2§f2f539e3@e74b4e55a4
Qd6 6 @dB ®f6 7 c5+ @e6
Better is 7...®c7 8 @c4 a6.

M/%/%a%% 8 b5! @d7 9 @c4 @c7

7”% //
lf3h42b5a53®g1$d6
More active is 3...?f4 provoking
lay such as 4 QdS QBb6+ 5 @c6
£3a4 6 b6 @xb6 7 @xb6 @g3 8
@1135 @gZ 9 @3122 @xfB 10 58g1+
@g2 11 @e2 with a draw.
4 @d4 @c5 5 55113 @e6+ 6 QM
@c7 7 @115 @b6 8 @d6 @xbs 9 f4
@a4 10 @e7 @1133 11 @xfl ®d4
Foreseeing 12 @xg6 a4 13 f5
@3f l4 @f QbZ 15 g6 a3 16 g7
32 17 g8=y a1=§, the rivals agreed
to a draw.
Sveshnikov-Sokolov
Moscow, 1991
Bad is 9...e4 10 @d4.
10 35 36!
The threat was a6, bxc6 and @b5.
11 b6+
After 11 bxa6 follows 11...?b8
12 @a2 @dS and 13...@a7.
ll...@b7 12 g3 h5?
He should not freeze his structure.
Correct is 12...®d7.
13 h4! 58d7 14 f4!
Closing the way for the king.
14...exf4 15 gxf4 58$ 16 @e2
Qg6 17 58M @xh4
After l7...@3xf4 18 @3f g6 19
5567 Black is in zugzwang.
18 @e6! @c8
To defend against @d8.
19 @xg7 @g6 20 @3t Black
resigned.

Torre-Portisch
Toluca, I 982
, W /x%z
%1/ {H2/
% // % % /
/ / / / ”1%
% ” %8/ %
//

/ % @/%
9% /&
White has the advantage on ac- / / %/%
count of his majority on the queen-
side and greater space, but the
52 Knight Endings

The endgame appears very diffi- Ilincié-Abramovié

cult for White but, with a pawn
sacrifice, he activates his king.
1 b6! axb6 2 Qc4 b5 3 Qe3+
@c5 4 f5!
Weaker is 4 Qg4 Qe6 5 @e3
Qc7! 6 Qf6 QdS, and the pawn
ending is hopeless for White. With
the sacrifice of yet another pawn
White creates his own passed pawn,
which is a principal factor in this
endgame.
4...gxf5 5 M f4 6 Qg4 f5 7 Qh2
Qc6 8 h5 Qe5+ 9 $92 b4 10 h6 b3
11h7 Q 612 Q13 b2 13 Qd2 @d4
14 913 C3 15 l+
Now Black cannot win the knight
on b1 because after h7-h8, Qxh8,
the king captures both black pawns.
He has to try and go with the king to
the h7 pawn.
15.. .Qd3 16 2a @e4 17 Qd2+
@d5 18 @e2 99.6 19 @d3 @f6 20
:d/ @g7 21 Qf3 @xh7 22 @b
h6
Tivat, 1995
/2 / /%
6/2 //Q@,/ L/
2,2/2%4
72% 2;
//
2,2 2
llllll
It looks like White has some prob-
lems converting his passed d-pawn
to a win, but really it is very simple.
l...g6 2 QM @fl 3 QcS!
Speculating on a transfer to a win-
ning pawn endgame, White clears
the way for his pawn and king.
3...Qb6 4 d6 @f6 5 d7 @e7 6
@e5 h5
Zugzwang—after 6. .Qc4+ 7 @d5
Qa5 (preventing $06) 8 g4 g5 9
@e5 the white king penetrates one
flank or another.
7 QM Qc4 8 @gs Qe5 9 M!
And because of zugzwang White
wins yet another pawn. Black

/ 7/2 2
2%2/222/
resigned.

Marié-Zaitseva
Tivaz‘, I 995

2 62 6 635 2
It is interesting to compare this 2 2 / 2x
ending with the two following ones, 2 2; 2 2
where doubled pawns were success-
62: 2 2
\

fully realised.
23 QCZ @hS 23 @d2 @g4 24
Qd4 Q85 26 m2 5817 27 m1 QgS
2 2 2 2 I ‘/////

26 <s Qe4+ 29 @gZ Qd2

And Black did not manage to real- 26262
ise his material advantage. /6/6/ /

In many cases 4:3 on one flank is
winning for the stronger side, but
here it isn’t so clear.
1 $13 f6
More or less necessary——White’s
plan was Qd4, g2-g4 and f4-f5
creating a weakness on e6.
2 @e3
Playable was 2 Qd4.
2...Qb5 3 g3 @e8 4 @d3 @d7 5
Qb4 fxeS?
Clearly better would be 5...Qa7 6
@e3 Qc6 7 Qd3 $67 with the idea
8...g5 and transfer of the king via
f7-g6 to f5.
6 fxeS @c7 7 $e3 Qa7 8 Qd3
Qc6 9 M g6 10 Qc5 Qd8
Otherwise after h4-h5 Black’s
structure would be completely
blocked.
11 g4 h6 12 g5 hn 13 hn
@c8 14 QM @c7 15 @d3 @b6 16
Qd7+ @c6 17 Qf8 @cs 18 Qd7+!
Rather dangerous was 18 n6
Qf7 when the e5 pawn is quite
weak.
18...@b5 l9 Qf8 QM 20 n6!
Now is the time.
20...Qf7 21 Qf4 n5 22 Qd!
Draw.
Alexandria-Marié
Tivat, 1995
5 /
/////
% /
/ But better would be 14. @134 15
./.%%/
//;%
Knight Endings 53
Usually such endgames with a
distant passed pawn give great Win-
ning chances.
1 Qe5
Black has a compact pawn struc-
ture and White tries to create some
weaknesses so he can penetrate with
his king, taking advantage of the
fact that Black must spend time to
win the a-pawn.
1...f6 2 Qc4 @b8 3 @122 @a7 4
$93 @1136 5 $e4 Qc7 6 Qe3 @bs
7s
/W
/W/
% Wé% §%%
% // /:
Wfi/Q%I%/
///
//%
/ // Qfifié
,,,,,
/%%;.,WW
/
7...Qe6
The pawn endgame after 7HQC6
8 Qxc7 @xc7 9 @d5 @d7 10 f4
@e7 11 f5 would be too dangerous
for Black.
8 Qxf6!
White has no other way of playing
for the win.
8...gxf6 9 $15 QCS 10 f4
The direct ap roach was 10 @xf6
Qe4+ 11 @g7 Qf 12 @xh7 $05
13 M @1514 h5 @e5 15 @g6 (15
h6 Q 4 A 16.. .Qxh6=) 15. .Qg4 16
@gS Qf6 l7 h6 @e6 18 @g6 @e7
19 h7! gave White winning chances.
@g6 Qg4 16 @gS Qe5 17 h6 Qf7+
and 18.. .Qxh6=.
10...Qd3 ll g3 @c6 12 @166
@d6 13 f5
01‘13 9g7 @6614 Qxh7 Q1515
h4l $16! 16 115 Qfl! 17 h6 Qg4=.
54 Knight Endings

l3...@65 14 h3
After 14 Qg7 h5 15 Qh6 Q67 16 ///,/
%// /
Qxh5 Qf6 17 g4 £313 18 M @eS /
the draw is obvious. /e///%// ////
l4...5Dd7+ 15 QgS Qe5 16 g4
5flf6 l7 h4
Or 17 Qh6 Qf4 18 Qg7 Qe5 19
//a%%/%
Qf7 h5 20 g5 @e4 21 g6 Qxf5 22
g7 Qf6;
17. .h6+! 18 Qxh6 n4+ 19
QgS @f6 20 Qg6 @dS Drawn.
/////a{:5
%///
//
9...Qc6?
Sermek-Hulak
The only chance was 9...g5! 10 g4
Slovenia, 1 995
(10 Qg6 g4 11 Qg5 Qc6 12 n4
Qd6=) 10...Qc6 11 Qg6 Qd6 12
n5 Qe7 13 Qg6 5803 14 Qh7
(14 f4 @d5 15 f5 Qf8=) 14...Qf6 15
f4 @e4 16 Qh6 Qf7 with a probable
draw. After the move in the game
White’s task is easy.
10 n6 Qd7 11 g4 Qe6 12 g5
5.13% 13 Qh7 Qe4 l4 g6 Qf6+ 15
Qh8 Qe7 16 f4 Q18 17 g7+ Qf7 18
g4 @g8 19 g5 Black resigned.

Beliavsky-Tratar
This is a very similar endgame to Bled, 1996
the previous one but here Black’s
pawn structure is weaker.
If now 1...@d5 2 @dl with the
idea 2...Qxa4? 3 @c3 +—.
1...Qb3 2 a5!
And here White 1S forced to sacri-
fice a piece—if 2 523d] then
2.. .Qxa4 3 Qe3 @305 4 Qd4 Qb4.
2.. .Qxb2 3 36 @b5 4 Qe3 Qc3 5
c /
Qxe4 Qb4
If the black king tries to go for the W//y /, / V
///// //
white pawns by 5. .Qd2 then White
creates second passed pawn by 6
g4! Qe2 7 f4 QfZ 8 f5 +—. White’s king is much more active
6 Qe5 Qas 7 Qf6 Qxa6 8 Qg7 than Black’s and this gives him
Qb6 chances to make progress.
Or 8. .@c3 9 Qxh7 @e4 10 f4 1 Qc6 Qe7 2 Qc7 Qa6+ 3 Qb6
@xg3 11 n6 +-. @cs 4 Qc6! @b3 5 55x34 ®d4+ 6
9 Qxh7 Qb7! @b5 7 @b6 @c3?

The way to the draw wasn’t easy:
7 @xa3 8 Qc6 @b1 9 ®c8+ Q16
10 @xd6 @C3 (10.. .QdZ 11 Qd7
Qg5 12 $66 Qf4 13 Qf6 @b3 14
@b7! Qxe4 15 d6 @d4 16 d7 @C6
17 5Da5!+-—) 11 Qd7 Qg5 12 Qe6
Qf4 13 Qf6 5.13a4 14 @b5 Qxe4 15
d6 @b6 16 5303+ Qd4 17 5334
®d7+ 18 Q67 ®b8=.
8 a4 @xe4 9 35 @c5+ 10 Qc7
Wrong was 10 Qc6? e4 11 @c4
Qd8! 12 Qd6 e3! 13 @xe3 @b7+=.
10.. .5Da6+
After 10...e4 11 @c4 Qf6 12
Qxd6 e3 (12...®b7+ 13 Qc7 @3a
14 d6!) 13 3311.23 £51m 14 Qc7
‘58a 15 d6 Q66 16 d7 @b7 17
@d5.
11 Qc8!
Once again 11 Qc6? was wrong,
because of 11.... e4 12 @c4 58b4+=;
ll Qb7 5305+;
ll...e4
Or ll...@b4 12 Qb7.
12 @c4 ®b4
Or 12...Qf6 13 Qd7.
l3 Qb7 53d 14 a6 Qd7 15 37
@c7 16 55b6+ Qd8 17 @d5 @338
18 QbS! Qd7 19 5316+ s 20
@xe4 d5 21 QCS d4 22 Qb7 d3 23
@xd3 Qd7 24 523c5+ Qd6 25 @3214!
@c7 26 @b6 Black resigned
Zotkin-Kudrin
Moscow, 1 965
2/ . In this lies White’s idea. After a
2 2 2/ ”/
,/2622 31?
6/77/3/22
\\\\“\
/ /1/32
// 21
2%
272/2
Knight Endings 55
Black’s task is to create yet
another passed pawn. And so there
followed...
l...fxg4 2 hxg4 h5
Now White has no time to win the
d3 pawn.
3 g5+ Qg7!
The king must retreat, otherwise
the white knight becomes highly ac-
tive: 3. .Qe6 4 58d4+ Qd5 5 5513.
On 5. .Qc4 6 f5 gf 7 g6 Qc3
Epssible IS 8 Qf4!. If 3.. Qf5, then 4
e5
4 @d4 h4
Necessary in view of the threat of
5 QB.
5 f5 h3
After this White himelf obtains a
protected passed pawn.
6M+Qfl
If 6...Q18, then 7 QB @e4 8
Qxd3 @xg5 9 @hZ Qf7 10 Qe3
Qxf6 11 Qf4, and Black is obliged
to let go of the h3 pawn.
7 9313 Qe6 8 Qd2 Qf5
The king hurries to support the
h-pawn, while the f-pawn will be
watched by the knight.
9 17 @d7
Of course not 9...®e6 in view of
10 ®d4+.
l0 Qxd3 Qf4 11 Qe2 Qg3
It seems it’s all over. The knight
13 is doomed, whereas the white
king is too far from the g6 pawn.
But White finds a study-like idea.
12 239.5 @318 13 @d3 h2 14 @f2
Qg2 15 @hl
few moves the players agreed a
draw. The question arises whether
Black could have won. Instead of
5...h3, played in the game, he had at
his disposal the more effective move
5...d2!, pointed out by I.Zaitsev. Af-
ter 6 f6+ Q17 7 Qe2 (7 Qxd2
®b3+!) 7.. .h3 8 @313 @e4 9 l
decisive 1s 9.. 2812+ 10 Qxd2 @g4.
 3 Rook Endings

Rook and pawn against rook Emms-Riemersma

Gausdal, 1993
The most classical and primitive
(though not for everyone) endgame
which should not, it seems, present %
//’/////
\‘
any particular problem for players
of grandmaster rank. But John Nunn ////V//
\\\\\\

wrote an interesting book about

\

§\§\
\\

these endings on the basis of com-

é/fl
2/
\
puter analysis where he gave quite a
%/
\\\

few complicated positions. However

in everyday practice it is much more
%%>
/
/// // 7 /%
\§

simple and tragic (or more confus-

\\‘

ing). Knowledge of precise posi-

tions and methods of defence here
have exceptional significance. l...§a1?
Correct is 1...§a3! 2 @gZ §a2+ 3
The edge pawns @fl a4 4 Eh3+ @c2 5 Eh2+ @b1 6
§h3 ECZ 7 @el fibZ 8 §a3 Eb4 9
The most frequent case—rook and @dl @b2 10 §d3 a3 11 §d2+ @b1,
pawn against rook. and the a3 pawn cannot be stopped.
2 @gZ?
People have already programmed But now White misses a draw by
computers for this type of ending, 2 EgS! a4 3 Eg3 @c2 4 fig? @d3 5
and so for the right method of play §g3 @e4 6 §g4 @f5 7 §b4 @e5 8
we should now turn to the Endgame @gZ @d5 9 §h4 a3 10 §h3l when
CD! White reaches a well-known
position.
Here are a couple of characteristic 2...a4 3 @f2 33 4 Eb5+ $32
examples.
 Rook Endings 5 7

There was a quicker win by Tosié-Gyimesi

4...§a4! Yugoslavia, I 998
5 @e2 Ebl! 6 EdS @b2 7 §d2+
<$b3 8 §d3+ @214 9 §d4+ Eb4 l0
Eds a2 7 @d3 @b3 White ////
resigned. / a //
% /// a
Bagirov-Kraidman ////// /
Grieskirch en, 1 998 %§/// ”4,
//// """ flgfl /%%
/// %% ///,, /
////////////7
fiflééfi 1 E35?
Simply driving back the king first
// /// by l Ec5+ @b2 2 ECS Exh4 3 @d2
//<§ %// draws.

//// / fl//
l...§xh4 2 E218
2 §a7 is also enough for a draw.
2...§b4 3 @dl?
The decisive mistake. 3 §c8+l
@132 4 @d2! Ed4+ (4...a3 5 Ec2+!)
l...§c5? 5 $e3 fidS 6 Eb8+l saves him.
Correct is l...§g5, holding the 3...@b2 4 EdB a3 5 §d2+ @b1
draw as in the previous example. and White had to resign.
2 @d4 Eel 3 316+?
Simpler is 3 a5 winning easily. Herrera-Vasquez
3...?g5 4 §b6 Eal 4 $36 Cuba, 1998
4 §b4 led to a draw after 4...@f5,
and the king rushes to CS.
/ /
4...?1‘5?
Again he should go for the draw-
ing mechanism 4...Efl!? and check
// /Va%%
on the f-file.
ya afia//
5 a5 §d1+ 6 @c5 gal 7 @b5
@eS %/
On 7...§b1+ there is 8 @c6 Ea] 9
¢// %
@b6 §b1+ 10 @c7 gal 11 §a8
@e5 12 a6 @d5 13 §d8+ QCS! 14 / a ' ////
Eda, winning.
8 fih6! @d5 9 @b6 Ebl+ l0 @c7
27
Ecl+ 11 @b7 Ebl+ 12 §b6 Ehl 1...331?
l3 §c6 §h7+ 14 §c7 fihl 15 a6 He can achieve a draw by any
§b1+ 16 @c8 @d6 17 a7 Black move except this, including even
resigned. 1...§e7, but best of all 1s1.§e8!.
2 a4 @f6 3 @d6, Black resigned.
58 Rook Endings

Vyzhmanavin-Lerner Simferopol, 1988, where the same

USSR (ch), 1984 mistake was repeated! 2 Ec6! @gS
3 ECS! §h7 4 §g8+ leads to a draw.
It is nearly always necessary to at-
/
W% / 2/W2 tack the king from behind, and not
in front, because in the end the king
2: %W227
W2// I,/
will get the better of the rook.
2...§h7! 3 Eel
, /7/2: In the above-mentioned game fol-
lowed 3 §h3 $g5 4 <$e2 @g4 5
éW/W WW /fi fihl h3, winning because of the

/ / // 2 W
poor position of the rook.
3...h3 4 Ehl @gs 5 9:33 @g4 6
$12
If 6 §g1+ QM 7 Ehl, then
1 @d2? 7...§a7 followed by 8...§a2, 9...h2
There was a simple draw by 1 and then 10...@h3.
@b2 §g3 2 @c2 @d5 3 @d2 §a3 4 6...§f7+ 7 fig] §a7 White
@e2 etc. resigned.
1...h2! 2 @92 gal! and White Since after 8 §h2 there follows
resigned. not 8...?g3? 9 §g2+!!=, but
8...§al+ 9 Q2 fibl! with a decisive
Vladimirov-Rashkovsky zugzwang.
Chelyabinsk, 1 975

Novikov-Lalié
Manila, 1992
W W é W
W / W 22%
/ / / / 2 2 / 2
2 / / / 2:2/ 2/W/
§\\

2M/ W ’////
\
\\“

,/ 2
\\\\\

//
\

W////
1 Ed?
Correct was 1 §g8+ @fs 2 fif8+
@g4 3 §g8+ @h3 4 EgS and after
the advance of the h-pawn a draw 1s
achieved since the white king is cut
off only by three files.
1...h4 2 fie3??
Interestingly, this mistake is typi-
cal for grandmasters. A mirror im-
age of this position was encountered
in the game Dvoiris-Kovalev,
W%//fl%%/
2% 2%2
It seems that it is not easy for
Black to defend himself, but he
finds his only saving resource:
l...§h8! 2 @b7 @d5 3 §g4
After 3 a6 QCS 4 Ebl §h7 the
draw is inevitable.
3...@c5 4 §g7 Eh6! 5 a6 §b6+ 6
@217 Ebl 7 §b7 Eal Drawn.
 Rook Endings 59

Kamsky-Karpov champion does not ‘go for’ for the

Linares, I 994 side.
1 @f4?
Correct was 1 @f6, threatening a
// check from g4. If now 1...a3, then 2
WWW/ §g4+ @d5 3 EgS! with a drawn
//////// pawn ending.
1...§e8! 2 QB @d3 3 Ebz fif8+
WW W WWW
/M Cutting off the king looks a
(W WW
/ deadly blow for White.
W/ W WWW 7
4 @g3 @c3 5 Eb7 Efl 6 EbS
After 6 @gZ Eal the position is
W
/W
W lost.
6...§al 7 QB fiv 8 @e3 fihZ 9
§c8+ @b2 10 Eb8+ @cl!
What system of defence should he Clearly not 10...@a1? because of
choose: wait for the approach of the 11 §a8 313+ 12 d a3 13 ad
black king to the 34 pawn or adopt with a draw.
the more active method. It is clear 11 §c8+ @b1 12 §b8+ EbZ 13
that the second solution is more ef- gas §b3+ 14 @d4 a3 15 @c4 QbZ
fective, but it requires accurate play. And in this theoretical position
1 §f6+! @e4 2 Eg6! $93 3 Eg4! White resigned.
The main thing is to maintain con-
tact with the opponent’s pawns. Rook endings
3...a3 4 §g3+ 99.4 5 EbB §a2+ 6 Several pawns on one flank
$g3
Obviously it is better not to go to This type of ending is met quite
the first rank. often in practice. With various posi-
6...?d4 7 §f3 gal 8 @gz tions of the pawns on both sides in
And the opponents agreed to a the majority of cases the weaker
draw. side should hold a draw even in
Euwe-Alekhine
those instances where his pawn
World Championship (m), 193 7 structure is weakened.

Beliavsky-Spraggett
Elista (01), I998
/ / W /
W W WW WW WW
W/W WWWW W/gé
// WlW
W W W W WWW W8
:W//
WWWWW/WW / ///
M M
WWW W //

White is in a dilemma, whether to

W 5W 4/

go forward or back, and the 'world

60 Rook Endings

1 M? 16...Qg2 17 Qe2 Qh3 Drawn.

Correct was 1 g4 hxg4 2 hxg4 and
then 3 g5, shutting in the black king, Bagirov-Berzinsh
whereas 1 B was also a possibility Riga, 1998
with the future transfer to another
type of classical position, looked at
in the next section.
1...Qg7 2 figs Qh6 3 g4 // 7
Now the plan with B promises
less since it is not clear how Black’s
/%/
/ ///‘
///,/ //%/
/////
position can be breached if White,
after an exchange on B, puts his /// W
own pawn on e5.
3. .hxg4 4 Exg4 395 5 Qfl Qh5 / ///
6 Ef4 §e6 7 Qe2 EeS 8 Qel!
White gains a tempo—on 8 Qd2 %/
there is 8mfif5.
8...Qh6!! Black demonstrates a clear and
Black has calculated the pawn instructive way to realise his
ending clearly. advantage.
9 Qd2 EfS! 10 Qc3 1...Ed5! 2 Q12 EgS 3 Q13 f5
Rather better was 10 Exe4 Exf2+ An important link in the plan:
11 Qc3, but even here there are no now White’s pawn fortifications are
real chances of a win. quickly smashed.
10.nfixf4 11 exf4 Qh5 12 Qd4 4 gf ef 5 Qf4 Et 6 §b7
Qxh4 13 Qxe4 figS 7 3217 Qh7 8 §b7 Qg6 9
§b6+ Q7 10 §a6 g6 11 QeS §g3
12 §a7+ Qe8 13 Qf6 h5 14 EM

/2/%%/r/
////%//
////// Also there is no salvation in 14 e4
fxe4 15 Ee7+ Qf8 l6 fixe4 g5.
l4...Qd8 15 337 M 16 §h7 h3
17 Ea7 M 18 Eh7 Eg2 19 QeS g5!
20 Qd6 Qc8 White resigned.
/////
////// //¢:§// Madsen-Hansen

///?3//.
/”/ corr, I 974

On 13 Qe5 correct is 13...Qh3!. /////

///
//
13...Qg4 14 Qe3 ////
Clearly not 14 Qe5 QB.
l4...@h3!!
2/
é/IVI' i/’///

Only this leads to a draw. //%7... Q? ”m/

l4...Qh4? is no use due to 15 f5!. //

15 $13 @112! 16 Qe3
After 16 Qg4 Qg2 l7 Qg5 there ///// / //
is the only move 17...Qf3!!.
 Rook Endings 61

The position to all intents and pur- 1 M!

poses concurs with the previous He cannot allow Black the possi-
one, and here 1 Ee4 Ecl 2 Qg4 bility of playing g6-g5.
should be played. But White l...Eb8
chooses another system of defence. After 1...e5 follows 2 Qg2 Ec2 3
l Eb4 Ecl 2 E34 Qg7 3 Eb4 g4 EC] 4 g5+ Qf5 5 Ef7+ Qg4 6
Efl+ 4 Qe3 Ef6 Qxh4 7 Exg6 winning.
Here already no help is 4 Qg4 h5 2 Qg2 Eg8 3 g4 g5 4 h5
5 Qh3 Eel 6 Eb5 Ee4 —+ This is more accurate than 4
4...Ef5 5 Ee4 hxg5+ Exg5 5 Qg3 e5 6 Eh6+ with
After 5 Qe4 g5 6 hn hn 7 some chances of victory.
Ebl Qg6 8 Ehl Ef2 9 Qe3 Ea2 10 4...Ea8 5 Eb7 Eh8 6 Qfl!
g4 Ea5 11 Qe4 Ea4+ 12 QB Ef4+ Now the king can proceed to the
13 Qg3 Ee4 Black wins a pawn. other side.
5...h5! 6 Qe2 g5! 7 hn Qg6 8 6.. EgS 7 Qe2 E38 8 h6 Eal 9 h7
Qe3 En 9 Q12 Qf5 10 Ee3 Ehl 10 Qd2 Qe5 ll Qc3 Eh4 12
After 10 Ef4 the pawn ending is Qc4 Eh2 l3 Qc5 Ehl l4 Qc6 Qf6
won. 15 Ed7!
10.. .Eg7 ll Eb3 Making a future path for the king.
Or 11 QB Eg4! 12 Qf2 h4 13 15...EhZ 16 Qc7 Ehl 17 Qd8
gxh4 Exh4 14 Qg3 Eg4+ 15 QB Black resigned.
Ef4+ l6 Qg3 Qg5 17 Eel EB 18
Ee4 Qg6!, and the king works his Beliavsky-Hodgson
way round to d5. In a previous ex- Cacak, I 996
ample the king also had the possibil-
ity of attacking the h-pawn in this
way
11.. EM 12 Eb5 h4! 13 gxh4
Exh4
White resigned, since the pawn
///%//%//7/%
///

ending after 14 Ea7 QxeS 15 Exf7

Ef4+ is lost. / %7 W %/
Huzman-Mikhalchishin
/. 44%
//
Lvov, I 985
&/ lllll lllI/I/

7//

////// l Ef6
Stronger would have been 1 h4!
Qg7 2 Ec6 f5 3 Bl, creating weak-
% 7////// nesses in Black’ s pawn structure.
1.. Q.g7 2 EM f5 3 h3 Qg6 4 g4
//////226
hxg4 5 hxg4 fxg4 6 Qg3 e3!
V///// Black’s only chance lies in the

flair/a creation of doubled pawns in

White’s position.
7 Exg4+ Qf5 8 Ef4+ Qg6 9 fxe3
Eb3 10 Qf2 QgS ll QB Ea3
62 Rook Endings

After 11...§b8 12 @e4 §e8+ 13 One extreme but typical case

@d4 §d8+, adopting a frontal at- occurred in a practical position.
tack, secures Black a draw.
12 EfB gal 13 @e4 §a4+ 14
@ds §35+ 15 @d4 334+ l6 @c5 Szabé-Keres
$215+ 17 @b4 E212?! Moscow, 1 95 6
Again correct was l7...§e5! 18
EB :68 19 @c4 §c8+ 20 @d3
§d8+ 21 903 368+ 22 @d2 §d8+ /
23 9e] §a8 24 @rz £34! with a
draw. / ’2 i‘
18 :12 §c2 19 e4 @gm
// 2 /
Rather better, though already in-
sufficient, was l9...@g6. /// $2/
20 e5 @gS
If 20...@g3, then 21 @b3! ECS 22
2/ E//&///
§f5 @g4 23 e4 with a win. 2 //2 2’2
///
21 @b5?
Correct is 21 §g2+1 @h6 22 @b5
§c8 23 e6 §e8 24 @c5 fixe6 25 Black has great problems and it is
@d5! with a theoretically winning quite complicated to transpose into a
position. drawn position with f and h-pawns.
21...@g6 22 e6 Ecs 23 EB l...h6
If l...hS, then 2 g5+ @f5 3 §c6+,
and there is no defence against 4

//22/ ////
Ef6.
2 2 Eh7! §d6 3 h5!
Possible is 3 Exh6 @g7 4 g5 fol-
lowed by 5 f4 and h4-h5.
2/3 22 3...? 5 4 §g7 QM 5 fin §d2+
22 6 @fl g3 7 §f61 332 8 @el fihZI
2 //////
2/: 9 @dl! Eg2 10 @c1 mm 11 @b1!

///// 2 The king moves further away

from its opponent.
ll...§g2! 12 Exh6 @xfB 13 §g6
@e4 14 h6 EhZ 15 g5 @d3 16 Eg7
23...?g7? @c3 17 h7 @b3 18 §b7+ Black
Missing the typical drawing resigned.
mechanism 23...? 5 24 e7 §e8 25
§e3 QM! 26 Ee6 f5. And here we see how difficult it is
24 e7 §e8 25 §e3 $17 26 @c6 even for a class grandmaster to
fixe7 27 Exe7 @xe7 28 @ds Black defend a classic position.
resigned.
 Rook Endings 63

Nikolié-Ftécnik Better than 14 fib6, played by

Pula, 1997 Capablanca. White’s plan, with the
rook on the 8th, is to to play f5-f6,
and Black’s defence, with the white
king on 12, is to play ...h6-h5.
/ l4...§b~3?
7/ / Correct is
f4-f5.
14...Ee3!, hindering

// /// A 15 @fZ?
Correct 1s 15 §d8l with the threat
of f4-f5-f6, 6g. 15. .Eb2+ 16 @g3
§b3+ l7 @h4 §e3 18 Ee8! and
////////%//
/
f4-f5.
15.333 16 Ed7!
16 §d8 is bad because of 16...h5
1 ECS! 17 g5 h4 18 §d7 @g6 19 §d6+
Correctly preparing for h4-h5 in @g7 20 §h6 h3 21 15 E35 with a
order to take on h5 with the rook. draw.
1...§3l 2 h5 §g1+ 3 Qt? @g7 4 l6...§32+ 17 $13 333+ 18 @e4
fi35 gt 5 Et 331 6 Eds Egl 334+ 19 EM!
Not yet a mistake but more This 18 the point of deploying the
technical is 6...§a4, cutting off the rook on the d-file. After 19 @f5
white king. @f8!, according to an analysis by
7 §d6 @h7?? Kopaev, the game is drawn.
A serious mistake—-—correct was to 19...§al 20 f5?
attack the e3 pawn by 7...§a1 8 @e4 Correct now was to move aside
§a3 9 §d3 EaS 10 §d5 E33 11 with the rook 20 §b4l fig] 21 @B
@d4 §b3 or 7...§e1 8 e4 Efl 9 @g3 gal 22 Eb8 §fl+ (after 22...§a3+
§g1+ 10 @h3 Efl, and the rook 23 @g2 the king goes via g3 to M)
does not have the 16 square. 23 @e3 Eel+ (23...§g1 24 f5!
8 e4! §fl+ 9 @g3 §g1+ 10 @h3 Exg4 25 f6+ @h7 26 e6!) 24 $12
Efl 11 Ef6! @g7 12 e5 §e4 25 QB Eel 26 §b7l $18 27
EM @g7 28 @f2 gal 29 §b8 h5 30
g5 M 31 $13! h3 32 Eb2 winning.
20...§el+ 21 @d5 figl?
// //'/
///, This position is from the game
/ Duras-Capablanca, New York 1913
(with colours reversed), and here
// 21...@f8! 22 $d6 331 gives a draw.
/ /’//
22 @d6 gal 23 §c4 E38 24 §c7
/ /////////
// 336+ 25 @e7 E34 26 e6! fxe6
Or 26...§xg4 27 @d6 winning.
// ////'7
%/
27 f6+ @g6 28 f7 EM 29 f8=
Ef 30 @1118 e5 31 fic4 Black
resigned.
White obtains a winning position
which is in every textbook. Even with a very lucky author and
12...§f3+ 13 9g £33 14 fid6! a passed e-pawn it’s drawn!
64 Rook Endings

Novikov—Beliavsky and the plan of the stronger side

Graz, I 997 consists of creating a passed pawn
on the d-file which promises him
great chances of success.
/
L WWW WA
// / Bagirov-Veingold

W a W, W Tallinn, 1981

/W W/W ‘\\
W
W,W /W/¢/gW W W / W/
WW 3W
WW WWW w/
W W W /{‘W
WW Wis/W
1...§c7 2 EaS @e6
He had to try the plan with the ap-
//
proach of the king via h6.
3 §e5+ @f6 4 E35 @g7 5 h3
@h6 6 fies §c2+ 7 @g3 @hS 8 §e7
Wé /WW
W /W/WfiW
/g
h6 9 fies §c3+ 10 @112 Eel ll g3!
After 11 @g3? Efl 12 EaS there 1 Ea]
is 12...g5! 13 fixf5 e3 with a win. With the idea of defending the
Now, however, despite the ‘cut-off’ pawn and freeing the king.
king from the first rank, Black does l...@g6 2 fifl fiaz
not succeed in improving his After 2...f5 3 gal fxg4+ 4 hxg4
position. the e6 pawn would be too weak.
11...§c2+ 12 @gl ECS 13 @g2 3 @e4 §e2 4 @es @g7 5 e4!
318 White’s plan is clear—after
How else to promote” .g6-g5? preparation to play d4-d5.
14 @fZ g5 15 @e3 g4 16 M Ef6 5...@g6 6 f3! §e3
17 @f2 @g6 18 §e8 336 19 fies h5 If 6...@g7, then 7 @d6 and d4-d5,
20 @fl @f6 21 @f2 §a3 22 §e8 when 6...§h2 is too late because of
gas 23 @e2 @f7 24 Ens @e6!? 7 d5 ed 8 exd5 §Xh3 9 d6 Ehz 10
But even the sacrifice of two Edll.
pawns does not help. 7 d5! ed 8 @d4! 333 9 ed
25 Et §a2+ 26 @e3 §a3 27 @f6 10 Eel!
$12 @d5 28 Exf5+ @d4 29 EeS Cutting off the king is more im-
e3+ 30 @fl §a1+ 31 @g2! Ea2+ 32 portant than a pawn!
@gl @d3 33 f5! §a1+ 34 @g2 e2 10...§xf3 11 d6 fifZ 12 QCS
35 f6! and it’s a draw! §c2+ 13 @b6 §b2+ 14 @c7 §c2+
15 s §c3 16 d7 Exh3 17 @e8
Rook endings of five pawns Ed3 18 d8=Ԥ fid 19 @d h3
against four are a rarer formation, 20 Eez Black resigned.
 Rook Endings 65

Ehlvest-Polugaevsky Complicated Rook Endings

Hanninge, 1990
It is interesting to see why grand-
masters cannot hold slightly inferior

2///2//’::‘;/ positions in which the method of

playing for a draw is well known.

2 2/ 2/ Dautov-Alterman
Germany, 1 998

/2
2//21’2
2/ y/ 32/ 2/ 2%
/ 23 22% 2 AK?”
2 2 2/12
1 h3! 22 // / //
Preparing ficS-c4 and g3-g4. 2 //// 2:32
1...f5 2 M!
It is necessary to fix the weak-
//2 ///g>§&2
nesses and not allow g6-g5 and W// ////
h5-h4.
2...@f6 3 E216
Unfavourable for White is 3 e3 In such positions ‘standing still’ is
e5 !, and the d3 pawn is weak. not to be recommended: there are
3...f4!? 4 @g2 §b2 5 QB fxg3 6 two active plans:
fxg3 §b3 7 335 §b4 (a) Petrosian’s plan l...f6!? and
Not allowing d3-d4. g6-g5, creating a weakness on M;
8 ECS e5 9 §c6+ @g7 10 §c4 (b) leaving the pawns alone by
§b6 11 E214 @f6 12 @e4 @e6 13 e3 l...§g2 2 @134 §b2 3 a4 §b4+ 4
There is no other plan for White. @e5 Eb3, and White does not suc-
13...@f6 14 E31 §b4+ 15 @d5 ceed in advancing the a-pawn very
§b5+ 16 QM §b2 17 336+ @H 18 far before the B and g3 pawns fall.
335 §c2+ In the game followed:
On 18...?f6 White’s plan would 1...§c2? 2 a4 332
be 19 d4 exd4 20 exd4 §g2 21 EgS, After 2...§c3+ 3 @e4 §c4+ 4
and then the advance of the d-pawn. $d3 EM 5 a5 §a4 6 f4 Ea3+ 7
19 @ds figZ 20 3217+ @f6 21 @c4 fixg3 8 §b6 §g4 9 a6 Exf4+
336+ $f7 22 @xeS Ex 3 23 Ea7+ 10 @b5 §f5+ 11 @c6 §a5 12 @b7
@g8 24 d4 Exe3+ 25 f6 §b3 26 White wins.
d5 fif3+ 27 $xg6 ($18 28 EM! §f4 3 <$94 333 4 35 3214+ 5 s
29 d6! @e8 30 @t §d4 31 d7+ E33 6 @c6 §c3+
@d8 32 EN §d5+ 33 $g6 §d6+ If Black takes the pawn 6...§xf3,
34 Ef6! Edl 35 h5 Black resigned. then after 7 §b6 Exg3 8 @b7! §a3
66 Rook Endings

(8. g5 9 a6 §a3 10 hn +—) 9 a6 Material is nothing—activity (in

f6 10 a7 Exa711Qxa7 g5 12 Eb]! View of the passed pawn)
gxh4 13 Qb6 Qg6 14 Qc5 Qf5 15 everything.
Qd4 Qf4 16 311+ Qg3 17 Qe3, 6...hxg3 7 hxg3 fin 8 a5 fig] 9
White wins, but also the move in the 3212
game does not save him.
7 Qb7 3213 8 f4 gal 9 E218 Qf6
10 a6 Q15 11 §f8 §b1+ 12 Q37 f6
13 EbS Eel 14 §b31
An important reconstruction. The
7; 2;;7/
7/
rock has two defensive functions: to
77/“7.7/
defend the g3 pawn and control the
b-file in order to flee the king. 7 7/ 7
l4...Qg4 15 Qb6 §e6+ 16 QaS
§e7 17 E213!
77 7 /
One more function: to support the E/ 7&7 7 /;'_/
a6 pawn. 7:7
_/
17...g5 18 fn fn 19 hn
§e5+ 20 Qb6 £n 21 a7 figs 22 9...§b1+?
§c31 §g6+ 23 Qb7 §g7+ 24 Qb8 Better at once 9...Qd6!, not allow-
Black resigned. ing in the white king, while on 10
a6 Ebl.
Mikhalchishin-Eslon 10 Qc5 fibs ll 36 g5 12 Qc6
Kecskemet, I 983 Qf6?
In such positions it is useful to
prepare to push away the white
king. 12. .Qe6 13 a7 fiaS 14 Qb7
/ 77777 Exa7+ 15 Qxa7 Qd5! with clear
/7 7 7 7 chances of a draw.
13 a7 E218 14 Qb7 311217 15 Exa7
77
/7 7 7 g4 16 Qc6 Qe5 17 QcS g3 18 Qc4
7 7 77 /7 Qf4 l9 Qd3 Black resigned.

//
33/; 7/7 /7
Keller-Mikenas
corr, 1992

l...§a$+? 7 7% /
In analogous situations where the 7/7 7/
pawn has not advanced to the fourth
rank, it is correct to activate the
,7///7/7
king, but the rook gives a frontal 7/
check.
2 Qe4 Qe7 3 QdS f5 4 Qc5 h4?!
7 7/7 7
Again correct is 4. W§a8 prepar-
ing to meet 5 Qb4 with 5.. .fib8+. E7 //7 /7’2
53?
5 Qb4 §e3 6 a4! Q77
 Rook Endings 6 7

Black must lose because of the 5 QdS 3213 6 a6 3x13?

weakness of the e4 pawn.
1 a6! g6
After l...h4 2 gxh4 Ea1+ 3 Q62
§a2+ 4 Qe3 §a3+ 5 Qxe4 Exh3 6
f4! Exh4 7 §a8 fihl 8 EfS and 9 a7
wins. In the event of passive de-
fence by Black, White pushes his
pawn up to a7, then the king goes to
e5, the Black rook must keep watch
from the a4 square. Then White
plays Qf4 and, on Qh7, QfS with
a W1n.
2 E218 Qf7 3 a7 Qg7 4 Qel Qh7
5 l Qg7 6 c Qh7 7 l 336
8 g4!
Determining the pawn formation
on the flank.
8... hxg4 9 hxg4 Qg7 10 g5 Qh7
11 Qb2 Qg7 12 Qb3 Qh7 13 Qb4
Qg7 14 Qb5 332 15 Qb6 Ebz l6
Qc6 §c2+ 17 Qd6 3212 18 fic8!
Exa7 19 §c7 and the pawn ending
is easily won. Black resigned.
Lerner-Dorfman
Tashkent, 1 980
7
/ ///7///%>%7
%7///
% 7 //
7 //
This is a decisive mistake-—after
6...Qf6 7 Q06 EXB 8 §b8 §a3 9
§b6+ Qf5 10 Qb7 Qg4 11 a7 Exa7
12 Qxa7 n3 13 §b4 f6! there is
an easy draw by ...g6-g5.
7 Ebs 333 8 §b6 Eni 9 Qc6
333 10 Qb7 g5 11 hn M 12 a7
h3 13 a8=w Exa8 14 Qa M 15
§h6 f6 16 Exh2 fn 17 fifZI!
We advise paying attention to this
move.
7 77
///%
777/
77/77
7/7/37
Now the white king avoids being
shouldered away and hurries over to
the opponent’s pawn, therefore
Black resigned.
Kon0pka-Shcherbakov
Pardubice, 1996

777/ Z? 77 7 /
7/ 7 /
7/”
‘7/
////
:6
/ 7/
1...Qg7?
Passive play—correct is 1...Qe6
——or 1...§a2 detaining the king on
f6.
2 a5 §a4+ 3 QeS 333 4 Qe4
234+? This position resembles the previ-
Simpler is 4...Qf6! 5 a6 Qe6 giv- ous one, but there are a number of
ing a draw. important differences. Black has a
68 Rook Endings

doubled pawn, and the greater num- l g4! g5 2 a4 Eal 3 a5 @gG 4 a6

ber of pawns creates more prerequi- QM
sites for a White victory. White also wins after 4...h5 5
l...f5 gxh5+ @t 6 §a8 @g6 7 a7 @f5 8
Worthy of attention is 1...e4!?, for @g3 with the idea 9 f3.
the present maintaining the pawn on 5 E218 @g6 6 a7 @g7 7 @g3
t7 so as to defend the king against §g1+ s @112 Ea] 9 @g2! h6
checks from the flank. After 9...?h6 White wins by 10
2 §b8 e4 3 h3 h5 4 M §b1+ 5 E18 fixa7 ll Exf6+ @g7 12 §e6
@e2 §b2+ 6 @fl Eb1+ 7 n fibz :34 13 @g3 $17 14 §h6 Qg7 15
8 b4 Eb] 9 b5 @go 10 b6 @h7 11 §d6 337 16 :66 §a4 l7 h4.
f4! exf3+ 12 @1113 @g7 13 b7 10 @h2 @h7 ll £13 @1137 12
On 13 @f4 there is 13...@f6! with fif @g7 13 §e6 £214 14 @g3 §b4
the idea ”EH. 15 M gxh4+ 16 @xh4 2214 17 @g3
13...§b4 14 Qez Eb] 15 @d2 §b4 18 fies @g6 19 is
§b3 16 @c2 fibS 17 @c3 §b6 18 With the idea of fifS-f4, f2-f3.
QM Ebl l9 @c5 §b3 20 @c6 l9...§b8 20 9“ E88 21 :65!
Ec3+ 21 @d7? Exes 22 @xeS @g5 23 @xe4 @xg4
Stronger is at once 21 @b5 §b3+ 24 @d3! Black resigned.
22 $05!.
21...§d3+ 22 @e8! Kozul-Nikolié
After 22 @xe6 Exe3+ 23 @f Sarajevo, 1993
EB+=.
22...§b3 23 @e7 Eb6 24 @d7
§b3 25 @c6 §c3+ 26 @b5 §b3+
27 @c5! Ebl 2 2 21/
If 27...@h7 winning is 28 $06
§c3+ 29 @d6 §d3+ 30 @e7 Eb3 31
2%2/.%
Qf6L
28 9%! Ecl+ 29 @d6 Eb] 30 % 2/2M2
@xeG fibS 31 Eds Black resigned.
/ 2 2y/ % /
/
Andersson-Hiibner
Ter Ape], 1997
1 a7
White sees no chance of going

/
2///” /2/
’4‘
with his king to the queenside to
help his a6 pawn. For this he has to
2 sacrifice the f2 pawn.
// 2 2 l...@e5
Also not bad is l...@g7 with the
/ 2é2g/ idea of2...f5 or 1...Ea3.
//2 2 13 E213 3 $12 Eal 4 @e2 @f6 5
722/
222/ f4 Ea3?!
The simplest way to achieve a
draw is 5...?67, and there is no way
for the king to approach.
 Rook Endings 69

6 Qd2 §32+ In the game after 7...g6? 8 §h8

A move such as 6...Qe7 can al- Black resigned.
ways be made. 8 @d2 £33 9 Qe3 :35 10 Qxd3
7 Qc3 333 8 Qc4 EaS 9 Qb4 §a4 ll Qc3 gal 12 Q04 §a2 13
Eaz 10 Qc5 gal 11 Qc6 Qc5 gal 14 Qc6 Ecl+ 15 Qd6
No use is 11 §h8 Exa7 12 Exh6+ Preparing to meet QeS with the
Qg7 13 fn Ea5. move Ea5, but now he threatens to
11.. .EaZ 12 QCS E216 13 Qd4 tranpose to a pawn ending.
$214+ 14 Qd3 333+ 15 Qez EaS? 15...§a1 l6 ECSE 3216+ l7 §c6
Completely losing his sense of fixa7 18 §c7+ Exc7 19 Qxc7
danger—why not 15.. .Qg7?
16 EhS! 2x37 17 Exh6+ Qg7 18
fn §a3 19 Qf3 E35 20 h4 EbS
21 Ef6 Black resigned.

Yermolinsky—Seirawan
USA, I 997

.//
WWWW/8W
IIIIII

///Q/ //WW/
3W W?a
//4/
[W/
f / Note an important fine point—the
tempo g2-g3 is decisive. With the
W/W// pawn on g3 it would be a draw.

// /a§/ /////
l9...Qe7 20 Qc6 Qe6 21 Qc5!
Destroying the opposition.
21...Qf6 22 Qd5 g6 23 hxg6
n6 24 Qe6, winning. But with
lufiv? the pawn on g3, 24. .QhS 25 Qf
Black does not suspect any danger is stalemate.
associated with the the white pawn
on hS. Therefore correct is 1...f4!. Vuj ala—Smith
2 f4! d4 corr, 1993
The other chance is to hide on h7
and wait with 2...Qh7 3 Qfl. But
then the white king enters the game
via b1, attacks the d5 pawn and as a
result of zugzwang captures the d5
W ...
and f5 pawns. lllll

3 Qfl d3 4 Qel! Qf8 [[[[[[[ //

Losing is 4.3n 5 §d7 §a2 6
a7 Qh7 7 l, and the white king
W/WW
captures the f5 pawn, and then goes / // // W
to b8. E/W. W ,,,,,.8543?
5 §a8+! Qe7 6 a7 Qf7 7 l W W jWQW
3214
70 Rook Endings

1 g4! not have enough tempi to save

Fixing the opponent’s weak-
nesses.
l...h6 2 QgZ E213 3 h4 Qf6 4 h5!
Qg7
Underlining White’s advantage.
4...Qe6 5
5 hxg6 n6
If 5...fxg6, then 6 Ec7 with the g4 Qe6 8 Qf4 there arises a theo-
idea Ea7,
6 f3 f6
Better is 6...Qg7 7 Qf2 Qg6 8
Qe2 Qg7 9 35 Qg6 10 Qd2 h5!=.
7 Qg3 £34 8 EdS! §b4
Now the continuation 8...Qg7 9
f4 Qg6 10 fid7! f5 11 Ed6+ Qg7
12 a6 §a3+ 13 Qh4 §a4 14 g5
leads to victory for White.
9 EdS! $214
It is too late for 9...h5 10 gxh5+
Qxh5 because of 11 §h8+ Qg6 12
§g8+ followed by Eg4.
10 §g8+! Q17 11 gas Eal 12 f4
Qg7 13 a6 3212 14 f5! gal 15 Q13
§a3+ l6 Qe4 §a4+ l7 Qd5 h5
There is also no salvation in
17...§xg4 18 Ea7+ Qg8 l9 §b7
and then a6-a7 +—.
18 gt 335+ 19 Qe4 Black
resigned.
Salov-Malaniuk
USSR (ch), 1988
himself.
2 §b4 Q35
At first sight 2...gxf3+ 3 QxfB
EgS seems sufficient for a draw.
Actually, in the variation 4 §d4
Qxb5 (4...Qc5 is useless because of
5 Ed7) 5 §d5+ $66 6 fixf5 Qd6 7
retical position where it is Black’s
move—and consequently a draw—-—
by 8.. H§a8 But White plays the
more refined 4 Eb3! (threatening 5
@114) 4.. .Eg4 5 fid3 @b 6 Eds
Qc6 7 Exf5 §g8 8 g4 Qd6 9 Qf4
Qe6. The same position, but with
White to move, 10 g5, and this is
already a win.
3 §b3 Qb6 4 Q12 338
Or 4...gxf3+ 5 Qxf3 §g8 6 Qf4—
Black does not stand his ground.
5 fxg4 fxg4 6 Eb4 318+ 7 Qe2
§e8+
In the variation 7...Qc5 8 Exg4
Qxb5 there is 9 314! §e8+ 10 Q13
Qc5 11 g4 §g8 12 Ef6.
8 QfZ Ef8+ 9 Qe2 §e8+ 10 Qd3
QCS ll §e4 §d8+ 12 Q63 Qb
13 Exg4 QcS 14 §g6 Qd5 15 Qf4
:18 16 Qg4 6:34 17 §e6+ @115 18
EM! with an advantage that is quite
easy to realise.

%,4:4
, , , ////7 Guseinov-Beliavsky
Pula, 1 997

//// //,
4/64 44 74/54/
4/4///// 454665
4/4/4
//
//%/%
1...Qb6
//
//@/
5/ 45474745
/ 4

5/ // /// 4.
// /,
//

/
The sealed move. It is easy to cal-
culate that on 1...§xb5 Black does
/////////
 Rook Endings 71

White has a weakness on C4, but No help is 3 gxf4 QfS.

this is small and Black needs to 3...Eb2+ 4 l fxg3 5 Ef6+
create another. It is necessary at least to force
1.. .h5! 2 Qf4 back the black king.
After 2 gt Et 3 Qg3 E05 4 5...Qg7 6 Exe6 Eh2!
£03 follows 4.. .Qd7! with the threat It is precisely this pawn which is
Qc7 and" .d5. important.
2...g5+! 3 QB h4 7 Ed6 Exh4 8 Ed
Now there is a weakness on h3. After 8 Qg2 Eg4! 9 Ed M fol-
4 Edl lows h4-h3 and g3-g2.
Preventing Ee6-el and hi. 8...Ee4! 9 e6 M 10 EhS Exd4 11
4...Ee5 5 Ee1+ Qd7 6 Ee4 Ee5! e7 Qf7 12 Eh7+ Qe8 13 Qg2 Eg4!
7 Ed4 Qe6 Zugzwang. l4 l g2. White
Now the king is included in the resigned.
attack on the weak c4 pawn.
8 Ed3 Qe5 9 Edl f6! 10 Ed2
Ee6 11 Ee2 Qd4 12 Eel Ee3+ 13 Mednis-Gurevieh
Qg2 Eg3+ 14 Qh2 Ee3+ 15 Ed1+ New York, 1982
Ed3 16 Eel E13 17 Ed1+ Qe5 18
Eel Ef2+ 19 l Ef4 White
resigned.
7////
Mikhalchishin—Khmelnitsky /E//EEE
Sibenik, 1990 E////EQE
7///
%EEK”/, . ///
E Ey/
E_H

/E E
E EE E
In this rather untypical position
there is only one possibility for
Black to save the game.
1...e5!! 2 b5
After 2 bc QdS 3 Qd3 E04 4
Qe3 Qe5! Black achieves a draw.
White can do nothing against the 2...Eb4 3 Eb2! Exh4 4 b6 Eh8 5
break ...f5-f4. Qe4 Qd6 6 Qb5 Qd7!
1...Eb1! 2 Qf2 It was still not too late to lose:
Bad 2 Qf4 Eel 3 Ee3 (3 QB Ee4 6...E08 7 Ed2+ QeS 8 b7 Eb8 9
with the idea ...f4) 3...Edl, and Ed7 Qe6 10 Q06 with a win for
there is no defence. White.
2...f4! 3 E13 7 Qa6 Ea8+ 8 Qb7 E34!
72 Rook Endings

b7, and White wins; or 13..§08 14

b7 §b8 15 @c6 with the same re-
/////////%
77 sult) 13 b7 §b3 14 §e2+l (only
////
7 7A 7 7%
now, when the rook stands on the b3
square!)
White wins.
14...@f5 15 @c7, and

Interesting methods of defence

in rook endings

Now if White plays 9 Ed2+, then 7/7

Black has the possibility to trans- / M
pose into a drawn pawn ending by
9...§d4.
/////////
9 ECZH c4 /////////
If 9...@d6 10 §d2+ Ed4? 11 7/ ////
/////
Exd4+ cxd4 12 @c8, and White
queens with check. 9...§a5 10 §d2+
///
is also losing. 7/ ///////
10 §d2+ @e6 11 @c6 c3

In defending this typical position,

//7/ // exceptionally important is the pres-
77/7 ence of the two white pawns—the g

/%//7//// or h pawns do not win, the others do

quite simply, by advancing them to
the seventh rank. Knowledge of
7/sfl7// these types of position is extremely
important in the different variations
\\

when transposing to a rook ending.

///
Chaunin-Friedman
12 §e2+ @f5 13 b7 Moscow, 1 951
If 13 @b5, then Black achieves a
draw by the manoeuvre 13...§a3 14
@c4 £216!
l3...§b4 l4 EhZ 7,7 7 7%2%7/ //7/7 /

Nothing is changed by 14 @c7=.

7/ % /7
\i\\\\‘\\\

14...§xb7! 15 @Xb7 @e4 16 @c6

\

@d3 Drawn
However at the end of the game
analysis showed that in the position
in the last diagram, 12 EcZ! de- ///////%H
served consideration. Now
12...§a3 (12...§c4+ 13 7135 s 14
if //////7
 Rook Endings 73

Here White can win easily by 1 Here, an analogous winning

@xg3! 2 h4l, after which the B method is demonstrated by Shirov.
awn goes to f7. However White 1 b6! cxb6 2 Ehs Black
decided that simpler was 1 hxg3?, resigned.
keeping his pawns connected, but it
turned out that after l...g4+! 2 fxg4 Mokry-Pribyl
he cannot win even with an extra Olomouc, I977
pawn.
WW /
Polugaevsky—Parma
Sochi, 1965 W/W/W
WW W WxW
WeW W
\
W
W / //W W/WsW W
W W W W
W WW W We /W
W
///fi
W/
W
//
W/WWWWV
W //;W// After l...@h7? 2 @c6 £212 3
EdS! E1137 4 §d7+ Exd7 5 @xd7
g5 6 @e6! Black resigned.
And here in a classic game Black The method which helped White
resigned since he saw 1 h6, and to win is typical of such positions. It
thought there was no defence. How- is possible that Black did not even
ever after 1.. .Qg6! and 2. .Qh7! think about the fact that it was still
Black calmly makes a draw. possible to give up even a centre
pawn.
On the correct defence—
Shirov—Kramnik
1. Ea] !—the operation to transpose
Belgrade, 1999 into a pawn ending is not achieved
since after 2 @c6 Black checks
along the file until the king IS forced
41 ////% away from the a-pawn.
example: 2...§c1+ 3 @d6 §d1+ 4
For

W// @e6 Ea] 5 fidS (5 EeS? §a6+!)

WnW/WW 5...§a6+ 6 §d6 Exa7 7 §d7+ Exd7
/W W W// 8 @xd7 @f6! 9 @d6 g5! 10 hxg5+
@n 11 @d5 e3! 12 fxe3 §g4 with
WW/W a draw.
/W W W The right method of defence was
m
not found in the following game.
 74 Rook Endings

Holmov-Timoschenko Here a mechanism like 2...§el

Pavlodar, 1982 does not work, since there follows
first 3 §a3+l and only then 4 Hv.
There is also no win for Black after
LEO] 3 §a3+ 303 4 Ev §c2+ 5
fic Qxc2 6 QB Qd3.
By comparison with the game
Mokry-Pribyl this pawn ending is
favourable for the stronger side. But
in this case the weaker side achieves
// a draw: 7 Qf4 Qd4 8 QB g5! 9
hn fxg510 e5!.
%% Also here the sacrifice of a centre
pawn saves him!
10...Qxe5 11 Qe3 Qf5 12 QB
White was convinced that he with a draw.
would make a draw but, despite Black could play 12...g4~—-but
thinking about this position for then arises a theoretically drawn po-
more than an hour, he did not con- sition—or else he must try to hand
struct that ‘fortress’. This perfectly over the move to White which he
appropriate example is of interest to will possibly succeed in doing.
the theory of rook endings. But even in this case White
The fact that the pawn stands on achieves a draw after 1 g4! hxg4 2
f6, and not on f7, is of no signifi-
Qg3. Therefore Black must take the
cance. Now simplest for White was pawn at once.
1 g3! Qe7 2 QgZ Qd7 3 338. In the 2...Qxe4
end Black reaches the following Reaching the basic position of this
position. ending.

/ //,/ ///,
///// //%/////
$9

/ //a/ /////m///
,//// ////////
2/% /a//// / / ///
'1\\\\\\\\\\§

1‘2}

// /Q/ z/ ///<a%
\
s‘\\\\\‘

7],
§\r\\\

,,:%5

With White to move, Black wins a It is of no significance at all what

pawn: 1 Qh2 Qf3 etc. With Black the move is here. It is necessary
to move, after 1...Qd3 2 Qh2 Qc3 only to bear in mind that Black
Black wins in the same way as in should not play g6-g5, since after
the game Mokry-Pribyl. Then sim- the exchange of pawns, a second
plest is to give up the pawn at passed g or h-pawn will be created,
once—2 2218! which does not win.
 Rook Endings 75

The only path to victory is to cre- with a draw since Black cannot go
ate a passed f-pawn, but how can to the c-file with his king: 18...?c7
this be done? If Black manoeuvres 19 hn.
with his king to the d4, c4 squares,
etc, then White gives check along We return to the game Holmov-
the file. But even here White needs Timoschenko.
to take care. For example, after 3
§c8+ @d4 4 §d8+ @es mistaken is
5 fie8? @f5 6 §a8 g5 7 Ea5+ @g6
8 hn f5!.
After capturing on g5 Black man-
V/Vfl
ages to create a passed f-pawn, // // / /
which wins. Instead of 5 fieS? it is / // // 7
necessary to continue 5 338!
Now the above-mentioned ma-
:7 //.:<//
noeuvre does not work: 5...g5 6 ///
hn f5 7 Ea6! M 8 g6, and already
Black has to think how to save
/ / /&//
_’/ ////

j
himself.
On 5 §a8 he will try to send the
king to g7, so as then to play l0 §a3+ @xe4 ll §a4+?
..g6-g5: 5.. .Qe6 6 §a6+ @177 Also here still possible was 11
§a7+ @gB 8 Ea8+ @g7 @hZ and then g2-g3.
On this follows a check on the 11...@e3 12 Ea3+ QM 13 §a4+?
rank—9 §a7+, and after 9...?h6 @cs 14 gas f5! 15 §a7 14+! 16
the manoeuvre 10 336 prevents the @hZ Qt“ 17 Ea4+ $65 18 §a3
advance ...g6-g5. QfS l9 §a6 @g4 20 fixg6+ @xh4
There is nothing else for Black, 21 Eats @g5 22 Eas 114 23 zgs+
besides 10...f5. Possible then is 11 @115 24 E218 113! 25 gxh3 f3 26 :33
§a7 g5 12 §a6+ @g7 @e5 White resigned.
Now Black can choose two paths,
each of which leads to a favourable
result for him. Let us look first at Activity is more important
the direct 13 hxg5 M 14 gxh4 f4 15 than material
h5 f3+ 16 $3 fihl 17 h6+! QM 18
§a7+ @g6 19 h7 Exh7 20 §a6+! It has long been known that in
@n 21 QXB with a draw. rook endings activity is more impor-
The second path also leads to his tant than material. This means maxi-
objective: l3 §a7+ @f6 14 §a6+. mum possible active deployment of
Black is at the crossroads. the king and rook in coordination
After 14...?e5 15 hn h4 16 g6 with one’s own passed pawns and in
he might even lose. the struggle against the opponent’s
But 14...?f7 15 §a7+ @eS pawns. And even the very idea of
(15...?e6 16 hxg5! M 17 g6) 16 the priority of material will be fatal.
§a8+ $d7 17 §a7+ @d8 18 Ea8! Here are a few striking examples.
76 Rook Endings

Larsen-Browne Arbakov-Gurevich
Las Palmas, 1982 Moscow, 1978

\\\
\\\\\\‘:
///////

& §
§ :\§§ s
/ /s/g%
/ 27/4
/ %,/ §&
2/4;

\\§
/

\\
// / // E// ///
The black rook is hopelessly pass- l...§cl! 2 QB §c4 3 @e3 e5! 4
ive and White’s plan is to go with @d3 EM!
his king to h5, place the rook on b6 Clearly not 4...e4+? because of 5
and break up the black pawns with Exe4.
the pawn march f4-f5-f6. 5 (£6 e4 6 Eg3 3212 7 Eg4
1 @g4! $96 Or 7 M g4 8 h5 §a3+ 9 @f2
After 1...g6 winning is 2 Eb6+ fixg3 10 @xg3 e3 with a win
§g7 3 f5 §h7 4 @f4 and 5 @eS. 7. 333+ 8 $12 §d3 White
2 15+ @e5 3 EM! g6 resigned.
On 3...?d6 follows 4 @hS.
4 fxg6 fxg6 5 §b6l @d4 6 fixg6 “Active positions of the pieces in
Exb7 7 fixh6 rook endings are worth a pawn"
At a necessary moment the b7 ——Smyslov. Here is a classic
pawn is given up in exchange for example.
the win of a pawn on the opposite
flank. This results in a theoretically Capablanca-Tartakower
winning position. New York, 1924
7...§g7+ 8 @f4 217+ 9 @gs @e5
10 g4 EfB 11 @115 an 12 g5 @f5
13 §h8 and Black resigned.

In the following position the

white rock is firmly posted, but it is
immobile and passive. Black finds a
plan with a transposition to a pawn
ending, exploiting the position of
the rook on g4.
 Rook Endings 7 7

l Qg3! Exc3+ 2 Qh4 EB? Lilienthal—Smyslov

Materialism in such positions is Moscow, 1 941
simply ruinous. This position
sparked a great debate in 1998 in
the magazine 64. Master Goldin
maintained that 2...a6, intending to
create 3 passed pawn with maxi-
mum speed, gave Black drawing
chances. However after 3 g6 b5 4
axb5 ab 5 Qg5 b4 6 317+ Qg8 7
Ef b3—-—master Barsky pointed
out 7. .Eg3+! 8 Qh5! (8 Qf6 Eg4!
leads to a blind alley—8.. .b3 9
fid b2 10 EbS Qg7 11 Eb
Exg6 12 §c2 fid6 13 304! Qf6 14
Qg4 c6 15 Qf3 Qf5 16 §c5+ Qe6 There followed:
17 Qe4, gradually driving back the l...Qe4! 2 Ec f4!!
black pieces) 8 Ed Qg7 (8...§c6 By sacrificing a third pawn, Black
9 Qh6 fid6 10 Ea5! Ed8 11 fibS! creates cover against checks to his
with a win) own king.
3 exf4
Or 3 Qfl §a1+ 4 Qe2 13+ 5 Qf2
%/
§

I f’I4 §a2+ with perpetual check.

:§‘\

IIIII
/
3...Qf3 4 h3 Eal+ with a draw
four pawns down.

/ /%
// a 21/gg/// Kozlov-Mikhalchishin
Vladikavkaz, 1978
W13
/ /’2”
08x

////?/2’ lllll
\

9 Ed7+ Qf8 10 d5! (but not 10 f5

/////
M/
\

b2 11 g7+ Qg8 12 Qg6 §c6+ 13 f6

bl=§l with check!) 10.. .Qe8 (after
10.. .b2 11 Qf6 Q68 12 367+ Qd8
&% / /’//
13 g7 §g3 14 Eel White wins eas- //;%/%//fl//
ily) 11 £117 b2 12 fih8+ $127 13 //// //
§b8 §c2 14 §b7 §g2+ l5 Qf5 §c2
16 d6! Qxd6 l7 Qf6, and White
wins. 1 e4!
3 g6! Exf4+ 4 QgS 39.4 5 Qf6! A passed pawn in conjunction
The f5 pawn protects the white with a centralised king is worth two
king, but not the black one. pawns. But where does the capture 1
5...gQ 863g7+Qh87§xc7§e8 Exb3 lead? After 1. .fixg3 2 Qe4
h5 3 Qf4 §g4+ with a further“ .g5
7 QfgEe4 8 Qf6 §f4+ 9 Qe5 §g4
10 g7+ etc. White won. White is in a deplorable state.
78 Rook Endings

l...§xg3 2 e5 @g6 3 e6 §d3+ There is a very clear path to the

After 3...§e3 4 @d6 h5 5 §b8 M draw.
6 e7 the white pawn queens. l...aS!! 2 ba §d4 3 gb :34
4 @e5 h5 5 EbS! 4 Eb3
5 e7 does not work because of Or 4 §b6+ @eS activating the
SWQH, but now this is a threat. king.
5...Ee3+ 6 @d6 @f6 7 Ef8+ @gs 4...§xa5 5 Q“ $214+ 6 §b4
8 e7 b2 9 Ebs @f6 10 318+! figs 3x213 7 §b6+ @g7 8 §e6 §b3 9
11 §b8 Drawn. Ee3! Exe3! 10 91:93 @fl 11 @d3
@e7 12 QM @d6 Drawn.
Taimanov—Chekhov
Kishinev, 1976 A classic activating of the king
was seen already in this game:

Owen-Morphy
/2//§.//
///”/// London, I 858

// /2r2 /W
/é//,2 / 2 //
2/222; /2 2fi//
//2% ”/
2 //
The threat is 1 64+, and then 2 a4,
3 @g4. Therefore correct is
1...§d3! 2 Exb6 §a3 3 EbZ Ea4
//2 2:22?
...followed by ...g5-g4 with a
draw. 1...?f7!
After 1...§d3 2 §c8+ $h7 3 §e8
A frequently met theme is the
White has great chances of a draw.
sacrifice of a pawn to activate the
2 §c7+ @f6 3 Exb7 §d3 4 @f2
rook.
Exd4 5 @e3 e5 6 b6 EM 7 §b8
Barlov—Schiissler 99.7!
Hanninge, 1988 Black chooses a plan to liquidate
White’s passed awn. Inferior is
7...Eb3+ 8 @d2 66 9 QCZ, and it

2:2////,/ is not clear how to win.

8 b7 @d7! 9 Eg8 Exb7 10 gxg7+
{”7
/x/
2;y/ / @c6 11 §g6+ @cs 12 Exh6 §b3+
l3 QeZ e4
/% E /%// After the pawn sacrifice all
/ Black’s pieces have become more
//%§ active
advanced.
and his pawn further
”/////
2/ / / 14 §h8 @d4 15 EgS Eb2+ 16
@d1 @d3 White resigned.
 Rook Endings 79

Schlechter-Lasker succeed because of 3.. .h3 4 a5 §g2+

Berlin, 1 91 0 5%f1f46a6f37a7h2,anditis
time to resign
3...h3 4 Eb8 §g2+ 5 @fl §d2
Quite possible is 5...§a2, but
/ / / / Black wants to use the rook to cover
2: 2/2/ against checks from the side, while

2/y 2 he simply pays no attention to the

’2 2 2y-
2/2/2///
// 7 Exh3, then 10...f3 11 Eh8 Ed3! 13
The best defence is...
1...§e4! 2 EcS @f6 3 Ea fic4!
4 3216+ @es 5 §a5+ @f6 6 E32
@es 7 gm §c3+ 8 @gz m6 9 @113
fic6! with a draw.
Kramnik-Beliavsky
Groningen, I 993
a-pawn.
6 @gl f4 7 §g8+ $13 8 §h8
Edl+1 9 @hZ 912
Black wants simply to promote
the f-pawn to a queen. If now 10
a4 @e2 14 Ef8 363! 15 a5 fie7! 16
a6 £2 with a win.
10 a4 13 ll 35 @f1 12 36 gal 13
gas
After 13 @xh3 f2 14 §g8 Exa6
15 @g3 Ef6! the win is straight-
forward.
l3...f2 14 a7 336!
%

2
2/ 2
2/2/2%/ ”2/,2
22%2//
\

2 / 2 2 2 2 /
2,2 2 2//////%
2 2 /
////
2Z2/2/2 2 2 2/2
\§

///
Correct was the natural 1 EbS, ac- The right idea. Now on 15 @xh3
tivating the rook and not fearing @g1 16 §g8+ t 17 §f8 §a3+ 18
1...§g3+ 2 $12 3x213, since the @h4 @g2! is reached a postion from
rook ending with the f-pawn is a classic study by Lasker, where
drawn. But White decides to defend Black wins by shouldering the white
the pawn and at a suitable moment king to the seventh rank, while on
to obtain counterplay with the help 15 §b8 follows 15...§xa7 16 §b1+
of the a-pawn. But this proves to be @e2 17 §b2+ $63 18 §b8 @e4 19
a decisive mistake. §b4+ Q65 20 §b5+ @e6 21 §b6+
I Eal? Eg3+ 2 $12 §g4 3 fibl @e7 22 fibl §a3l 23 Efl EB fol-
Here also he had to go back; lowed by the approach of the king.
counterplay with 3 a4 does not 15 @h1 h2! 16 §b8
80 Rook Endings

If 16 @t, then 16...§h6+ l7 Boj kovié-Kakhiani

@g3 @g1 18 m8 Eg6+ 19 @114 Erevan, 1996
fia6, again reaching Lasker’s study.
16...§xa7 l7 §b1+ @e2 18 §b2+
@e3 19 §b3+ @e4 20 §b4+ <$d3
/4
21 Ebl §f7
/‘/,/,
% %
Also good is 21...Ee7 with the // //%fl/
idea of 22...§el+.
22 Efl @e2 23 Ef+ @f / W %
White resigned. % %/,8 ////
/////,///
Activating pieces also means acti- E7
vating the king to coordinate with a
passed pawn.
Let’s try to evaluate the position
Smagin-Naumkin —White has an extra pawn, but
Moscow, 1983 Black has a strong passed pawn on
g3 plus an active rook plus the pos-
sibility of activating his king—it is
// % //
y //4/ Black who is playing for the win.
1 Egz Eh3!
Usually the rook is placed behind,
477/ 4% but here Black does not allow 2
@c2 to be played because of 2...§h2
winning.
///;
‘/// ”7%
2 fig] @d6 3 QCZ @e5 4 @d3
%/%/%%J% g2+!
If 4...?f4, then 5 e5! @xeS 6 Qe3
/ / /© with equality.
5 QM?
The endgame with f and h-pawns Correct was 5 @e2 §g3 6 @fl
is drawn, but it is necessary to trans- §b3 7 @xgz fixb2+ 8 QB §b3+ 9
pose to it at once! @e2 fixa3 10 Ebl with a draw.
1 @g4 $17 2 (figs Eb4 3 M fibl 5...fig3 6 c @xe4 7 b4 $d3?!
Black reluctantly decides to give Stronger is 7...§g6 with a win.
up a pawn—which he should have 8 @b6
done earlier. But now, psychologi- Or 8 b5 §g6 9 a4 b6 10 %d5 @e3
winning.
cally, he is not ready for defence.
4 Exam ‘QgS 5 f5 Egl+ 8. .gfi 79a4®c410b5$b411
Correct 1S 5. g’h7. a5 §g6+ 12 @xb7 $a 13 b6
6 §g4i Hal 7 @g6 332 8 f6 Eal Exb6 14 @c7 §g6 15 @d7 @b4 16
9 17+ $18 10 h7 Eh] 11 916 Black @e7 QM 17 $17 §g3 White
resigned. resigned.
 Rook Endings 81

Eliskases-Levenfish Zugzwang—bad is 19...@f4 20

Moscow, 1 93 6 f6=.
20 @f7 @f4 21 @g7 @g5 White
resigned.

// / / / Azmaiparashvili-Kupreichik
///// 4 Kuibyshev, 1986

{3/6/ 7/2 // /// / /

2s %//
/ //// / // /4;% /1%
./ g %% / @aw i
1% / %/%
White has a pawn more but Black
//////a
has an active king and a far ad-
vanced passed pawn. White must a;M /////. eta
(”V/AW:
play 1 Ee7! c3 2 a4 gal 3 §c7 §a3
4 a5 $65 5 a6 @d4 6 a7 @d3 7 f4
/E% %
c2 8 §d7 with a draw according to Readers will ask—what has all
an analysis by Smyslov. However in this got to do with activating in a
the game he quickly played: rook ending. This is what it has to
1 $215+? @e6! do with it. White has the exchange
Sacrificin yet another one. for a pawn but Black threatens after
2 §a6+ d5 3 Exh6 c3 4 §h8 ...§d7 to seize the initiative. So
gal 5 §c8 fixa3 6 @h3 White decides to transpose to a rook
After 6 f4 g4 7 f5 @e5 8 ECS+ ending a pawn down, but in the
@d4 9 §c8 gal 10 f6 Efl 11 §c6 process activating all his remaining
@d3 12 §d6+ @c4 13 §c6+ @b4!, pieces.
and he has to give up the f6 pawn. 1 fixc5!! @c 2 Exes bc 3
6...@d4 7 @g4 @f1 @f6 4 @e2 @e6 5 @d3 @d5 6
Bad is 7 f4 gxf4 8 gxf4 c2 9 @g4 @c3!
§c3 +—. Now 7 §d1+! is threatened fol-
7...§a5! lowed by @c4.
‘Building a bridge’—the main 6...c4 7 QM! Ec7 8 §d4+ <§e4 9
manoeuvre in rook endings. b6 §c6 10 @b5 fies 11 fid7!
8 f4 fies 9 §d8+ @e3 10 Edl With the threat of fic7.
No help is 10 §e8+ $12 11 §a8 11...§b8 12 Ed! @d3 13 Ec6!
c2 12 gal gxf4 13 @xf4 c1=y 14 Here is the key move—defending
Excl Excl 14 g4 §c4+ 15 @f5 the b6 pawn and keeping in his
@g3! 16 g5 §h4 17 g6 @h5 18 g7 sights the c4 pawn, White wants to
Eg4 +- take on a4 and obtain connected
10.. .c2 11 Eel gxf4 12 gxf4 @d2 passed pawns.
13 Eal c1=¥ l4 Excl Excl! 15 l3...f5 14 9x214 c3 15 @b5 g5 16
@gS @e3 16 f5 @e4 17 g4 @e5 18 34 f4 17 gxf4 gxf4 18 as e4 19
@g6 §c6+ 19 @g7 £36! QM!
82 Rook Endings

The last finesse—now on l9...c2 White has a distant passed pawn,

follows 20 @b3. but his cut-off king, indeed also the
19...?e2 20 a6 e3 21 37 318 21 centralised black king, gives Black a
fxe3 Black resigned. decisive advantage.
1 §e7+ QM 2 Ee6?
Van der Doel-Klovan Correct is 2 EH !.
Gelsenkirchen, 1998 2...d5 3 Exh6 @e3 4 §f6
Also bad is 4 §e6+ @xfB 5 Eel
d4.
4...d4 5 Ef d3 6 §e5+ @xfB 7
2/ 2 // 2 §f5+ @e3 8 §e5+ @d4 9 §e8 d2
{/2 10 Ed8+ @e3 11 §e8+ $13 12
218+ @g4 White resigned.
72/
46

/1
/ 2;2 There will be cases when it is nec-
essary to sacrifice all one’s pawns
// for maximum coordination of all the
42 pieces.
22/2 2/
Each of White’s remaining pieces Pelletier-Rozentalis
are clearly more active than his op- Erevan, I 996
ponent’s. Therefore any delay by
Black would be equivalent to death.
1...§c8!!
F0rcing the capture of the pawn, 2 2 2 2
which allows Black, through the ,/ 2 22
opening of the d-file, to invade the
opponent’s position. 2 2%2/2 2
2g 2 2/// 2
2 3x216 Eds 3 §b6 EdZ 4 Eb
3f 5 a4 fin 6 35 f5+! 7 @xeS
13 8 EM Drawn.

Smyslov-Epishin
”2/2/2fi/
Rostov, 1992
l...g4!!
Inferior is l...f4 2 gxf4 gxf4 3

2:2 /%2 Exb4 e3 4 fxe3 fxe3 5 Ea4 §b2 6

b4! @e4 7 §a8 @d3 8 §a3+ @d2 9
2 W2,2 §a1!=.
2 /2 2 2 2 fixb4 f4! 3 3214 fib2 4 gxf4 g3!
5 fxg3 e3 6 f5 @e4 7 f6 @f3 8 Eal
2 2,2 Eg2! 9 f7 e2+ 10 @el @e3 White
//&”2 2 2, resigned.

5% V2 #752 The rule of the two weaknesses——

the possibility of transferring an
 Rook Endings 83

attack from one weakness to another Fercec-Mikhalchishin

until the time comes that the Nova Gorica, 1997
opponent cannot defend all his
weaknesses—is an exceptionally
universal one. /2

Nikolié-Movsesian Q46
Polam'ca Zdroj, 1996

2%.,4,“ Black clearly has the more active

////// 7/
/% /// king and rook, and the method of
//2/ realising the advantage lies in the
creation of two weaknesses and at-
tack on them.
1...Qe5
Less logical is 1.. g4, and White
Even here the two weaknesses on can avoid the drawn pawn ending. 2
a6 and c6 would not mean anything :14 Exf4 3 exf4 Qf5 4 Qe3 gxh3 5
were White not to have chances of gxh3 c56a4b67b3 a5 8Qf3 d49
organising another weakness on the cxd4!? cxd4 10 M g6 11 Qg3 Qe4
other flank. 12 Qg4=. It is equally a draw after 9
1 h5! gt c4.
2 b3
If l...g5, then after 2 Qe4 the king
More principled is 2 g4 g6 3 b3 f5
breaks through to the black pawns.
2 fit EgS 3 EM! 4 §g3 f4 5 EB c5 i.
Weak is 3 Exh6? Exg4 4 §Xf6 2...g4! 3 Efl
§g3+ and the endgame is drawn, 3 fig3? loses at once because of
while the passivity of the rook is 3...f5. The attempt 3 hxg4!? de-
provisional upon the arrival of the serves attention, but even then
white king to the defence. 3...Exg4 4 g3 b5 5 b4 §g6 6 a3 §h6
3. .Qd7 4 Qe4 Qe6 5 Qf3 Ehs 6 with the idea 7...§h3 leads to a win-
Ens §h7 7 Qg3 Ed7 8 E215! 2217 9 ning position for Black.
Qh4 Qf7 10 QhS Qg7 11 f5 Qh7 3...gxh3 4 gxh3 EM 5 E13 b5 6
12 fies Ec7 13 34! Qg7 Qe2?
After 13...§b7 14 nc6 Exb4 15 A decisive mistake. Better is 6 b4
Exf6 White wins easily. g5 7 §g3 Qf5 8 Ef3+ Qg6 9 §g3 f5
14 b5 21b 15 ab Eb7 16 bxc6 intending 10.. .g4 —+.
§c7 17 Eel §c8 18 c7 Q17 19 §c6 6...g57Qf2 358§g3
Qg7 20 Qh4 Q17 21 Qg3 Black If8 Qg3 §e4 9 QfZ b4— +—.
resigned. 8...c5 9 QgZ
84 Rook Endings

If 9 a3, then 9...a4 10 @e2 axb3 This was the last time White had
11 cxb3 §h8 with the idea 12...§a8 the possibility of activating his rook:
—+. 4 §c6l Ev 5 M! gxh4 6 @xh4
9...b4 10 cxb4 axb4 11 EB §h8 with equality.
12 if] §a8 13 gal $134+ 14 @fZ 4...h5! 5 §c2
f5 15 @132 §h8 l6 Ehl EM! Now there is already no saving
It was still not too late to let the himself:
win slip. l6...g4? 17 M. (a) 5 M hxg4 6 hn f5—+
17 fig] fixh3 18 Ex 5 E112 19 (b) 5 gxh5+ @t 6 M g4 7 312
l @xe3 20 §g3+ d4 White f5 8 @f4 §a4 9 @xf5 3——;+
resigned. (c) 5 xh5 @xh5 6 c2 f5 7 Ed2
f4+ 8 £n QM intending. ..fic3,
Typical mistakes in rook endings a7-a5-a4--,a3 Ec3-c1-b1 -b—2 —;+
(d) 5 h4 hxg4 6 @xg4 f5+ 7 @g3
In rook endings there are a great g4 8 m2 @h5 9 Efl Ea4! 10 fxg4
many typical methods of play, but Exg4+ 11 QB @xh4! 12 Ecl @g5
also, naturally, also a great many 13 a3 §a4 14 §c3 @f6 15 §b3 a5
typical mistakes. We acquaint you —+, Levenfish.
with the most typical of these. 5...h4+ 6 $12 a6 7 EbZ §c3 8
@gZ a5 9 EfZ 3213 10 @fl @f7 11
Making Passive f4 gxf4 12 fixf4 @g6+ White
resigned.
Ilivitsky-Taimanov
USSR, 1 955 Vaganian-Schlosser
Germany, 1994

///
////
7/ /1/
//// /
/////”/,
/////,/
///£S&
7//

l 13?
He should play 1 h4!, boxing 1n 1...Eb7?
the black king. An analogous mistake. He should
1.. .g5! 2 @g3 @g6 3 ficZ? activate his forces at once by l...h5!
Again White sticks to waiting tac- 2 g4! @g7 3 @e2 §e7+ 4 $13
tics. The correct path was 3 Ed6+l §c7 5 h4 h6?! 6 @g3 §c3+ 7 13
f6 4 h4! gxh4 5 ®xh4 Ev §c7 8 @f4 §b7 9 h5 Eb4+ 10 @g3
(5.. .EXB 6 §a6=) 6 f4 §a4 7 15+ gt
§g7 8 Ed7, maintaining equality, If 10...§b7 White gains the ad-
Levenfish. vantage by 11 hxg6 fxg6 12 f4 fol-
3. .f6 4 EhZ? lowed by §a5-a6, f4-f5 i.
 Rook Endings 85

11 gt EM 12 Qg4 Ebl 13 f4 1...§e5 2 h4 h5 3 §c4 EfS 4 §e4

After 13 Exa7 Black activates Q36 5 Ee7 g5 6 hn fxg5 7 §e8
himself by l3...§g1+ followed by M 8 338+ Qb7 9 §g8 Qe6 10
EgS. §g6+?
13...§b7 14 $15 §e7 15 a4 EC7 A mistake. He should activate the
16 Qg4 2117 17 3216 £137 18 a5 king by 10 Qc4l.
Q7 19 f5 zc4+ 20 Qg3 fies 21 10...Qd5 11 Exb6 QeS?
Qf4 §c4+ 22 Qe5 §c5+ Now Black meets mistake with
Worthy of attention is 22. EM“? mistake. Activity of the king is the
(Schlosser) 23 Exa7 fit 24 a6 main thing in rook endings. Black
fihl 25 §b7 Eel+ 26 Qd6 Ea] 27 maintains the advantage by shoul-
a7 Qf6. dering the opponent’s king with
23 Q84 EC4 24 QdS £114 25 f6+ ll...Qd4!. Now however it’s a
Qh7 26 fixa7 Et 27 Qc6 Qg6? draw.
Better is 27...§f5!=, Schlosser. 12 Qc3 Qf4 l3 §b4+ Qg3 14
28 36 3215 §g4+ Qf2 15 Qd2 Drawn.
White wins also in the event of
28...§h1 29 §d7 gal 30 Qb6.
29 QbG gal 30 §d7 §b1+ 31 Abramovié-Nikolic’
Q07 @1116 32 EdSII §c1+ 33 Qb6 Igalo, 1994
Qg6 34 a7 §c8 35 gas 115 36
8=W Exa8 37 E3138 Qf5 38 Q65
Q84 39 Qd6 Qf4
///W/W
On 39...f5 winning is 40 §a4 with W; W
the idea of Qe5 +--. W W WA/ /
40 EhS! f5 41 QdS!
W WW/WW W1
41 Et? would be a mistake be-
cause of 41 ...Qe4 with a draw. W
/W//£/£
41...Qe3 42 §e8+ Q13 43 Qd4 f4
44 Qd3 M 45 2118 Qg3 46 QeZ W
Qg2 47 §g8+ t 48 Qf3 Black W W W W
resigned.

Griinberg-Brunner 1...§f6?
Germany, 1992 Allowing White to create counter-
play on the king’s flank. He should
W/ restrict this by 1.. .f5! 2 g4 hxg4 3
fxg4 fxg4 4 Qg3 Qd8 5 n4 Qc8
W_%W W .2W* 6 Eg7 b5 with advantage to Black.
W 4 W 2 g4 Qd8 3 Qg3 Qc8 4 §e7 b5 5
,gW/WWW/ EeS §b6 6 gt gt 7 Et b4 8
Eds
8 305 is rather worse.
W W/ /& / 8...b3 9 Edl b2 10 Ebl Qd7 11

W/4 / W %W Qf4 Qe6 12 QgS?

White does not exploit the oppor-
tunity presented by his opponent.
86 Rook Endings

The only chance for him was 12 Shirov-Morozevich

@e4! @f6 13 @d4 @g6 14 @c3 Amsterdam, 1 995
@hS 15 Eb Eb 16 @b @xh4
17 @c3 @g3 18 @d4 @718 19 @e5!
with a probable draw. Now, how-
ever, Black has no difficulty realis- / WW W WW
ing his advantage. I

12...§b5+ 13 @h6 @f6 14 h5

§b3 15 QM Ebs 16 f4 Eb3 l7 W WWW?
@gS EbS 18 h6 @g6 19 f5+ @f6 20
@hS §b3 21 @gB Eb7 22 @hB
@gs White resigned. / W W Ws
//,/@
WW W
Marié-McNab
Hastings, 1995 1 Eg4?
Better is 1 fig] with the idea of
connecting his pawns in the rook
ending by f2-f4-f5.
WW WW1" /
1WWWW
W% W
/// //0‘
{W l...§e7 2 334? E218 3 f4 @xe6 4
@gfi @ds 5 @g4 @c 6 @gs @bs
This tempo was made possible as
a consequence of 1 fig6-g4.
/ 7 Ea] §g8+ 8 @f6 gm 9 @e5
E WW W/ §e8+ 10 @d5 Ed8+ 11 @e5 §e8+
WW/ fly WW 12 @d5 §d8+ 13 @134 §e8+ 14 $13
fins 15 Eh] @c5 16 15 @d6 17
WWW @f4 34 18 @gS @e7 19 f6+ $17 20
Eb] Exh3 21 §b7+ @fs Drawn.
l...35 2 @hZ?
Correct was 2 f3! a4 3 @g3 Ea] 4 Vaulin-Voikhovsky
f5! @f (if 4. .g5 5 §a6+ 9x56 Russia, I 997
§a5+ with a draw) 5 Ef @e5 6
367+ @d4 7 §a7 a3 8 @f4 a2 9
§a8 with equality.
ZufifB! 3 336+ @f5 4 Exa5+ WWW@/
/ /
@g4! 5 9g Exf4 6 337 @xh4 —+. // W
Analysis.
/// W 1/
Technically weak play
in rook endings

In the following position White,

W/WWW1W
/ WW W
7/%///%W
of course, has a winning position
but, it is not so easy as it seems at 1 e4? 3213+ 2 @gZ fxe4 3 §f4
first sight. @e5 4 Exg4 @d4 5 §g8 @e3 6 §d8
@e2 7 Ebs
 Rook Endings 87

The a-file is inaccessible. with an easy draw and therefore

7. .EaZ 8 Eb] e3 9 @g3 gas 10 played
Eb2+ @113 11 2133+ @d2 12 Eb2+ l...§a2?,
@c3 13 EM §e8 White resigned. But he did not reckon on
2 h4!,
Krasenkov-Iskusnik after which White has a winning
Russia, I 996 endgame, whereas he could have
made a draw by 1...§b1! 2 h4 §g1+
3 @h3 Eh1+ or 1...§b8!, preparing
WW against h3-h4.
W W W W
/W W/ W :éW Materialism instead of activation

/
/ / WWWy Schmitdiel-Mikhalchishin
Berne, 1 994

/W@W
W WW W W
WW WW WW WW
\

W/W/W W
1...?f5? After the simple l...§f4
/’/”WWWW
there is an easy draw. 2 c6 Qe6 3
§c2+—. WgW/WW
\

A narrow spectrum of noticing

the opponent’s threats
JWW/ /% /,
/W//W/W
(one threat he sees, the other not)
l Edl?
Kozul-Mikhalchishin, He should think about the liquida-
Bled, 1 996 tion of his opponent’s activity (re-
stricing the mobility of the pawns)
by l EaS! @f8 2 @g3 §e7 3 @B
@d7 4 @e3 @c6 5 QM with excel-
W WW/ / / lent chances of a draw.
W W W W 1...?f8 2 Exd6 a5 3 @gfi?
White’s only chance was to strug-
W W W W
W/ W// ////
gle against the a-pawn by activating
his c-pawn: 3 c5! @e7 4 EdS! a4 5
06 a3 6 07 a2 7 Ed], with a

W//W
’W W
draw—clearly better is 4...?e6!
with chances
however...
of victory. Now,

3...a4 4 @f3 a3 5 Edl @e7 6 @e3

Black saw only one threat 1 §f5+ a2 7 fial @d6 8 QM §a4 9 f4 h5
and prepared to repulse it by 10 g4 h4 ll g5 g6 and White
1...§b3+ 2 @gZ Eb2+ 3 @fl §b3 resigned.
88 Rook Endings

Mikhalchishin-Stangl A well known drawing mecha-

Dortmund, 1992 nism without the b2 pawn—the
rook goes to the sixth rank (f6) and
WWW upon the approach of the king to b5

/EW;//MW begins an endless checking se-

quence. But in the game there
1/ WW followed...

WWQW//’ /A l...§b4? 2 §c7 334 3 §c8+?

For what reason? Why not 3 §c6
/,//W
/ @xh7 4 @b, winning easily.
W / 3...?xh7 4 §c6 Eb4?
After 4...@g7 5 @b 9H 6 @b?)
W W Eal 7 @b4 @e7 8 @b5 @d7—
draw.
5 a7 §a4 6 §c7 @g6 7 @b @f6
l...§d3? 8 @b3 Eal 9 @b4 @e6 10 @b5
A technically incorrect attack—— @d6 11 §c6+ @d5 12 §a6 §b1+
necessary was l...§dl+ 2 @gZ fial 13 $35 @cs 14 §c6+! and Black
3 §a7 Exa3 4 b5 a5=. could resign.
2 a4 a5 3 ba §d4 4 36 31:34 5
a7 @h6 6 @fl g5 Inaccurate technical execution
Also no help is 6...§aZ 7 @el,
and the king goes to bl. Ribli-Mikhalchishin
7 @e2 EaB 8 @dz @g6 9 @c2 Germany, 1993
gxh4 10 gxh4 @f5 11 @b2 §a6 12
@b3 Eal l3 §c7l
Preventing 13...@g4 because of W W
W / / W
14 §c4+ and 15 Ea4.
13...§a6
resigned.
14 @b4 and Black

/
W/W//W/
Ignorance of typical drawing
WW “ W’W
mechanisms
7 W&/W
Ivanchuk-Lautier EWVW W
Horgen, I 996 g/W
l...§b2?
Occupying the b2 square, which is
needed for the king—correct is
1...§c2!, and Black has no difficul-
ties. Now, however, he lacks a
tempo.
W 2 h5 a2 3 @gz @c3 4 $g3 @b3 5
W W f4 Ebl 6 f5 al=¥i 7 Exal Exal 8
W W g6 @c4
8...hxg6 9 fxg6 @c4 10 Qt?! +—
9 f6 hxg6 10 f7! Black resigned.
 Rook Endings 89

Typical Rook Endings

\\
In the endings it is difficult to find 6:9?! ”

\
anything new—everything has
already been played. But among
various positions resembling one ////////:<//
another it is still possible to find
great differences and great simila-
/ // / /
rities at one and the same time.

Our attention was attracted to the

/ /%/ /%/ /
ending from the game
11 §f8!! Exf8 12 exf‘8==%’+ @xfB
Hfibner—J.Polgar 13 @d6 with a won pawn ending.
Dortmund, 1996 2 E36 @f7 3 §f6+ @g7 4 e6 Edl
5 §f7+ @gs 6 @134 figl 7 EB
/// §e1+
After 7...?g7 8 e7 fie] 9 §e3 an
IIIIIII ///////// easily winning pawn endgame is
////fi//‘f/ again reached: 9...§xe3+ 10 @xe3
@fi 11 $64 $68! 12 @d5 Qd7 13
y// e8=y!, etc.
\\\\\\\:

8 @d5 @g7 9 317+ @g8 10 @d6,

/ ////g//
/ and Black wins.

////////
\\\\

This reminds me of a very similar

endgame that I had a year earlier.

Barle-Mikhalchishin
Here Black resigned.
Slovenia, [995
1...@f8
After 1...§b4 Hfibner gave the
following variation: 2 §a6 (also /%%
winning is 2 fig7 Eb6 3 @e4 §a6 4
////
§c7 @d8 5 305 Ea4+ 6 @d5 §g4 7
@d6 Exg5 8 §a5 +-) 2. .@f7 3 /////
§f6+ §g7 4 e6 EbS 5 @f4 §a5 //,// /
(5.. .Eb4+ 6 @es §b5+ 7 @d6 2n
8 fif7+ @gS 9 Ea7 figl 10 §a8+ /§z¢////
@g7 11 e7 +—) 6 e7 §a8 7 $65 //a/é,
§a5+ 8 @d6 §a6+ 9 @c5 Ea5+ 10
@c6 EaS
7/.///fi/
\
90 Rook Endings

There followed: Qf6 obtaining a draw. Correct is 1

1...g4+ 2 hxg4 hxg4+ 3 Qe3 Qe5! §e7+ 2 Qf4 Ea7 3 §b6 §c7 4
After 3 QfZ §b2+ 4 Qfl arises a §f6+ Qg7 5 e5 reaching an easily
position which 13 very similar to the winning position, known from the
previous one, with the following previous examples.
variations: 4. .Qe4 5 Ec4+ Qe3 6 l...§d7!
§c3+ Qd4 7 §a3 §b4 (7. .Ec2l?) 8 The only chance—the threat was
Qf2 e4 9 Ea2 e3+ 9 Qe2 Qe4 10 2 e6, followed by QeS, fidS and
§c2 Ed4, again threatening a Ed7.
transfer to a winning pawn 2 Qe4 Edl 3 Eb7+ Qe6 4 Eb6+
endgame. Qe7!
3...§b2 4 §c8 EgZ 5 §f8+ Qe6 6 To achieve a draw it is necessary
Qe4 to give up yet another pawn.
After 6 §e8+ Qf6 7 Ef8+ Qe7 8 5 Exg6 §e1+ 6 Qf4 fifl+ 7 Qg3
§f5 Qe6 Black wins easily. En
6...§xg3 7 Ee8+ Qf6 8 Exes In the game 7. .fiel 8 §f6 §e4 9
White reckoned on 8...§e3+? 9 Ef4! was played, with a win.
Qxe3=, but more tenacious was 9 8 QfZ §g4!,
318+ Qg7 9 §a8 (9 an Qg6 10 And according to an analysis by
Qxe5 :13 -+) 9...:13 10 mm Ef6, M.Yudovich—it’ s a draw.
and then the black king goes to g6
and g5, winning. More complicated variations with
9...Ef3 10 fies QgS 11 Ens £17 the addition of the h-pawns are met
12 Qe3 g3 White resigned. in the game

A similar endgame was analysed a Akopian-Almési

very long time ago. Ljubljana, 1995

Filipov-Kopatsny
USSR, 1 968
7 / 7
//// / 7 7
///
:7 e77 / 7 /
7///// 777
/:/ 7 7
/7 7237 7
777
77/”
://”/

7%7/7/7 77 /
There followed:
//// / 1...Qf4
Inferior is 1...§h1, since after 2
There followed: §a4 he cannot play 2...§xh2? 3
1 e5? l.
If 1 §b6 Ea5+ 2 Qd6 En 3 e5 2 EcZ
figl 4 Qd7 (4 §b7+ Qf8, and there After 2 g3+ hxg3 3 hxg3 Qg4 4
is no win) 4. .EdH 5 §d6 Eel 6 e6 §e2 Qf5 5 Qg2 £133 6 $12 §d3 7
 Rook Endings 91

QgZ e3 the threat of fidZ is on g3, which also, however, leads to

unstoppable. a draw.
2...?1‘5 15 $215+ @d4 16 §a4+ £04 17
On 2...§h1 Almasi gave 3 g3+ Exc4+!
hxg3+ 4 hxg3+ @g4 5 ECS (5 §c4 The transfer to a pawn ending is
§h2+ 6 9g] §e2 —+) 5. .6fih2+ here the simplest way to a draw.
@gl EeZ 7 §e5 §e3 8 @hZ (8 $12 17...?xc4 18 99.3 @d5 19 @e2
§B+ 9 @e2 Ef5! —+), but this po- @d4 20 @d2 e3+ 21 @e2 @e4 21
sition is not winning——-for example, @el $13 22 @fl e2+ 23 @e1 with a
8...Ee2+ 9 @gl e3 10 an E12+11 draw.
@g 1 EB 12 @hZ!
3 fics+ @g4 4 Ecz fibsz, Both of the young grandmasters
Preventing g2-g3. played rather inaccurately, and par-
5 @el h3 6 E312 ticularly White who allowed his op-
Weaker 1S 6 gxh3+ @xh3——+. ponent to advance with threats
6.. .fibl+ 7 @f2 3111 8 EeZ against the g and h-pawn. The right
Bad IS 8 §a4 @f4 9 g3+ @fS —+. way to achieve counterplay had al-
8...@f4 ready been shown 1n old games.
But not Smfit? 9 Exe4+ @f5
because of 10 @g3l. Tiets—Forsberg
9 g3+ @e5 10 $9.3 Eel 16th World corr.ch. 1985-91
There is nothing in 10...§g1 11
/(:’I
§a2 §g2 12 §a5+ @f6 13 @xe4
Et 14 $13.
11 3212 §c3+ 12 $62 @114 13 / y /
§d2+ $65 14 EaZ /////// 21/
%///M/
// 6%
y/

/%%//% 2 //% //
///
%'/// 7///
/ %%/
//
/ l...§d5 2 Ef6 e4+ 3 @e3 §b3+ 4
$12 §b2+ 5 @fl
/ This passivity is forced, since af-
ter 5 $g3 h4+! 6 @h3 e3 the pawn
cannot be stopped.
l4...g4? 5...@d4 6 §g6 @d3
After the correct 14...?f5 15 $12 After 6...?e3?! 7 §g3+ @d2 8
@g4, it seems that there is an irre- $12! §b8 9 §a3 Ef8+ 10 @g3 e3
sistible threat of ...§d3, e3 and EdZ 11 §a2+ it is very difficult for the
transferring to a winning pawn end- king to esca e the checks.
ing—however the pawn ending is 7 §d6+ e3 8 Edl §f2+ 9 @gl
drawn, and, secondly, with the EdZ l0 Eel+ @d3 11 @fl e3 12
white king on e2 there is no other gal 312+ 13 @gl um 14 an
plan besides the sacrifice of the rook @d2 15 g3
92 Rook Endings

White defended in the following

way
WWW/W 1y§e7 Ed3 2 g4!
W // An attempt to create very quickly
a passed pawn on the king’s flank.
/W W/ /
/W WNW W
/// %///////,///
2...§d8
After 2...Ed4 3 QB EdS 4 §a7l
the position is analogous to the
game.

WWW //Wfi 3 Qf3 Ef8+ 4 Qg2 §f4

After 4...e2 5 §d7+ Qel 6 Ed6!
§f2+ 7 l EfS 8 h3 the threat to
Here Black has several different take on h6 saves White.
plans: 7 E217!
(a) 15...§c2 16 h3 Qd3 17 §a3+ Transposing the game to a well
Qd2 (after 17...Qe4 18 EaS QB 19 known position with a rook attack
§f5+ n3 20 Et with a draw) frosm the long side.
18 gal §b2 19 l Qe2 20 Qg2 .e..2 6 @32+ l 7 331+ Draw.
§d2 21 g4 Qd3+ 22 Qfl Eh2 23
§a3+ and a draw was agreed. The correct plan of defence for
(b) 15...Qd3, and now this class of position was demon-
(b1) 16 l? Qe4 17 §a8 §b1+ strated in the game
l8 Qg2 e2 —;+
(b2) 16 h3 Qe4 17 §a8 QB 18 Portisch-Pietzsch
§f8+ n3 l9 §e8 EfZ+ 20 Qel Madrid, 1960
QB 21 §f8 Qg2 22 Ee8 EB 23 h4,
and the win is not easy;
(c) 16...h4!?, and, for example, 17 W%
gxh4 Qd3 18 l Qe4 19 h5 §b8 WE W/WW
/
20 h6 §g8+ 21 Qfl Q13 22 §a3!,
and White maintains equality.
/ WW
WWWW/
An analogous ending is W/W/
/ ///
/ WW
Chiburdanidze-Galliamova
Groningen, I 997
flW
lufibl! 2 h4 Efl+ 3 QeS Egl 4
§c7+ Qg6 5 §c6+ Qf7 6 Qf4 §f1+
7 Qg5 h6+!
Weaker is the more passive
7...§el 8 §c7+ Qf8 9 Qf5 fifl+ 10
Qe6 §f6+ ll Qe5.
8 QhS §f4 9 e5 fie4!
The black rook has gained maxi-
mum activity, but the position still
requires accuracy.
 Rook Endings 93

10 (36+ @f6! $64! 6 §c4+ @d3 followed by

Not 10.. .Qe7 11 §c7+ @f8 12 ".316 and the advance of the a5
§c8+ @e7 13 EgS +—. pawn.
11 g5+ $e7! 4 @g2 §d7 5 ExeS Ed2+ 6 @gl
Again not 11...?f5? 12 §c5+ @113 7 figS?
@xe6 l3 §c7 g6 14 @xh6 +— or White sees a defensive idea, but
11..2M 512hxg5®e713® 6+—. his execution of it is inaccurate. Af-
c7+ $18 13 Ec8+ e7 14 ter 7 Ee6! Eg2+ 8 @hl Exg3 9
§c162 fixg6! Black can give up trying to
After 14 EgS hxg5 15 hxg5 EXCG wm.
16 fixg7+ @1‘8 draws. 7...§g2 s @111 312:! 9 @g1 Ef6!
14.. $18 15 §b6 $67 16 §b7+ Now, however, it’s zugzwang.
$13 17 an @g8 18 29,7 @111 19 10 gas 213 11 g4 gg3+ 12 @111
gxh6 gxh6! @xg4 White resigned.
Clearly not 19.. .Qxe7? 20 hxg7
§e5+ 21 ®g4 +-.
20 §f7+ @e8 21 Ef6 $67 22 Bogolj ubow—Rubinstein
Exh6 §e5+1 and a draw. London, 1922

Trabattoni-Barlov
La Valetta, 1979 //
/ % / ;%
% 71M
/ y / é////
/
// %%%‘/
/%
/
%§///
A/// we///%%
fl/ /
/ // Rubinstein is considered the
//// ‘king’ of rook endings, but his play
in the following ending leaves us
1...§b7! profoundly bewildered.
The main task is to drive off the 1...g6?
rook from the sixth rank and cover Correct is l...g5!, transposing to
the king against checks along the the position in Barle-Mikhalchishin.
f-file. Then follow preparations for2 §b8 Ea2?
the advance of the CS pawn. Clearly better is 2...§a3+, gaining
2 @216 EN 3 gas some tempi.
Another possibility would have 3 firs §g7 4 §e8 @f7 5 EbS §h2
been 3 §a4 Ef6! 4 §b4 §a6 5 §c4 What is he doing? He can’t take the
e4! 6 Ec5+ @g4 7 §g5+ @h3 8 @fl pawn!
§f6+ 9 @e2 3131, winning. 6 §c8 332 7 M Ea7? 8 QM @f6
3...Qg4?! 9 fif8+ @g7
An interesting plan, but better And a draw was agreed. Rubin-
looks 3...§f6 4 §b5 §a6 5 305 stein’s worst endgame!
94 Rook Endings

The fact that matters are not quite gal @g6 22 fibl EeS 23 QM em
so simple is shown by the game 24 Eel?
Stronger was 24 §fl+.
Fischer-Geller 24..£35! 25 Exe4?!
Curacao, 1962 Again stronger is 25 if].
25. sets! 26 §e8 @g4! 27 @e3
@g3!
//// And in this theoretical endgame
yfi/7/ 43/ White resigned.

The last game did not answer sev-

/ // / eral questions of defending the end-
%/%/%% game with the e and g-pawns. In
g %/ / %/ modern chess there is one more very
interesting example.
% / / %
Balashov-Ulibin
Uzhgorod, 1988
1 g5+?!
An extraordinarily crucial deci-
sion——it would be simpler to wait
with 1 §c7 or 1 Eb5. ////%////////
%% / /,,//
1...hxg5 2 hxg5+ @g6 3 §e7 §e3
4 $12?
Not to this side—it was necessary //%/%
to exchange the g5 pawn for the e6
pawn and not stick the king’s head
out. Therefore 4 @hZ! 265 5 @h3
§'/ %/fi
leads to a draw. // // /<§/
////// ///
4...§e5 5 $13 Efs+ 6 @e3
After 6 @e4 E17! 7 fixe6+ @n
the win is easy. 1 E18
6...e5 7 @e4 En 8 EeS? The preliminary 1 912 does not
Preferable is 8 Ea7! @hS 9 gal threaten 2 e5 because of 2...Ea5 3
with the idea after 9...§g2 to play Ee4 @fS.
10 @e5 g5 11 @e4 though 11...?g4 1...§a2+ 2 $13 Ea3+ 3 's
wins. The general idea to win with §a2+ 4 @e3 $213+ 5 QM Ea4+l
the g and e-pawns consists of giving Not 5.. .Exg3?? 6 e5 +—.
up one pawn and transposing to a 6 @d5 §a5+ 7 9w §a6+ 8 QbS
theoretically winning endgame. 3213! 9 §f5+
8...gfi 1! 9 @113 Efl+ 10 c$g3 §f5 On 9 65 §e3 10 368 @f5 11 g4+
11 En @g5 12 §e8 @f6 13 318+ @xg4 12 $05 @f5 draws.
@e6 14 §e8+ @d5 15 338 §f7 l6 9..Qg6! 10 g4 §e3 11 e5 §e4 12
@g4 Ee7! @c6 fixg4 13 §f6+ @g7 14 @d7
When the rook is placed behind, E214
thenlhalf the business is done. A drawn ending is reached with
16 fiaS+ @e6 17 3216+ Q17 18 the weaker side’s king on the
$13 Ee6 19 gas e4 20 <36 g5 21 kingside.
 Rook Endings 95

15 §c6 E38 16 §c8 §a7+ 17 §c7 easily winning and described in

$218+ Drawn.
How many roads lead to Rome?
We have already repeatedly said
that in the endgame there are usu-
ally two paths (moves)—right and
wrong. But at times the number of
paths is surprisingly greater and it is
exceptionally difficult to find the
right one.
every book on rook endings;
(d) 1.. Qd3! (this was the only
possible continuation) 2 Efl (after 2
EeS Qd4 3 EbS Qc4! the rook is
continually hounded) 2. .Qe2 3 §f4
Qe3 4 §a4 (there is nothing else)
4...§xf5 5 a6 §f8 6 a7 §a8 7 Qxh3
Qd3 (similar to variation (b), but
the cut-off king has proceeded one
rank further which is of decisive
significance) 8 Qg4 Qc3 9 Qf5
Qb3 10 gal Qb4 ll Q66 QCS

Beliavsky-Azmaiparashvili
Portoroz, 1997 /%
/
{29//
% /% / 7
/%//
IV ’%%%//
7 /% / 7/
//i /// /% y
//
/%//
%
a Q/////
/ ///
////

,Q'o,
2 / %
I’ll/l/

(After 11...Qb5? 12 Qd6 a posi-

:1!) tion is reached from variation (b),
while exerting control over the d6
The black king has six(!) possible square is the key to evaluating the
moves, but only one(!) leads to a position) 12 Qd7 Qb6 13 Ebl+
draw. Qc5! (the point) 14 §b7 §h8!, and
(a) 1...Qf3 (f4) loses because of 2 a theoretical draw has arisen.
Ea] §a6 (he cannot draw by taking
on f5) 3 f6 winning easily; Ward-Baburin
(b) 1Q12 2 gal £f 3 a6 §f8 4 Isle ofMan, 1997
a7 Eat? 5 §a3l (a very important
moment to cut off the king) 5.. .Qe2
6 Qh3 Qd2 7 Qg4 Qc2 8 Qf5 Qb2
9 §a6 Qb3 10 Qe6 Qb4 11 Qd6! /E
// // /
% % //
(shouldering away the king and
preparing to set up a mating net)
// M
11...Qb512 EalQb613 §b1+ Qa6 %/ / / W
l4 Qc7! Exa7 15 Q06, and White 71/
wins;
(0) 1...Qd2? 2 EeS Qd3 3 a6! C// %
Exa6 4 Qxh3 Qd4 5 E86, and % %
Black resigned, the rook ending is
96 Rook Endings

White has an extra pawn but Branicki-Sefc

White’s cut-off king plus the strong Prague, 1 955
passed e4 pawn and centralised
Black king means that it is only
/%

©\\
Black who can play for the win.
Baburin assessed the position as /////% /

§
7/%/
winning for Black. Let’s have a a6
/// ..... zzzzzz

look.
(a) 1 fibZ §c7l 2 @b3 @e5 3 Ecz /;§/’///-"//,
fid7! 4 @c3 e3 5 f6 @e4 6 EgZ
§c7+ 7 @b2 EH 8 QCZ Exf6 9
§g3 (On 9 @dl, 9...?d3 is unpleas-
%% // / 6/”?
,,,
ant. Instead, without the h2 pawn,
//
//// / %
White saves himself by means of
the stalemate 10 §d2+!) 9...Ec6+ 10
l @d3 11 M 2216 12 @c1 §a1+ Here there are several ideas for
13 s Ehl l4 h5 Et 15 @cl+ White to realise his enormous ma-
Ehl+ 16 am @dz White terial advantage, but only one of
resigned. them leads to its objective.
(b) 1 f6 @e5 2 £12 EN 3 @b3 e3 (a) l Ea6 fidl 2 g4+ @f6 3 d7
4 Efl @e4 5 @c2 e2 6 fig] @e3, @g7 4 §a7 @f6 5 @g3 @eS!
and there is no apparent defence (shouldering away the white king,
against Exfl or @fl; now after 6 d8=§ fid 7 §f7 §g8
(0) 1 §b2 §c7 2 §b5 e3 3 §b3 a well known draw is reached) 6
(on 3 f6 there is 3...e2 4 Eb] §a7+, QB §d3+ 7 @e2 §d4 8 @e3 Edl 9
and an exchange of rooks) 3...§e7 4 §b7 @f6! (on 9...§d6 there is 10 g5
§d3 (if4 f6, then 4... e2!) 4...?05 5 @f5 11 g6!) 10 @e4 §e1+? and
Ed] @c4, and again there is the now after 11 @d5 §d1+ 12 @c6
threat of 6...a2; @e7 there is no defence against
(d) 1 h4! (logical, it is necessary 13...Ed6+!, capturing the d7 pawn.
to urge the passed pawn on) 1...%>e5 This is how the game went.
2 h5 @f (after 2...?f4 3 h6 e3 4 (b) l g4+ @g6 2 d7 @g7 3 §b7
§h2 §h7 5 f6 @g3 6 f7 Black can- @g6 4 §a7 @h6 5 g5+ @g6 6 @g4
not win) 3 h6 §h7 4 §h2 e3 (On §d4+ 7 QB @g5, and a draw;
4m®g6 there is 5 EM e3 6 :64) 5 (0) 1 d7! §Xd7 2 g4+ @635 3 g5
@b2 @f4 6 @c2 @g3 7 §h5 e2 8 and after 4 §f6 a well known theor-
@d2 QfZ 9 §h2+ with a clear and etically winning position is
uncomplicated draw. obtained.

Nevertheless material is material. And now a few examples on the

theme of choosing the best continu-
There was a wide choice for ation. The solutions are given at the
White in the following game. end of the book.
 Rook Endings 97

How the “one-legged” Viktor

Lvovich tested the youngsters in
the endgame

\\ \\ \§W \
® s :\
WW W W W Viktor Korchnoi’s play has al-
ways been characterised by the
highest class and technique. Before
WWW/W W the start of the Beme tournamnent
”/é/
W7 //W W and his match with Lucas Brunner,
W77 W WW} W the veteran broke his foot and the
organisers offered to postpone the
W W7 W@ match but to their greatest surprise
W WW WW W the ‘patient’ had not even thought
about refiising to play! From chess
What is correct: history it is a well-known paradoxi-
1 @195; 1 $65 or 1 @di? cal fact that grandmasters with bro-
ken limbs play very strongly! We
2 mention just two examples—Jan
Timman, with a broken foot, won
brilliantly at the super-tournament
in London 1983, while Alexander
Beliavsky, with a broken hand, won
the board one prize at the Thessalo-
WWWW’ niki Olympiad in 1984! Incidem

gag/a
tally, both breakages were sustained
W7 7/ playing football. It was rather un-
usual to see the active Korchnoi sit-
ting motionless for all his games
7W//W/ and only at the end with difficulty
moving away on crutches. But he
played splendidly, gaining particular
How does Black make a draw?
success in a couple of rook endings.
3
Brunner-Korchnoi
Berne, 1296
WE
aza/ae
W W 4
W W A W
W7W7/s
W ,
///W
//W//
/I///

W W W77 W7 W /
W W W WW WW//7WEW//
//7
WW
W7W7/
W W 6
What is correct: l...f3,1...§e1 or W
1...§c1?
98 Rook Endings

There followed 1 EdZ? e3 16 §e8+ e2 17 §e3 follows

Correct was 1 §d7 f6 2 l §e2 17...f4! 18 gxf4 fid3!) 15...@e5 l6
3 h3, and it is difficult for Black to §e8+ @d4 17 §d8+ @e3 18 Ef8
carry out his plan. $12! 19 Exf5+ 9g], and White
1...®h7 2 @gZ? cannot defend against mate.
Here it was still not too late to re- Well, now we return to the game.
turn to 2 §d7!. 7...@f5 8 §h8
2...g6 3 fxg6+? If 8 Ea3, then 8...Ed2! (with the
The last chance was 3 f6 g5 4 QB threat of 9...§d3+! 10 fixd3 64+ 11
§a4 5 EeZ g4+ 6 @f2 e4 7 Ef2+ §e3 exd3 12 @xd3 Qg4) 9 @e3
and 8 §f4 with some chances of Ed4 10 §a6 §b4 11 $13 Eb3+ 12
holding the game. Now a 3:2 end- $12 e4 with the unpleasant threat of
game is reached, which it seems is ...@f5-g4
practically winning for Black. 8...e4+ 9 @e3 §b3+ 10 $12 @g4
3...n6! 11 §g8+ QM!
This is stronger than taking with
the pawn, since the passed e5 pawn
must be supported by the f-pawn.
4 §a2 h5! WW W/W W/
Threatening to create a weakness
on g3 for White after h5-h4. 4/ W W
WWW WWW W/WW
5 M Eb4 6 E218 §b2+ 7 $13
After 7 @h3... W/W/
W/W WW5?

/W /W W
WW/W/W
WWWW W/ %W flW/

W/
WW W?
/W/W///
3%
Usually such an approach of the
king leads to a decisive outcome.
12 EgS
W W// /W There were rather more chances
remaining with the preliminary 12
EWW W "WW Eg7, when Black can choose be-
tween 12...f5 and 12...§f3+ 13 @e2
Exg3 14 Exfi @xh4.
is obtained practically an identical 12...§f3+ l3 QeZ f6! 14 §g6
copy of the famous game, Smyslov- On 14 Et Korchnoi intended to
Gligoric, Warsaw 1947, (only with play 14...?xg3 15 §h6 @g4 16 h5
White to move and the black pawn f5 17 §h8 <ags 18 h6 @g6 19 h7
on e4). There the very instructive fih3, obtaining two connected
continuation was 8 §e8 §e2 9 §e7 pawns.
f5 10 §e6+ @g7 11 Ea6 (If 11 14...f5 15 figS @g2 16 9e] fifZ!
§e7+, then 11.. @f6 12 Eh7 @g6 17 §g8
13 §a7 £12 -+ 11.. £12 12 §e6 If 17 @d1, then 17...?f1! and
$17 '13 fixeS f6 14 :68 §d2l e4-e3-e2.
(zugzwang) 15 §f8+ (after 15 §h8 17... e3 18 §g7

Or 18 §g5 f4 19 gxf4+ QB 20
Egl fieZ+ 21 l @212 22 Efl Qe4
23 fig 1 Qd3 +.
18.. .f4! 19 gxf4 Qf3 20 Ec7
For 20 Egl see above.
20.nfia2 White resigned.
Korchnoi-Kengis
Berne Cup, 1996
7079/
/ ///‘
77/
/ 7/ //"//
77/77M7
7/
7/
At first sight the endgame looks
completely drawn—only the passed
a-pawn is potentially stronger than
the passed c-pawn. Now Black can
calmly transfer his king to d6, but
he decides to display activity on the
king’s flank, since White obviously
intends to move his king over to the
a3 pawn.
1...g5 2 hn hn 3 Q13 3214
It is always useful to cut off the
enemy king.
4 Qe2 Qg7
Again it is worth trying to go to
d6, though here White can prevent
this by 4...Qf‘8 5 §e3 while on winning plan.
5...§d4 6 §d3 §c4 7 Qd2 Qe7 8
303 :34.
5 §c3 f6(?)
The authors would prefer 5...c5!?
6 Qd2 Qf6 7 Q02 c4 8 Qb2 Qe5 =,
but Kengis probably feared 6 Exc5.
6 Qd2 c5?!
Again preferable IS 6.. .Qf7 7 Qc2
(7 E33 f5 8 Q02 f4 =) 7.. .Qe7 8
Qb3 §a6 9 a4 Qd6.
Rook Endings 99
7 Qc2 c4?!
The question is where is the pawn
best placed—on c4 or on 05.
We throw in the variation 7...Qf7
8 Qb3 EaS 9 §e3 f5 10 Qc4 Qf6
11 Qd5 f4 12 gxf4 xf4 13 303
Q15 14 f3 Qg5 15 Qc6 Qh4 16
Qb6 §a8 17 Qxc5 Qg3, and White
should not win.
8 Qd2
/ 7 ///W
7/7 7'7
7774/
78/7
7
/
7&7
I
8...Qg6?!
After 8...Qf7 9 Qe3 Qe6 10 Qd4
Qd6 11 g4 (11 KB c3!=) 11.7126
12 QcS QeS 13 QbS EaS 14 a4
Qd5 Black has his own counter-
chances.
9 Qe3 QfS 10 m4 Qg4 11 as:
Qh3?!
It is not quite clear where to go
with the black king. Better looks
ll...§a8 12 Qxc4 f5 13 Qb3 §b8+
14 Qa2 §e8 15 :02 EaS, and it is
very difficult for White to find a
12 QbS gas 13 a4! $32?
After 13...f5!? 14 a5 §b8+ 15
Qxc4 §b2 16 a6 Ef 17 §a3 §c2+
and 18...Ec8 Black holds the draw.
If 15 Q06, then 15...§b2 16 a6 gaZ!
l7 Qb7 fixfl 18 §a3 §b2+ 19 Q06
03 20 gal 02 21 a7 fibl.
14 f4! Qh3 15 fn fn 16 a5 g4
17 a6 Qh2 18 Ea3 Qh3 l9 Qxc4
Ec8+ 20 Qbs Eb8+ 21 Qc6 §c8+
1 00 Rook Endings

22 @b7 fihs 23 a7 §h7+ 24 @b6 It looks like Black has defended

§h8 himself—White cannot queen since
then a drawn ending arises because
of the distant white king. However
Korchnoi finds a path to victory.
£3? // //////
//////// 25 @c6 Ens 26 Eb3! §f6+ 27
y// @b5 fifs 28 $35 §a8 29 @a6 £18
30 §b8 an 31 a8=y §a1+ 32 @b5
% // Exas 33 ExaB @xg3 34 QM Black
// ////g
”// resigned.
// at? The young grandmasters made a
surprising number of mistakes as
/ 7// 7// Korchnoi demonstrated by clear-cut
play.
 Rook Endings 101

Exercises:
Rook Endings

/,/ f/// /, ,
e
/// /4e
//
/&V ///

% //// //
E;\\\\

/ ////
////////V///
\

////%/8?
/// //// %h %e
Demonstrate the correct What 15 correct: 1...@f4 or 1...a4 ?
plan ofdefencefor Black.
4

%/ ///%
//////
/%¢///
7//%////
////
////%
////,//
///
What ts correct: 1 §d6 or J §e8+ ? Evaluate the posztzon andfind the
right plan ofdefencefor Black.
102 Rook Endings

5 8

W%
”W% / W W // W / 8/
W / WWW W W /”W
W W W/ WEW3 WW
/ / / &/W W/W // WW 4
$/W/////g ///////////
W W ”W” W
Find the right method ofdefence Find the right plan for White to
for White. realise his advantage.

6 9

/ / %W

”/// ”W/WE? /////////

/”W”W
W W?” ”W”W W
/ W / W
What IS correct: 1 @c6 or 1 c6 .7 Find an accurate order ofmoves
for White.
7 10

/W W W
WA/iff/é
sWW/ ///’W

WW
W/
Find the right continuation for Black. Find the right continuation for White.
 Rook Endings J 03

11

WW%/
W//%/¢/%
/W / /W /
M W W WW
/ / // W:
W W W
WW
What plan should White choose? Choose the correct continuation
for Black.
12 15

W/ W /W
W W W WW
W
[III

/ / WY? W
/W/WWW
&

/W W” W
W/ W W W
W////@ W/
WWW// WW W W W/W
How should White conduct Find the right continuation
the defence? for White.

l3 16

W//
W W/W/W
WI!” 1/ ,W
/
/ W W W,W
W: ;W W W8
W?
:W/W/W/W %W%W/W
_W W W/
W W WWW/W
///W/
W /W W
What continuation should What is correct: I §d5 or 1 Ed] ?
White choose?
104 Rook Endings

l7

,/<W/W
W04

///,..

W W7 /%////W%
/// W
/%/%///
/ W W /W/ /%/
WW W W WW
What ts the best wayfor White White to play and win.
to achieve a draw?
18 21

/ W WA/ / / W WW
//;WMW} 3W W
W///WW/
W /WgW / W W W
W W/ / / W //
//W fl/g
/-—-.
.—
.—
WW W
//W%
How does White win? What IS White s winningplan with
an exactly calculated variation.

19 22

2W
WyWyW/W W W /
/W W W%/
W/

WzW/QWWW
WQW
W W WW
W%W /%W W
WWWW
W W W/
/ W/W W
2% // ///...///

W/
///// W
What is correct: I...Exh4 or 1...Eg1+ ? How does Black defend?
 Rook Endings I 05

23 26

WW,/ WJWW
WWW W //
/W//% //////

/ W W
W/WW
W // W/
W W W Wy/W
W
White to play and win. What is correct: 1...a8=g or I @137 ?

24 27

2% %
\

W? W W/ W W W W
7929 /92W
//%999,9 W
/ /9,
/W/W%Wfix

9 9 99 \\
W/ W /a W? W W W
WW W/W W W%/ W

What is correct: 1...@e3 or LUQxfl? White to move. How does he win?

25 28

W% W/ W???é/A
/ / WW WWWW
W W WEW WWW;
\

W/WW/ WW/W/Wis,
\:\\1§
\

\\\\

/
% W 9% /E/W/ ///@%
\

W/W//
W /W W W
/ /9/W
// /

What 18 correct: 1...h3 or 1...@h1 ? What IS the winning move: LUQgZ

or 1...§a1?
1 06 Rook Endings

29 32

7/ 7 7 /
//
%/'¢4/ fi/‘
‘1/
/
7777 7H 7 7 7
“7
7/23 33
557/ 7
/77727/
7/7//7 / ////%7
How does Black win? Can White make a draw?

30 33

7/2. 7% V7 7 7,7
7/777 7 7/ ”7777777
7 7 7 7
%//7/ 0/7 28A
7/ 7 7 7%
7/7/7
/ /7 ///
7%7%/'
/7 7 7 7 7 7
How does White win: I h6 or 1 @g4?

31

llllll

7 7 7 7
77 7 7/7:7
/
7 7 7/7
Can Black save the game? What gives chances ofa win:
a)!§xa4orb)1a’7?
 Rook Endings 1 07

38

W? W W W/Wk//
W/ / / //
W W WE/ZW W
/£3¢ /%
M W W W W W W
W W;
W W WW
Threatened by 1...§g3, how can Where is the clear drawfor Black?
White make a draw?
Calculate the variations.

36

{UH/é
,/// //

W%W W/
///%/// / ////%
How does Black make a draw?

37 40

Ey/W/
W%W//%W W7W W WV
//////// ’////,

W /e//,MW /W W W
LW/fiW /
/

NW WSW W
,WW WyW W/W W W
:2//% WaW //W
White to play and win. How did Blackplay?
108 Rook Endings

41 42

fl 2J4
/” 9: ///
%1/ /// // // /
a /% %
/ /// //
§///’
/// %7////

//V // fl 7 /,/
/&% / /: ////////¢8é%
//////
///////%’
How does White win the game? Can Black make a draw?

Point out a wmmng plan for Black.

 4 Shouldering:
the struggle of the kings

A reciprocal attack by a piece is

not usually restricted by either time
/
or, more so, by space. However,
when we find ourselves talking 2 2
about kings then the situation is
cardinally changed—the king can
7
neither attack, nor, more so, capture 72 77:7
/2 //2

\
the enemy king. Every king creates
around itself some strong squares in
which can enter any enemy piece,
2M2/272/
except the king, of course—this
feature makes its biggest impression 7 7 /7 /7
in pawn endings.

Manukovsky-Pustovalov 1 @e6! @c3 2 @d5! @b4 3 QCG

Rosija, 1992 $35 4 @b7 wins.

22 Also possible is double

2/2;2// shouldering.

2 2 2 Zinar

///2 272
1 984

2 2 272 7///
2 2 2/2 2i2 2 2
/7/7/7//
l...®f4!
Only so. INQXB 2 @f5 leads to a 7 7 7 2
draw.
2 222 77/
////7/7
\\\\

After the text Black resigned.

The mixture of shouldering and
zugzwang is indeed deadly.

Or yet another classic study by 1 @e2! @gZ 2 a4 $g3 3 @e3

Maizelis. @gZ 4 35 @fl 5 @d4 wins.
1 10 Shouldering

However, it is possible to define L.Mitrofanov

shouldering, particularly in pawn
endings, as a form of zugzwang.

Mikhalchishin-Azmajparashvili
42/:
Tbilisi, 1980 22/,
222257
2/2/ //
2/2, %//
22//
//
l Ed7!
2/2
2/// Bad is l Ef7+ @e4! 2 Ee7+ @d5
227/ 3 Ed7+ @e6 4 Ed8 Ec5+ and
5...Ed5.
/E 1.2134 2 @g4!
The only move—if 2 @g6, then
2.. .Ec6+ 3 @g7 Ec7 with a win.
It is obvious that White is strug- 2...Ec4! 2 Ed 96+ 4 @gS
gling for the draw, and the question @d 5 114 @e3 6 h5 Ec5+ 7 @g4l!
is whether Black’s king will succeed This move was made before, but
in getting back. There followed nearer to the opponent’s king it se-
1 @xfl al=y 2 Exal Exal 3 f5! cures a draw, while the‘ more ac-
It is dangerous to advance the tive’ 7 @g6? loses because of
other pawn. 3 b5? @c4 4 b6 @d5 5 7. @f4 8 h6 Ec6+ 9 @g7 @g5 10
b7 Ebl, and White has problems. lg Ec7+ ll @g8 @g6 12 h8=5b+
3...?c4! f6.
If 3...?xb4, then 4 f6 @c5 5 @g7,
and there is no defence against Sometimes even the greats make
f6-f7. mistakes in the endgame.
4 @e6! Alekhine-Bogolj ubow
Here is the shouldering—king
World Championship (m) I 929
against king, shoulder to shoulder.
There is nothing else—if 4 f6, then
4. .QdS 5 Qg7 @e6 6 f7 Ea7 etc. %:é //
4.. .Eel+
If 4...Ea6+, then 5 @e5!——the ”2222/
22 //
king goes to the other side, but the
main thing is not to allow in the 2 2 222%
enemy king.
5 @d6! Efl 6 @e6 Eel+ and he 2”2 2 2
had to agree a draw. 2 2 2 2
This idea was expressed simply
2,2 2
brilliantly in one study.
 Shouldering 1 1 1

l...§g4?? 2 b7 f5 3 b8=w Exbs 4 Exbs f4

Forgetting about the shouldering. 5 @d5 13 6 QM f2 7 £18 @g3 8
l...$e4 2 b7 f5 3 b8=yl Eb 4 @e3 and Black resigned.
fib f4 5 E68"!- Qd3 6 §f8 $63 7
@d5 B would draw.
 5 Complex Endings

Playing to exploit a pawn fif! 3xg3+ 8 @xgil gxf6 9 c6

majority on the flank ye5+ l0 QB f5 11 @c8+ $17 12
l§b7+ Black resigned.
One of the most important factors
in the endgame is the possibility of
A classical example.
creating a passed pawn—~and in the
majority of cases this is formed
Botvinnik-Rabinovich
from a majority, i.e. from a pawn
Leningrad, 1934
advantage of two against one or
three pawns against two. From this
follows that the majority is the pre- /®/
cursor of forming a passed pawn
and in itself is an important posi-
tional factor in the endgame. The
fewer the pieces, the more important
,/
éa,g%
2 ‘?
‘/ ,// , , ,
this factor, and the greater the role it
plays in the position.
% s%.a.a
£W;%% 7,/
Playing to exploit the passed
pawn in the endgame is the most ex-
2f} , , , ,/§M/
treme case of the majority.

Portisch—Ribli The presence of open lines and

Skelleftea, 1 989 White’s pawn advantage on the
queen’s flank determines his su-
periority. Among his tasks now is to
create a passed pawn.
/%///4¢4 1 a3 EM 2 Ebz £g6 3 b4 @18 4
1%? /////,.//// / @fZ @e7 5 @e3 @d7 6 .932 @c7 7
//A
/‘ fiy/fi
////
b5 §b8
After 7...a6 8 a4 ab 9 ab
/////// V/// White controls the a-file and in-
%% vades the opponent’s camp.
8 a4 f6
% //4/ II (It , 8...?b6 follows 9 fidZ.
9 a5
More accurate is 9 f4!.
1 ml! 34:722gf3 h5 9...e5 10 fidZ .917 11 f4! exf4+
@67! a5 5 @112 66 12 @xf4 §e8 13 9.13 £e6

After 13...§.xc4 winning is 14
£02 3d3 15 §x05+ @b8 16 .936
§d8 17 201! with the idea of Eel.
14 acts g5+ 15 913 fins 16 ads
ids 17 $63 fies
The pawn ending is hopeless.
18 E212 £b7 l9 EdZ §e8+
On 19....Q.c8 follows 20 §d3 with
the sequel @e3-d2-c3, h3-h4xg5,
35-a6, Ed3-h3.
20 @a gas 21 g4 nos 22 fid3 3.d 20 @d Exa3(!) 21 §e5
firs 23 §e3 f5 24 gf Exf5+ 25
@gB
White’s position is winning due to
the threats b5-b6, a5-a6.
Botvinnik-Kan
USSR, 1 955
y/
/fl
”/2
7
//g//
////
From what we have said before it
follows that White needs to ex-
change as many pieces as possible.
And he sets about realising this
plan.
1 §d4l f5 2 EeZ! g5 3 fixe4
523x“ 4 @dz Qxd2 5 @d!
The king must head towards the
majority.
5...§h6 6 e4 £f4+ 7 @el fxe4 8
fixe4 c5 9 £c3 a4?!
Stronger is 9...§a6.
10 bxa4 §b6 ll §e2 @d7
On 11...§b1 would have followed
12 235.
Complex Endings I 13
12 @d3 Ebl l3 Eel §b6 14
@c2!
Now White needs to defend the
b1 square.
l4...§d6 15 §e4 b6 16 a3!
It is better to give up the pawn on
the third rank—prophylaxis!
l6...§g6 17 35! EgB
There is no saving himself by
17...bxa5 18 $.a §a6 19 £d2
with a decisive advantage for White.
The rest is forced.
18 a4! $18 19 E92 ba 20 @d3
§d8 21 §e4 Qc6+ 22 @e2 @b6 23
gel! icl 24 g3 hxg3 25 fxg3 §h8
26 M gxh4 27 gxh4 £b2 28 £d2
.9..d4 29 Qf3 £b2 30 igS fif8+ 31
EM §g8 32 fifS $d4 33 h5 @c6 34
h6 @d7 35 @g4 @e6 36 §f4 §b8
37 §e4+! @f7 38 §e7+ @g8 39
@f5 with a win.
Vukié-Pietzsch
Sarajevo, I 96 7
l...®e8
Intending to transfer the king to
the queen’s flank to neutralise the
opponent’s pawn superiority.
2 58M! e6 3 @g2 @g7 4 Ed3
521% 5 Eedl @xd7 6 Exd7 h5 7
@f1 @f8 8 @e2 $98 9 Exc7 Exc7
By placing his pieces in good
positions, White forces his opponent
[[4 Complex Endings

to exchange off, thus 1ncreasing the Georgiev-Khalifman

significance of the majority factor. Pardubice, 1994
10 §d4 f6 11 f4 @e7 12 b4 b6 13 {I

@d3 §c8 14 @c3 e5 15 fxe5 fxe5

16 §d5 @e6 17 a4 a6
. 7Q/
17.. .fit‘Sl? would create more 77 7177
problems for White.
18 a5 b5 19 ECSH Ec 20 bxc5
l
7/;/ %/
77177 7 13/
QM?
And here it was still not too late to
escape with a draw by 20.. .b4+! 21
@xb4 §d7==.
/7773.77
777777 8
21 cb ab 22 M! Black 23,7 74/77/77//7
resigned. 7 7:77 7&7
1 f5!
Martinovié-Yudasin The d4 pawn is well blockaded
Krynica, 1998 and White has a serious advantage
on the king’s flank.
IMQI'S?
y// Correct is 1...Ee8 2 M Ee 3
@e £d6 4 h5 $35!, defending
/ the black king.
// 2 h4 §e8 3 h5 Ee 4 @e 9.115
%,/ 7A 7 Better is 4....Qd6 5 f6 gxf6 6 gxf6
7,8 7/ 4,7// fies 7 ih6+ @gS 8 £g7! with the
idea h5-h6-h7.
77 $7 5 f6 gxf6 6 gxf6 @gS 7 3% b4 8
// /7//
//%7 %% "/77 b3! £e6 9 QB idS 10 @f2 axb3
11 axb3 3xb3 l2 2g7 $.18 13 h6
with a winning position for White.
Whose majority is better——
White’s or Black’s? After 1 @e2 Kovacevié-Tosié
@d6 2 @d3 e4+ problems with the Belgrade, 1998
b3 pawn appear for White. Best was
1 f4 @d6 2 $13 with equality. In /@7 / «
the game, however, there followed
1 g4?
27
He should only play on his weak
flank when it is possible to create
weaknesses for his opponent.
1...Qd6 2 gxf5+ gf 3 c5 bc 4 / 7 77
bc @b5 5 c6 f4 6 @c2 @d6 7
@e2 Qxc6 8 Qdfi @dS
7
And Black has a winning
endgame. ’77 7

Against doubled pawns, the natu-
ral strategy is to exploit the majority
on the other flank.
1 h3! a4 2 $13 §d7 3 Exd7
<$xd7 4 g4 hxg4 5 hxg4 b5 6 b3!
Black cannot create a passed
pawn.
6. .21b 7 axb3 @e8 8 ®h4
earing f4-f5.
f7 9 @e4 g5 10 $33 gxf4 11
@xf4 $18 12 @d2 @h8 13 @gS
$17 14 @h6! @g6 15 @313 @f4 16
QM c6 17 ®g5+ @e7 18 @e4 c4
19 bxc4 @dB 20 g5 Black
resigned.
Psakhis-Bonsch
Tmava, 1988
// / 4.x”:
//y’// z /
/$/ ”%
/£///$3// //
/ // ///
//'/o'//
//_:=?//,
1 @fl!
The main trump in White’s posi-
tion is the pawn majority on the
queen’s flank. Psakhis exploits this
advantage very instructively.
1...Wc8 2 @e2 @d7 3 @dl! $18
4 @c2 g6 5 $.c3 $.g7 6 213! $18 7
b4 cxb4 8 axb4 Qb6 9 @b3!
White has 1n effect an extra piece:
his king actively supports the pawns
on the queen’s flank.
9..We8 10 @b2 Wd7 ll $.d4
ales 12 $.e3 @e7 13 was: wet;
Black is forced to exchange
queens and the remainder becomes
a matter of technique
14 gm Qxc6 15 @d3 $.g7 16
$34! QB
Complex Endings 1 1 5
On 16...®xe5 decisive is 17 ['ee5
$.xe5 18 @a5, and if 16...$.xe5,
then 17 b5 axb5 18 cxb5 $.d4 19
bxc6.
17 $.c5+ @e8 18 $.d6 f6 19 exf6
3111.6 20 523c5 e5 21 @1136 e4 22 b5
QeS 23 $.c5 @d4 24 $.xd4! $.xd4
25 @113 @d7 26 c5 h5 27 g4 hxg4
28 hxg4 g5 29 QM e3 30 @d3 and
Black resigned.
If 30...?e6 winning is 31 b6!;
while on 30...?c8—31 c6 $.b6 32
@b4 and 33 @d5.
The following game serves as an
example of a bad majority (i.e. a
case when the majority can become
an object of undermining and
attack).
Kan-Keres
Moscow, [952
z/"”/// ’//
21/19
”6‘/‘
/ /
2 2 fly/ 9 Ixx //
/
1.. 21.5!
Trying to create a weakness.
2 2d Ed 3 Ed] @d5 4
$3d+ $.d 5 c4?
Creating a bad majority.
5. .$.e4 6 Ed @xd8 7 $.d1
91:7 8 an @116 9 f4 e5 10 g3 .2111
11 a4 exf4 12 gxf4 @e6 13 $12
QfS 14 @e3 115! 15 $13
If 15 M $.e4!, zugzwang.
15...h4! 16 $63 99.6 17 2. 4+
If 17 $13 .215 18 @gz f6 19
@hZ $.e4! and Black has a decisive
advantage.
116 Complex Endings
17....91'5 18 $13 @f6 19 $.f
$f 20 ees f6 21 $13 g5 22 fn
fn 23 @g2 g4 24 hxg4+ @xg4 25
$12 QM White resigned.
Ljuboj evié-Ivanchuk
Buenos Aires, 1 995
W / W
White has a pawn majority on the
queen’s flank, therefore he can
boldly improve the positions of his
pieces which only increases his
positional advantage.
1 c4! $.n 2 Exd7 Exd7 3 Exd7
@xd7 4 $n ye6+ 5 @gl f5 6
gha! @e4 7 b4 e5 8 b5 f4 9 gxf4
exf4 10 bxa6 bxa6 11 @xa6 yel+
12 @g2 and White has a decisive
advantage.
Hiibner-Spassky
Candidates (m), 1985
,/,/®/
/£WW W W
/ , ,L.
WWW,
W
W W
W,
White’s extra pawn on the king’s
flank proves a decisive factor.
1 @f2 @f7 2 g4! @e6 3 @xe6
@xe6 4 @e4 d6 5 5! fxg5 6
@n+ $15 7 @xh7 xf4 8 @f6
£c8 9 @e8 £15 10 c3 Rbl 11 a3
$.32 12 b4 c5 13 @xc7 3&4 l4 h4!
cxb4 15 axb4 @e4 16 @e8 @e5 17
@e3 .917 18 @c7 QgG 19 @b5
s 20 @1137 @c4 21 @c6 Black
resigned.
Toothill-Heemsoth
corres, 1986
WWW‘
42%
// W
g%/
W
W W///
W / ////”’
My /WW 8%, //
W W
lufifs!
With the unequivocal intention of
driving back the opponent’s king
from the centre.
2 Edz f5+ 3 gxf5 gxf5+ 4 $13
c4! 5 a4 a6 6 £14 EdB! 7 Exds
Taking into account the previous
comment, more chances for White
rested with 7 302, retaining as many
pieces as possible.
7....Q.xd8 8 £33 b5 9 ab 21b
10 Rel 2c7 11 h3 @d5 12 @e3 b4
13 §d2 b3! 14 f3 f4+ 15 @f2 Res
16 3c] £d4+ 17 @e1 9&3! White
resigned.
Mutual Majorities
Each side has his majority, and
the main problem is how to limit
mobility from the opponent’s side.
With level pawns, if one of the
rivals has a majority on one flank,
 Complex Endings 11 7

then for the other it is the opposite Yates-Alekhine

-—this is a fact. But often it will The Hague, I 921
happen that one majority is better
than the other, then we speak of a
qualitative majority.

Karpov-Yusupov
Dortmund, I 997

/,//%/
.\\\\\

?mu
”I”/if"
//fl%
,, ”W
,
423/ 4/g/ .
\\

A classical position—White loses

it without a struggle.
2/%&
//’% l g3?
Incorrect prophylaxis!
l...§f7 2 c5 @f6 3 Qc4 $38!!
A terrible move. For Black the
1 58“! bishop is more important—so it will
It is necessary to keep in his sights stand firm on the 08 square.
the opponent’s weaknesses—— 4 a4?
White’s plan consists of an ex- Here it is worth playing 4 f4!?
change of queens and advance of exf3 5 @fl §d2+ 6 @xfB fit 7 b5
pawns on the king’s flank. with counterplay.
1..5Dd72b3f53@f1 4...g5 5 b5 f4 6 @fl §d2 7 gxf4
Also good is 3 g4 g6 (3.. .fxg4 4 gxf4 8 @el EbZ 9 $32
hxg4 QB 5 Wf5) 4 gxf5 gxf5 5 After 9 Edl £g4 10 §d6+ @e7
yc3+. 11 Ed4 $.13 12 QdS §b1+ 13 @d2
3...?f7 4 f3 @e7 5 $e2 We6 6 63+! 14 fxe3 Edl+ 15 @c3 Exd4
ya @f6 7 @fZW 16 @xd4 QdS -+.
The king will be needed on the 9...?e5 10 c6 bxc6 ll Exc6 $.36
other side. 12 fidl fibl 13 §c5+ QM 14 £02
7...@d7 8 g4 @c6?! e3 15 fxe3 fxe3 16 §c6 3g4 17
Better really is 8...fxg4 9 hxg4i. §d6+ @e5 18 h3 £h5! White
9 @e5! resigned.
Now the knight ending is But correct prophylaxis was
hopeless. possible:
9mgxe5 10 @xeS @d5 11 523M (a) 1 f4! (intending @gl-fZ-e3)
fxg4 12 @xb6+! @c6 13 @c4 1:13 §d2 2 c5 @fi 3 £c4 £xc4 4 Exc4
14 55a @d5 15 @xf!’ g5 16 g) $66 5 b5 @d5 6 c6 bxc6 7 bxc6=;
h5 17 @d2! @e5 18 e4 @e8 19 @e3 (b) 1 f3! (to create a weakness on
@c7 20 @c4+ @f6 21 @fZ! @a6 22 64) 1...e3 2 f4! Ed4 3 Eel fixf4 4
@g3 @b4 23 M! @c6 24 a5 @b4 Exe3 3c4 5 £xc4 fixc4 6 §e7 with
25 @d2!i. a drawn ending.
1 18 Complex Endings

Bronstein-Rantanen Portisch-Kramnik
Tallinn, 1975 Biel, 1993

/
A’
.....%A/W/x//
%/
/// // /'/%
//. ”/67; /
2/ A
’// /’/
\

/7
1 g4!
Advancing on the other side—but
here it is important to force back
and restrict the black pieces.
1...36 2 g5 @e8 3 a4 §a7 4 M
§b7
Better is 4...?f‘8!?.
5 §d3
Now it is important to control the
d-file.
5...§c5 6 Eel §d7 7 Eedl!
Tactics. If 7...e5 there is 8 58155:
7...g6 8 @e2!
Typical.
8...§xd3 9 Exd3 b5 10 cb
ab ll §d7 @1’8 12 35 ECG l3
§b7 b4 14 §b8 fies 15 @g3 Black
resigned.
A classic example of exploitation
of a pawn superiority on the queen’s
flank is presented by the following
position from a modern tournament.
Here Black has a pawn superiority
on the queen’s flank, typical for the
Grfinfeld Defence, and therefore
White should keep as many pieces
as possible so as to maintain the
tension.
a/
1 §c3
Ifl @a3 @c4 2 %4 b5 3 a4
@b6! 4 axb5 @dS! Black maintains
a small advantage due to his control
over the central squares.
l...b5 2 fifcl
Bad is 2 Ea3? because of the
tactical blow 2...¢3c4 3 fixa6 QM.
2...®c4 3 ’@e2
On 3 @d2 there is the effective
3...a5!; also in the event of 3 e4
there is 3...f5!.
3...®b6!
With the idea of capturing on c3.
4 §c7
4 $35!? deserves attention.
4nfle6 5 figs
And now the only chance was 5
$35!? with a slight advantage for
Black.
5...5Ad5 6 §7c5 h6 7 £h4 b4 ¥ 8
m2
There is no saving himself by 8
@c4 EXCS 9 yc ®c3 T.
8...5Bc3 9 Exes Exes 10 @hl
QM”
Prophylaxis. 10.. .a5 11 a3 @a4 12
Ec cS 13 gm Wcl+ 14 @hZ
Exa3 15 d gave White
counter-chances.

11 Ea]
Other moves are also no help. 11
m3 @v! or 11 a3 @a4.
11.. .a5 12 ybfii @b 13 axb3 g5
14 £g3 a4! $
In complicated endings
usually to be found complicated
plans (many stages), various pieces
and asymmetrical formations.
Therefore it is necessary to pen-
etrate deeply into the position, tak-
ing care over the exchange of
pieces.
Balashov-Korchnoi
USSR (ch), I 969
11111111
/£filfi
/% %t%
/ % g/V/
/ flafi/'/ //
//¢ %
’/”
5% //’V
White seems to have a very cen-
tralised position and a solid pawn
formation, but Korchnoi begins to
undermine the white structure.
lufifsl 2 @d2 f6 3 fifl
White wants to exchange rocks to
neutralise the pressure.
3....Qd8! 4 En £c7
Increasing the pressure on the
centre—improving the position of
the bishop.
5 5Dfl 3b6! 6 §d2 fxe5 7 fixes
EH 8 QeZ d6 9 £c3 £c4+! 10
@e1 fixfl!
A surprising exchange of the
strong bishop, but now it is more
important to create a passed pawn.
11 @xfl e5 12 @gZ e4!
Complex Endings 119
Not 12...exf4 because of 13 EB.
13 EeZ d5 14 Res RC7!
An exchange of strong pieces.
15 ixc7 Exc7
But now the weakness of the b2
and g3 pawns is felt.
are 16 @fl §b7 17 c3 @b5! 18 ECZ
QM 19 QeZ Eb3! 20 @e3 d4+!
A blow!
21 §xe4 dxc3 22 bxc3 Exa3 23
g4 hxg4 24 EgZ Eb3! White
resigned.
Mikhalchishin-Beliavsky
Moscow, 1 981
/.~M
Q?
we
”I
/’//
% /
’I
Black has a weak pawn on c4, but
together with this there is strong
counterplay on the d-file. White
should combine the improvement of
the positions of his pieces with
prophylaxis.
1 @c5! ids
If l...§d2, then 2 if]! fiv 3
5Dxe6 fxe6 4 9104 with material
advantage.
2 f3 §c8 3 @3214 .936 4 e4
White’s cunning manoeuvre has
cut off Black’s play along the d-file
and restricted Black’s setup with an
ideal structure f3-e4. Only now does
his king make an appearance in the
centre.
4...¢3d7 5 @f2 @b6 6 556!
The knight is best placed on a
blockading square.
 120 Complex Ending-5'
6...f5 7 @e3 fxe4 8 fxe4 $18 9
Ebl! _ prove his position, therefore he re-
Now he has time to improve the
position of his rook—the threat is
Ebl-bS-aS.
9....3d7 10 @d4 @e7 11 EM
EdS
If 11...?d6, then 12 e5+ @c7 13
a4!, taking from the knight its last
good outpost on b6.
12 fixc4! £h3+ l3 @d5+ 523d
14 ed .2.n 15 @e5!
Pawns are equal but the activity of
each of White’s pieces is clearly
greater——this is also a typical case of
domination.
15...§d7 16 fibs! .9.f3 l7 §g8 g6
18 d6+! fixd6 19 §g7+ $18 20
§f7+ Black resigned.
Fantastic technique in a compli-
cated endgame was demonstrated by
Ivanchuk in the following game.
Dao Thien Hai-Ivanchuk
Moscow, 1994
/////7////14
///;Q
”/
74 7 7 7
737 / 7
7&7 7g//
a 7
777
At first sight it seems incredible
that Black should be playing for a
win—White has a pawn majority on
the queen’s flank. First of all Black
improves his structure on the king’s
flank and thereby restricts the oppo-
nent’s pieces.
1...f6! 2 @e3 @fl 3 @dl
It is very difficult for White to im-
sorts to such unnatural maneouvres.
But 3 h4 deserved serious attention.
3...e5 4 .932 @gS! 5 ficl @e6
With each move Black improves
the positions of his pieces—the
same cannot be said of White.
6 @c3 f5 7 a3 3c6! 8 b4 .934 9
@bS?!
White did not think that his oppo-
nent would so readily part with his
bishop, therefore better was 9 idZ.
9....Q.xb5! 10 cb @f6 11 3M
9&3!
Preparing the d4 square for the
knight, while preventing a3-a4 or
ficl-dZ.
12 @fl l93d4 13 f3 f4!
/////
/7
7 f
7
A decision by a grandmaster of
extra class—weakening the white
squares rather too much, but in re-
turn restricting to the maximum the
white king and bishop c1.
14 @fZ h5 15 g3
If White waits, then Black pro-
ceeds with ...-g7 g5-g4 and then
...-@f6-g5-h4 with the threat of
...g4xf3 and" .Qh4 h3.
15...g5 l6 @gZ 53d!
Surprisingly, after the knight
move many threats appear—for
example, not possible is 17 3b3
because of 17...®e1+—this is a

consequence of the restricting ma-
noeuvre ...f5-f4
17 @fl RbZ!
One cannot leave out the tactical
element—it has been well known
since the time of Capablanca that to
achieve the maximum one must
transpose positional factors into
tactical ones.
18 $.b
If 18 £b3, then 18....Qxc1 19
£xc2 fxg3 20 hxg3 9&3, and with
the presence of the passed pawn on
the h-file Black ought to win easily.
18...5/3e3+ 19 @e2 @xc4 20 gel
@f5 21 @d3 @d6 22 a4 g4 23
fxg4+ @xg4!
Precisely this—the main thing is
to break through to the h2 pawn.
24 gxf4 exf4 25 £b2 13 26 $d4
Qh3 27 a5 @xh2
Not everyone would pay attention
to such ‘trifles’ as the b6 pawn.
28 axb6 axb6 29 Rxb6 @g2 and
White resigned.
Here is a very complicated re-
alisation with pawns on just one
flank.
Romanishin-Nunn
Debrecen, 1992
E77
77/7 /
// 7 7‘7
//
/ %7fl7%7§7
77 7%7/7/
//‘@
//
7?7 7'
l g3 h6 2 $14 @lfl 3 fidZ §c6
Stronger is 3...§c5 and 4...h5.
Complex Endings 12 1
4 EdS @g6 5 g4 .916 6 @gZ RgS
7 £g3 .937 8 M §c4 9 §d7 £16 10
f4
Black already has some trouble
—White activates his phalanx to the
utmost.
10...§c5 ll §d6 @g7 12 9.12
EaS l3 §d3 EbS l4 @g3
Now that the rock defends the
third rank, White improves the
position of his king.
l4...§aS 15 £b6 EbS 16 9.33
£b2 17 913 gal 18 fidl £c3 l9
Eel §b3 20 9%!
Beginning a very important
stage-—-activating the white king.
20...§b4+ 21 @d3 .216 22 g5
hn 23 hn £d8 24 Ec6!
Preventing 24...f6, on which fol-
lows 25 §d6 937 26 §d7!+—.
24...§b3+ 25 @e4 f5+?
It is not clear why—the fact is that
White does not threaten the move
26 f5 because after 26...§xe3+! 27
@xe3 £xg5+ Black has a certain
draw.
26 $13 £e7 27 §c7 $18 28 @e2
§b2+ 29 @d3 Eb3 30 @d2 @e8 31
9&5! £d8 32 §g7 $235+ 33 @e2
Eb2+ 34 @d3 §d2+ 35 @c4 fiez
36 g6! Ee4+ 37 @d3 Exf4 38 §e7+
@d8 39 g7 §g4 40 EN! Black
resigned. A fine piece of work.
Nezhmetdinov-Roman ovsky
Moscow, I 95 7
@E/
////‘ ///
7 7 7 7
37,7 1ӎ
 [22 Complex Endings

What plan should White choose? l fies!

1 g4!!
With the intention of playing
further 12--f4 and e4--,eS gaining
ground, therefore Black’s reply is
forced.
l...hxg3 2 £xg3 @318 3 @h2 @e6
4 f4 b5 5 Egl g6
White has threats along the g-file,
therefore Black is forced to weaken
himself
6 55h6+ @g7 7 ®g4 f6
How otherwise to deny the knight
access to e5? If 7...£.f6, then 8 e5
and 9 f5 with a strong attack.
8 f5!
Reminiscent of the play in a
famous Lasker-Capablanca game.
8...5Dg5 9 nl c5 10 e5! .2.f
On 10...fxe5 White plays
£xe5+ @h7 12 fid! fid 13
fxg6+ @xg6 14 Ef6+ @h7 15 §h6+
with decisive threats.
ll exf6+ $18
On 11....@.xf6 there is 12 @xf6
®xf6 l3 £h4.
12 @eS! b4 13 @ds Rxc2
More stubborn was 13...5/3f7.
14 f7 §e6 15 58d7+ Black
resigned.
Lputian-Tukmakov
Moscow, 1983
’l/
/%// 2/
§\\
Clearly not 1 a4 because of
l...58b4, and the black knight jumps
to freedom.
1...§e6 2 @e5 g5 3 Rg3 @g7 4
@313 h6 5 M!
Such moves are always unpleas-
ant. White latches on to the weak-
ness and wants to ‘occupy’ it. If
5...f6, then 6 h5 91W 7 §d3 with the
idea 5.13M, though Black can of
course defend himself for a long
time.
5...g4 6 @el!
It turns out that the knight is head-
ing for the f4 square.
6....9.d6 7 ixd6 Exd6 8 §c4 h5 9
99.2 .
The exchange of bishops has not
11 eased Black’s position very much.
9...5Db8 10 ficS @h6 11 @d3
@d7 12 @215 36 13 g3 Eb6 14
58M!
By covering the file, White pre-
vents the activation of the black
rook.
14...Ee6+ 15 @fl 9 7
If the passive 15... b8, then 16
@d3 §b6 17 5.13M §b5 18 Exb5
axb5 19 @e2 @a6 20 ®d5l with a
winning knight endgame. Now
however any capture is quite good.
16 E11136 E93
If 16...§e4, then 17 @d3 5565 18
5AM! 558 19 53t+ @f8 20 fia8+
$e7 21 §e8+!!, exploiting the

tactics.
//,/// 72A 17 @d5 @c5 18 §c6 @e6 19 §c4
§b8 20 a4 §b2 21 @f4 §b1+
The rook ending is hopeless, since
///£///
///// /// // the king simply goes to his passed
22/ / pawn.
22 @e2 @f6 23 53t+ @e5 24
a??? fixg4
\\'\

//22 (£7 And Black soon resigned.


Salov-Khalifman
Candidates (m) 1994
5.3/ /
//§@4
/&/ ,/%3
///£a V//
/
//////%
%
White has a majority plus a space
advantage.
l M! f6 2 h5
Spoiling Black’s majority.
2...a6 3 5/36 @e5 4 58d5+ @ds 5
b3 b5!?
Without this move Black would
not have any real counterplay.
6 cb ab 7 b4! £g4 8 a3!
@cS
After 8....§.xh5 9 9366+ attacking
the g7 pawn.
9 @e7+! @c7 10 53 6 .9.t
Or 10...@xg6 11 xg4 5865 12
@3+:
11 @xeS! fxe5 12 @xeS g5
After 12....9.g4 13 5866+ £xe6 l4
@xe6 the pawn ending is simply
lost.
13 @e6 @b6 14 @f6 .932 15 ®g6
Rfl l6 g3 @c6 17 @xh6 @d5 18
’93n .932 19 @gG QM 20 523e6
@c3 21 @fs @b2 22 c 91:33 24
Q3216! and there is no defence
against the march of the g3 pawn.
Exploiting a space advantage...
Complex Endings 123
Nezhmetdinov—Luik
USSR 1 950
White has the advantage on the
king’s flank plus a more active de-
ployment of pieces—there followed
1 b4!
With the threat of 2 QCS, whereas
1....9.xb4? is impossible because of
2 §a8+ §.f8 3 2.05.
l....§.e8 2 Res @fl 3 g6+!
Completely cramping the oppo-
nent——-not possible is 3...hxg6 4
fxg6+ @xg6 5 @xe5+.
3...hxg6 4 fxg6+ Q18 5 h5 £xc5
Already he has to allow White to
create a passed pawn.
6 bxc5 §c7 7 @218 Ec 8 h6!
Here also an advantage on the
king’s flank expresses itself in a
tactical form.
8...gxh6 9 QM §c6 10 @f5 §e6
ll Ebs
White plans to go with his king to
h5, therefore he places his rook to
hold up the b-pawn.
11...h5 12 @g3 b4 13 @h4 b3 14
$t b2 15 QM @g8 16 Exb2
And as Black is in zugzwang, he
resigned.
124 Complex Endings
An entertaining endgame motive
was produced in the game
Vaganian-Smirin
USSR (ch), 1988
7/
////,
/’//
/,¢/
% to the e-file.
//// '//,,:
/////
/ £g3 Qxc5 20 9392 $.13 21 @gl
1 c6 is striking with the intention
of winning in the variation l...£.xb4
2 07 @b6 3 @c4. But Black plays
1...!93b6! 2 c7 (or 2 b5 £d6 3 QgZ
Qf8 with equality) 2....§.d6! 3 08=W
58c 4 £xc8 £xb4, and is in no
danger. The game, however, went
1 Qg2
Surprisingly for White, Black is
saved by a piece sacrifice.
l...523xc5!? 2 bxc5 $.c 3 523g4
Qd4 4 $.06 Qg7 5 M QCS 6 Qg3
£d6 7 QB .937 8 h5 gt 9 ¢3f2
f5 10 @d3 Qf6 11 @e5 2d6 12
Qd7+ Qe7 13 QbS M
The draw is obvious.
Neverov-Dreev
Lvov, 1985
It seems that there is no hint of
danger in White’s position, but...
1 Rdz @e5 2 Qg3 2&6 3 f4
@d3! 4 £33 95 5 fxe5 fxe5 6 QM
Qf6 7 Qg3 Q96 8 QM Qf6 9 Qg3
35 10 l a6 11 $.93
Black’s two doubled pawns
successfully counter White’s three
thanks to the excellent position of
the knight on d3, while the passed
65 pawn has become dangerous,
since the white king cannot get over
11...Qe6 12 QM fihl 13 Qg3
Qd7 l4 @dl @b4 15 @c3 Qc6 16
QM @c2 17 .912 e4 18 QgS e3 19
Rdl 22 h3 h5 23 n6 hxg4 24
hxg4 Rxg4 25 Qg5 l 26 Qf5
Qd4 27 Qe6 @b4 28 fiel 53v 29
.9.a @c1 30 b4 Rg4 31 Qd6 e2
White resigned.
Space plus control of an open
line.
Botvinnik-Balashov
Hastings, [96 7
/ %Q/
A/ WA”
'/ ////
2/ ////
/€1s/
4:?
7/
/ //
/ éé /”
1 34 f6 2 35
White increases his space advan-
tage and thereby drives back the
black pieces.
2....§.d8 3 fies Q17 4 e4 §d7 5 b5
2&7 6 Rdz £d6 7 gas b6 8 a6!
Now not only is there a weak a7
pawn but, when the opportunity

presents itself, also the manoeuvre
Ea8-b8-b7.
8....@.c5 9 £33 @e7 10 h3 @d6
11 §e8 §c7 12 95+! @d7 12 EgB!
There is no sense in winning a
pawn at the cost of exchanging
rooks.
12...f5 13 fixg7+ s l4 §g8+
@d7 15 §h8 9.9.7 16 .364 Ec2+ 17
@d3 Eaz 18 3218+ Black resigned.
Tikhomirova-Morozova
Kiev, 1966
White’s plan is simple—to push
the pawn to a6 and still further
hamper the opponent, while in the
event of an exchange on a5 the CS
pawn will be very weak.
1 53c]! 586+ 2 @d3 @e8 3 34
@d7 4 @b3 QM 5 g3 @315 6 35
@d8 7 a6 @c7 8 @bdZ
The knight has done its business
and now transfers to the other flank.
8...@b8 9 523M! ®h6 10 @de4
§c7 11 58e8+ @d7 12 ®4f6+ fixf6
l3 @xf6+ @c7 14 h3 @f5 16 g4
Now it is time to expand on the
other flank.
16.. .hx 4 l7 hxg4 939.7 18 939.4
@b8 19 e1 @g8 20 3M @c7 21
@e3 @d7 22 @B @c7 23 @d6 f5
24 gf gf
Somewhat better IS 24.. ..ef
25 figS @d7 26 @g3 @137 27
QM!
Complex Endings 125
The invasion of the king is decis-
ive—this is more accurate than
fixe7 and 58c8+.
27...®g6+ 28 @hs 039.7 29 £xe7
@xe7 30 @g6! Black resigned.
Playing to exploit a weakness.
Korchnoi-Pinter
Reggio Emilia, 1987/88
//’//
%
1 @e2!
Hurrying to activate his pieces
would be punished. 1 Eb3 Qc5;1
@e3 @e5.
1...§c8 2 §c3 g6 3 EM Ec7 4
@f4 fihc8
The threat was 5 @d5.
5 35 @d6 6 fibl 939,5 7 h3 @d7 8
M
Gradually occupying space.
8...h5 9 99.3 @e5 10 ficZ @e7 11
@d4! $d6
On 11...§d8+ White had prepared
12 ®d5+l (12 @xeS f6 mate)
12..@e613 $63.
12 §d2 Ed7 13 @c3+ @c7 14
Ed <$d6 15 a6
Securing a bridge-head for an in-
vasion along the b-file.
15...§.dc7 l6 §b3 @d7 17 §d1+
@e7 18 @d3 §d8 l9 Edbl Edc8
20 Eb7 @d6 21 f4 58b6
A mistake in a difficult position.
He should not lose control of the e5
point.
126 Complex Endings
22 @e5 Qe6
The threat was 23 Qxfi.
23 @c4 Exb7 24 axb7 §b8 25
@a5!
Leading to complications was 25
@Xb6? Exb7 26 13+ gf 27 exf5+
Qf 28 Efl+ Q 4.
25...Qd6 26 b4 Qc7 27 Ecl
Qd7 28 Qa3 f5 29 e5 @d5 30 e6+
Black resigned.
Weaknesses in the endgame
and their exploitation
The presence of weaknesses in the
pawn structure is a great minus, and
the correct technical exploitation of
them is the most important factor in
realising an advantage.
Topalov—Beliavsky
Linares, I 995
///
2222
222
2222/ . . .
2%2%2/2/
222
Black has two passed pawns but
at the same time these are serious
weaknesses which need looking
after.
1 c4 ig3
Bad is 1...g5 because of 2 £f6!.
2 Qc2 .914 3 £16 g5 4 Ee7+
Inferior is 4 §g8 ECS! with the
idea of.. 205-f5
4.. .Qc8 5 §g7 §c5 6 Qd3 35 7
£37 EfS 8 c5!
White activates his pieces to the
maximum and also the passed pawn,
and in the process threatens to enter
with the king on b5.
8...Qd7
He must activate at the cost of a
pawn.
9 £xg5+ Qc6 l0 £xf4 Exf4 11
Qa7 Ef3+ 12 Qc4 §f4+ l3 Qc3
§B+ 14 Qb2 Qc 15 Exa5+
Connected pawns in the majority
of cases win against a single pawn.
15...Qd4 16 gas e5 17 a5 §f6
If l7...Qc5, then 18 EbSI, cutting
off the king.
18 §e8 e4 19 b4 e3 20 Qb3 if]
Or 20...§f5 21 a6 EeS 22 ExeS
QxeS 23 a7, and in the queen end-
ing White wins the enemy queen.
21 §d8+ Qe4 22 a6 gal 23 b5
e2 24 §e8+ Qd3 25 Qb4 Qd2 26
Qc5 e1=y 27 Exel Qxel
/ ///2 / /
////§/7/2/////
%%/
/ //
2 2/2 /
28 Qb6!
A typical means of advancing the
b-pawn to queen—the only way to
victory.
28...Qd2 29 Qa7 Black resigned.
As in rook endings a single weak-
ness will usually be defensible, to
achieve victory it will often be
necessary to create a second
weakness.

Kasparov-Andersson
Belgrade, 1985
/ / /
/EW
//
/7 A
cfl/ %//
/// / fiQW
W/WWW
W
All Black’s pieces are forced to
protect the b7 pawn, while White
must create a weakness in the oppo-
nent’s camp. To the question—-
where?—the reply is simple——on
the g6 square!
1h4! @fl 2 h5 @g7 3 @gz §e7 4
§b6 m7 5 9d5 Efd7 6 fil Ee7
7 @g3 fied7 8 QM Ec2 9 @gS!
Exfl 10 hxg6 hxg6 ll Exg6 @1'8
12 9b3 5817 13 @f6 f4 14 e4 Eb2
15 e5 f3 16 e6 12 17 9c4! Black
resigned.
Commentary here is superfluous
since White’s play was direct and
logical.
Play in equal positions
In equal positions first and fore-
most the task is to create weak-
nesses in the opponent’s camp. This
is usually achieved through the
greater activity of one’s pieces,
which, with the help of an attack,
force a weakening of the pawn
structure.
Complex Endings 12 7
Hiibner-Ftécnik
Polam'ca Zdroj, I 995
1 g4! .9415 2 94:2 g6 3 ®g3
@e7?!
The first dubious manoeuvre—
better is 3...9.d6 4 5564 9.e7
4 ®e4 @ds 5 g5!
Beginning a weakening of Black’s
position.
5...fn 6 58n h6 7 @e4 @fi
But not 7.. .5De6? because of 8
@c3
8 @c5 9d6 9 @e4 9.e7 10 936
9.e6 ll @g2 g5?!
Another weakening—better is
l 1...®c6
l2 9e4 §d6 13 9302 94:4 14
@gs @136 15 @132
With the threat to go to h5
15...9.d5 l6 @g4 9.xe4 17 am
9317 18 @115 b6?!
Stronger is 18...?f5 19 @xd6
@xd6 20 @xh6 g4, but White right-
ly plays 19 58g3 @f6 20 f4.
19 @g6 9.18 20 9.c3 @h8+ 21
$h7 9317 22 @gs .937 23 f3 h5 24
@g7 g4 25 fxg4 hxg4 26 @g6 a5 27
@hS QdS 28 @d2 axb4 29 9.xb4
9.xb4 30 axb4 @c6 31 @xg4 QbS
32 91.5 @xh4 33 @g6 @118 34 $16
+—
128 Complex Endings

Endings with a pawn sacrifice Barlov-Seirawan

Zagreb, 1987
In principle, a material advantage
in the endgame is the main factor,
but it can happen that other factors,
particularly the presence of a passed ///‘/ /x 7/,’//,
pawn, are decisive. ’%
Rublevsky-Shariazdinov
/ / //%/ 4r
Elista, 1 996 3152’s , ”/¢€?8
\\\\0

///:% //W/
\ \‘3
\\\\
\\

Though Black is a pawn down,

/,//” // the American grandmaster tries to

fi/%
§

exploit the active positions of his

///
\\
[pa

/§./ / //”4/
pieces and does this, as we shall see,
successfully.
///// / l...f4 2 gxf4 gxf4 3 l93c4 §d3 4
Eal h5 5 M @f5 6 3218 $g4
Black obviously does not rush to
capture the white pawns, for the
1 a5! EM 2 fidS £216 3 @c7! time being improving the position
The threat is b4-b5. of his king.
3....@.d8 4 @e8! 7 Eg8+ @xh4 8 §g6 Ed 9 $13
If 4 @xa6? bxa6 5 £b7 £g5! EM l0 Egl 9&5 11 @e2 RM 12
with the idea 6....Qd2=. $13 QcS 13 @e2 §e4+ 14 @f3 d5
4...@h7 5 ®g2 15 @d2 §e8 16 Ed] @g5 17 @b1
5 @xd6? .937 would be bad d4 18 @d2 @f5 19 Ehl §h8 20
materialism. @e4 £b4 21 fig] h4
5...g5 Slowly the black pawns crawl
On 5...®g6 there is 6 h4!. along to their queening squares.
6 EH! 22 §g5+ @e6 23 §g6+ $f7 24
Zugzwang! Egl d3 25 $xf4 d2 26 @e3 §d8 27
6....@.e7 7 @c7 £d8 8 @xa6! @e2 §e8 28 f3 h3 29 fihl §d8 30
Now that the g5 square is l
inaccessible to the bishop. If 30 @312 Black has 30...h2 31
8...bxa6 9 QM @g7 10 £216 d5 Et d1=§+ 32 523d §d2+.
11 ed £37 12 .938 £xb4 13 a6 30...§a8 31 QeZ h2 32 52312
QCS 14 $13 @f6 15 d6! h5 16 @e4 If 32 fit?, then 32...d1=§+ 33
$.d4 17 @ds $17 18 .915 @f6 19 @xdl §a1+ and 34...§a2+.
@c6! Black resigned. 32...§e8 33 @e4 §d8 34 @312
fics 35 @d1 figl —+.
 Complex Endings 129

Chekhov-Karsa 1 @f2 e5 2 EdZ @a4 3 $.11 @b6

Lvov, 1983 4 @c2 24:4 5 589,3 £b3 6 g4 M 7
g5 £36 8 @g2 113 9 ®h4 Rb3 10
ixh3 QM ll fide2 Eds 12 f4
fidl l3 §e4 93b
The strategy is completed and the
game transposes to tactical
channels.
14 @gl §d3 15 in Exc3 16 f5!
/,, 7 W, .. ,’ QCZ 17 f6 fies 18 ExeS fixeS l9
7// %% ”/ fixes fins 20 §e8+ @117 21 firs
// £b3 22 51313 334 23 @e5 £64 24
58xf7 §g4+ 25 $12 $167 26 Exfl+
@g8 27 §g7+ $18 28 fixg6 §f4+
29 @e1 §e4+ 30 @d2 QC4+ 31
1 b5! cb 2 a6 £c4 3 2.x“ 2.x“ Exc4 32 Eh6 Black
bxc4 4 586 QcS resigned.
Also bad is 4....9.b4 5 @d5 QaS
because of 6 @c7! and 7 a7!, Domination is a superiority in the
deciding the game. functional activity of each piece. In
5 .9.c @3c 6 a7 3218 7 @ds the endgame this increases many
c3! 8 @c! times over.
Black returns the pawn and White
must not fall for 8 @xb6? fixa7! 9
Exa7 c2 10 Ea] ®b3 with a draw. Morozevich-Balashov
8...®b3 9 EaG @d4 10 @d5 @c6 Novgorod, 1997
11 fib! @xa7 12 :36! Black
loses a plece, therefore he res1gned. ///

irov— an e I // ‘////
Eisterlzm, 39'} W/%//{;
///g/ 6%
”//g2%/
§.M/M//
%§%/% %’//m:
%/// y %% a
Z736 /W/ 7/ 1 114! @e6 2 W6 h5
{:39 £33?%/%/& %/ / There is no other apparent
43/ ///// ////8//, , defence against @g4.
7”/,
2%,;-:- 3 @ds ycz
Defending against @e7+ and
Black has achieved the ideal 58xg6.
structure for the Sicilian Defence, 4 5897+ @fs 5 QcS! We4+ 6 @gl
but now he begins a principally in- Wd4
correct advance on the king’s flank. If 6...®g8, then 7 @d6.
I 3 0 Complex Endings
7 wen @gS 8 @d6 @dH 9 @gZ
55f4+
The last chance.
10 @112! @02 11 f'H @hs 12
We8+ @h7 13 Wd7+ @hS l4 Wc8+
QM 15 %7+ @hS
Or 15...@h6 16 523m @g7 17
@g5+.
16 Wbs QM 17 yxafl @g8 18
Wfl+ @118 19 Wf6+ @h7 20 @gZ!
Black resigned.
 6 Defence in the Ending

A complicated, important and Fercec-Cvitan

exceptionally large theme. We can Zadar, 1995
touch upon it only partially.

Playing for stalemate / y 7 é

.. .
How strange it 18 that this method,
//// / M
// AW//
for example, rather frequent in pawn / // ////’/;/ //
endings with an extra pawn, is sel- // / /
dom met in its purest form. fl; // /I‘

Beliavsky-Gelfand
Belgrade, 1997

1...nZ? 2 Exg33! @xg3 3 EM

/fl% / / %// $38 4 Wh8+ @f7 5 @e8 Drawn.
//// /
e/e/
/ //£// @W/
/ Exploitation of a lucky chance

Gelfand-Lautier
// , I
%/ ///// Belgrade, 1997

/ Wr 3% /
W éW /% W 1 4V
/ % /@/
The threat is ...@h3-h2 and then ‘%
////_/ / //
so /
g21+,fif3!! g2+ 2 @xgz £x13+ 3 //%// ,fi%//
@f Drawn.
I32 Defence in the Ending

Here winning are 1 Eel or 1 @d2, Difficult practical defence

but Gelfand decides to play more in the ending
simply.
I ECS??, Beliavsky-Mikhalchishin
Now 1...b4! is winning. However USSR (ch), 1984
after...
1....9.c4? 2 @d2 Black resigned.
m/
x/ //
Shirov—Lautier
AWéM //
/, /
I

/
a ...l

Belgrade, 1997

/W 72/ / /
////, 37/ / //
/7/ fl/fi

/ 4r W /W
/’//'
”Wl/
.,/ g
//// é/// //// fl”
W //W/
Thinking a long way ahead,
Lautier went for the variation
1 Exe4 @xe4 2 f6
Reckoning on 2....§.e3 3 @h4
3h6 (3...?f5 4 @t @e6 5 @g6)
4 @t 9.18 5 @g6 d5 6 M d4 7
fla5 d3 8 h5 @e3 9 h6 with a win.
However there followed
2....@.xf6!! 3 £xf6 d5 4 $h4 d4 5
$t $13! 6 @gS
Or 6 b3 g3 7 hxg3 @xg3 8 @gS
QB 9 @f5 d3 10 £33 @e2 11 $65
d2 12 $.d @d 13 @d5 @cZ,
reaching the pawns.
6...d3 7 .933 c4! 8 @f5 @e3 9
@xg4 d2 10 £xd2+ @d
And a drawn ending with a c-
pawn is reached.
//////
Black has two pawn weaknesses
on e6 and b7 (For the present Black
is coping with defending them) and
a rather worse deployment of
pieces. The question is—how can he
improve his position? Correct is
1...g5!, commencing play on the
king’s flank, but Black played
l...e5?!
It is not usually worth moving a
weakling.
2 f3 @d5 3 5834 @eS?!
Too intricate—simpler is 3...?d6!
4 @b6 @d5 5 @e4 @f6+! with a
draw.
4 5136+ @e6 5 @e2 523d6 6 §b4
@d7
Black wants to defend the b7
pawn with his king, though also not
bad is 6...§c7
7 §b6 @c7?!
Very strong was 7...§f6!, and
then 8...g5 and 9...§h6, obtaining
counterplay.
8 @d5+ @d7 9 b3 h5?!

This is only a weakness—better
was the transfer of the rook via f8,
c8 to c6 The text just weakens the
g6 square.
10 @e3 @c7 11 58M! §d7!!
After 11...@xc4 12 bxc4 the rook
proceeds to g6 and then to g5.
12 EM
If 12 ’93xe5, then 12...§e7! 13 f4
58137 with a draw, while the pawn
ending after 12 @xd6 Exd6 13
Exd6 @xd6 14 @e3 @c5 15 g4 b6!
was also drawn.
12...e4! 13 52)xd6 exf3+ 14 @xf3
Exd6 15 QM b6!
Completely equalising by ex-
changing the important a5 pawn.
16 §c4+ @d7 17 b4 ba 18
ba Ec6! and drawn.
A quite rare means of defence is
without pawns—against a central
pawn and bishop he succeeds in
constructing an impregnable
fortress.
Gretarson-Magerramov
Groningen, 1993
// //
//g/AVZ'V/
7/9 ///
//////
77%, 43“.,{/H/
7/ / end!
In endings with same-colour
bishops very often an extra passed
pawn ‘will not run’.
Defence in the Ending [33
l QM h4!
Cutting down pawn material is al-
ways useful for the defence!
2 @d3 hxg3 3 hxg3 .9.d8 4 @e3
9g5 5 @fi! @e8 6 9d6 @d7 7 9e5
@e8 8 @gZ @d7 9 QB @e8 10
@e2 @d7 11 @d3 9d8 12 QM
9.b6 13 f3 .912 and White cannot
improve his position. Therefore it’s
a draw.
Rozentalis—Christiansen
Groningen, 1992
/////
1/{/71
/17/
/</
///
//// / /%%
///
7 /$
'/
1 f4
He should try 1 g4 .9.b1 2 g5, pre-
paring an attack on the h7 pawn.
l...@d6 2 @f2 39.7!
Not allowing the white king into
the centre.
3 9f3 h5 4 36!? 9g4!?
An interesting decision—the ex-
change of bishOps weakens the d5
pawn.
5 9.9.4 9f5!
Seeing the idea through to the
6 9f3 9g4 7 9e4 .9..f5 8 axb7
fixb7 9 §c6+ @e7 10 9.f gf 11
@e3 Eb 12 @d4 Et 13 @e5
Ee2+ l4 @f §e3 15 §c7+ @d6
16 fixa7 fixg3 17 Exfl @d
I34 Defence in the Ending
fl? / fl
%,/
fl//M%
% / %
// /M%
/%
An interesting rook ending——
Black must play exceptionally
accurately in order to hold the
position.
19 §h7 @d6!
The king must come closer at any
price.
20 @f6 Eh3!
But now he needs to advance his
pawn.
21 §h6!? @d7 22 §h8 fihl 23 f5
M 24 @fi h3 25 f6 h2 26 §h6 @c7
27 §h3!?
White wants to transfer his rook
to the second rank, and then, with
the rook on e2, try to enter with his
king. But Black too activates his
king.
27...?d6! 28 §d3+ @e5 29 §d2
QM Drawn.
Black heads for the h2 pawn and
then plays Efl.
During defence, mistakes due to
tiredness from protracted defence
can occur.
In the following game, over the
course of the last 50 moves, with the
same correlation of forces, White
had stubbomly tried to improve the
positions of his pieces and pawns,
while Black defended technically
and cooly.
Gurevich-Andersson
Leningrad, 1 987
////M, @2/
4% M/
\
%,/f
\
%'@ 78
\§
/ /M//%
/
\
l g5
Finally White goes over to the
attack.
1...hn 2 fn @d7 3 @e4
Wb3+ 4 @h4 @318 5 ®e7+ @hS 6
WM 58M
Leading to a quick defeat. It
seems to us that at the very last mo-
ment Black could still surprisingly
save himself, in a truly fantastic
way, by playing 6...?h7ll. Now 7
fif5+ g6 8 gxfB leads to an im-
mediate draw in view of 8. .wg3+
(yxh3+) 9 @xg3(h3) stalemate!
Meanwhile Black threatens after
_7...5/3g6 to bring the game down to a
drawn queen ending, therefore it is
necessary to accept the knight sacri-
fice: 7 @xfB yc4+l (but not
7...%4+? 8 @hS nor 7.. .wa4+? 8
@g3 ya3+ 9 @g4! Ea4+ 10 W4
Wd1+ 11 $3, and White wins) 8
Qg3 fid3+l (again avoidin a crafty
pitfall: 8...yc3+ 9 $8 éefi 10
@g4 e7 116+! mating) 9 gt?)
(On 9 §g4 e4+ 10 ‘Qf4 fixe7
White lands his queen in a poor po-
siiton—ll g6+ already does not
lead to mate. On 9 @gZ follows
9.. .W while 9 @f4? is im oss-
ible because of 9...Wf1+) 9.. .éd6+
 Defence in the Ending 135

10 @g4 m4+l (this is the point of You should always take the op-
Black’s idea: he refrains from the portunity to spoil the opponent’s
immediate capture of the knight—in pawn structure.
view of 11 g6—and with a correct White has two extra pawns and, it
intermediate check disrupts the seems, an easily winning position.
coordination of White’s pieces) 11 But...
@hS fixe7 12 Wf5+ @g8 13 Wc8+ 1...h3!! 2 gxh3?
@137 with a drawn queen ending. How could he not take a third
7 @g4 WCZ 8 EB Black pawn? Correct was 2 g4 ¢3d2 3
resigned. Ed4! Qxe4 4 Exe4 fixc3 5 @g3,
and by defending the B pawn with
the rook White then wins the h3
Gelfand-Shirov
pawn.
Munich, 1993
2...®d2 3 3d3 Exc3 4 §d4 §c6
5 <$e3 @b3 6 fidS QCS 7 QbS
§e6+ 8 $12 @b7 9 M @d6 10
4 / /, // £d7 §f6 11 Rg4 @e8 12 h5 g5!
The pawn must hold two white
\\\

pawns.
l3 §d8 @d6 14 @e3
2: 44/2/41
’/////
The last chance would have been
14 §h8+ @g7 15 §a8 §h6 l6 fiaS
\

/////8 @e8 17 @g3 followed by h2-h4.

14...513f5+15 $9.4 @h4 16 @125
’////
// // / Ef4! 17 §d3 ®g7
Black has blockaded everything
excellently—therefore it’s a draw.
 7 The isolated pawn
in the Ending
The isolated pawn is a weakness, 7 Qf6 Qc4 8 n6 Qd3 9 e4
but this weakness is felt in different Black resigned.
ways. The strategy everywhere is
similar-blockade by utilising
strong blockading pieces or direct Yurtaev—Temirbaev
attack on the isolani itself. Elista, I 998
We begin with the simplest
example—a pawn ending.

Ehlvest-Rausis
Riga, 1995
///¢:¢//
/%//
Sooner or later Black will run out
of usefill moves, and he will be
forced to allow the white king into
his camp via the e5 or c5 squares.
l...b6 2 a4 Qe6 3 13 Qd6 4 a5
ba 5 ba Qc6 6 Qe5 Qc5
After 6...Qb5 7 Qxd5 Qa 8
Qc5 White also wins.
/ _/
//r ’/ //
//
// 45:7
// // may
Here the b6 point is firmly de-
fended and White does not succeed
in penetrating via the CS square.
1 h4 h5! 2 a4 Qe6 3 a5 Qd6 4 a6
Qe6 5 e4
Sooner or later it is necessary to
try and exploit the slightly better
position of the king, but this is
insufficient.
5...dxe4 6 fxe4 fxe4 7 Qxe4 g6 8
Qf4 Qf6 9 g3 b5 10 g4 hxg4 11
n4 Qe6 12 Qf4 Qf6 l3 Qe4
Qe6 Drawn.
 The Isolated Pawn in the Ending 13 7

Felling-Miles 1 .915! £a6 2 f4

Ilfora', I 974 It is important to exploit his ma-
jority on the flank.
2...®g8 3 h5 @f6 4 $13 Qb7 5
///W//, Rg6 @e7
%// //3
i//%”
Bad is 5....9.c8 6 g5 £g4 7 @g3
£xh5 8 gxf6 £xg6 9 fxg7 with a
3/W/e/W/W
/// winning position.
6 g5 939.4 7 9xe4!
The knight here is undoubtedly
, 3/ //y
stronger than the bishop.
7
// ”/3 7. .dxe4+ 8 @e3 fidS 9 b4 axb4
u I//’,//

10 axb4 .917 11 435 @1’8 12 @g3!

£b3 13 @xe4 @f7 14 @e5 hn 15
fn £c2 16 g6+ @g8 17 QM ibl
1...a5! 18 @gS £d3 19 @f5 @18 20 Qd4
Weaker 1s 1.. .h5 2 b4! f6 3 h3 g5 £c4 21 h6 gxh6 22 @t $.32 23
4 fn fn 5 @e3 e5 6 dxe5 @xe5 @g5 @g7 24 935+ and White won.
7 @d3 @d5 8 @e3 @c4 9 @e4 with
a draw. Panno-Donner
2 b3 34! 3 bxa4 bxa4 4 h3 Palma de Mallorca, 1967
After 4 g4 g5! White is badly
placed, while on 4 @c3 there is
4...@e4 5 @b4 @xd4 6 @xa4 @c4!
7 <3a5 f6 8 @b6 65, and the queen % /m%3{
ending is easily won since Black
manages to exchange queens.
// // « 3W/
4..115 5 h4 f6 6 $6 $e4 7 @c4 ////s/ /1
f5! 8 @CS @e3 9 @b4 $13 l0 @c5! / /% % /
@ni 11 @d6 @xf4 12 @xe6 @g3, // /ZԤ
and Black just as easily won the
queen ending. ”3/ /
W/W/WM /
Rausis-Faragé
Germany, 1996 How is it possible to win this
position for White?
1 fibS 589.4 2 f3 @ecS 3 @f2
///
// 4/3/ @e6 4 RC3 @c7 5 $.34
The problem is that while for
w

/‘%a ;/
Black there is no active plan—for
White there is.
5...@e6 6 Rb4 @e8 7 @e2 @d6 8
£b3
//ae;/
IIIII

There appears the first threat of

/ / a/ a3-a4.
8...f5 9 M!
Now already arises a real threat to
exchange the dark-squared bishop,
138 The Isolated Pawn in the Ending
after which all Black’s weaknesses,
riveted on white squares, will be
exposed.
9...5De5 10 icS ®d7
After 10...®dc4 11 £a4 @d6 12
3xd6! @xd6 13 $38 @e7 14 f4!
9317 15 ixf7 White wins in the
pawn ending.
11 £d4 @bS?
Better in fact was 11...5Df6 con-
ceding the d5 pawn.
12 9.112 @cs 13 .932 @d6 14 g4!
hxg4
Also no help is l4...®e6 15 gxh5
gxh5, since he cannot hold the two
weaknesses on f5 and h5.
15 fxg4 @e6 16 h5! ¢3c7 l7 hxg6
Black resigned.
Szabé-Penrose
Bath, 1973
//. '=<'/,
// ”fig/é
%o /
////
//’
a?
//
W//
//
White plan consists of three parts:
(a) transfer of the king to d4;
(b) transfer of the bishop to 13;
(c) transfer of the rook to the
eighth rank and an advance of the
pawns on king’s flank.
1 :34 a6 2 §b4 §d7 3 @c3 @e7
Or 3...d4+ 4 Exd4 §c7+
(4.. .Exd4 5 @xd4 2xa2 6 £c4+!
with a winning pawn ending) 5 @b4
$.v 6 b3 fie] 7 <$a3 fibl 8 $b2,
winning.
4 QM @d8 5 9.9.2 9:7 6 $.13 b6
7 §b3
Two parts of the plan have been
completed, White goes over to the
third.
7...§d8 8 §c3+ QM 9 g4! Ed6
10 a3 35 11 M .917 12 f5! §d8 13
ficZ!
A prophylactic move, eliminating
the threat of Ed6-c6.
l3...§d7 l4 g5! hn 15 hn
fn 16 EgZ §d6 17 En g6 18
fxg6 Exg6 l9 .9.d+ Black
resigned.
Nikolié-Portisch
Ter Apel, I996
/ 7%‘4?
/ / /
//,///
{27
/£¢
1 ECZ!
% ’4?
A very important prophylactic
move—the a2 and f2 pawns are the
main target for Black's counterplay
and therefore ought to be defended
to the maximum.
1...¢3e4 2 @e5
In these kinds of positions block-
ading the isolated pawn is not so im-
portant—the main thing is to attack
it, as Bent Larsen said about such
situations!
2...§b7 3 @fl @fs 4 @e2 @e7 5
g458f66h3g6713h5
As the defender is supposed to do,
Black tries to exchange pawns.
8 g5 439.8 9 @df! @c7 10 EcS d4
Black has to sacrifice a pawn,
since 10.. .Qd6 11 §c6+ @xeS 12
f4+ leads to mate.

11 exd4 5239.6 12 E35 @3n 13
@xg6+ @f6 14 @f4 and Black with
a clear conscience can resign.
Magerramov-Kohlmeyer
Bad Worishofen, 1993
a .
923% /
, /a%
An instructive example of the
struggle of two knights against two
bishops, with rocks on the board,
this being a rare case of the com‘-
plete superiority of the knights.
l 9334
As Nimzowitsch said, weaknesses
do not tell unless they are attacked!
1...§d6 2 Eacl fies 3 Exc8+
Also 3 i3!? deserves attention
~—Magerramov.
3....@.xc8 4 @b5 §d8 5 @d4!
Worse is 5 @xb6?! £g4 6 Ed
§b8 with strong counterplay.
5...Ed6 6 13 3d7 7 @c3 h5 8 @f2
$18 9 @e2 RCS
Upon 9...?e7 10 @d3 $38 11
Ecl nothing good awaits the black
king in the centre.
10 b4! axb4 11 axb4 £a6 12 b5
Fixing a second weakness on b6.
12....@.b7 13 gal $16 14 @dB
§d7
The only move. If 15...@e7? l6
§a7 §d7 17 Exb7! Exb7 18 @xd5+
White wins.
15 @334 Rd8 16 Eel
Successfully combining threats
along the a and c-files.
The Isolated Pawn in the Ending 13 9
16...Q.c7 17 @306 f6 18 f4
With the threat of @d4, h3, g4
etc.
18...§f7 19 @d4 $86 20 113 £117
21 @c3!
Play over the whole board!
21...§d7 22 g4 hxg4 23 hxg4
QdG?!
More stubborn is 23...§h7 24
58d Ed7, in spite of 25 f5+! gf
26 gxf5+ @f 27 ®e7+ @e6 28
Ec7 Exc7 29 @xc7+ $67 30
58d+ @d6 31 58f6li, Mager-
ramov.
24 5834 £c7 25 Ehl! §g7 26
@c3 §d7 27 §h6 §g7
If 27...§.xc6 28 bxc6 §d6 29
Exg6 White also wins.
28 @3d §d7 29 f5+!
Black resigned in view of
29...gxf5 30 Exf6 mate.
A case of an isolated pawn with
bishops was given in the book
Winning Endgame Technique.
Bishop against knight is also a very
well known endgame, and the side
with the isolated pawn has to main-
tain the equilibrium.
Very interesting also is the case of
rook endings with an isolated pawn
for one of the sides.
Bareev-Faragé
Rome, 1990
Zy/ '7
g M‘
H [j//%
'///”/
140 The Isolated Pawn in the Ending
l...h5?
He should immediately take
measures on the queen's flank by
l...a5, though after 2 g4 White still
maintains the advantage.
2b4 @e63Qd4f64h4Qf55f3
@g6
If 5...g5 6 hn fn 7 EfB Qg6 8
$65 there is no defence against 9
@e6.
6 34! @fl 7 a5 @e6 8 a6!
The weakness on a7 is now
permanent.
8...@d6 9 b5 $96 10 g3 g6 11
§e8+l @d6 12 318 $96 13 ECS
@d6 14 §c6+ @e7 15 g4 EN 16
gt gt 17 Ec8
Now it will be very difficult to
hold on at once to the three wea-
knesses on a7, d5 and h5.
l7...@e6 18 §e8+ $d6
If 18...@f7, then 19 §b8 and 20
Eb7.
19 Ens fic7 20 §d8+ @e6 21
fid f5 22 e4 and Black can
resign.
 8 Rook and bishop against
rook and knight
In general the line-up of rook plus
bishop is rather stronger than rook
and knight (to a large extent this de-
pends on the structure), but a few
grandmasters, for example Evgeny
Sveshnikov, think the reverse. Rook
and knight is stronger when there
are fixed weaknesses or, for
example, when there are doubled
pawns.
We look at a few cases of the ad-
vantage of rook and knight, since
with this line-up the methods of
play are more complicated.
Sakaev-Novikov
Moscow, 1 998
// ”7/
///// / with an easily winning position.
//%/
9%/
////
/'/,...=,/ //
, ”7/ 4,/////,, 5%
/ a. — /E
1 @d2 b4 2 Qa4!
There is no sense in White open-
ing the a-file—-but here it is possible
to open the b-file.
2...bxa3 3 bxa3 §b7 4 @c3 $18
5 @c5 EbS 6 @b3!
White begins a plan associated
with an attack on the a5 pawn or the
forcing of the advance ...a5-a4.
6...@e7 7 Ea]! @d6
After 7....Qd7 8 a4 §b4 9 55a
fixa4 10 Exa4 fixa4 11 @b4 and
then 12 @c5 White wins the d5
pawn.
8 a4 §b7 9 £3a §c7+ 10 @d2
£d7 11 @b3 §c4 12 a5 RC6
No help is 12....Qa4 13 a6!
winning.
13 a6 Eb4 14 @215! £38 15 @c3
§b8 16 @b3 §e8 17 @d3 @c7 18
ales @b6 19 Eb1+ @217 20 @d2!
Preparing an exchange of rooks
—technically the simplest way to
win.
20....Qc6 21 §b3 §e7 22 §e3
Mikhailov-Volchok
corr, 1995
/////;//
///C/”
//// /’//
&/
///4 //
Z;,/”/ ”/4
142 Rook and bishop against rook and knight

Black has a majority on the king’s Milov-Pelletier

flank, but the knight has to carry out Biel, 1 997
an unpleasant defensive function
and, in addition, the pawns are ham-
pered by their white counte arts.
1 g5! Eds 2 figl @c6 3 Ed §f5
4 Ehl!
With the interesting threat, after
4. 4hx 5, of 5 ixg7! @xg7 6 h6+—.
d7 5 @d3 fibS 6 gxh6 gxh6
7 .9.t @f6 8 figs 43M 9 EM!
V
Again prophylaxisu-utilising the
rook to the utmost. a%//
9...e5 10 QM $96 11 a4 §b8 12
EeZ f6 13 £h6 EgS 14 @c5 9M8 15
@b6 Black resigned. l...g5! 2 a4 §a3 3 E216?
Better is 3 £b5 Eal+ 4 QhZ
55g6+.
Alterman-Chernin 3. .3211 4 @g2 @g6 5 fibS 58h4+
Beersheva, 1992 6 @112 5mm 7 @gZ 55h4+ 8 @112
e5! 9 £c6 f5 10 f3 §a3 ll §a7
93x13+ 12 @gZ @d4 13 3d5+ @hB

4/4949”
éV/Qfi/fl
// ///
14 35 e4 15 36 e3 16 3C4 f4! 17 h4
13+ 18 @h3 e2! White resigned.
It is very important to control the
moment of exchanging rooks or
minor pieces, which can often alter

m.4425/, the assessment of the position. At

times, with pawns on different
/ / 99 flanks, the knight can even prove to
be the stronger piece, though
\

usually this is the more long-range

bishop.
If we remove a pair of pieces
Gausel-Agdestein
from the board, then White is better.
Reykjavik, I996
With the interchanging of threats the
advantage lies with Black.
1 @gl §c3 2 fie] ®g4 3 e4 @f6
4 h3 @e5 5 Efl+ @e7 6 Eel h5!
Improving his position!
7 fieZ M 8 Eel 55g6 9 @f2 §f6
10 Ee3 gel 11 §B+ @f4 12 g3
hxg3 13 Exg3 Ehl 14 a4 Ebl! 15
§c3 @e5 16 35 Ea] l7 Ec5+ @d6
18 §b5 §a3 19 @g3 @c6 20 §b6+
@c7 21 @g4 fia 22 §b4 @d3 23
§d4 @e5 24 @hS g4! —+.
 Rook and bishop against rook and knight 143

Not only are White’s chances not
better, but Black simply has a clear
advantage.
1...Ee4!
A surprising exchange of rooks.
2 Exe4 @xe4 3 9.c2 @c3 4 g4
@d6
It was also possible at once to fix
the opponent’s queenside pawn
structure by 4.. .Qb5!? 5 34 @d6.
5 @el QCS 6 @d2 @b5 7 34 @c7
8 @c3 @d5+ 9 $b2 @b4 10 g5 b5
11 3b cb 12 .9.e4 @c3 13 .9.c6
White wants to obtain counter-
chances by doing away with the
black f7-pawn.
l3...5Dd1+ 14 $32 @f 15 .9.e8
@d3 16 9.xf7 @eS!
Now the knight dominates the
bishop, while the king transfers to
the other flank.
17 .938 c 18 @b2 b4! l9 @c2
@d5 20 h5 gt 21 9.t $94 22
@d2 g6 23 ..9.d1 @f5 24 @e2 @n
25 @f2 @h4 26 @g2 @d3 27 9.9.2
@cs 28 9d] @g5! 29 ®g3 @f6 30
@g4 @e5 31 9.c2 @xb3! 32 9.xb3
@e4 White resigned.
Beliavsky-Neverov
2 58h3! 9&8 3 QBgS $f6 4 5513
@e6 5 EbS!
For the time being activating the
rook.
5....9.d7 6 §b2 @f6 7 @es 9.34 8
§b8 @g7 9 @d2 @f6 10 @c3 @g7
11 @b2
Also ossible was 11 d4:
11... M 12 $33 9.c6 13 Qb3
@g7 14 34 .9.b7 15 35 9C8 l6 @c3
@f6 17 §b2 g5 18 EhZ! gxf4 l9
gxf4 e6 20 d4 cxd4+ 21 @xd4 @e7
22 c5! 9.36 23 §h8 9f] 24 §h7+
@d8 25 fihl 9g2 26 Egl 9.d5 27
Eg8+ @e7 28 36 @f6 29 §g6+ @e7
30 fih6!
Zugzwang!
30...9.g2 31 523g6 @e8 32 93M!
9.d5 33 @3f Black resigned.
Genba-Irzhanov
Russia, 1998
3 3: 3
3 3 3
,3 6x3
313 //7/”’/
\\\\\\\\

3 3
Koszalin, 1998
\W‘s

/ /'3
/// 63 3
33 333 6V
// // // /3
333M Sometimes chessplayers them-
3 3 / selves do not know how to exploit
3 £3 '3 the possibilities of their own pieces.
/ ///“/
363,3 1 9x36?
Driving the bishop to a passive

7/63 position. He should play 1 a4! §a8

2 b5, creating a assed pawn.
l...§38 2 b5 d7 3 E34?
l...f5? Better really was to leave the
Better was l....9.f5 2 g4 9.d7 3 g5 pawn and activate the bishop by 3
.9..f5 4 @e4 .9.xe4 5 @xe4 f5+ with 9b7.
chances of holding the rock ending. 3.3536 4 333 @f6
144 Rook and bishop against rook and knight

Now the black king enters the Kamsky-Cvitan

game with decisive effect. Palma de Mallorca, 1989
5%g2ée76éf3d4I7Qe2
<$d6 8 Qd2 @c7 9 M @b6 10 @d1
firs 11 @e2 Eds! 12 E13 d3+ 13
@d1 @a4!
Nobody wants the bishop on a6!
%
//// //
mgWy
///‘z,...' / //
14 @d2 @b2 15 a4 @c4 16 @d1
// ///
d2 17 §b3 h5 18 35+ @1135 White
resigned. is? /z/Q//
With a struggle on one flank the
advantage of knight over bishop in
conjunction with some other pieces
can be very great.
1 5235+ @gé 2 §g8+ @f6 3 @e3!
QCS 4 ®d5+ @e6 5 Ee8+ <§d6 6
Lastin-Alferenko §d8+ @c6 7 §c8+ @bS?!
Moscow, 1998 Obviously better was 7...?d6,
though even here after 8 §c7 Black,
has serious problems.

7 y a 8 Ed!
After this move everything
a? /.l becomes clear.
8...@c4 9 Exfl §a2 10 EB! @d4
//
/////
11 9316 h4 12 gxh4 £e7 l3 h5 §a6
l4 @gB! figs 15 Eg3 £14 16 §g6
///aé 7
Black resigned.

// / Gulko-Sveshnikov
//// Volgodonsk, 1983

/£ /,
1 $13 §c7 2 §b5 3M 3 936
.937 4 ®d3 .916 5 EM! e: / //////
Guaranteeing the centralisation of
his king.
4 /
/ ...
5...§d7 6 $e4 fidS 7 §b7+ @113? ////////
Slightly better is 7...?g8.
8 f4! exf4 9 @xf4 365+ 10 $13
EfS 11 e4 gas 12 @xg6+ @g8 13 /&/%/gfi/
QM! Black resigned. 722/ // I/II// (If!
 Rook and bishop against rook and knight 145
Activity of the pieces always
comes before material advantage.
1 §b7 Ev 2 Exb6 @g7 3 §b7+
@h6 4 §e7
White tries to limit the activity of
the black knight.
' 4...Eb2 5 £93 Eb] 6 9g EbZ 7
@h3 @115 8 §d3 55%!
Again trying to control the e4
square.
9 §e3 @d5 10 313 @f6 11 g4
hxg4+ 12 £xg4 g5!
The transfer to the rook ending is
lost: 12...5n4? l3 @xg4 fit 14
§h3+.
13 fn @n 14 .936 559,4!
Finally the knight achieves its ob-
jective—the pawn is not enough but
the activity of the scattered pieces is
sufficient to achieve a draw.
15 Ef5+ @g6 16 EbS @d2 17 b4
@311 18 315+ @f6 19 ads Exh2+
20 ®g4 an 21 Ecs ®e3+ 22 @g3
fidZ 23 .234 §d4 Drawn.
However you cannot always man-
age to achieve such coordination
with a rook and knight. Very often a
rook and bishop compliment each
other splendidly.
Karpov-Hort
Tilburg, 1979
/s/
W/Wz/ 1 W4
W
1W W/
W W/WWW
Z/W,
WW /%
W/
1 §b1!@e7 2 g4!
Preventing both. ..-f7 f5, and also
58.65-8--d6-f5
2...@d8 3 b4 §e7+ 4 @d3 axb4 5
Exb4 <$c7 6 Eb] @f6 7 a5 @d7 8
gal!
With the threat of 9 a6 bxa6 10
Exa6 attacking the c6 pawn.
8...!523b8 9 M!
The knight has gone over to the
opposite flank. There he can attack
on the other side.
9...®a6 10 Ebl @b8 11 2.13
@d7 12 gal §e8 13 Efl §e7 l4
QgZ @b8 15 39.4 58216 16 Ebl
@b8 17 .915 @d7 18 gal g6 19
£xd7! Exd7 20 if] @b8 21 Ef6!
@a7 22 h5!
Black’s weaknesses start to
crystallise.
22...Qa6 23 g5!
The passed pawn comes first!
23... hn 24 h6 $a 25 h7 §d8
26 E1117 b5 27 cb @b 28 §b7+
@a6 29 §g7 §h8 30 @e4 and
White has a winning ending.
Balashov—Vaganian
USSR (ch), 1972
IIIII
WW
/l//Il’{/
/W / W WW WW W
1 @c2!
White plans to reconstruct with
his king on d4 and knight on e3.
Now Black should go over to
counterplay by 1.. 3.135 2 @b4 2.36
3 @d4 f6 with a tenable position.
146 Rook and bishop against rook and knight

1....@.e6? 2 @e4! 915+ 3 @d4 18.nfic8 19 EbS fieS 20 @xc4

fie? §e4+ 21 @d5 EXM 21 3x35 <$g7
Vaganian makes the mistaken 23 b4 fihl 24 b5 h4 25 C4 113 26
decision to play a rook ending, Eaz h2 27 @c5 $17 28 fieZ Black
though even after 3. .fie6 4 @e3 resigned.
305 5 f5 .9.f 6 @xc4 §c6 7 @e3
£e6 8 c4 things are not easy for
him. Tal—Balashov
4 EXCZ ECS 5 362+ @f6 6 fieS! USSR, I974
Now it would be a mistake to
transfer to a pawn ending: 6 EeS?
ExeS 7 fxe5+ @f5 8 @1104 @g4! 9
@135 @xg3 10 @1136 @xh4 11
Qxb7 <$g3 12 c4 h4, and Black is
quicker.
6...a 7 Eez fies 8 a4 g6 9 fies
§c7

// /a /
at; ////i
/ a 1 a4!

51/11
/’//7
Depriving Black of the possibility
of ...a5-a4.
l...b6 2 @d2!
Now White improves the position
//V
4/ of the knight, whereas Black’s fol-
lowing activation assumes an unreal
character.
Now White needs to find a plan to 2...§.e5 3 13 318
play for a win, and Balashov dis- Or 3...§d4 4 @122 and 5 523e4, and
covers a splendid transfer of the if Black does not exchange rooks,
rook to b5, where the rook attacks then 6 Ehl and 7 M.
the b6 pawn and does not allow the 4 @e4 @c7 5 Ehl!
king to penetrate to the fifth rank. It is necessary to open the h-file,
10 EeS §c6 11 EbS! @126 12 <$e4 whereas the d-file has no
$f6 13 f5 g5 14 Eds gxh4 15 gxh4 significance.
@e7 16 @eS! 5....@.f4+ 6 @e2 @d7 7 M gxh4 8
Provoking ...f7-f6, so taking this EXM @e7 9 5812!
square away from the opponent’s Now the best place for the knight
king. will be d3.
16...f6+ l7 QM §c8 18 fibS Ec6 9...§d8 10 ®d3 figs 11 Eh]
16 a5! ba 17 Exb7+ @1'8 18 e52?
QdSI A vain attempt to obtain activity
All White’s pieces are active, and with a pawn sacrifice.
Black’s passive. 12 Edl @e6 13 5N2!
 Rook and bishop against rook and knight 14 7
White again wants to exchange
rocks, but Black is forced to avoid
this exchange.
13...§d4 14 @394 £14 15 $36!
White’s main objective is the b6
pawn.
15....@.g5 16 @d5 ids 17 fid3
$17 18 @c3!
It is necessary to drive the black
rook away from the important d4
square.
18....@.e7?
Better is 18...§.g5!, which was the
only chance.
19 @b5 e4 20 Ee3! §d7 21 Exe4
$.16 22 f4 £g7 23 f5 $36 24 §e6+
Black resigned.
The following ending serves as
yet another example where the
knight is stronger than a bishop
when there are rocks on the board.
Pelletier-Arencibia
Ubeda, 1998
{I
a}
7 /1%s
2/7/61 /1/
% / //
/ /
% /
/’// /
fl /&§§//
7/
In order to maintain the balance
Black must activate his rooks along
the d-file and then subsequently at-
tack White’s king’s flank.
l h3 .935 2 @e2 fiab8
Already Black does not succeed
in organising counterplay along the
d-file because of the exchange of a
pair of rocks, which is in White’s
favour.
3 b3 @g7 4 §a6 Ebcs?!
He should defend by 4...§b7 5
@c5 Ec7, though after 6 b4 (but not
6 Exe6? in view of 6....§.b2) and
Black has no counterplay.
5 EcS!
The best way to exchange after
which the white knight appears on
c5.
5...§xc5 6 @3c £b8 7 b4
Also here an exchange on e6 was
not possible because of 7...§c8.
7...?f6 8 335!
The rock not only cuts off the
black king along the fifth rank, but
also improves its own position: its
objective—the b7 square.
8..Eds
Hoping for 9 EbS $d6 10 @e4
@e5 11 @xd6 @xd6 12 §b7 a5
with probable equality.
9 a4
Now already the above-mentioned
variation is threatened, because at
the end White has b4-b5.
9...§e7
On the more tenacious 9...?g7
White intended to play 10 @b3!
Ed7 (the exchange of heavy pieces
leads to a quick win for White:
10...§xa5 11 @xa5 a6 12 @c6 .937
13 b5 ab 14 a5! etc) 11 @d4 $d6
12 @c6 Ec7 13 §a6 with the future
march of the king to a4 and b4-b5
and the win of the a7 pawn. On the
other side of the board Black does
not get anything
10 EbS £d6 11 Eb7+ @e8 12
@e4?!
Probably, sufficient for the win is
12 @a6 §g5 l3 g4 h5 14 gt
Exh5 15 Exa7, but simplest was 12
@d3 a5 13 b5.
12....§.e7 13 3x217 fies
Or l3...f5 14 5803.
148 Rook and bishop against rook and knight
14 §a8+ <$d7 15 §a7+ @eS
If the Black king stepped to d8,
then 16 53:2 fixb4 17 Exfi would
decide.
16 @d3 Rxb4?
It would be better to play the rock
ending after 16...f5 17 @305 $.c
18 bc EXCS 19 Exh7 g5 (l9...§a5
20 EM), though 20 Ea7 was suffi-
cient for victory because of the very
bad black king.
17 5Bf6+ @d8 18 §d7+ @cs 19
EH7 EfS
l9...h5 20 @d7!? EdS 21 $04
.935 22 5305 also loses.
20 g4! :13 21 @e2 fixh3 22
§Xh7 Exh7
After 22...§g3 23 @f2 the black
rook is in a ‘hole’.
23 @xm @d7 24 @d3 ea 25
@gs @d6 26 @313 and Black
resigned.
Adams-Almasi
Dortmund, 1998
1 @d5! @d 2 ed
White is obliged to transfer to an
ending where for him the knight 06
controls all the important squares
around itself, while even the dark-
squared bishop, controlling the long
diagonal will feel redundant.
2.. .g6 3 fiabl £g7 4 c4 3.33 5
fiecl $16 6 @fl §e4 7 g3 figs 8
ECZ Eaes 9 @bB!
For the present beginning an at-
tack on the black pawns.
9...a5 10 @c6 <$Pg7 ll 58a £38
12 fibS .216 13 §a2 RC3 14 @c6
Exc4 15 a5
The a-pawn, supported by the
knight c6, can only be stopped at the
cost of material loss.
15...§xa6 16 ECZ ficxc6 17 dxc6
fia 18 §b7 Eal+ l9 n $.35
20 E217!
And Black resigned, since there
is no defence against the manoeuvre
ECZ-bZ-bl
The knight is particularly strong
in attacking various weaknesses,
whereas a bishop is incapable of de-
fending light and dark squares.
Van Laatum-Mikhalchishin
Eeklo, 1991
//W /
74% %1//
/ // / .t
/
/// ,/
% //§//
l...¢3f5
There was another, and perhaps
more unpleasant plan for White:
1...f5!? followed by 2...?f6, 3...f4
and 4...?f5.
2 §g2
After 2 fidS En 3 £e4 §g4 4
.9.f gxf5 it is not easy to defend
the endgame.
2...£De3
Interesting was 2...¢3d4 with the
same plan ...f7-f5-f4.
3 332 f6!?
 Rook and bis/70p against rook and knight 1 4 9

A new idea: ...g6-g5-g4 and 7 31:8 @d6 8 3d7 §d3 9 £36

...Qe3-fl. @e4 10 fig g5
4 £36 f5 There was also another plan:
After 4...g5 White brings the 10...§f3 and 11...5Df2.
bishop into the defence by 5 hxg5 ll hxg5+ $n 12 317 @g4 13
hxg5 6 ih3. 39,6 §e3
5 d4 @f6 6 d5? Avoiding the trap 13...@xg3? l4
This pawn should be ‘discarded’ £xf5+!.
by playing 6 QCS and 7 £b7, trans- l4 £c8 §c3 15 .9436 Ed3 l6
ferring the bishop to the long Q17?
diagonal. It was necessary to give up the d5
6...®c4! pawn, if only by the move 16 $c8
Now the knight shows itself in all followed by 17 .336.
its splendour. 16...¢3g5! White resigned.
 9 Rook and two pawns against
rook and knight (many pawns)
It would seem that having more
pieces should confer an automatic
advantage, but the knight has such
limited mobility in a battle with
pawns that quite frequently nothing
can be done with it.
Mikhalchishin-Sveshnikov
USSR (ch), 1978
, / f// /
/z/
4/%,%/
//
////§//
%/7
l §d2
The decision to exchange rooks
was not easy. Other possibilities
were 1 @d2 or the roundabout way
1 EfZ! with the idea of 2 Ef8.
l...@b5 2 Ed4?
After 2 Ed cxd5 3 @d2 a3 4
@d4 <$a4 5 bxa3 bxa3 6 @b1! a2 7
®c3+ @b3 8 @3v the pawn ending
IS won.
2...a3!
Bad is 2...Exd4? 3 $xd4 a3 4
Qd6+I (m4 5 bxa3 bxa3 6 $03
with a win.
3 bxa3 Exd4 4 @xd4 bxa3 5 @c3
a2! 6 @b2 QM
After sacrificing a pawn, the black
king rushes over to the opponent’s
pawns, it seems without success“
7 ®d6+ @d5 8 @f5 @e4 9 @xg7
@f4 10 h3 @g3 11 Qf5+ @xh3 12
@xh6 @h4!
But this is the point—and typical
for the struggle of the knight against
glawns—there is no defence against
g5, therefore a draw.
Smagin-Mikhalchishin
Moscow, 1989
%/ / %g/s
%4 /4/////
%1/ /////
/ /
V /% g
%fi/&é
l...¢3b4?!
Clearly Black wants to attack the
pawns, but stronger seems 1...h6,
though it is not usually recom-
mended to play on the side where
the opponent is stronger.
2 a3 @d3 3 g5 fieG 4 Edz @cS
After 4. fie3+ 5 @g4 58e5+6
@f4 Ed3 7 Exd3 @xd3+ 8 @e4
 Rook and two pawns against rook and knight I51

@xb2 9 @d4 the white king Or 6 @d4 gxh3+ 7 @xh3 @f4.

threatens to break through to the 6...§xb3 7 fid @xg4 8 @xh6+
pawns on the queen’s flank. QM Drawn.
5 ®g4 @xb3 6 §d7+ @gs
After 6. .Ee7 7 Ed6 there is
equality on the board. Korchnoi-Lj uboj evié
7 h5, and a draw was agreed, Bath, 1973
since the threats EdS, h5-h6 give
White sufficient play.

Mikhalchishin-Savchenko g
Postojna, 1991
/,%/%g2,, ,
////

/ 2g
Black has two plans of play-the
best one was the plan to advance the
passed pawn. 1...§dl!? 2 5/3c5 (2
§a2 d4! 3 Exa6 §d2+ =) 2...d4 3
@xa6?! d3 4 §f6 h5 5 Ed6 §d2+
with distinct counterplay. But Black
decides to exchange pawns on the
queen’s flank, which is clearly
weaker.
1...a5?! 2 @g7?!
//%$a
/§g//%//
gg//7
////
Lufibl?
After the correct l...@f‘8 2 QB
Eb] Black is in no way worse,
whereas now problems start to ap-
pear with the f7 pawn.
2 @d3 @g8 3 §b7 @g7 4 @e5
@f6 5 @3n Eel 6 @xh6 §c2+ 7
$13 b2 8 @e4! 382+ 9 @d4 §h2
10 g5+ @e6 11 §b6+ @e7 l2 @c3
$18 13 fib Exh4 l4 §b7 §f4 15
@d3 Black resigned.
Hertneck-Narciso
Berlin, 1998

g/
But White decides to play for
mating threats. whereas after the
correct 2 §d2l Black would en-
counter serious difficulties.
Zufibl 3 h3+
If now 3 §d2, then after 3...?f4 4
533e6+ @e3 5 Ed §b2+ 6 @g3
.}///%/@g
///////////
//// fig
Exb3 7 fia @e4+ 8 $g4 §b2 9 /%

4/£3%
,-/

h3 fig2+ 10 @hS §g3 Black elimin- ,,.,/:/23

ates White’s last pawn. /%
3...?h4 4 EdZ g4! 5 @f5+ @gS 6
hxg4
152 Rook and two pawns against rook and knight

Here Black is three clear pawns 14 5mm g3 15 Qe3 g2 16 $12

ahead, but after 1 @e5! with the Qe4 l7 n2 Qd3 Drawn.
idea of Qf4-e4 and then 202 seri-
ous problems arise for Black, since
he cannot create 3 passed pawn. But Romanishin-Markowski
in the game there followed... Polanz'ca Zdroj, 1993
1 Qf4? Q17! 2 QeS Qg6 3
fixd4?!
A primitive decision, more inter-
esting is 3 Qe6!? or 3 ®h2!?. W
3.. .Exd4 4 Qxd4 QgS 5 QeS
Qh4 6 5312 Qg3 7 @e4+ Qxh3 8
W W23W

\&
58g5+ Q 4 9 @xh7 Q13 10 Qd4
Qf4 11 f8 g5 12 ®e6+ QB 13 //,W,EW
, , ,%
/WW/%/
fi/W/

\\\\
5&8
WWW/W
W///
1/ /W 1 e5! dxe5?
MW
W Also not winning is 1...§d4 2
WW exd6 Ed 3 h4! gxh4 4 §e4 §d4 5
d7! with a draw. But necessary was
1...§e4! 2 exd6 Exe6 3 dxe6 @xd6
W8/ /W/ 4 Qg2 @f5! 5 QB 58h6! (against
h2-h4, so as to have g5-g4) 6 Qe4
Qf6 7 QdS @f5 8 h3 ®g7 with an
W/ W WW/ easy win.
2 d6 52)“ 5 ExeS Qf6 4 d7 @c6
l3...g4! 4 fies §d4 5 EcS!
By sacrificing yet another pawn, This is the idea—to take out the
Black diverts the opponent’s king. knight—now it’s a draw.
 10 Two minor pieces against a rook
Endings with an uneven align-
ment of forces are some of the most
complicated in chess—and also
comparatively little investigated.
Among them can be included end-
ings where a rook struggles against
two minor pieces.
In the middlegame, particularly in
positions bearing a closed character,
two minor pieces will be stronger
than a rook even where there is a
deficit in pawns. However, in end-
ings where the rook has room for
action, the situation often depends
on the arrangement of the pawns.
Thus, when there is a distant passed
pawn on the board, the rock can
prove to be stronger than two
pieces. However with a balanced
pawn formation the situation may
be reversed.
When playing such endings, it is
very important for the side having
the minor pieces to put right their
coordination, and then the field of
activity of the minor pieces may be
sufficient to struggle successfully
against a rook.
According to the generally held
opinion, in standard situations with
two pieces one can count on having
an advantage. But rooks are not
without their own chances too, par-
ticularly when there is a passed
pawn on the board. However, first
and foremost, as we go over to a
study of positions of this type, we
will turn our attention to the possibi-
lities of the defending side. In terms
of defensive resources the defender
should strive for an exchange of
pawns, bearing in mind the possibil-
ity of transferring to a theoretically
drawn ending, exploiting the stan-
dard methods of: (a) from the rook’s
side—sacrificing the exchange; (b)
from the minor pieces’ side—sa-
crificing a piece.
Let’s look at some possibilities in
positions where the rook is playing
against passed pawns. The rook’s
Specific characteristics depend on
lengthening lines of communication.
The broader they are, the more
chances the long-range rook has in
its struggle against the less mobile
minor pieces, and it is even possible
to think about playing for a win.
You see, whereas the minor pieces
are riveted to the blockade of a
passed pawn, the rook can be uti-
lised also on another flank to organ-
ise new objects of attack.
To counteract the rook, one of the
postulates of play for the side hav-
ing the minor pieces is to put right
their interaction while controlling
the opponent’s passed pawn. But the
various plans of activity of the
forces on the board produce play of
a complicated character, and the
exploitation of one’s resources is
not always successful even where
highly rated chessplayers are
concerned.
Naturally, the outcome of the
struggle in great part depends on the
arrangement of the pawns: the more
154 Two minor pieces against a rook
compact they are, the shorter the
communications of the rock, and
naturally the narrower their
possibilities.
Beliavsky-Miles
Thessaloniki (01), 1984
/ c3 7 @e3
///// / // 7%
777/ 7/
77/7
771/ /
7% W / 7
"7a7 7/
Despite the far advanced passed
pawn, supported by all Black’s
forces, the chances of White’s
pieces are higher. They firmly
blockade the pawn, while the quite
narrow range of activity between
the flanks allows White to fight for
the creation of a passed pawn of his
own.
1 2&2 Efl?!
If 1...g5? 2 53d1+ @a3 3 $03 the
passive rook is unable to prevent the
win of the c4 pawn: 3...§cl 4 @e3
and 5 @xc4.
Therefore Black activates his
rook, attacking a pawn. But this is
inaccurate: on the f-file the rook is
in the sphere of activity of the king.
Therefore stronger is Infigl, de-
fending the g6 pawn and riveting
the white pieces to the passed c-
pawn. Here, after 2 @dH c$a3 no
good is 3 $03 Efl or 3 @e3 Egz 4
£34 @b3. But also on the active 3
f4 Eg3! the rook succeeds in stand-
ing up to the pieces. For example: 4
52383 EB 5 @d5 03+ 6 @e2 §h3 7
@e3 (7 £xg6? Eh2+) 7.. .Eh2+ 8
$d3 (after 8 913 Black forces a
draw: 8...§xc2! 9 55c @b2)
8...?b2, and Black prevents White
from putting right the coordination
of his pieces.
2 523d1+ @213 3 @e2 Eh]
There is no saving the pawn:
3. .Egl 4 @e3 c3 5 $12 EgS 6 f4.
4 axga 3112+ 5 @es QM 6 am
White has won a pawn without
losing control of the opponent’s
passed pawn.
7...@b3 8 £d3 @b2 9 @e4
It is useful to pay attention to the
interaction of the white pieces——
now the time for White’s passed
pawn has come.
9...@c1 10 f4 §h8 ll fibS §b8
After 11...02 12 .934 Black loses
a pawn and cannot derive any bene-
fit from the temporarily uncoordi-
nated white pieces: 12...@d2 13
58x02 §a8 14 £36! or 12...§a8 13
2b3! §a3 14 $.c @d2? 15 $304+.
12 Rd3 @dZ
Or 12...Ee8+ 13 @d4 @d2 14
55c4+.
13 55c4+ @dl 14 $13 EhS 15
£34 @el 16 @333 @d1 17 f5
And White, maintaining the
blockade of the opponent’s passed
pawn, realised his own one.
Morovic-Yusupov
Tunis, I 985
7/
§§DN
/// 7/ /E7%
////
//
7// 77/
177/.
\
////
77
\\\\E
727/77
\
\\\\
‘
//7/7
 Two minor pieces against a rook I 55

In this example the sphere of ac- the king penetrates to the weakened
tivity of the rook is quite broad, and queen’s flank.
this means the more possibilities it l0 ®fl+ Qd5 11 b4 g3 12 hxg3
has. In fact, despite material equal- hxg3 13 @d3
ity, the advantage is on the side of Also after 3 ®g4 Qe4 14 @e3
the rook. It is necessary only to axb4 15 £xb4 §h6 the passed pawn
create a distant passed pawn, while must decide the game.
not forgetting to prevent it being 13 axb4
blockaded by the minor pieces. This It was also ossible to win the ex-
is achieved by the advance of the chan e: 13.. e4 14 bxa5 g2 15 a6
pawn pair after the preparatory gl= 16 .3.n §c2+ 17 l
1...§f5. Qxd3 18 a7 Ev 19 c b4 20
But in the game there followed... l b3 21 £f2 Qc4 22 £e3 Qc3
1...g4?, (A.Yusupov), but Black prefers to
Breaking the coordination of the retain the passed g-pawn.
pawn pair, which allows White to 14 9.9.3 §c2+ 15 l §c3 l6
blockade it with the move 2 fiell, aez m4 17 tacs+ Qfs 18 @d3
without upsetting the coordination Qe4 19 @c5+ QdS 20 @d3 §c2+
of his own pieces. However he is 21 l EhZ!
deflected from a direct attack on the Now the g-pawn cannot be
d4 pawn. stopped.
2 3M?! Qe6! 22 fig] Ehl 23 @f4+ Q94 24
No expense spared for creating 3 @e2 Qf3 White resigned.
passed pawn.
3 586+? In the context of defensive possi-
It was still not too late to return to bilities and technical realisation of a
the blockadin plan—3 £f4 h5 4 distant passed pawn, we present a
3g3! QfS 5 3M. The bishop con- classical example from the praxis of
trols the el-h4 diagonal, while the Yuri Razuvaev.
king has nowhere to go along the
light squares, since on any ...Qf3 Razuvaev-Kirov
there is @eH. Moreover, also poss- Sofia, I 981
ible is the energetic 3 .Q.f2!?, taking
aim at the d4 pawn.
3...Qf5 4 @d3 §c7 5 Qd2 §c6 6
3g7? 7///
//////
/”/s
Now already the creation of the 7/ 7/e7/
passed pawn cannot be prevented,
whereas 6 9f4! Qe4 7 Qg3 Qf3 8 /// ///
523e1+ would have set up a solid
”//,
/////
blockade.
6...h5! 7 £xd4 Qe4 8 24:5 ///////
8...§d6 was threatened. ”/
8...h4 9 Qe2 aS
More energetic was 9...g3 10
hxg3 hxg3 11 b4 :08, imposing an With such a specific arrangement
original zugzwang on White: 12 a3 of pawns on the king’s flank, the
§c6 13 @e1 Ef6 14 Qd2 Qd5, and bishop itself can resist the rook
1 5 6 Two minor pieces against a rook
when there are no pawns on the
queen’s flank. It goes without say-
ing that here the king of the weaker
side should be close to the pawns.
Therefore to achieve a draw it is
enough for Black to sacrifice the
knight for the passed a-pawn. He
should also avoid any unnecessary
weakening of the position. White
should take this into account and or-
ganise offensive action. Black has
succeeded in constructing a defence:
his minor pieces are well coordi-
nated, preventing the advance of the
passed pawn and preparing a break
on the king’s flank. In addition the
bishop does not allow in the rook
since on §b7 he can cover up by
Rel Here, however, Black need-
lessly weakens his king’s flank with
the move
1...h6?
...and creates objects for attack,
thus increasing the attacking possi-
bilities of the rook. Now, as well as
the h6 pawn, it is possible to add the
invasion point f5, which can be
created by undermining the g6 pawn
with h2-h4-h5. After 1...@d6 2 h3
(more accurate than 2 @g2, on
which Black quickly organises an
attack on the passed pawn by
2.. @e5 3 h3 58d7 and @c5) Black
has two possible plans of defence:
(A) Active—2.93% (preparing
the attack a4 after transferring the
knight to CS) 3 @g2 @06 4 EeS
@b7 5 EeS $05 6 §h8 @b4 7 Exh7
@c5 8 h4 $xa4 9 h5 gt 10 gt
QgS, and Black, after giving up his
bishop for the h-pawn, reaches a
theoretically drawn ending ‘rook
against knight’.
(B) Passive—2.. .@e6 3 $g2 @d6
4 @g3 (White presently controls the
5th rank in order to prevent the
move. ..-h7-h5 exchan ing a pair of
pawns) 4.. .Qe6 5 h4 £d6 6 h5 @e6
7 §b7 £37 8 @f4 Qd6 9 @e4 @e6.
Now the king cannot break through
to the 04 square without loss: 10
@d3 5De5+ 11 @c3 523xg4 12 hxg4
hxg4 13 @c4 (or 13 a5 939.5 14 a6
@c6) l3...5De5+ 14 @b5 £d8.
Upon the attempt to proceed to c4
via the b3 square it is necessary to
reckon with the threat of ...@a5.
The rook’s intended penetration to
the 8th rank by 10 §c7 @d6 11
fic8? £d8 ends in its capture.
Also nothing is offered by 10 EbS
£d8.
2 @g2 @f6?.
As before, Black senses no danger
and breaks the coordination of his
pieces since now the position of the
bishop is limited by the blockading
function of the knight. Better was
2...?d6 3 §b7 £c7 with the idea
...5Dc6-a5.
3 §c5 9335 4 @f3 $e6 5 @e4
£b6 6 §c8 916 7 M @e6
///.//
, , , , /e214
7/2/
%////a
/.,//////
\
22%
\
’//// /
8h5
It is important to fix the weak
pawn. Too hasty is 8 §h8 @b7 9
fixh6 5305 10 $8 @fl 11 h5 gt
12 Exb6 hxg4+ 13 c$xg4 @xa4
with a drawn ending. Now, how-
ever, in the event of 8...gxh5, 9 §h8
is unpleasant.
8...?” 9 @d5 xh5 10 gt @f6
11 EbS .933 12 e4! .Q.d2 13 am
391 14 §e2 QM 15 Eg2 @fi

The weakness of the h6 pawn is
self-evident. Combining an attack
on it and the knight, White breaks
the coordination of the Opponent’s
pieces. On 15....Q.i‘8 follows 16 §g6
@f7 17 §a6, and the a-pawn starts
to move.
16 §g6 fidZ 17 Eda RgS 18
EdSI
An ideal square for the rook.
18...!93c6 19 aS §e6 20 36 £33
Trying to blockade the a-pawn on
the final frontier, but the mechanism
of breaking the coordination of the
pieces again comes into operation.
21 §b5 igl 22 Ebl £12 23 §b2
RCS 24 §b7 @217 25 EM
And White wins the h-pawn and
the game. A very fine ending!
And here is an example where
pieces blockade the passed pawn
from more distant ‘remote-control’
squares.
Karpov-Kasparov
World Championship (m), 1985
//%
////”
\
//%//
//// M/
§
\\\\
A///////////
\\\\‘D>§
M i
\\\\\\
{’4
i§W
fl///’:///¢//,’
//
7/
Here the bishop will not allow the
rook to get behind his lines via the
queen’s flank (due to the covering
£b4 and RC3), whereas the squares
of invasion along the central files
are covered by the king. There
followed:
1...§b8 2 £b4 EdB 3 @e2 33?!
Two minor pieces against a rook 15 7
While no additional objects of at-
tack have been created on the king’s
flank, the advance of the pawns to
vulnerable squares is premature.
More logical was the exchange of
the f-pawn after 3...g6, which forces
the white king to remain far away
from his pieces and increases the
scope of the rook.
4 .26 f6
Also here, stronger is 4... g6.
5 RM
The pawn structure on the king’s
flank is stabilised, and Black intends
to attack White’s pawn weaknesses.
We look at Black’s possibilities on
the basis of comments made by
G.Kasparov in his 1987 book Two
Matches.
5...?1’7?!
Here the king does nothing, since
there are no prospects for it in the
centre. Much more accurate was
5...@h7. In this case the method of
defence applied by Karpov in the
game would place White under
threat of defeat. For example: 6 2.03
§b8 7 £b4? EbS! 8 g4 fibS. A zug-
zwang position has been reached
and White is forced to allow the
rook into his camp: 9 @d3 §d8+ 10
@c2 hxg4 11 hxg4 §d4 12 Rxa3
Ea4 13 @b3 fixg4 14 c §g3+ 15
@c3 if?) 16 @c2 Ef 17 @d3. A
simple evaluation of such a position
is not possible, but in practice Black
would have excellent chances. But
I after 7 58b4! White holds on:
7...§b5 8 g4 §b8 9 @d3! §a8 10
@a2 §a4 ll £b4 @h6 12 3d2, and
the fortress cannot be breached.
6 @c3 §b8 7 @332 EbS 8 g4 EbS
At first sight there is no basic dif-
ference from the above-mentioned
zugzwang position, but...
9 @d3 §d8+ l0 QM fidl 11
21:33 Eal 12 @b3 Eh]
Not 12...h4? 13 5801.
158 Two minor pieces against a rook
13 gt Exh3+ 14 @c3
This is where the position of the
black king tells: after 14...Exh5
@e4 the f5 pawn is inviolable.
l4...Ef3
But now he cannot, without loss,
cope with the h-pawn.
15 Eel! Ef 16 h6 g6
Right up to the draw there re-
mains the sacrifice of a piece for
two pawns. On 16.. .g5 this is
achieved by 17 @e4 @g6 18 h7 EB
(18.. .@xh7 l9 @xf6+) 19 @c4 Eh3
20 @xf6.
17 @e4 Eh5 18 EbZ
And the f6 pawn is lost, since on
18...f5 follows 19 h7. Drawn.
If, together with the rook, there
are no passed pawns then roles are
reversed. It remains only for the
rook to counteract the opponent’s
attempts to create his own passed
pawn. But none the less the defen-
der’s resources are quite substantial.
Short-Kasparov
Belgrade, 1989
// 7
///////§%//
/'/// %
/: /
,/////
%/////
In this position Black’s chances
can only be linked to the creation of
3 passed g-pawn. Therefore he en-
deavours to win the h2 pawn. How-
ever to do this it is necessary to
break through the line of demarca-
tion—the 3rd rank—which the rook
allows only ‘in exchange’ for a
counterattack behind enemy lines.
All this is not easy. Let us look at
the possibilities of the two sides.
based on G.Kasparov’s analysis.
1 Eg3 58d5+ 2 @fZ QfS
The king moves closer to the h2
pawn.
3 Ef3+ @g4 4 Eg3+ QM 5 $13
White strives to activate his king,
since he cannot hold on by the pass-
ive 5 Eb3 Qf4 with the threat
...®d3+. Now 5...5Dxb4 1S not poss-
ible because of 6 Eg4+.
5...@f6?!
Stronger is 5....Q.d7!, keeping
under threat b6, while on 6 @e4 the
knight manages to attack the h2
pawn: 6...@f6+! 7 @e5 55g4+ 8
@d6 QbS 9 EgZ @e3, and the black
king draws closer to the h2 pawn.
6 @gz @hS 7 Ee3 @f4+ 8 @g1
@g4
Now, with the white king driven
back, Black puts right the interac-
tion of his king and pieces.
9 Eg3+ $5 10 E13 g4
Not letting out the king, since in
the event of 10...?64 11 @f2 @d3+
12 @g3 @xb4 13 Ef8 White again
activates his forces.
11 Ee3 @dS!
Again insufficient is 11...¢3d3 be-
cause of the activating of the
rook—12 E67! (12 @gZ? @f4+ 13
Ee7 3c6 in Black’s faVour)
12...5Dxb2 13 Ef7! @g6 14 Ef8
@d3 15 @gZ @xb4 16 Ef4.
12 Eb3 @f4 13 @fZ
Only the b3 and g3 squares re-
main accessible to the rook, but now
he should encircle the king. How-
ever, on 13 @gZ? @e3 14 @f2
58d1+ 15 Qel £a4 16 E33 5/3b
17 E212 QBd3+ 18 @d2 EbS White
loses a awn without compensation.
13... f6 14 E33 ®e4+ 15 @gZ
.932
 Two minor pieces against a rook 159

The manoeuvre 15...¢3d2 16 303 @c2 '93t 33 Et g3 and Black

@c4 17 Ecl! '93b 18 §c8 allows won.
the rook to escape to freedom.
16 2113 211+ 17 @gl 21:4 18 One more example from the
333 @g5 19 §c3 QdS 20 333 classics.
$34?!
The a6 pawn is indirectly de- Réti-Bogolj ubow
fended: 2] 3x367? @h3'l‘ 22 @fl Bad Kissingen, 1 928
Rd3+, but this move, losing time,
also allows the rook to escape to
freedom. Stronger was 20.. .5Df3+ 21
®g2 @e5 22 @117. ic4. /1 / fit
21 @123 9&7 t/ //// /
/;’// // géy/ /

47/
/,

/%
/% %w
//
/,/' /%
//
7/
/ 9

W / /%W
%/
/ passed pawn quickly.
22 fid3?!
White misses his chance to acti-
vate the rook by 22 EaSL threaten-
ing to exchange the b-pawn with
tempo—22...5/3e4+ 23 @gl @d6 24
EcS, and, it seems, maintaining the
equilibrium.
22...®h3+ 23 @e1?
After moving the king further
away from the h2 pawn the game
already cannot be saved since now
Black cuts it off from this pawn.
Necessary was 23 @fl 9.06 24 303.
23....@.c6! 24 §d4+ $13 25 §d3+
@e4 26 §g3 QM 27 E213
More tenacious is 27 £03 ibS 28
308, though after 28...?f3 29 §f8+
@g2 30 figS @312! White also loses
the h2 pawn.
27....§.b5 28 Ec3 @g5 29 §c8
@e3! 30 fins @313 31 @d1 £e2+ 32
1...b6
For the side having the rook it is
very important to create a distant
2 axb6 Exb6 3 @e2
A loss of tempo. He should quick-
ly put right the coordination of his
pieces, combining an attack on the
85 pawn with detention of the a-
pawn. For example: 3 @c4 fibS
(3.. .fib4 4 QxeS fixe4 5 @c6 in-
tending @a5 and RdZ) 4 £d2 §c5
gBQJaS $17 6 @e2 Qe6 7 @d3 and 8
c4.
3...§b4 4 913 @f7 5 9J4 fibl 6
93m $96 7 Rg3
This threat is a sham, since it has
a tactical refutation.
It is useful to stop the pawn as
soon as possible. For this purpose
the blockading move 7 £d8 is
appropriate.
7..ficl! 8 @a5
8 @xeS is no good in view of
8...Ec3+ 9 @f4 g5+ 10 @n
Exg3+.
160 Two minor pieces against a rook
8...Ea1 9 @c4 334 10 @e3 a5 11
Eel §a3 12 @e2 a4 l3 @c2 §b3
14 RM
At last White manages to put right
the coordination of his pieces and
blockade the passed pawn. Trying to
obtain additional objects of attack,
Black organises pawn pressure on
the king’s flank, supported by the
active rook.
14...h5 15 3.18 g5 16 h3 ECB 17
@d2 §b3 18 @e2 Ebz 19 @d2 Ebl
20 @e2 Eel 21 @d2 fifl 22 Res
§f4 23 @d3 EN 24 £e3 §d7+ 25
@e2 Eg7
2% ///
2 /
2?:
2 22e2 2
11111111
/ / 2 2
;2 2g/ 2
\
\\.\\\\\
27 2 %//. £3
§
@a2e2gé
26 g4?
This superficially active move,
pursuing the objective of blockade,
turns out to be the decisive mistake.
White creates in his position a weak
h3 pawn, which the rook can attack,
exploiting the fact that White’s
minor pieces are tied down to the
threatened advance of the a-pawn.
After 26 @a3 g4 27 hxg4 Exg4 28
QB §g8 29 2.05 White defends
himself successfully.
26...h4! 27 Rel?
The only move was 27 QdZ! so as
to meet 27...§c7 with 28 @d3, and
27.. fib7 with 28 RM.
27.. £07 28 @d3 §b7 29 $.n
§b3+ 30 $04 Exh3
And so the main idea of attacking
with the rook is seen in a clear light:
Black has two distant passed pawns
and the opponent’s pieces are in no
state to contain them without ma-
terial loss.
31 @b4 a3! 32 @xa3 Exa3 33
ixh4 §e3 34 QM Exe4+ and soon
White resigned.
Let’s look at yet another example,
illustrating the strength of the rook
in conjunction with an outside
passed pawn.
Miles—Van der Sterren,
Wijk—aan-Zee, 1984
IIIII
//j////
2-/ 2
\\
, , , ,,,, 2 2//
2/z
\\\\s\
2//'2”2
/%2 // %/&/ 2/2
lllll
Material advantage is on White’s
side. He has a passed pawn, har-
monious deployment of pieces and
can reckon on playing for a win.
However the strength of the passed
pawn in conjunction with an active
rook is such that Black’s counter-
resources prove sufficient to obtain
full equality. For a start he ties the
minor pieces down with the threat to
advance the a-pawn.
1...33 2 @d3 gas 3 .232 @g7 4
@c3 @f6 5 @313 EbS
The rook transfers to an active
position. Weaker is 5...g5?!—6 e5+
@f5 7 @d4+ §g6 8 e6 @f6 9 @b4
§a8 10 @c6, with advantage to
White.
6 e5+ @e7 7 @d4 §b2 8 QdS
EfZZ 9 58d EeZ 10 @b4

White hopes to realise his material
advantage and therefore rejects the
draw after 10 @xa3 Exes.
10...@d7 11 QM g5
We are acquainted, from the
previous examples, with the method
of exploiting a ‘quality advantage’
(in the words of A.Nimzowitsch).
12 h3 h5 13 g4?!
As shown by the endings previ-
ously looked at—this is a risky
weakening of the position. However
White plays for the win and does
not want to reconcile himself to the
‘drawing zugzwang’: 13 58a2 Ed2+
14 @e4 Ee2+ 15 @f5 Ed2 etc.
l3...h4 14 3&4
AAAA/
//////A,/A
/A /
;,/W %a%
7//
/%// /
//%%/%/
14...Ee1!
Black should not let the king go
on to c5, since his own king cannot
hold back the onslaught of the white
pieces. For example: 14...Eb2 15
QCS Eh2 l6 e6+ @eS (l6...@e7 17
5306+ @e8 18 Ed3) l7 523d5 Ec2 18
@d6 Exc4 l9 ®f6+, and White
wins; or 14...Eh2 15 @c5 Exh3 16
e6 Ee8 (16...@e7 l7 @d5+) 17 @d5
with the threat of 18 Eb5+ and a
mating attack.
15 .933?!
Continuing to play for the win,
White suddenly breaks the coor-
dination of his pieces, and the initi-
ative passes to Black. Could White
have improved his position?
Two minor pieces against a rook 161
Analysis shows that in fact he al-
ready has to worry about maintain-
ing equality. For example, in
Black’s favour is 15 @d5 Ee3!?
(sufficient for a draw is 15...Ec1,
but Black is striving for more) 16
Eb5+ @e7 17 @c6+ @f8! 18 Ec4
E62 19 @b4 Eb2 20 5532? [20
9C5! A 20...Eh2? 2.0 .QCW) 21
@d6 Exh3 22 @d5 h] 23 e6 Eel
24 e7 Exe7 25 @xe7 h3 26 @f5+-—]
20.. Eh2 etc or 15 e6 @e7! (but not
15.. .Qd6?—l6 5802 Eh] l7 @xa3
Exh3 18 l93b5+ @e7 19 @a7 and 20
5Dc8+, and White has the advan-
tage) 16 @c2 Eh] l7 @xa3 Exh3
18 @b5 Eg3 etc.
It seems he should reconcile him-
self to a draw by playing 15 @cZ!
Ehl 16 @xa3 Exh3 17 @302! Eg3
18 @e3=.
15...Ed1 16 QM EdZ 17 .234
EbZ 18 @312 E112 19 $15 Exh3 20
$n Ehl
An inaccuracy. Far stronger is
20...Eh2 with the threat of
21...Exa2 and 22...h3. White cannot
at the same time hold the two distant
passed pawns (there you are, don’t
you recall the move 13 g4?!). For
example: 2] @f5 Ef2+ 22 @g6
Exa2 23 .9.v @e7, and one of the
pawns queens; or 2] @f4 Ef2+ 22
@e3 Ec2 23 EdS h3 24 @b4 Eb2
25 523d3 a2 (or) 25...h2), with a de-
cisive advantage.
21 @f6 h3 22 @c3 Eel 23 .236
@d8 24 Eds h2 25 g5 Efl+ 26
@e6 Ecl 27 @332 hl=y 28 .9.t
Exhl 29 g6 Egl?
Having achieved a winning posi-
tion, Black makes a mistake——and
victory escapes him. After 29...@e8!
White’s passed pawns are neutra-
lised. For example: 30 @f6 Ef1+ 31
@e6 @f8 32 @d6 EfZ 33 e6 Ed2+!
etc. or 30 g7 Eh6+ 31 @d5 @fi 32
@c4 Eh3, with a winning ending.
162 Two minor pieces against a rook
30 Qf7 En+ 31 Qe6 §g1 32 $17
Efl+ 33 Qe6 Egl Drawn.
The opportunity for the side hav-
ing the rook to create a distant
passed pawn quite often appears as
a leit-motif of the struggle.
Novikov—Kaidanov
Vilnius, 1984
// 4’61
// 4/4/
/&// %%fi /
// //
2a?
Q?%////
1...§c4
Black immediately exploits the
break in coordination of the oppo-
nent’s pieces. In view of the threat
2...§xc3 White is forced to allow
the creation of a passed pawn.
2 @c6 a3 3 5.13M
The piece sacrifice 3 bxa3 Exc3 4
a4 does not work because of 4...a6!
3...§c5! 4 Q13
The knight cannot move: 4 @c2
ab, or 4 @b3 Exb5.
4...a2 5 55c2 Eb 6 @331?
He cannot hope to hold the posi-
tion with passive defence. Stronger
is 6 Qe4 with the idea of transfer-
ring the king to 04. For example:
6. .Eh5 (on 6.. .Qf7 possible is the
prophylactic 7 g4 followed by 8
Qd4) 7 h4 g5 8 hxg5 (dangerous 1s
8 gxf6 because of 8. .gxh4 9 gxh4
Q17 10 £d4 Exh4, with the cre-
ation of a second distant passed
pawn, while 10 .268 loses after
10...§c5) 8...fxg5 9 fn (also poss-
ible is 9 b3 with the threat of 10
@b4, but risky is 9 f5?! because of
9...§h2 10 @al Q17 followed by
the creation of a passed h-pawn)
9...Exg5 10 Qf4 and then b2-b3. By
reducing the number of pawns,
White can count on holding the
position.
6...Qf7 7 Qe3
Why not to e4?
7...Qe6 8 g4 Qd5 9 Qd3 EcS 10
b3
Allowing a tactical solution to the
position. As seen from the previous
examples, the exchange sacrifice
does not achieve its objective after
10 b4, but in this case the break-
through with the king decides—
10...§c4 11 QdZ §d4+ etc.
10...§xc3+! ll Qxc3 Qe4
Black has reached a technically
won endgame, since White’s king is
riveted to the a2 pawn and the
knight will not succeed in contain-
ing the pawn break on the king’s
flank. We present the main vari-
ation: 12 g5 Qxf4 l3 gxf6 gxf6 14
Qb2 Qe4! 15 QCZ f5 16 @a3 f4 17
9304 Qd3 18 523e5+ Q62 19 Qxa2
f3 20 @7113 Qxf3 21 Qa3 h5, and
White is lacking one tempo to save
himself. With the pawn placed on
b4 it would be drawn.
The game ended in the following
way:
12 f5 Qf4 13 Qb2 n4 14 @c2
Qh3 15 @d4 Qxh2 16 @e6 Qg3 l7
@xg7 Q418 @eB Qf 19 @d6
Qf4 20 gets Q 5 21 @d6 h5 22
®e4+ Qf4! 23 xf6 M 24 @d5+
Q13 and White resigned.
Let’s look at a more complicated
example of the creation and realisa-
tion of an outside passed pawn.

Beliavsky-Dolmatov
Minsk, 1979
%/
/////./g,§t
% 9. .Qf6) 9...Qf6 10 ®f4 Ef2+ 11
,/ //
7/%/%fi
/
///%////
///
White has a material advantage,
but the presence of a weak pawn on
a2 and the lack of cooperation be-
tween his pieces allows the Oppo-
nent to count on creating a distant
passed pawn. The position bears a
concrete character and in the event
of a premature fixing of the weak
pawns, similar to the previous
examples, White will manage to put
right the coordination of his pieces
and prevent the invasion of the
rook. For example: l...a3 2 @f4
Ec8 3 @e2.
However Black breaks through to
the a2 pawn in a tactical way.
1...c4!! 2 Exc4
After 2 bxc4 Eb8 Black organises
the advance of the passed b-pawn: 3
58B b3 4 a3 (or 4 axb3 a3 5 Ed]
Exb3 etc) 4...b2 5 Rb] Eb3 6 $62
Ec3 7 @e4 Exc4 8 @d2 Ec].
2...Ec8
In the event of 3 QfZ? there is the
threat of an exchange sacrifice
together with a pawn break——
3...Exc4! 4 bxc4 b3 5 axb3 a3.
3 Ed3 a3! 4 @e3
The most tenacious continuation.
The attempt to place the knight at
once loses quickly. For example: 4
Two minor pieces against a rook 163
Qf4 Ec3 5 @e2 Exb3 6 Ec4 Eb2 7
@d3 b3; or 4 52312 Ec15 Ee4 Ea16
®d3 Ev 7 @xb4 (White should
also lose in the event of 7 h4 EfZ!)
7...Exh2 8 EdS f5 9 @e3 (on 9
.9.f winning are both 9...Eb2, and
@e3 Eb2 12 Ec4 @g5 13 Exf7
@g4 14 .936 @xg3, and each of the
passed pawns will cost a piece.
4...Ec1 5 @f4 Ea] 6 55d5 Exa2 7
Qxb4 Exh2 8 £32 Eh3 9 $12
Not allowing the rook to assist
with the advance of the a-pawn,
which would have been the case
after 9 @f?) Ehl 10 3c4 Ea]
(otherwise White can blockade the
a3 pawn by 11 @a2 and 12 b4) 11
@c2 Ea2 12 @b4 Eb2 etc. However
Black finds a new resource.
9...f5!
In the event of 9...Ehl White suc-
cessfully solves the problem of re-
stricting the rook by 10 if] Eh2+
11 ig2 Eh8 12 @a2 Ec8 13 .934.
Now however the threat of the break
...f5-f4 is irresistible. On 10 Rf]
follows 10...Eh2+ 11 EgZ Eh8, and
the rook breaks through on the
queen’s flank.
10 @332 f4 11 gxf4 Exb3 12 9&4
Eh3! 13 31']
Black is preparing a breakthrough
with his king, whereas White cannot
force the rook from the 3rd rank
without material loss. For example:
13 @g2 E63 14 $12 Ee4 etc
13...Eh2+ l4 £g2 @f6 15 @g3
EhS l6 Efl Ec5 l7 Ed3 Ed5 18
£36 Ed2 19 £c4 Ed4 White
resigned.
An analogous idea of an ultra-
sharp breakthrough was exploited
by Kramnik against Ivanchuk.
I 64 Two minor pieces against a rook
Ivanchuk-Kramnik
Linares, I 998
A A A
AA
11111
§ /
/A
\
A A
A//,
./%
\Ԥ
\\\\\\.“
\
I?
A/QA /
$§
AA / ,; £A///
\§
A
1...c4!! 2 bxc4
On 2 3xc4 follows 2...§d2+ and
3...§xc2 and 4.2v, and Black
has a passed pawn. However it is
impossible to avoid the loss of a
pawn.
2...§b8 3 (:5 EbZ 4 c6 $67 5
§xg6+ @d6 6 @e5 Ev 7 523c4+
c7
After 7...@xc6 8 ixf5 there is no
threat of 8...§x02 because of 9
$.c a2 10 @a5+ and 11 55b}
8 ®g3 Eal 9 @xa3 3x33 10 QM
Drawn.
In positions where the side having
the rook cannot actively support the
passed pawn, the minor pieces fre-
quently prove to be stronger than
the rook. Therefore an important
task for the defender is to ensure the
activity of the rook and king.
With a small number of pawns a
good defensive resource is served
by the possibility of an exchange
sacrifice with the elimination of all
the pawns or with a transfer to a
position which, although without
pieces, is still a theoretical draw.
Tavadian-Tseshkovsky
Irkutsk, I 983
A% /
A?/%E A
/ /
Q
\
/
x/ ////
\
671/7//
\g‘wf‘“
\\\\\~
”/3 'c..%
A 7/
\
To realise his material advantage
White should adjust the coordina-
tion of his pieces in such a way that
the threat to advance the f-pawn
hampers the activity of the rook and
king and wins the pawns on the
queen’s flank. Together with this it
is important to firmly blockade
these pawns.
l 03% §d3+ 2 @f2 @gs
Black would readily exchange the
b-pawn for the f-pawn, but 2...?f4
does not work because of 3 @d4 b4
4 $362+ Qf5 5 £xb4.
3 @b4
White does not find a clear-cut
plan to attack the b-pawn. As a re-
sult there still remains an interesting
struggle for tempo, leading to a de-
cisive increase in material advan-
tage. As R.Tavadian pointed out,
there was a win here with the
brilliant manoeuvre 3 @d4! @f4!
(3...b4 does not work because of 4
@e6+ @f5 5 @c5 §x03 6 bxc3
bxc3 7 @e4 02 8 @g3+ and 9
5/362+-) 4 @e2 §e3+ (above all
4...b4 does not work because of 5
RdZ etc) 5 @fl! §d3 (he cannot

play 5.. .Ee7? because of 6 .262
@g3 7 586+ or 5. .Ee8 6 712 fib8?
7 QdZ @e5 8 $306+) 6 712, and an
elegant ‘dance’ of the king along the
‘triangulation’ e2-f1-f2, typical for
pawn endings, culminating in the
win of the b- awn.
3...§d7 4 93 @fs 5 f4?
It was still not too late to return to
the above-mentioned plan. However
White’s careless advance of the f-
pawn removes his control of the e4
square, allowing the rook to develop
great activity with support from the
king. In effect the material advan-
tage is now devalued.
5...Eh7 6 fies §d7 7 Qt} fidZ! 8
m3 §d7 9 @a2 §d3+z 10 @e2
§b3
The white king is cut off along the
lst-3rd ranks, and when the Oppor-
tunity presents itself an advance of
the b-pawn is threatened. The
chances of the two sides are even.
11 @d2 §d3 12 @c2 E13 13 @b4
@e4! 14 @216 @d5 15 ®c7+ @c6
16 @e6 @d5 17 @gS Eg3 18 .26
§g2+ l9 @c1 EfZ 20 fidz b4! 21
3xb4 fixf4 and the game soon
ended in a draw.
Here is a good illustration of the
failure to exploit one’s chances.
Miles-Adorj an
Wijk-aan-Zee, 1984
777 7g
//
7 7/,7
7(787 7
.7777 777
7 7 7 7
7 77/
7 7
Two minor pieces against a rook 165
Superficially the impression is
that Black’s advantage is of a
decisive charcater. However
analysis shows that, even though the
d5 and g7 pawns are doomed, his
defensive resources are in no way
exhausted.
I EgZ @gB 2 @gS?
In such endings it is very import-
ant to activate the rook. This is
achieved by 2 EgS £34 3 f5 @3d
(if 3...f6 4 Eg6 @3d 5 @g4 the
chances are already on White’s side)
4 f6! @xf6 5 Eb n7 6 §b6!,
with the idea of 6...d5 7 @gS. The
position is simplified and White ob-
tains definite counterplay, asso-
ciated with the possibility of
creating a distant passed pawn on
the queen’s flank.
2....Qe4 3 §g3 93d 4 @h6?.
White again misses the opportun-
ity to activate the rook, though in a
less favourable light. Correct was 4
f5 @xg7 5 §g4 58f6! (5....Qf3 looks
tempting on account of 6 §d4 f6+ 7
@h4 @e7, winning the f5 pawn,
however, by playing 6 §g3!, White
is forced to repeat the position:
6....§.e4 7 fig4, since it is dangerous
for the bishop to leave the bl-h7 di-
agonal because of the threat f5-f6) 6
Egl £c6 7 h4! (it is useful to hold
the check to the king in reserve:
after 7 @f4+ @h6 it is not easy for
White to hold the position, for
example: 8 Edl @d5 9 @g3 @g5,
or 8 Eel @hS 9 fig] @dS, with
advantage to Black; nevertheless
possibilities of resistance were of-
fered by 8 h4! @h5 9 @g3l, but not
9 Eg7? QdS—H 7. .Qe4+ 8 @f4
®h6 9 Eg8, and with an active rook
White can
himself.
successfully defend
4...58e7 5 Ec3 ®f5+ 6 @gS @xg7
7 ECB @d4 8 EeB @e6+ 9 @g4
£c6 10 Ebs @d4 11 f5?
1 66 Two minor pieces against a rook

Now also the f-pawn is lost. We 1...?f8 2 @dz @e7 3 @b3 §b4 4
mention that on 11 §d8 Black re- @3c5
groups his pieces by 11..”9.3+ 12 And so the coordination of the
@g3 .934 followed by 13. W5Df5 and knights is put right. They defend
the d-pawn quickl advances. one another and therefore are
11...@f6 12 M xf5 l3 h5 unassailable.
Belated activity by the passed 4...f5 5 @gl EbS 6 @fZ @f6 7
pawn. @d7+ @e6
13...@e3+ l4 QM d5 15 §d8 d4 His hopes of driving away the
16 §d6+ @e5 17 h6 ®f5+ 18 @hB knight from the a6 pawn are unreal.
@3t l9 fixh6 d3 20 §h5+ f5 Stronger is 7...®g5, trying to keep
White resigned. back the king and create a passed
f-pawn.
Let’s look at an example where 8-Qb6!
the minor pieces have to hold back
an onslaught by an active rook when
there is relative material equality.
m7//717/:
/7
Keres-Szabé
2/
//§/i
Candidates (1), Ziirich, 1953

@/
/QA/
//
//1//,
/////&///7 // Q //
/ 77/
//7/ The knights have regrouped and
again are unassailable.
7/7// 8...?e5 9 @g3 Eb3+
777 Worth considering is 9...f4+, but
even in this case the king is not ob-
77 liged to step back. For example: 10
§g4 f5+ 11 @gS Eb3 12 55d7+
The initiative lays with Black: his (weaker is 12 @c4 because of
rook is very active, whereas the 12...®d5 13 @bd6 13! 14 ‘gxf3
knights are far away from each fixfB, and the knights are “hob-
other and, what is no less important, bling”, while Black is able to create
they have no support points. Besides a passed pawn) l2...@d5 13 €3c
this White needs to reckon with the f3 14 gxf3 3x13 15 @xa6, with a
possibility of a transfer to a pawn probable draw.
ending. White’s main task is quickly 10 QM §c3 11 581:5!
to put right the coordination of his A tactical solution to the problems
knights. Let us proceed with of defenceu-typical for such end-
P.Keres’ logical and beautiful fili- ings. White exchanges pieces and
gree manoeuvre which solved the transposes the game to a drawn
problem of the defence. pawn ending.

11...§xc5 12 55d7+ Qd6 l3
@3c Qc 14 Qg5 QbS 15 Qh6
And the game was drawn after a
few moves.
With the pawns placed on one
flank the task of the defending side
still remains quite complicated.
Miles-Kindermann
Bath, 1983
/, // a: /
% % / /
/ ”46/27/47
%////g%
/%///
Black has the advantage, since he
has the possibility of creating a
passed pawn. But the small number
of pawns gives White a defensive
resource—at an appropriate moment
to sacrifice his bishop and obtain a
theoretically drawn position.
1...Qe6 2 3&3
The best practical chance! In the
event of 2 £37 §c2+ 3 Qb6 (after 3
Qb5 f4 4 3d8 Et White loses a
pawn, since he cannot play 5 .9.n
because of 5...§h5) Black plays not
3...f4——4 $.d8 Et 5 $.n with
possibilities of defence, but first of
all 3.. .,h6 with the idea 4...f4 5 h4
xgh4 6 .if4 h3- +.
2. .Exh2 3 £d5+ Qe7 4 Qd4
Qf6 5 Qe3+ Qg6 6 3e] f4+ 7 Qd4
h5
Two minor pieces against a rook I 6 7
A speedy advance of the distant
passed pawn should lead to a win
for Black.
8 QeS M 9 £e4+ Qf7?
Carelessness, after which White
saves the game with a surprising
tactical blow. The natural continu-
ation appears to be 9...Qh6, and the
h-pawn swiftly advances. For
example: 10 Qf6 h3 ll £b4 §e2 12
£g6 §e6!, or 10 Qf5 Eg2 ll £b4
h3 l2 Qf6 fieZ—in both cases with
an easy win.
//
/i§/ /
///. /Mx”/
%
//////§
7/
¢=/,,
h
10 £31114”
Now it all comes down to a theor-
etically drawn ending.
10...gxh4
The situation is not changed by
10...§xh4 ll Qf5 §h6 l2 n5
§f6 13 £f5 and 14 Qxf4, with a
draw.
11 Qxf4 §g2 12 .915
White would have replied with
this move also after 11...§h1. There
followed:
12. .Qf6 l3 £h3§ 1 l4 £g4
Ebl 15 2.113 §b4+ 16 e3 Qe5 17
£g4 §b3+ 18 Qf2 Qf4 19 Qg2
Eb2+ 20 Qh3, and Black soon con-
vinced himself of the futility of
playing for a win.
I 68 Two minor pieces against a rook

When there are pawns on one 14 $13 Eb4 15 Eds Ea4 16 $.17
flank for both sides, the rook can Eb4 17 @112 Eb7 18 Ec4 Ed7
successfully contend with the minor The king breaks through in the
pieces. centre, and Black changes the
rook’s horizontal resistance to a
Dan-Pyte] vertical one.
Austria, I 978 19 @1’3 Ec7 20 EdS Ee7! 21 ECS
Ee8 22 E114 Ec8 23 EM Ee8 24
99.3

\0
\‘\\
Black has prepared against the
breakthrough g4-g5. For exam 1e:
/% / 24 g5 fxg5 25 fxg5 Ed8! 26 g
/7/ /M// @gS, defending against 27 gxh6.
Now in the event of 27 gxh6 gxh6
\

or 27 g6 Black is ready to sacrifice

the exchange for the dark-squared
//
/%/@fi7¢
\

// I/ I/ bishop, reaching a theoretically

/ /%
//’///
/%
ZEé
Material advantage is on White’s
side, but the inevitability of pawn
exchanges when pawns are ad-
vanced, gives the defender the
possibility of exploiting the ex-
change sacrifice motive to reach a
theoretically drawn position.
1...f6!
The more active 1...f5 creates too
much space for the bishop’s
activities.
2 g4 Ea3
One of the most important el-
ements is playing for the restriction
of the king’s movements.
3.9.12 h64h4Eb35h5
White plans the transfer of the
bishop to d4 and organises the break
g4-g5.
5.. EM 6 @g3 Eb3 7 Ed4 Ed3 8
Eb2 Eb3 9 Eel Ec3 l0 Ed2 Ec2
More principled IS 10...Ed3, hold-
ing the third rank and preventing
White’s coming manoeuvre.
11 Ee3 Ec3 12 @12 E33 13 £94
Ea4
Now defence by the rook moves
on to the fourth rank.
drawn position.
24...Ee7 25 EcS Ed7 26 EfS
Ed5!
In preventing the attempted
breakthrough of the king to the d5
square, Black once agains includes
the rook for defence along the rank.
27 Ed4 Ea5 28 @e4 Eb5 29 Ee6
White can attempt to break
through with the king, by playing 29
Ed7. Here are some sample vari-
ations: (a) 29.. .Ea5?! 30 Eb6 Ea6
31 EcS Ea5 32 @d5 @h7 33 @d6
@hS 34 Ed4i‘; (b) 29.. E.b7 30 3.06
Ec7 31 EdS Ed7 32 Ee6 Eb7 33
ECS Ebl 34 Ed7 Ecl 35 EfBi; (c)
29...Eb1! 30 Ec5 Ecl 31.9.18 QgS
32 Ea3 Eal 33 Ee6+ §h7 34 EcS
Ecl, and Black succeeds in defend-
ing himself.
29...Ea5 30 Ec4 Ea3 31 Ee6
As before 31 g5 fxg5 32 fxg5
does not work in view of 32...Eh3
33 gxh6 Eh4+ 34 @e3 Exd4! with a
draw.
31...EaS 32 Ed7 E32 33 g5
Reconciling himself to the inevi-
table draw. On 33 QdS the balance
is maintained by 33...Ed2.
33...fxg5 34 fxg5 E216
 Two minor pieces against a rook I 69

Also possible 18 34. 3.112 35 gxh6 king to f6 or h6. However it should

§h4+ 36 @e3 Exd4 37 @xd4 gxh6 not be enough to win.
with a draw. 11.. .fie7 12 58M 39,8 13 £g6
35 g6 §d6 §b8 14 @d3
Drawn. It 1s difficult to reach the h6 pawn,
On 36 Rg4 there follows since Black does not allow the white
36...§xd4+!. king admission to f5. All his hopes
lay in ‘humanitarian aid from the
West’.
Ponomariov—Plaskett
14...@d5 15 @e2 @d4 16 @d2
Hastings, 1 999 f3?
He can t keep himself waiting!
/,///
/ 17 @e1 §b2+ 18 @c2 @e5 19
@e3 Eb3+ 20 £63 §c3 21 @e1
Eel 22 523xf3+
And White won shortly.
///Zf
/// Changing the material balance
(obtaining a position with rook and
// /¢3’//4% pawns against two minor pieces) is
///a<é very often exploited by chessplayers
of the highest rank.
////
Petrosian-Tal
The well-known Scotch ending Moscow, I964
with two white minor pieces against
a black rook. The game smoothly
transferred to this drawn position, /, 4
there only remains to make a solid 5’
move before the time control...
1...§a2?
In accordance with his active
style.
2 .9.d Ed2
2."?d 3 ®b4+.
3 3&4 @e6
Despite the win of the pawn, as
before it is not apparent how White
wins.
4 M @d6 5 .915 39.2 6 $13 §e8 l...®xe3!? 2 £xb7 @3d 3 Exdl
7 3g6 §e7 8 58d Eel 9 @d3 Ee7 fiadB 4 @c3 fixdl+ 5 @3d EdS 6
10 3e4 §e8 11 h5 $.13 §d2
Not an obvious decision. It was For the two pieces Black has not
possible all the same to exchange on quite enough material, but activat-
g5, attach himself to the black ing the rook gives Black sufficient
pawns, place the bishop on f5, the play.
knight on e4, and try to run with the 7 @fl a5 8 £32 9.5! 9 a4 £d4!
1 70 Two minor pieces against a rook

After creating his passed pawn, Formanek—Mikhalchishin

Black can rivet White’s forces to the Hastings, 1985
struggle against it
10 fixd4 exd4 11 @el 3212 12 g4
$18 13 f4 @e7 14 5/3f2 Eal+ 15 E/E%W 39%
l §a2 16 M f6 17 h5 @e6 17
@d3 g6 18 hxg6 hxg6 19 .932 §c2 W/t/i/
20 9d] §c3 21 @d2 g5
And Black quietly held the draw. 3% ///
g// //// //
Beliavsky-Geller
Erevan, 1975 /aa/
g/

/ /xx 1...®xb3! 2 e7 ! @xcl! 3

/ a / a: Exf8+
There' is nothing else for White.
3...?xf8 4 Excl 31132
/7////// Black has a rock and two edge-

a %//a%, pawns for two pieces. To win it is

sufficient for him to exchange the
a/
/%/// ”2.371
Q? bishop.
&

5 l93g5 f6! 6 @e4

/e%” After 6 @xh7+ 9H 7 @xf6 fixc3
1...§b4!
8 Exc3 @xf6 the edge pawn must
decide the outcome of the game.
Black stands to win the a2 pawn,
and White finds himself faced with
roblems beyond his control. Now 2 /
d2 does not work because of
%/ fl:
fl!
2. .EbZ. Therefore White has to
confine his bisho to passivity.

//// ////
2 .933 §a4 3 ébl g5!
A pawn advantage on the flank is
only real when its exploitation gets /// ////
///
under way. / % 8/“
///
4 @e2 $g7 5 @d2 f5 6 @d3 EM
7 @e2 //
"/
/
/
éifi
”3:03,’,
After 7 f3 §h4 8 h3 h5 and 9...g4 ”I

a weakness arises for White on g2.

7...§h4 8 h3 2214 6...f5 7 @f6 E37! 8 @xh7+ Exh7
And a draw was agreed. 9 £b4+ ficS! 10 £xc5+ bxc5 11

EXCS E217 12 Rc4 £xc4 l3 Exc4
a4
Again there has been a transform-
ation of the advantage. This time the
game passes on to a rook ending
with positional advantage for Black.
Two minor pieces against a rook I 7]
l4 §c2 33 15 2212 @e7 16 @fl
@d6 17 @e2 QCS 18 @d3 Qb4 19
@c2 e5 20 M e4 21 @b1 §d7 22
@c2 §c7+ 23 @b1 @b3 24 §d2
§h7 White resigned.
 11 Rook against knight and pawn

The rook is a little stronger than a @g7 13 @xf4 @156 14 3+— or

knight, and it would seem that with 10.. .nZ 11 @g5 13 12 a2 and
pawns on one flank the knight, in then 13 :12, winning the £3 pawn,
View of its capacity to jump, could and then also those on g6 and h5.
compete with a stronger piece. 11 E218 Qg4+ 12 $17 @117 13
However, practice shows that in this £37! @116 14 @131
type of position the knight is even Avoiding 14 @gS Qf6+ 15 @h8
weaker than the bishop in the g5!
struggle against a rook. l4...Qf6 15 £17 Qg4 l6 §g7z
Black resigned.
Faragé-Csom
Budapest, 1984 We don’t need to talk about a big
alignment of forces, when
miniatures provide brilliant
examples...
y / y
% / Romanishin-Rodriguez
Moscow, 1 985
/ / /
// ////Q
%%%// %
%/'/%/ / ’ '
/;/,. .. x//

/ // //
1 @hfi! ///, %
Black is in a ‘little zugzwang’ and
he is forced to let out the white
///// /
king. //
1.. .Qd6 2 §c7 Qe4 3 E37 Qf6 4
9% Qg4 5 e14 Qf6 6 e95 Qg4+
/ / // /
7 e6 f4 Is it possible, perhaps, for a
Bad IS 7. .g5 8 @f gxh4 9 §a6+ grandmaster to lose this position?
@g7 10 @gS Qe3 11 Ea2, and As it turns out, yes—it is possible!
White attacks the pawns. l...h5
8 @fl QM 9 ZaS! Qe3 10 @f6! Simpler is 1...Qg7 and 2...Qf5,
@hG creating a typical drawn setup.
There is no saving himself by 2 99.4 Qg5+ 3 QM Qe6+ 4 @e3
10.. .QfS 11 fif! gxf5 12 @f QgS

Now, however, after 4...5bg7 5
@e4 5/35 6 @f4 and 7 §a6+
weakening the g5 square where the
white king will infiltrate.
5 M @e6?
Black loses his nerve. After the
correct 5...5Df7 6 @f4 58h6 7 §a6+
@g7 8 @e5 @g4+ 9 @e6 g5! it is
still a draw. Now, however, a rook
against pawn ending is reached.
6 @e4 g5 7 Ef5+ $g6 8 QeS
gxh4 9 @xe6 h3 10 @e5 h2 11 if]
@g5 12 fihl @g4 13 Et M 14
§g2+ Q13 15 EaZ Black resigned.
Hulak-Beliavsky
Bled, 1998
h'\\\
%///% // 1/
/,//z
\‘\\§\
£7";
//////// ’/
’ /
7
////V/////%%fi/
/ fi/ //
Defence in such positions is not
easy and requires, in the first in-
stance, an understanding of the best
distribution of the functions be-
tween his pieces. Therefore White
should place his knight on h3 and
wait for the black king. Instead of
this he decides to post his knight on
g3, which proves to be incorrect.
1 @e2 fial 2 58g3 @es 3 @gZ
E214 4 f4+?
Too active, but, you know, this is
not a rook ending. He should prefer
4 $13.
4m§d5 5 @f3 f6! 6 g5?
Rook against knight and pawn I 73
When defending one should re-
duce the amount of remaining ma-
terial, therefore he should prefer 6
h5!.
6...f5 7 h5 §a3 8 h6
White intends to attack the h7
pawn.
8...?“ 9 @e2 §b3 10 5AM
Trying to provide a covering de-
tachment for the black king.
10...§b2 11 @f3 @d3 12 58f2+
@d2 13 e4 §b3+ 14 @gZ fxe4 15
@xe4+ @e3 l6 @c5
After 16 5/3f6 @xf4 l7 @xh7
Black plays 17...§b8! followed by
h8, and all the white pawns are
attacked.
l6...§b4 17 ni Exf4 18 @d7
§f3+ 19 @g4 Efl! 20 @f6 §g1+ 21
@h3 $13 White resigned.
Incidentally, Vidmar lost the same
classical endgame against Alekhine.
Salov-Kamsky
Buenos Aires, I 994
//,//
///1§
//%//
/
’/
/
/ // /Mf/
/ //
// ///€t§
1...5Z)h7!
A correct transfer of the knight to
g6 in order to defend the M pawn
and attack the e5 point.
2 @hfi
After 2 g5 (93xg5 3 fixh4 the
knight all the same goes to h7!
174 Rook against knight and pawn

2. .5818 3 §e4 ®g6 4 g5 Qg8 5 Beliavsky—Kupreichik

Qg4Qh76Q13Qg77Qe3Qf88 Budva, 1995
Qd4 Qe7 9 Qc5 Qd7 l0 Qb6 Qe7
ll Qc6 Qf8 12 Qd6 Qg7 l3 EeZ
Q18 14 £12 Qg8 15 §f6 Qg7 and E / /
White cannot break Black’s de- %Q/
fence, therefore draw. % % %%//;/
It is well known that with pawns /
on one flank the knight is rather % 71/8/
stronger than a bishop. But against a
rock the knight proves very weak,
/ /Q /

since it often cannot escape from the / 7 W %% /

pursuit of a more mobile rook.

Beliavsky-Short
Linares, 1995
1...Qe6?
Despite all the suffering, he
should allow 1.. .Qxe3 2 fibS Qf6 3
Qg3 Qg6 4 EeS or 3.. .QfH 4 Qf2
%// / £23112 5 EfS! followed by 6 Qf2--.g3
2 Eb3 QdS 3 §b5+ Qc6 4 En
/ , Qxe3 5 Qg3 @d5 6 Efs e3 7 Q13
Qd6 8 Eff} Qe5 9 E218 Qe6 10
EaS! Qd6 11 g5 ®e7 12 §a6+
Black resigned.
% / ////
%/
/
/ Suba-Chiburdanidze
Dortmund, 1983
1 Ed!
Inferior is l Qe3 g5! with the
threat of ...f5-f4.
l...e4 2 b6 d2 %/%/
% /m
Forced.
3 Qe2 exf3+ 4 Ext? Qc6
If4.. @e4, then 5 §b3l,
5 Ef @3134 6 EeS! @n+ 7 /
Qd g5 / %
After 7. .Qxb6 8 Qe3 g5 9 hxg5
hxg5 10 En 5/3f1+ White plays 11 % //
Qd3 followed by figZ and QeZ. /
8 hxg5 hxg5 9 Qd3 g4
On 9.Qxb6
9.. again follows 10 One of the few positions where,
En 58H 11 Eg2. with equal pawns, the knight can
10 366+ Qb7 ll QC4 QB 12 hold the position against a rook.
fifG ®e3+ l3 Qc5 Black resigned. 1...Qh5 2 E34
 Rook against knight and pawn I 75

On 2 §a6 there is 2...®f5, while A.Petrosian-Rashkovsky

on 2 @h3 5/3e4 3 Ea5+ g5 4 g4+ Minsk, I 976
fxg3 5 §e5 @d6! 6 hn @317 with a
draw.
2...@g4 3 @gl @315 4 $12 ®g3t
Of course not 4...@xh4? because % %///
of 5 g3!.
5 §b4
////A,///
If5 fia6, then 5... g5! 6 hxg5 5/3e4 ///////@
and 7...5Dg5 with a drawn setup. ////
5...5Df5 6 §e4 g5 7 11n @n 8
fies ®g4 9 §e8 QM
Also possible 18 9. ..¢3d4 /%/%/
10 §g8+ QB 11 @e2 @g6! 12 ///
@d3 (936+ 13 @d4 13! and a draw!
Is there a basis for Black to play
for a win? To this question the
Stein-Bobotsov
grandmaster himself answered in
Moscow, 1 96 7
the affirmative and to start with
went after one of the pawns. There
followed:
/ /// 1...§d3 2 e4 3213 3 @fZ Eh3!

/ / / Black suggests that his opponent

leaves the g-pawn. His argument ap-
pears to be 4 @g2 Eh4 5 @e3
/ /’fi”/'/ Exg4+! 6 n4 @xg4 with a win.
The Erevan grandmaster agrees with
/ //% the evidence and tries to organise a
é’XI/é”? defence with an equal number of
pawns.
/’”/ / 4 $92 §g3 5 @fZ Exg4 6 $13
EM 7 @e3 Eh3+ 8 QM E213 9
IMQfS 2 EM b5 3 ab ab 4 QM 3216 10 @d3 @g4 11 @e2
fidS $64! On ll...@g3 there is the reply 12
Activity! @e3, and it is not apparent how he
5 Eb ®d3+ 6 @e2 13+ 7 @dz can improve his position. Rash-
f2 8 @e2 @d4 9 §b8 @c4 10 @fl kovsky in fact allows the second
@d4 11 §b7 argument, the essence of which con-
In the event of 11 b4 ($04 12 b5 sists of transferring the turn to move
QCS 13 b6 @b4 Black holds on. to his opponent.
11...?04 12 @e2 @d4 13 §d7+ 11...fiaZH 12 99.3 Ea3+ 13 @d4
@e4! 336 14 <$e3 @g3 15 52M §a3+ 16
The only, but sufficient, move. @d3 §b3 17 @e2 EbS! 18 @e3
14 §b7 @d4 Drawn. §e8 19 @312 f5 20 $111+ @112! 21
1 76 Rook against knight and pawn
(9312 Qg2 22 Qf4 Qxf2 23 ef
318!
And here is the last argument,
bearing in mind that the ending has
assumed a study-like character.
24 QgS Qe3
On 24 Qe5 Black would have
proceeded on the other side—24
...Qg3.
25 f6 Qe4 White resigned.
Chernin-Georgiev
Dortmund, 1991
/ //
////§/ H
/,/ /// "/fi////
fl?
// " as
//////////
7// /%/////
The defence, even with play on
one flank and an extra pawn, isn’t
easy. Here the most unpleasant
thing is zugzwang.
1 fieS! 55b2!?
Or 1.. .Qd6 2 §g8 5365 3 Qe4 g5
4 Qf5 winning easily.
2 §d8+ Qe6 3 EM
Now White’s idea is to hunt for
the runaway knight.
3...Qf6 4 Qe4 h5
After 4. .Qe6 White wins by 5
EM @d1 6 Eb6+ Qf7 7 fib3! 4812+
8 Qe5 Qg7 9 EB t 10 Qf4 5+
11 Qg4 Qg6 12 if] h5+ 13 f3
and the knight lS caught.
5 Qe3! Qe5
Or 5...g5 6 Qe2 followed by 7
fib4.
6 h4! g5 7 hn M 8 Exh4
And Black resigned because of
8...58d1+ 9 Qd2 @f2 10 g6 Qf6 11
Ef4.
The zugzwang also plays an im-
portant role in the next example.
The extra pawn on the other flank is
usually lost after a few moves.
Adorj an-Morozevich
Alushta, I 994
/,,2 A
% /%
///////
/ ’/”//
If;
//” // é/9
/////M§
1...§dl! 2 @g6+ Qh7 3 @e5
Efl!
A ‘little zugzwang’—White has to
release the black king or start push-
ing his pawns somewhere.
4c4g65h4Qg7653g4h57
58e3 Ehl!
The white king is riveted to the M
pawn and his black counterpart is
read to start his ‘promenade’.
8 d5 Qf7 9 @e3 Qe6 10 Qf4
White tries his last chance.
10...Exh4+ 11 g4
Or 11 Qg5 2124 12 @3115 Exc4--+
11...Ehl 12 gt fit 13 Qg4
Qe5 l4 c5 Qe6 White resigned.
Now two very similar positions
with just one small difference—the
white pawn.
 Rook against knight and pawn I 77

Sokolov-Khalifman 55g6+ 2 @e4 @gS 3 §g7 @h6 4

Minsk, I 986 §a7 @gS 5 §a3 @e5 6 §g3+ @h4 7
§f4 5506 8 Ec3 @e7 9 §d3 @hS 10
d7.
l...§h4 2 §h8+ @gs 3 @d5?
7 7 7 7’77
7/7>// ’7
This king transfer isn’t correct.
/7 The king has to go towards his own

7 / 7 7 pawn when White has serious win-

ning chances.
7% / 7/7 3...5D13 4 @e6
////7 7:7.777&7,, / It was still possible to return to his
pawn.
7 4...5De5 5 @d5 58$ 6 @d6 @e5 7
7 7 7
III ’I”’/

/ $9.6 583 8 $e7

Black played
1...?g5
...trying to activate his king. The
game continued
2 $13 @fs 3 gal @gs 4 §a5+
®e5+ 5 @e4 QM
Black tries a counterattack on
White’s pawn, but he hasn’t enough
time.
6 @f5 @hfi 7 £32 ®d7
7...5Df3 doesn’t work because of 8
fia3! @gZ 9 3x13 @xffi 10 h4+—.
8 337 Black resigned.
White has definitely decided not
to go back to his pawn.
8...5Z)e5 9 £38 58g6+ 10 @f7 @f4
11 333 f5 12 Ed @xh3!
Black has calculated this known
endgame very carefully
13 Exh3 @g4 14 Eh6 f4 15 §g6+
Qh3 16 §f6 @g3 17 @e6 13 18
@e5 f2 19 99.4 @gZ Drawn.
A knight is generally more useful
than a bishop when the pawns are
all on one flank, but against a rook
the knight has a more difficult time.

Dokhoian-Shirov Horvéth-Sherzer
Klaipeda, 1988 Brno, 1993

E/ 7 7
7/ 77// 7 7 /
7 /.7. /7/7/%/
7 7 f 7
\Efie

7 7 7 77 /7 7
7/7 7 7& 7/7/7/7s
7y/’/%,
\§

777 g
The game continued 1 @gB $e6 2 @213 g5
1 QM?! Black has to build some sort of
This is not the most exact way to wall against White’s king, but it
win—more direct would be 1 @f4 cannot be done.
I 78 Rook against knight and pawn
3 gas Q17 4 Q13 Qg6 5 Qe3
QhS 6 336+ Qg7 7 Qe4 Qf4 8
Q13
Another good possibility was 8
§a2 Qf6 9 M Qg6 10 h5 Qf4 ll
g4 and after a check the king ob-
tains the very important f5 square
8. .Qh5 9 Qg4 Qf4 10 g3 Qe2 11
§e6
Another method would be
fid6, 12 §d2 and 13 Qh5.
ll...Qd4 12 Ee4 Qb5 l3 Qh5
Qd6 l4 §e7+ Black resigned.
Two instructive and I would even
say amazing endings played by an
international master from Argentina
against experienced grandmasters.
Sorin-Alterman
Erevan, 1996
///fl/%/
/ For the exchange White has three
/////// a
//Q/
//
/i
White threatened to play
Eg5-g3--c3 and, after capturing the
b2 pawn, winning the well-known
ending of rook against knight with
three pawns on the flank. But Alter-
man finds an elegant way to save
the game.
1...Qe3+! 2 Qxb2 Qd5 3 Qc2
If 3 g3?, then 3.. .Qf6, and the
rook 15 in a trap after 4. h.6.
3.. .Qxf4 4 Qd2 Qf6 5 figs Qf7 6
§g3 h5!
Activity comes first!
7 m3 114 8 E13 n2 9 $12 113
10 Qg3 Qel 11 §c3 Qg2 12 Qxh3
Qf4+ 13 Qh4 Qf6 14 £33 Qg6+
15 QhS Qf4 16 @114
There is nothing in 16 Qh6 Qd5,
followed by the advance of the
f-pawn.
16...Qg6 l7 Qg3 e5 18 M Qf4
19 §a6+ Qg7 20 $3 $17 21 gas
11 Qf6 22 3216+ Q17 Drawn.
Zaichik-Sorin
Erevan, 1996
/ //
//
\\\§\\ \\\\
\s
”/ ////////
// //afl/Q
.‘\
t‘\‘
/,// /
K\\
/////fl/////
pawns and an easily winning
position—for example:
1 Qg6 §a6 2 f5 §b6 3 f6+! Exf6
4 h6+ Qf7 5 h7 Exg6+ 6 Qh5 §g8
7 hxg8=W+ n8 8 Qg6.
But Zaichik did not see this plan
and chose another one.
1 Qf5 Ef6+ 2 Qe4 £216 3 g5 Ea]
4 Qf5
Not bad was 4 f5 §a4+ 5 QB
§a3+ 6 Qf4 §a4+ 7 Qg3 333+ 8
Qh4 §a4+ 9 Qg4, and there is no
defence against f5-f6+ and g5-g6
4. .3212 5 g6 Qh6!
The threat was 6 Qg5 and then 7
h5-h6.
6 Qd7
6 Qg4 led to roughly the same
thing.
6...§a5+ 7 Qf6 $216+ 8 Q17 @217
9 Qe7 335 10 QeS
 Rook against knight and pawn 1 79

Better really was 10 @f6, prepar- Martinovié-Mjkhalchishin

ing the return of the king. Igalo, 1994
10...§a7+ 11 Qf6 336+ 12 Q17
$217+ 13 Qf6 336+ 14 Qf5 Qt
15 g7 §a8 l6 @g4
There is nothing in 16 Qf6 Qh6.
l6...§aS+!
Now Black’s main task is to keep
the king as far away as possible
from the f4 pawn.
l7 Qe4 334+ 18 Qd3 Ea3+ 19
Qe2 §a2+ 20 Q13
If the white king heads towards
the rook, then at a necessary mo-
ment follows Ea8, @f6 Qg6, g8=y
fin, @3n QfS, and a draw.
20...Ea3+ 21 93e3 3218 22 @f5 1 b5?
Qg6 23 g8=¥+ fixgs 24 ®e7+ Qf7 Nothing is gained by 1 Qfl!
25 5/3n n8 26 Qe4 Exb4 2 Qg2.
1...s 2 5815 Qe4 3 ®d6+
There is no saving himself by 3

7/ //
@e3 Qf3.
3...Q13 4 l
//
/ 7/7.
In the event of 4 5/368 §b1+ 5
Qd2 Eb 6 @Xf6 after 6...h5 with

// 7Q% 7/
the idea of 7...§f5 Black has a de-
cisive advantage.
/////
7/ 4...§xf2

7% //’/
7/.
26...Qf8!
A clear-cut move—and a draw!
In the following position White
has good compensation for the ex-
change, but he should immediately
start building a fortress for his king.
4...Qxf2? 5 @e4+ QB is mis-
taken because of 6 Qxf6 fib 7
@xh7.
5c
The result is not changed by 5
@e8 n3.
5...§32! 6 b6
Or 6 @e8 f5 7 58156 h6.
6...§a8 7 Qc2 fibs 8 b7 n3 9
Qd3 Qxh4 l0 Q64 Qg4 ll QdS h5
White resigned.
 12 Queen Endings

In queen endgames the main prin-
ciples are slightly different from
other endings. And, to be precise,
though material advantage (for
example, a pawn) naturally has
great significance, much more im-
portant is a passed pawn. Thus the
main principles can be placed in the
following order:
(a) Passed pawns
(b) Activity (approach) of the king
(c) Activity of the queen
(d) Combinational exploitation of
various factors
Passed Pawns
Mikhalchishin-Gufeld
Nikolaev, 1981
2 gxf7 Ee2+ 3 @cl @xf7 4 h5!
White must exploit his chance to
the utmost—the passed h-pawn is
also ready to be exchanged for the
b5 pawn.
4mgc4+ 5 @d2!
Inferior is 5 @b1 b4, and after the
advance of the b-pawn the white
king is threatened with mate. White
does not fear the transfer to a pawn
ending.
5...Wc6 6 WhB!
White wants to advance the h-
pawn filrther.
6...Wd5+ 7 @cl Wc4+ 8 @112
Wd4+ 9 @b3!
Meeting the pawn half-way!
9...Wa4+ 10 <$b2 Wb4+ 11 @a2
Wd2+ 12 @b3 Wd3+ l3 $b4
Wd4+ 14 @215 $214+ 15 @b6
The white king is up in the
‘clouds’ and Black reconciles him-

//// /@ self to a draw.

//// / 7
/ Ed.Lasker—Marshall
New York, 1923
/£
/ //
/
/
/ /7

// /,
/// gy/
/// /
/._~
/._h

There followed:
1 g6+! Qg7
Ifg l.. .n6, then 2 Wg5+ with
perpetual check or 1.. .fxg6 2 We7+
with the same result.

A difficult struggle with a passed
pawn hoping to queen but with the
king unable to get over to help it.
1...WeZ 2 gel g5!?
Striving to create counterplay.
However 2. .We5 deserved the pref-
erence, intending to transfer the
king to the queen’s flank with an
exchange of the h7 and g6 pawns
for the pawn on a5
3 35 f4 4 gfl W32 5 @b5! @g6 6
fib6+ @hS 7 a6 g4 8 a7 g3 9 fxg3
fxg3 10 fic5+ @g6 11 Wd6+ @f7
12 fif4+ @e6 13 ye3+ @d7 14
@fl! Ea6+ 15 @e1 @c8 16 @xg3
@x37 17 Wg8+ @b7 18 yxh7+ and
19 @1137 +—
Taimanov-Gligorié
Santa Fe, 1960
//
2 2//%
2(2/2
/
’ 2
/
”//
a////2/
l Edi”
Defending against ...yc2+ and
preparing a4, therefore Black de-
cides on a pawn sacrifice with the
aim of activating his forces.
1...@f6 2 yd ¥c2+ 3 @h3
gel
Chances of saving the game re-
mained with 3. .WfZ! 3 f4 Wfl+4
@gZ fidl.
4 gd4+! c$e6
Not possible is 4...?137 because of
5 @f4+, nor 4...@g5 because of 5
§e5+.
5 we4+ @d7 6 QM @112 7 yeS!
Queen Endings I81
ain not 7 fixg6? in view of
7.? h2+!. It IS important to activate
his king via the g5 square.
7...yaz 8 @gS $12
No help Is 8. .%H 9 @f6!.
9 WM gb6 l0 ygf'7+ c ll
@xg6
Black can resign since 11...§e3+
12 @t @xffH is no use because
of 13 @g4+.
Activity (approach) of the king
One of the most important factors
in queen endings.
Ribli-Hertneck
Dortmund, 1986
/ 2
%/ 2 2 %
/g212
2 /.2M 2
22 2 2
2 2
1 h3 h5
Otherwise White will play g3-g4
obtaining real winning chances.
2 ¥d7+ @f6 3 @e8 @a6?
The only defence was 3. .@g7 4
ye7+ @hS 5 y@f‘S @h7 6 gf7+ @h6
7 WgS m2 and White must return
to the plan with g3-g4, creating a
passed e-pawn.
4 Wf8+ 99.6 5 $13!
Suddenly there has appeared the
idea of penetrating to the king’s
flank with the white king—very
typical in fact.
5...@a4 6 ygB @f6 7 @d8+ @e6
8 @d4!
182 Queen Endings

And Black resigned because 1 @c3! Wc5+ 2 WM ye3+ 3 QM

there is no defence against @f4-g5.
For example SNWaZ 9 @g7 or
8...?02 9 e4!.
Mikhalchishin-Nestorovich
Budva, 1994
¥d2+ 4 @cs Wa5+ 5 @d4 fidz+ 6
$95 Ems 7 WM fihs+ 8 @e4
yhl+ 9 @d3 yfl+ 10 @d2 Wg2+
11 @c1 @fH 12 @112 WbS 13 @d8
Black resigned.

Hui-Mikhalchishin
/
¢ 91/ Budapest, 1989

// / 9/
/, //l
///////// 7/% 29
Mi /

% / % £3 %%
// //
/////
% //
//////é? /
;%/
//w//
1...@e7?
In queen endings the king must
// // «a
either be ultra-safe or ultra-active!
Better here would be Luya3! cut- The king usually approaches its
ting off the white king. own pawn, but very often it also
2 e5 @b6 heads towards the opponent’s king
If 2...wa3 then 3 Wd6+ @e8 4 to create mating threats.
e6+—. 1 HfB+ @g6 2 gg8+ @fSB
3 $13 m7 4 «an gm 5 @gs Wh7+ @es 4 ye7+ @d4 5 Ed8+
@e8 6 @hG! Black resigned. @e3 6 We7+ @f2 7 ¥a7+ @fl 8
n+
Rittner-Bernstein As a result of the approach of the
corr, I 968 black king White has had to mis-
place his queen.
8. .Qe2 9 WM c2 10 @g4 @d3
11 u 3+ @d2 12 Wf4+ @cs 13
/l%
/%////
/ % 7 412
gel d6+ 14 @h3 yd3+ 15 g3
Wd7+ 16 g4 WM!
The black queen is centralised to
the utmost, there is no threat of per-
aa %% petual check, and the march of the
%§%/%
////
// a6 pawn is decisive.
l7 @gfi We5+ 18 @h3 35! 19 a4
%/
@b3 20 @gz Em 21 fle3+ @212
///// White resigned.
 Queen Endings 183

The active position of the queen Marshall-Alekhine

Bradley Beach, I 929
This allows the king to be utilised
in the struggle against enemy

/ / / fie
pawns.

Gufeld-Minev
/,% /¥0
Sofia, 1 967
//
/&/%/ fi/
% /% %%
/%//31 ///%fi%fi
/”//
/¢//
, ,y/
/
7 /g/g/ """
%?;7{
%%/ 1...@a6+! 2 @g2 @d6

// //
a“ In the first instance Black block-
ades the opponent’s pawn.
3 WM @gs 4 EM @fl
It turns out that the white d5 pawn
The white queen controls the can be attacked by the king.
position and attacks all Black’s wea- 5 Wbs @ds 6 f4 Ed 7 @f3
knesses. It is only necessary to in- yc3+ 8 ®g4
clude the king. Clearly not 8 @e4 %4+!—the
IQfl f5 2 M Wc7 3 @e2 M6 4 pawns in this formation usually win
Hfi+ @h6 5 b3 Wd6 6 WM @h7 7 for Black.
@el WbG 8 M5 9d 9 @d2 yeS 8...@e7 9 WM h5+! 10 $t
10 @c2 gal 11 yc $212+ 12 yh3+ 11 @gs wg2+ 12 $5
@c3 gm 13 yc7+ @g8 14 QM @d+ 13 c$g6 Wd6+! 14 @xdG
fibz+ @xd6 15 f5 34 16 @xg7 a3 and
If 14...yd2+, then 15 @e5 e3+ Black won.
16 @f6 with a decisive invasion.
15 E6 gfl 16 @a yxg3 l7 Keres-Aronin
wd8+ @117 18 35 gel 19 ”EM f4 Moscow, 1 951
20 Ed @g8 21 a6! Wd2+
After 21.. .gxe3+ 22 @e5 one
//
advanced passed pawn is quicker
than two.
22 @e5 %2+ 23 @e6 b3+ 24
//.%e/’
.a¢/;«}/
@f6 %2+ 25 yes Black resigned.
% /W
In the following position, who
stands better? He who has the
further advanced pawn or he who
ara/
fi /
// /
advances it further? Let us see.
184 Queen Endings

1...Qe6? Combinational exploitation of

Better is 1...h5 or l...e4, trying various factors in queen endings
somehow to exploit his pawn major-
ity on the king’s flank. Ballon-Mikhalchishin
2 g3 @dH 3 Qg2 ¥d7 4 M Berne, 1995
§b7+ 5 QhZ @d7 6 W18! h5?
A decisive mistake. Correct is
5. .f5, endeavouring to advance his ///
pawn. / / / A7////
7 @g8+ QfS 8 c5 QM 9 yc8+! /
Qe4 10 Qg2 Qd3 ll Ea6+ Qd2 12
@d6
Also not bad is 12 fixfd
/}%g%/%
12...Qe2 l3 d4 exd4 14 c6 d3
15 c7 d2 16 c8=¥ dl=¥ 17 fie6+
Qd3 18 gd7+ Qe2 19 ge7+ Qd3
20 Wxa7 and White won easily.
Black has a protected passed
Romanovsky-Stahlberg pawn plus a much better pawn
Moscow, 1935 structure.
1...n! 2 l Wdfi!
By speculating on a transfer to a
winning pawn ending, Black
% % threatens to advance his passed
/%//// % pawn.
3 QM c3 4 @g4!
/ Z W/ White wants to put the black king
%/ in a box, but it boldly ploughs on.
§////§@W 4. .f5! 5 gh5+ Qe6 6 ge8+ QdS
7 Ee5+ Qc4 8 d5 c2 9 ¥f4+ Qb3
/ / l0 QhZ yc4 White resigned.
//// Botvinnik-Donner
l d! Noordwijk, I 965
After this centralisation White
threatens to transfer his king to e4,
therefore Black must hurry with his
counterattack. /, A
1...gb2+ 2 Qe3 Qc1+ 3 Qe4 A 77%
////// //,g
¢
l§e1+ 4 Qf5 @xg3 5 c5+ Qg8 6
gc4+ Qh8 7 @g4! %V%/%%/
With the threat to exchange White
obtains the g6 square for his king.
7...@e1 8Qg6! Wcl 9 WM /8% 78fig
%
Qg1+ 10 @817 l17 11 yf5+ g6 12 / /
WM Black resigned. ”/////

Black has the better structure, and
therefore White’s only chance lies
in the creation of a passed a-pawn,
which will be quicker than the
passed h--.pawn
1 Qd7+ @gS 2 @c8+ @h7 3
fixa6 gg5+ 4 @fl ‘9t 5 gd3+l
@gB
A trap—after 5. .Wg6? 6 @e4! the
pawn ending 18 winning for White.
6 a4 whl+ 7 @e2 h5 8 Wd5+
@hB 9 @dl!
Typical—by speculating on the
transfer to a pawn ending, White
forces back the opponent’s queen.
9...@h2?
Better is 9mgh4, though even
then White’s position is preferable.
10 b4 h4 ll gfl!
The queen is trapped!
ll...g5 12 aS ba 13 ba g4 14
a6 Black resigned.
Korolev-Palm
corr, 1980
%/,
%¥/
///¥ / evitably go on to queen.
// /%
/ ;/ //
/ /
// /%2‘3///
%/// 7
Thanks to his far advanced passed
c-pawn and strong queen position
on d6, White has a slight advantage.
1...wb2?!
On 1...g5 there would have fol-
lowed 2 Wd3 @f6 3 @d4+ @fi 4
@hZ @07 5 @gl with a slight ad-
vantage. Stronger is 1.. .ge4!, cen-
tralising the queen with chances of a
draw.
Queen Endings 185
2 c5 e5
On 2...@xa3 would have followed
3 Wm @f6 4 c6 wow 5 @112 65 6
@d6, leading to a position occurring
in the game.
3 yc7+ @e6 4 gd6+ $17 5 c6
The pawn is already on c6 and
there is no apparent perpetual check.
5...6§c1+ Qh2 Ef4+ 7 @gl
ge3+ 8 @fl6gf4+ 9 m2 ge4+ 10
@d2
The king will boldly go behind
enemy lines, the pawn already has
no significance.
10...@f4+ 11 @c2 ¥e4+ 12 @b2!
‘yxg2+ 13 @c3 @212 14 l§d7+ $16
15 c7
White has achieved his objective
—a pawn on 07. Now he needs to
shelter his king from the checks.
15...%xa3+ l6 @c4 wcl+ 17
@b4 gel+ 18 @c5 $215+ 19 $04
Of course, not possible is 19
$06? because of 19. .Wb5+ with a
perpetual check.
l9...@a2+ 20 @b4 gb2+ 21 @215
gay 22 @b6 §e3+ 23 QM firm
24 c film 25 @d5 Black
resigned, since the c7 pawn will in-
Karpov-Timman
Buenos Aires, I 980
/////,///
/// //// ///
//// // / 4
\\
\
(EV/9
fi/,/////// 626’s
,4 /, /,,// %% //
White has problems since the
Black king is in play while White’s
186 Queen Endings
is placed far from the scene of
action.
1...b6 2 Wh7
If White had sacrificed a pawn to
create a passed pawn by 2 h4, then
Timman gave the variation 2. .gxh4
3g5h34g6h25We3+Wf36
Wg5+ QB 7 Wd2+ We2 8 Wf4+
@g2 9 Wg5+ @h3! 10 Wh6+ @g4
11g g7 Wd1+ 12 @212 Wd5+ 13 b3
Wg2+ followed by 14. .h1=W
2.. .Wd4
The exchange of the a7 pawn for
that on h3 favours Black.
3 WM Wd1+ 4 $32 Wd5+ 5 W31
a5 6 Wh7 Wd4 7 Wh6 Wdl+ 8 @a2
Wd5+ 9 @211 b5
Threatening a mating attack after
...b5-b4-b3.
10 h4 gxh4 11 g5 b4 12 axb4
axb4 13 b3
After 13 g6 b3! 14 We3+ @g4 15
n+ @hS! Black wins.
13...We5+ 14 @212 We2+ 15 l
We3 White resigned.
Yusupov-Gerusel
Moscow, 1981
22, 2.2
12////
2%
212/ 2/22
2 4 Wf4 Wc6 5 g4
22 2/ &/////
22 222/
Despite the material equilibrium,
Black has problems in connection
with the excellent positions of
White’s king and u.een
We6+ @111 2 g6 Wg2+ 3 Wg4
WaZ?
Correct is 3...Wb2 4 Wf5+ c3e7 5
WcS+ @e6 6 Wc6+ We7 7 We4+
§d6 8 f4, and all the same White
has a dangerous passed pawn.
4 Wc8 We7 5 Wc7 @e8 6 @161
Fantastic domination by the white
pieces, leading to a winning pawn
ending.
6...Wb2 7 We5 Black resigned.
Marosi-Yudasin
Budapest, 1982
2 Q/
%/’//
66222/ ‘22 l;
//W2 22
22
27/2
/2§/2///
/ 2
l...Wd5!
This centralisation is better than
simply playing to eliminate White’s
pawns.
2 We8+ WM 3 Wb8 Wd7
He should not give the o ponent
counter-chances after 3... xf5 4
Wxa7 Wb1+ 5 @g2 b2 6 f7.
White’s only chance to change the
course of the struggle is to open up
the position of the opponent’s king.
5...aS6h4b57g5Wc48Wg3
We4! 9 gxh6 gxh6 10 Wc7 fS 11
Wa Wb1+ l2 @g2 b2
As a result of the exchange of
blows Black’s positional advantage
has been transformed into a material
one.
13 Wc7 b4! 14 @fl Wbl+ 15
@e2 We4+ 16 Wd2 WfS l7 Wb6
 Queen Endings I 8 7

On 17 yg3 Black achieves his Sometimes a complicated method

objective after l7...yd5+ 18 @c2 has to be adopted to create a passed
gc4+ 19 @b2 yc3+l 20 @xc3 pawn.
bxc3+ 21 @xc3 @g6 followed by l h6! yc4+
$115 and. @xh4. l...@xh6 2 @08+ and 3 @xc6
17.. .gf4 18 Qd3 t4 winning.
Further comment is unnecessary. 2 $13 fixa6 3 e5! yQa3+ 4 @g2
19 $63 Wm 20 @217 b3 21 QCS @xb4 5 exf6 xh6 6 §c8+ QM 7
@g6 22 $18 @xfl 23 flg8+ @115 yd7+ @g6 8 g7+ @115 9 f7 we4+
White resigned.
// gr]; III/.-

Mikhalchishin-Cvetkovié 7%? /
//
&;
//’Iro
x
/’///

Tmava, 1988
7%
/_//
777%7/
M / // 7
777
///7
/ gym 77/7
//%/7/
//2:¥/%%%//// 10 @h2 yf4+ 11 @gl yc1+ l2

/ 7 % 6/7 @gZ!

7 77%
As a result of the ‘triangulation
dance’ Black has no more checks,
and so he resigned the game.
188 Various Endings

Exercises:
Various Endings

Q Q&\Q
Q//Q////////
Q ////‘/4
g

Indicate White ’5' winning plan. Find a plan to realise White ’5

extra pawn.

2 4

/
Q Q,Q
Q Q
QAQ 4 Q/ / Q//
g/ /g/ Q / Q
QQQQ / Q
Q Q Q&/ Q
Find the right continuation Can White win?
for Black.
 Various Endings I8.

W W
/ WW / / / W W:
/, W/W}@
W/ WZW W3
W/ % $W/ W”W y
W W Z;ZW W W /&W

\
3/ W
WZW//
/ / W W%/%W
What is right: [H.319] or 1....9.e6? How does Black realise his extra
pawn ?

6 9

W W W
WW W / WW W W
/
aWWWZWZ/ ZW /y/ W
/&Z / ZWW WaW/
WQWZ/ W /W; W%W

How does Black defend after Can Black make a draw?

I @xa5 ?
10

1 /1.1;
W//W
W? W WsW WW/W/
WW/W/
W W5
llllll
//

/W/é}§
4§W%Zn/WZW Z”

W WWW W WWW
\

W W /%WW
W%1W W W/
\

WWZW
How does Black continue? How does Black defend?
190 Various Endings

ll

W
WWW W W
/2W%W/W
<§\

/y ,,,,,,
/ W/W//
\\‘

W/WaW/W

\
.5 \
W W W
///////////
:\\

W/////
Demonstrate a plan to realise
the extra pawn.
12 15

W
WflW/
W / W W
/W W W W
W//// /// WW /8
W///W'W
W W W
//%/%/’
W/WW/Wé/VWW
///W% W// W
W WW:WW
What is White '3 winning plan? Point out the correct plan of
defence.
13 l6

, WW W W
W’W / /
WW WNW/W
// WgW/ W
X

W W: W 4”? W W ”ti/M
\ \ \*§‘“

WQW
§

W}/ WW/ /
W/////W%/
®W§

W fig]
/W//W//
WWW/£3
\Q

W/WWWWW
White to move. What should be Can White win by 1 g3 hxg3 2 h4,
the result of the game? creating a passed pawn?
 Solutions to Exercises

PAWN ENDINGS

1 d1=¥ 7 ¥c8+ and a queen ending

was reached with an f-pawn for
1212 @f5 2 m3 @es 3 g4 White, Janvarev-Schcherbakov,
hxg4+ 4 @xg4 99,4 5 h5 f5+ 6 Moscow, 1994.
$h3! f4 7 h6 f3 8 h7 f2 9 @gZ 1-0,
Botvinnik. 5

2 (a) 1...g6? 2 9:24 @d6 3 h5 g5 4

White has a winning position, but
in the game he lost.
1 9e??? (1 a4 @d6 2 a5 @d 3
a6! +- 1...@d6 2 $e4 c4 3 a4 c3 4
@d3 xd5, 0-1, Guliev-Tukmakov,
Nikolaev, 1993.
3
(a) 1 gf? gf 2 @e2 @e7 3
@d3 h5 4 @xd4 h4, l/2-1/2, Yermo-
linsky-Ivanov, USA (ch) 1996
(b) l g5! @f7 2 @e2 @g7
(2...@e7 3 @d3 @e6 4 @xd4 @d7 5
$04 @d8 6 e6 @e7 7 $d3! @xe6 8
$14 +-) 3 e6! (3 @e2 h5!=) 32.93
4 @e2 @e8 5 @d3 @e7 6 @c4
@xe6 7 Qxd4 +-.
4 @b3 5 @g6, with a draw. The f6
1...hxg4?——Black thought that he
could achieve a draw as he pleased,
but correct was 1...fxg4 2 fxg4 hxg4
3 h5 @e5 4 @g3 But suddenly
there followed 2 f4!! @c4 3 h5 d4 4
h6 d3 4 h7 g3+ 5 @xg3 d2 6 h8=fi
@f5 b5 5 cxb5 @d5 6 b6 $06 7
$64 @xb6 8 @d3, 1-0, Dreev-
Anastasian, Tbilisi, 1989;
(b) l...@f6! 2 @e4 (2 h5 g6 3 a4
g5=) 2...?e6 3 h5 em 4 QM g6 5
@e4 gxh5=.
6
1 g4! hxg4 2 h5 @e6 3 a3! g3 4
ea: (4 @137 f5!) new 5 $13
@g7 6 @xg3 QM 7 @g4 @h6 8
@f5 @g7 (8."?t 9 @xf6+—) 9 e5!
fxeS 10 $n +-, Evreinov.
7
White resigned (7) because he did
not see 1 @f5 @xc4 (1...@b4 2 c5)
2 M! @b4 3 @g6 @xa4 4 @t
pawn helps White.
8
(a) 1...?e5? 2 h3!! (2 M? @fs 3
@g3 @g6=) 2...@f5 3 h4 @e5 4
@g4 @e4 5 h5 f5+ 6 @h3! @e3 7
192 Solutions
h6, 1-0, Padevsky-Latinov, Sofia,
1973;
(b) l...QgS! 2 Qe4 Qg4 (2...f5+?
3 Qf31+-) 3 Qe5 f5 4 h3+ Qg5 5 M
Qg4 6 h5 f4=.
9 14
(a) l...g4? 2 Qe3! f4+ (2...gxh3 3
QfZ) 3 Qe2! gxh3 4 QfZ QfS 5
l Qe5 6 Qh2 Qd4 7 Qxh3 Qc4
8 Qxh4 Qxb4 9 Qg4, 1-0,
Rufenacht-Orseth, corr, 1996;
(b) l...Qf6! 2 Qe3 QeS 3 f4+
gxf4 4 Q13 Qd5 5 Qxf4 Qc4 6
Qf Qxb4 7 Qg4 Qc5, and the
king hurries to f8.
10
1 c4!! cxd4 (l...dxc4 2 dxc5 bxc5
3 Qe2 +-) 2 cd b5 3 Qe2 b4 4
a4! QeS 5 h5 Qe4 6 d6 b3 7 Qd2,
1-0, Degraeve-Hansen, Germany,
1998.
11
(a) 1 Qf4? g6=, Ilyin-Zhenevsky-
Botvinnik, Leningrad, 1938;
(b) 1 Q13! g6 (l...Qe7 2 Qf4 Qe6
3 g3) 2 hxg6+ n6 3 Qf4 h5 4 g3
+-
12
Yes, l...c5!! 2 Qg2 cxd4 3 exd4
exd4 4 Qf2 Q97 5 Qe2 Qe6=,
Kozakov-Asparuhov, Pernik, 1972.
13
(a) l...Qe6? 2 Qc5 f5 3 h3 fxg4 4
hxg4 d4 5 Qxd4 Qd6 6 f5, 1-0,
Nimzowitsch-Chigorin, Carlsbad,
1907;
(b) l...Qc6 2 h4 2 h3 Qd6 3 h4
Qc6 4 h5 (4 g5 fxg5 5 fxg5 hxg5 5
h5 g4 6 h6 g3 7 Qe3 d4+ 9 QB
d3=) 4...Qd6 5 g5 fxg5 6 fxg5 Qe6!
7 g6 Qf6 8 s Qg7=. 2...Qd6 3
h5 Qe6! 4 Qc5 f5 5 g5 hxg5 5 M
Qf7! 7 fxg5 f4=.
Yes, but not by 1 axb4 a3 2 Qc3
cxb4+ 3 Qb3 QdS, and White re-
signed, Sofia Polgér-Smyslov, Lon-
don, 1996, but 1 c4!!, creating an
impregnable fortress.
15
1 h6! (1 Qc6 f5 2 Qd5 M4 3 a4
Qg3=, Filipescu-Citron, Romania,
1955) l...Qxh6 2 Qc6 Qg5 3 Qd5
f5 4 34 f4 5 Qe4 +-; l...gxh6 2 a4
f53a5f44a6f3537f26a8=¥
f1=§ 7 ygSH +-.
16
(a) 1 Qe3? d4+! 2 cxd4 Qd5 3 b5
Qc4, 0-1, Havsky-Yuferov;
(b) 1 Qd2 d4 2 04 d3 3 c5 Qd5 4
a5=.
l7
1 Q96! (1 Qg6 Qg3 2 Qf5 Qf3=)
1".t 2 QfS! @114 (2...g6+ 3
QgS! +-) 3 c4! g5 4 c5 g4 5 Qf4 g3
6 Q13 Qh3 7 c6 g2 8 c7 +-, HOCh,
1972.
18
(a) l...e4 2 fxe4 g4 3 Qxf7! g3 4
f6 g2 4 Qe8 g1=w 5 f7 +-;
(b) l...g4 2 fxg4 e4 3 g5 e3 4
Qxf7 62 5 g6 +-;
c) l...Qg8 2 f6 g4!! 3 fxg4 e4 4
g5 e3 5 g6 e2 6 gxf7 Qh7 7 185%
el=§+ 8 Qf7 @e6+!! stalemate.

19
l...@e2! (1...@f2 2 @e5 +-) 2 h6
@d3=.
20
(a) l @gS? a4 2 h6 a3 3 g7 a2 4
h7 a1=w 5 g8=§ @CH, 0-1, Wells-
Hector, Oxford, 1999;
(b) 1 Qf4!!=.
21
(a) 1...?h5? 2 b4 @xh4 3 b5 5 4
a5 g4 5 b6 axb6 6 axb6 g3 7 £3!
+-;
(b) l...@g7!=, Uhlmann-
Robatsch, Marienbad, 1965.
22
1...axb5? (after the correct l...aS!
Black holds. On h3-h4 he replies
...h6-h5, while on g2-g4 forced is
...g6-g5!) 2 ab @d7 3 QB @e7 4
h4 h5 5 @e4 @d7 6 @d4 @c7 7
@c4 @d7 8 @b4 @c7 (matters are
not changed by 8...@c8 9 @215 (9
c6) 9...?c7 10 g3!) 9 Qa5 @c8 10
$b6 e5 11 fxe5 fxe5 12 @a5! l-O,
Speelman-Saltaev, Las Palmas,
1998.
23
1...?b5? (l...h5! 2 f3 [2 h3 @b5 3
g4 hx 4 4 hxg4 §c6 5 g5 fxg5 6
fxg5 d6=] 2...@b5 3 g4 hxg4! 4
fxg4 @c6 5 M @d6=) 2 g4 @c6 3
g5 e5+ 4 fxe5 fxg5 5 f3 h5 6 @c4,
Solutions 1 93
1-0, Anastasian-Khalifman, Erevan,
1996.
24
1 c4? led to defeat. White should
give preference to 1 b4! cxb4 2
cxb4 @fl 3 b5 @e7 4 g3 $17 5
@d6 @f6 6 @c7 @e7 7 @b7 QM 8
@xa7 @c7 9 g4! e5 10 gt gt
11 $38 e4 12 a7, and a draw.
The game ended 1...@f7 2 @d6
@f6 3 $06 @671 4 @b7 @d7 5
@xa7 @c7 6 @a8 e5 7 g4 hxg4 8 h5
gt 9 b4 cxb4 10 C5 b3 11 a7 b2
12 cxb6 $xb6 13 @b8 b1=§ 14
a8=W Wh71, 0-1, Mortensen-Plueg,
Hamburg, 1997.
25
Here White resigned, calculating a
variation with a necessary exchange
on f6 However, correct is not to
take the pawn.
1 QB @xc4 2 M! @b4 3 @g6
@xa4 4 $xh5 @b3 5 @g6, and the
pawns queen at the same time, while
the f6 pawn ‘helps’ White,
Klovan-Elistratov, Moscow, 1963.
26
(a) l...@xb4 2 a6 @c3 3 Qel!
@d3 4 a7 f2+ 5 @fl e2+ 6 @xfz
@d2 7 a8=y el=§+ 8 @gZ yxh4=,
Papendieck-Geiger, Austria, 1998.
(b) 1...?c4! 2 36 @d3 3 @e1 e2 4
37 @e3 —+
194 Solutions
TRANSPOSITION TO A PAWN ENDING
No,
l...@e5??
1...5Dd6 2 g5 (2 M 58e4) 2...fxg5
3 58n 04+ 4 bxc4 @xc4 5 93B
@e3 6 g4 Qg2=
2 @xes fxe5
2...Qxe5 3 M Qf4 4 h5 Qg5 5
Q02 Qh6 6 Qb2 Qg5 7 g3 Qh6 8
a4 bxa3 9 Qxa3 A Qa3-b2-02-d3
b3-b4 +-
3 g5 e4+ 4 Qe2! Qe5
4...d3 5 Qd1!+-
5 M Qrs 6 g4+ Qf4 7 g6 d3+ 8
l e3 9 g7, 1-0, Ivanov-
Sagalchik, Kramatorsk, 1989.
2 2 @xeS QxeS 3 Qg5 b5!
No. l Exfi+? Qxf7 2 Q13 Q16 3
Qe4 g5! (3...Qg5 4 Qxd4 Qg4 5
Qe5!= 4 b5 d3 5 Qxd3 Qe5 6 Qc4
g4 7 c5 QM s Qc6 Qd4!!, 0-1,
Renet-Olafsson, Reykjavik, 1993.
Instead I §c4 §d7 2 QfZ d3 3
Q61 d2+ 4 l Qf6 5 ECS Ed?) 6
§a5 draws.
3 l-O, Ricardi-Sorin, Buenos Aires,
1...§d5 2 332 Qg6 3 EfZ f5 4
Edz Exd2 5 Qxd2 Qg5 6 Qe2 Qg4
7 Qf2 Qf4! 0-1, Balashov-
Tiviakov, St.Petersburg, 1993.
4 1...f4!
1 c5! fixeS (l...bxc5 2 Qd3 A
Exa5+-) 2 Ec bxc5 3 Qd3 e5 4
Qc4 f5 5 QXCS h5 6 b4 axb4 7
Qxb4 f4 8 35 e4 9 Qc3 e3 10 Qd3,
i3; Kuzmin-Petrosian, USSR,
1.
5
After 1...b5? 2 Eb fib 3
axb5+ Qb 4 e4 Qc6 5 e5! fxe5 6
g5 hn 7 f6!, Black had to resign,
Averbakh-Bebchuk, USSR, 1963.
6
No, it is not worth it. 1 £g2? (1
fix 6 n6 2 b5! Qf5 3 Qd4 +-)
1...QeS 2 9.8 b5=, Adams-Xie Jun,
France, 1994.
7
No. 1 4313?? (1 Qf4=) 1.... Qe5
—+,
Bronstein-Timman, Tallinn, 1973.
8
(a) l...§.c5!, transposing to a
drawn rook ending;
(b) 1.963? 2 Exam Ems 3 94:5
Q67 4 Qe3 Qe6 5 Qxd6 Qxd6 6
Qd4 Qe6 7 Q05 h5 8 g5 QfS 9
Qxd5 Qg4 10 Qe5 Qxh4 11 f5,
1995.
No, he cannot.
l...§gxd7?
2 Exd7+ §Xd7 3 fixd7+ Qxd7 4
f4! g4
4...gxf4 5 QB Qe6 6 Qxf4 Q16
7 g3 +-
5 g3 gxh3 6 gxh4 Qe6 7 Qg3
Q16 8 hS! Qg7 9 Qxh3 Qh7 10
Qh4 Qh6 11 b3!, 1-0, Estrin-
Gusev, Moscow, 1963.

10
(a) l...@g5? 2 £f5 @f6 3 @e3
@e6 4 .9166 @xe6 5 @e4 +-,
Petrosian-Larsen, San Antonio,
1972;
(b) 1...5Z)e5 2 @e3 @xg6=.
11
1...g5? (1...£.e6!?) 2 @3d $d
3 g4! (a typical breakthrough)
3...gxh4 4 gt $96 5 @gz @f5 6
f4!, Black resigned, Deak-Horvéth,
Hungary, 1994.
12
There followed 1...£.c4+?? 2
@xc4+ gxc4 3 Wd3+, 1-0, Tal-
Suetin, Sochi, 1973.
Correct was 1...Wf4!=.
l3
1...§d5? (1...@f6 2 Exb6+ @125 3
§c6 Eel 4 b4 fixe3 5 b5 §d3=) 2
Ed $d 3 @213! (going round
the flank) 3...@c5 4 $34 g6 5 h4 h5
(5...g5 6 hxg5 hn 7 g4 +-) 6 g3
@c6 7 b4! (7 @b4? b5 8 @a5? QCS
9 @a6 @134 -+) 7...@c7 8 @b5
QM 9 @c4 @a6 10 @c3! @b7 11
@d4 @c6 12 @xe4 @b5 13 @d4
@xb4 14 e4, 1-0, Ryumin
Ilyin-Zhenevsky, USSR (ch), 1931.
14
l...§x 2+? (l...hxg2 —+) 2 Exg2?
hxg2 3 xg2 a5, 0-1, Valvo-Levitt,
Chicago, 1992;
Necessary was 2 @xh3! Exg4 3
@xg4 a5 4 @f5! @c6 5 e5 a4 6 e6
a3 7 @g6!=.
Solutions 1 95
15
1 @f2? (1 Qg2! §c2+ 2 En
fixf2+ 3 $f figS 4 @g3=)
l...§c2+ 2 Ee2 £c5+ 3 .933 3xe3+
4 @xe3 ExB+ 5 @xfB Exe2 6
@xe2 @g6 7 @d3 @f5 8 @d4 a5,
0-1, Almasi-Ponisch, Budapest,
1994.
16
1....214! 2 £33 @gs 3 Exf4 Exf4
4 3xf4+ @164 5 b4 b6 —+,
Szelaig-Pinski, Poland, 1997.
17
No, it is not possible.
1 936+? @xf6 2 f8+ @xfs 3
exf6 g5! 4 f4 @gB 5 fn @h7 —+,
Adorjan-Chemin, Budapest, 1995.
18
(a) 1,215? 2 @134 tz 3 c5i,
Cruz-Seirawan, Moscow, 1994;
(b) l...§aS+! 2 @b3 3x32 3
@xa2 @d6 4 @b3 @c5 5 @c3 e5 6
@d3 15 7 @c3 e4 (7...h5 8 @b3 f4 9
@c3 f3 10 m3 @134 —+) 8 Qb3 115
9 @c3 f4! 10 gxf4 e3! 11 fxe3 h4
12 f5 saw (12...g3? 13 hxg3 hxg3
14 f6 @d6 15 C5+!) 13 §b4 @eS!
- 14 c5 @f 15 c6 @e6 16 @CS g3
17 hxg3 h3! 18 @b6 h2 l9 c7
@dfl, and the pawn queens with
check!
19
(a) 1...®xe4? 2 fxe4 @d6 3 e5+!
@c5 4 @c7=, Mestel-Chekhov,
Tjentiste, 1975;
(b) l...@d6 2 @xa7 @cs 3 acts
43H -+.
196 Solutions
20
(a) 1....§.c4? 2 £xc4 dxc4 3 64
@b5 4 f4 @cs 5 g4 @b5 6 65! fxeS
7 15!! @c5 8 g5 +-, Heinig-Liebert,
DDR, 1979;
(b) 1...d4+! 2 exd4 @d6=.
21
1 flb! ixb6 2 axb6 axb6 3
@f2 99,6 4 @e3 @d5 5 c4+ @cs 6
g4 b5 7 f5 gf 8 gf @d6 9
@xe4, Solozhenkin-Anchesi,
Reggio Emilia, 1998.
22
1...?65? 2 2xd6+! @xd6 3 QM
@d5 4 b5! ab 5 a6 $06 6 @xe4
b4 7 f4, 1-0, Larsen-Uhlmann, Las
Palmas, 1971. l...@b5!=.
 Solutions 197

ROOK ENDINGS

1 18 @b6 305 19 EhS E08 20 §h6

In the game followed 1...§a1? 2
§d5+ $66 (On 2...?06 follows 3
§g5+) 3 fid4! Efl 4 Ee4+ @d7 5
fixa4, 1-0, Gurevich-Rechlis, Israel,
1989.
He should play l...§f4! 2 @g7
§g4+ 3 @f6 :14+ 4 @g6 @116 5
Ea6+ @e7 6 Ea7+ @e6 7 @g7
Eg4+ 8 $18 @d6! 9 @e8 §e4+ 10
@d8 §f4 11 336+ 96 12 @e7
Exflfl l3 @xfi @b4 with a draw.
2 §e8 §b1+ 7 @a7 §a1+ 8 @b7 Eb]
(a) l Ed6? @e3 2 §e6+ @f2 3
§d6 662 4 366+ 9d] 5 @gS <12 6
$14 EaZ, 0-1, Topalov-Kasparov,
Geneva, 1995;
(b) 1 368+: $13 2 218+ @e2 3
§e8+ @d1 4 gas d2 5 a7=.
3 In the game followed l...§b8?. 2
(a) 1...?1‘4? 2 h4 §a2+ 3 @gl
ECZ 4 h5 305 5 h6 EhS 6 h7 f5 7
Ea7, 1/2-1/2, Leko-Akopian, Ubeda,
1997;
(b) 1...a4 2 M 332+ 3 @g3 a3 4
h5 gal 5 £37 32 6 @g2 @g5, and
the f-pawn advances.
4
(a) 1...@e7 2 §e6+ <$f7! 3 Exd6
2:14 4 eggs @e7 5 366+ $17 6 fies
@f6 7 E115 @g6=; §g4+ $18 3 mm EdB? (3...?g8 4
(b) 1...?e7 2 Ee6+ @d7? 3 §f6
@e7 4 315 §c3 5 h4 §b3 6 h5 §a3
7 h6 §a8 8 h7 EhS 9 EhS @f6 10
@g3 @g7 11 @g4 @g6 12 Eh] @f6
13 QM @g6 14 964 @f6 15 @d4
@g7 (15...@e7 16 @c4 @d7 17
gal! +-) 16 964 268+ 17 @b5 @118
+-, Lesiege-Ivanov, Bermuda, 1995.
5
(a) 1 §b6?? §a4 2 <$h3 fixg4 3
Exf6 §h4+!!, 0-1, Gulko-Gurevich,
Parsipanny, 1996;
(b) 1 EM QM 2 §c6=.
6
1 @c6! (1 c6? e5 2 fxe5 fxe5 3
@b6 e4 4 §d7+ @f6! 5 §d8 e3 6
9 $08 Ee1!=, Short-Topalov, Nov-
gorod, 1996) l...e5 2 fxeS fxeS 3
@d5 $16 (3...e4 4 §d4 +—) 4 c6 e4
5 Ef2+ $97 6 EeZ §d1+ 7 @xe4
@d6 8 ECZ +-.
7
Correct was 1...§c8! 2 c3 @f6 TL.
§b3 Exb3 After 2.308 3 C3 d4 4
Eb7+ <$f6 Black holds the rook
ending without problems. 3 axb3
@f6 4 @gl @es 5 @fZ <$e4 6 QeZ
@d4 7 @d2 aS_(7...@e4=) 8 @e2 e5
9 @dz e4? (9965:) 10 c3+ ecs
11 @e3 @c6 12 @d4 @b5 13 h3 h6
l4 h4 @c6 15 c4 1-0, Kupreichik-
Zheliandinov, Gomel, 1968.
8
1 @f6 @gs (INQCS 2 e6!+—) 2
§a8+ @h7 5 Ef8 Edl 6 $167+ @gS
7 §a7 Efl+ 8 @e6 3e11, with a
theoretical draw) 4 fia7? (4 EM!
@gS 5 @e7 EdS 6 §g4+ @h7 7
@xfi +-) 4...:118? (4...§g8 5 am
Ed] 6 Ea7, again draws) S $117+
@gB 6 §g7+ 918 7 e6 §b6 8 §a7
I 98 Solutions
Eb8 9 Eh7 Qg8 10 Eg7+ 1-0,
Kirov-Kosic', Informator 59.
9 (2...Exh4 3 Qd3 Et 4 Qc4 f5 5
1 Eb7!! (1 Qg6 Eg3+ A 2...Eb3)
1...:b 2 Qg6 Q18 3 M Ebl 4
Eb8+ 1-0, Levenfish-Lisitsin,
Moscow, 1936.
10
1 f5!!—-a typical breakthrough
——1...exf5 2 e6! fxe6 3 n6 A
h5=.
The game continued 1 Qf6? Qb5
2 Eal a4 3 f5 ef 4 e6 fxe6 5
n6 f4 6 h5 13 7 h6 e5! 8 Eel (8
$15 Eh7 9 Ehl 12 -+) 8...a3 9
Exe5+ Qc4 10 Eel a2 11 M Ea8
(11...Exh7? 12 Qxh7 Qb3 13 £11:
or 11.12 12 :11 Qd3 13 Eal!) 12
Qg7 (12 Eal f2 13 Exa2 Exa2 14
h8=w Ea6! -+) 12...12 13 Eal
Qb3 l4 Efl al=§ 15 Exal Exal
16 h8=§ Egl+ 0-1, Lasker-Leven-
fish, Moscow, 1935.
ll
1 Qb6? (l Ee3! Ee8 2 Qc6 +-)
l...e4 2 b5 Ee8 3 Qa7 e3 4 Ehl e2
5 Eel Qxh6 6 b6 Qg5 7 b7 Qg4!,
1/2-1/2, Bologan-Kramnik, Germany,
1994.
12
1 117? (1 Ee8 m7 2 Qc3 $12 3
@113 $117+ 4 Qc2 e2 5 mm Qel 6
Ef6=; 1 E18!?) l...Ef7!, 0-1,
Lautier—Yermolinsky, Tilburg,
1993.
13
(a) 1 g5? fxg5 2 hxg5 h4 3 d6
Ea2+! 4 Q13 h3 5 Qg3 h2 6 Ea8+
Qd7 7 Eh8 Qxd6 8 Exh2 a4
1/2-1/2, Karpov-Yudasin, Madrid,
1992;
(b) 1 gt £113 2 Ea 15!
QcS fxe4 6 Qc6 +-) 3 ef Exh4 4
Qe3 Qe7! :, Yudasin) 5 E36!
Et 6 Qe4 +-.
14
1...Ed2? (1...Eb2!=) 2 b4! Ed4
(2...Eb2 3 Ee4 +-) 3 3133 @127 4 b5
Qd8 5 b6 Qc8 6 Ec3+ Qb7 7 E07
Qxb6 8 E7117 Ed3+ 9 s Ed4 10
Qf3 Ed3+ 11 Qe4 Eg3 12 f6
Exg4+ l3 Qf5 Egl, 1-0, Timman-
Van Wely, Amsterdam, 1995.
15
(a) 1 g4? Edl! 2 Ec8 Ebl 3 E18
Exb4 4 Exf7 E04 5 Exh7 b4 6 Ehl
b3 7 Efl b2 8 Ebl Eb4 9 f5 Q07,
l/2-1/2, Kasparov-Short, Novgorod,
1994;
(b) 1 Ec8! f6 2 gxf6 Exf6 3 g4
Ec6 4 Eb8+ Q36 5 f5 h5 9 Eg8! +-.
16
(a) 1 Ed5+? f5!! 2 Exf5+ Qh6
(b) 1 Edl a2 2 Eal Ea3+ 3 Qg2!
Ea4 (3...n4 4 h6=) 4 Q13=,
Permiakov-Petkevich, Riga, 1988.
17
(a) 1 @142 b4 2 Qd3 @135 3 <3(12
b3 4 E01 Ec4! -+, Piampuu-
Lancava, Leningrad, 1992;
(b) 1 E32! b4 2 Ec2=.
18
(a) 1 d6? Qg7! 2 b5 Qf6 3 d7+
$67 4 Ed6 @d8 5 Ef6 Qxd7 6
Exf7 Qe6 7 Exh7 Eb3 8 Eb7 EbZ!

9 b6 e410 §b8 e311@f1 @f6 12
M @g7 13 g4 @h7 14 h5 gt 15
gxh5 @g7 16 §b7+ $h8 l7 h6 QgS
18 §g7+ @118 19 b7 3131+! 20 @122
§b2l 21 @xe3 §b3+ 22 @d4
§b4+=, Yudasin-Kramnik, Candi-
dates (m), 1994;
(b) 1 fies! @g7 2 b6 Em 3 d6 +-.
19
(a) 1...§xh4? 2 En §c3 3
fidSH @d3 4 c §h1+ 5 @b2 Eel
(5...@e3 6 $02 §h2+ 7 l! d3 8
Ee5=) 6 EdS, Draw, Dreev-
Beliavsky, Odessa, 1989;
(b) l...§gl+ 2 @e2 (2 @d2 E 2+
3 @el g4 4 @fl Eh2! 5 Exg4 £03
6 @gl §a2 A d4-d3 —+) 2...d3+ 3
@d2 §g2+ 4 @dl g4 5 h5 @c3 6
§c8+ @d4 7 h6 §h2 @e3 9 398+
QB 10 §h8 @g3 11 h7 Eh“, and
there is no defence against the
march of the g4 pawn.
20
1 @f6! §c6+ 2 @es §c8 (2...Ec5
3 @d6 §C8 4 Eel+-) 3 Ega! @117 4
Ec6! EaS 5 @f6 A fie6, Cvitan.
21
1 g6+ @fs 2 :18 $95 3 f6!! (3
EH? §a1=) 3...Exf6 4 EH! @e6
(4...§f5+ 5 @g4 £16 6 @gS +-) 5
Exg7 fifl 6 3217 +-, Bayer-Polasek,
Luxemburg, 1986.
22
1...§a5! (1...@g7? 2 a5 @f6 3
@f4 $66 4 $64! h5 5 @f4! @f6 6
@g3 @g5 7 §a4+-, Holmov-Hasin,
Minsk, 1962) 2 an @g7 3 m4 115
4 @d4 114 5 @c4 h3 6 @b4 §h5 7
Ea] 112 8 3111 $16 9 35 @f5 10 a6
§h7!=.
Solutions 199
23
1 @c4 Ea4+ 2 @b3 333+ 3 @c2
§c3 4 <$b2! +-, Salwe.
24
(a) 1...?e3? 2 @g3 gal 3 fieS! (3
a7? §g1+ 4 @hZ @flfl 3m®d4 4
@f4 Exa6 5 §d8+! $05 6 @g5, 1-0,
Ulibin-Nevednichi, USSR, 1986;
(b) l...@xf3!! 2 a7 EaB!!=,
Gelfand.
25
(a) 1...h3? 2 §g3 +-, Bykova-
Rubzova, m, 1958;
(b) Luéhl! 2 @g4 h3 3 @xh3
fixg7!, stalemate.
26
(a) 1 a8=y Ea 2 Ea @f5 3
§h8 @g4 4 @c5 h3 5 @d4 @g3 6
$63 @g2!=, Dammes-Sosonko,
Leningrad, 1963;
(b) 1 @117: 317+ 2 @216: am
(2mm; 3 EbS! h3 4 Ea3! +-) 3
@b5 21‘s 4 as=¥ 31:38 5 311218
$15 6 @c4 h3 7 @d3 +-.
27
(a) 1 @f4? @134 2 $65 @d3! 3
@f6 @e4 4 @g7 §a8 5 h7 @f5 6 g6
@g5=, Gutman-Tseitlin, Riga, 1976;
(b) 1 §g4 91:4 2 @115 @ds 3 g6
+-.
28
(a) 1...?g2? 2 §f6 f3 3 g6 fiaS 4
if], l/2-1/2, Morovic-Agdestein,
Havana, 1998;
(b) 1...§al+ 2 @d2 fig] 3 g6
Eg3!! A 4...@g2 —+.
200 Solutions
29
(a) 1...f5? 2 M! a6 3 e3 e4+
(3...a5 4 64 f4 5 gxf4 exf4 6 65) 4
fxe4 fxe4+ 5 @xe4 Exc4+ 6 Exc4
@7104 7 965:, Wirthensohn-
Mikhalchishin, Lenk, 1998;
(b) l...h4! 2 gxh4 f5 3 e3 36 4 h5
(4 §b3 2x04 5 Exb6 64+!) 4...gxh5
5 M a5 —+.
30
(a) 1 h6? @d3 2 $13 fih3+ 3 @g4
Ehl 4 @f5 62 5 @f6 §h3l 6 @W
Ef3+ 7 @g6 Ee3=, Piket-Sokolov,
Dortmund, 1995;
(b) l @g4! @d3 2 g6 e2 3 @gS
el=¥ 4 Exel Exel 5 g7 +-.
31
(a) 1...Eb3? 2 §g7+1 @118 3 £137
Exb4 4 @g6, 1-0, Salov-Yudasin,
St.Petersburg, 1997;
(b) LEBH 2 mes 3133 3 @d6
Exb4 4 e5 §d4+ 5 @e7 E2143! 6
Em @g7 7 e6 gas 8 $d6 ms 9
@d7 @g7=.
32
Only 1 §a3!! makes a draw.
(a) 1...gxh4 2 §c3+ @d5 3 Ecz
@e4 4 @hS 913 5 §c3+ Q92 6
§c2+ @d3 7 §h2=;
(b) l...g4 (1...@d5 2 §a5+ @e4 3
§a4+) 2 §c3+ (2 @g5? @c5!)
zugds 3 Ed m4 4 @gs $13 5
§c3+ @gz 6 @xg4 Egl 7 £33
al=w 8 filial Exal 9 115:.
33
(a) LEM? 2 ECS fixa7 3 §c7+!
3x07 4 @xc7 @h7 5 @d7, 1-0,
Benké-Gereben, Budapest, 1951;
(b) l...§dl+! 2 @e6 Eal 3 §d8
Exa7 4 Ed7+ mm 5 @xd7 9117:.
34
(a) 1 Exam §e3+ 2 s £5! 3
gf @f6 4 Ea6 @f 5 @c7 g4 6
d7 E87 7 @d6 fixd7+ 8 @xd7 g3=,
Olafsson—Tal, Portoroz, 1958;
(b) 1 d7 §e3+ 2 s a3 3 EaB!
+'.
35
(a) l @gl? §g3 2 $12 $66 3 EaS
Exg4 4 §a4 @fs 5 3114 §e4 6
2135+ $14 7 §b6 f5 8 §d6 g4 9
§d8 g3+ 10 @fl h3, 0-1, Ioseliani-
Onischuk, Luceme, 1997;
(b) 1 ms: §g3 2 d5 Exg4 3 d6+
@e6 4 fidS! @d7 5 £15 §f4 6 EH4
gxf4 7 @h3 $xd6 8 @xh4 @e5 9
@h5!=.
36
(a) 1.2938? 2 g4 @ds 3 @g6 @128
4 @f6 @d8 5 §g7 §b5 6 Exg5
n5 a5 8 @h6 a4 9 g5 a3 10 g6 a2
11 g7 a1=w 12 g8=w+ 917 13
yg7+, 1-0, Lein-Suetin, Bad Wild-
bad, 1997;
(b) 1...§a3!! 2 @n (2 g4 Exh3
3 @n §g3 4 @f6 @g8 5 g5 26+
6 @g6 Ef8!=) 2...§g3+ 3 @115 @gs:
4 EH6 Exg2=.
37
1 Ed! @e6 2 §c3!! a3 3 17 21:17
4 293+ @ds 5 9x17 a2 6 Ea3 1-0,
Konstantinopolsky-Fridman, Lvov,
1940.
38
(a) 1...§b1? 2 @h5 fig] 3 g5 fxg5
4 f5! QfB 5 f6, 1-0, Zaitsev-
Hiibner, Busum, 1969;

(b) l...§b4! 2 f5 fibl 3 @115 Egl
4 @g6 fixg4+ 5 @xf6 §g1=,
Larsen.
39
(a) 1...?h4 2 ECS §h7+ 3 @e6
Exd7 4 @xd7 g4 5 $66 g3 6 @f5
g2 7 $14 +-;
(b) l...@g4 2 @115 mm 3 99,6
§d8l 4 §d5 (4 §c8 Exd7 5 @xd7
@fS! 6 §c5+ @f4 6 @e6 g4 7 §c4+
@gS 8 @e5 g3=) 4...@f4! 5 §f5+
@g4 6 EH @hB 7 @fs g4 8 @f4 g3
8 $13 @h4, 1/2-1/2, Keres-Mikenas,
Stockholm, 1937.
40
1...d3!! 2 cxd3 (2 @xe4 dxc2)
2...§c4!! 3 bxc4 c2 4 9x“ cl=¥+
5 @e4 gdll, 0-1, Munios-Salazar,
Dubai, 1986.
41
(a) 1 @fl? b5 2 ab (2 a5
§a3ll=) 2...axb5 3 @e2 $a31! 4
@d2 b4 5 $02 §c8+l 6 @d2 EhS,
Draw, Levy-Peev, Cienfuegos,
1973.
Solutions 201
(b) l §h4+l @xb3 2 as +-.
42
(a) 1...§b7? 2 §g4l @eS (2...b3
§a4 §b8 4 §f4 @e8 5 d7+ +-)
m4 Ebs 4 Ec7 b3 5 fixg7 @fs
317+ @gS 7 d7 b2 8 Efl @g7
gbll §b6+ 10 $67 §b7 11 $61
1-0, Gelfand-Lerner, Norilsk, 1986
(b) Lun! 2 Exg7 Ee2+ 3 @d
b3 4 Eb7 b2 5 g4 @c8 6 Eb
@d7=.
43
It seems that White has success
fully completed a difficult defencc
and Black must agree a draw afte
1...a3 2 @g6 §g1+ 3 @h6 Ehl-
with a perpetual check or 1...Ef1
fixb7 EfS 3 Ea7. But, as it turn
out, there is a third possibility
There followed the disheartenin
1...§b6!!, and, playing on throug
inertia, 2 @g6 c5+ 3 @f5 cxd4
Ed dxe3, White acknowledge
defeat, Lopyshnoi-Dreev, Maiko;
1998.
202 Solutions

VARIOUS

l 5

(a) 1 @153? tabs” 2 @d6+ (2 (a) 1....9.e6? 2 @e4! A 3 035 +-,

@3c @306 3 @d3 @d4+ =) 2...?e6
3 @e4 5306 4 gxf4 53d4+ 5 @c3
@xf3 6 @3c @155 7 fxe5 @xeS 8
@d3 55x04! 9 @xc4
Sale-Mikhalchishin, Sibenik, 1990;
(b) 1 @d2! @b6 2 @d3 @d7 3
Q6 @e6 4 @b3 @b8 5 gxf4! @c6
6 55c @f5 7 @d3 ®d4+ 8 @c3
43x13 9 fxe5 £8t 10 c5 +-.
2 1961.
(a) l....@.xg2! 2 ab @e6 3 @xfl
@xfl 4 @n 99,6 5 96 @d5 6 b6
@c6 7 b7 @xb7 8 $94 4 -+;
(b) 1....@.d3? 2 ab e7 3 55c8+
@d7=, Van der Wiel - Larsen,
Brussels, 1987.
Vukié-Vujosevié, iacka Banja,
1998;
(b) 1....9.b1!=.
@g4=,
6
1 @3a sz 2 QM! (2 b4
@a3!=) 2....Qxa5 3 b4 £b6!! 4 a5
.912 5 a6 (5 b5 @a3=) 5....@.a7 6
@b5 96:, Parma-Gligorié, Bled,
7
l...g5+! 2 hn M 3 @c6 58g6+ 4
@e4 h3 —+ Markowski-Gdanski,
Ksiaz, 1998.
8

3 He should choose l...f6! 2 mm

@hS 3 yg7 h6 -+.
lynfl®fl@%2@fi@y3 In the 'ame White played l...f5?
g4 hx 4 4 @xg4 @xc4!) 1...5n4 2 2 @e7! §d2+ 3 @g3 @eH 4 @h3
M4 f6 (2.3% 3 e5+!! taxes 4 fih1+ 5 @g3 fig1+ 6 @h3 f4 7
@xeS dxe5 5 $64, and the awn yf8+ @hS 8 @xf4, l/2-1/2,
ending is won) 3 e5+ dxeS 4 1:95 Mikhalchishin-Kavcié, Slovenia,
@d6 5 @xg6 @d+ 6 cd c4 7 1997.
@e4 cxb3 8 @es, 1-0, Razuvaev-
Ostojié, Berlin, 1988. 9

4 (a) l...gcl? 2 e6 fixf2+ 3

@113 Wm 4 @114 @hH 5 @113! +-,
(a) 1 .933? @c7 2 QCS g2 3 ®d4 Polugaevsky-Bronstein, Tallinn,
$06 4 @e4 @b5! 5 QB $04 6 1964;
@n QdS 7 £f4 @e4! intending (b) 1...¥f5!=.
...<§f5, f6=, Kudrin-Cebalo, Beme, 10
1988;
(b) l 9.35” g2 2 §b6 @b8 3 $06 l...e5? (l...fxe5 2 yt
QCS 4 @d6 +-. yxa5=) 2 9g3+! $17 (2...Wxg3+ 3

n3 +-) 3 WES fxeS 4 Q13 Qe7 5
Qe4 Q66 6 f3 Qf6 7 f4 +-.
11
1...Qc8? 2 yg4 and 3 Wdla
Ehlvest-Topalov, Novgorod, 1994
l...Qc7! 2 We7+ Qc6 3 ge8+
Qc5 4 Mn Qd4 —+.
12
1 117: Exh7+ 2 Qg6 Em 3 f5+ (5...@f6 6 Qxf5 @3t 7 QgS +-) 6
@d6 4 @116 £111 5 5517+! Qc7 6
@gS Qd6 7 f7 fins 8 Qg7 gas 9
f6! (9 f8=y+i2 ExfB 10 Qxf8
Qe5=) 9...Qe5 10 5.13M §a7 ll
Qg6, 1-0, Krumpachnik-Maksimen-
ko,Pnn,1998.
l3
1 Exes ExeS 2 g3, Black re-
signed, But he could have saved
himself by 2...Qg6! 3 ExeS Qh5
and after 4...f4—draw.
14
The game continued 1 $165+?
ef 2 Qf Qxh6 3 Qe6 @b8 4
Qd6 @c6 5 Qc7 $.38! Black’s only
chance, which White underesti-
mated. Losing is 5...¢3xe7 6 Qxb7
Qxh5 7 Qxa6 @c6 8 Qb +-. 6
Solutions 203
QCS Qt 7 Qb6 Q 4 8 Qxa6
Q13 9 $b Qxe3 10 b6 5mm,
1/2-1/2, Chiburdanidze-Maric, Bel-
grade, 1996.
But better is 1 2&2 $38
(1...Qxh6 2 fidl 3c6 3 £g5+ Qh7
4 h6 A 5 £h5 +-) 2 e4!! dxe4
(2...fxe4 3 fldl Qxh6 4 £g4 9&6 5
£g5+ @117 6 2xe6 +-) 3 m3 ads
4 .9.d ed 5 £d6! (5 3g5?
@b8! 6 Q65 523c6+ 7 Qd @xb4+
8 Q66 @c6 9 d5 @xa5=) 5...Qxh6
Qf Qt 7 Qe6 e3 8 £f4! e2 9
QdZ ($318+ 10 Qd Qg4 11 Qc6
$13 12 d5 @132 13 @b7! +-.
15
(a) l...e5? 2 3x06 exd4 3 exd4
Ed3+ 4 Qf4 Edl 5 Qe4 Qg6 6 d5
§e1+ 7 £33 Qg7 8 d6 Edi 9 9.1%,
and at last he resigned. Osterman-
Mikhalchishin, Bled, 1995;
(b) l...e6! 2 Exc6 £d8! with the
idea QdS-aS-el with a sufficient
defence.
16
No, because of 2....9.a4!! 3 EeZ (3
h5 £xb3 4 h6 .2.c 5 h7 §b3+ 6
Q62 g2 -+) 3....9.xb3 4 3d3
Exe2+ 5 Qe fid1+!!, 0-1,
Timman-Salov, Amsterdam, 1991.
204 Solutions

CHOOSE THE BEST CONTINUATION

IN ROOK ENDINGS

l Alatortsev-Chekhover, USSR, (b) INQhS? 2 3xf3 3h2 3 @gl

1937: 3xh3 4 @gZ +—-.
(a) 1 @b5? 3xa7 2 3xa7 $xf2=; (c) l...@g5!! 2 3xf3 3h2 3 h4+
(b) 1 @d5? @f‘l 2 $66 3xa7! 3 @g4=.
3xa7 @f=;
(c) 1 @CS! 368+ (1...3xa7 2 3 Cuartas—Zuidema, Skopje,
3xa7 @xfl 3 @d4 +—) 2 @b6 368 3 1972:
@C6! (3 @b7 367+ =) 3...?fl (a) 1...f3? 2 67 f2 3 3d2 A 4
(3...3h8 4 6137 3117+ 5 @136 3h8 6 3b2+-.
362 +—) 4 @b7 367+ 5 @b6 368 6 (b) 1...361? 2 67 f3 3 3d8+ Q67
362! @gZ 7 @b7 367+ 8 @b8! 4Qb7+—.
368+ 9 308. (c) 1...3cl! 2 c7 364” 3 3d5
$67 4 @b7 $66 5 3b5 @d7 6
2 Karner-Renter, USSR, 1955: 3d5+ @6673 Karpov.
(a) 1...3h2? 2 M QhS (2...f2 3
3f3+-) 3 3a5+! @g4 4 3g5+ @h3
5 h5 3a2 6 h6 3a6 7 h7, 1-0
 Index of Players and Composers
Abramovié-Nikolié 85 Beliavsky-Short 174
Adams-Almési 148 Beliavsky-Spraggett 59
Adams-Lautier 17 Beliavsky-Sveshnikov 17
Adams-Lutz 11 Beliavsky-Tratar 54
Adams-Xie Jun 31 Benké-Gereben 106
Adorj an-Chemin 33 Bogoljubow—Rubinstein 93
Adorj an-Morozevich 176 Bojkovié-Kakhiani 80
Akopian—Almési 90 Bologan-Kramnik 103
Alatortsev-Chekhover 97 Botvinnik 25
Alekhine-Bogoljubow 110 Botvinnik-Balashov 124
Alexandria-Marié 53 Botvinnik—Donner 184
Almési-Portisch 32 Botvinnik-Kan 113
Alterman-Chernin 142 Botvinnik-Rabinovich 112
Anastasian—Khalifman 29 Botvinnik;Simagin 41
Anastasian—Romanishin 19 Branicki-Sefc 96
Andersson-Hiibner 68 Bronstein-Rantauen 118
Arbakov-Gurevich 76 Bronstein-Timman 31
Averbakh-Bebchuk 31 Brunner-Korchnoi 97
Azmaiparashvili-Kupreichik 81 Bykova—Rubzova 105
Azmaiparashvili-Novopashin 44 Capablanca-Tartakower 76
Bagirov-Berzinsh 60 Chaunin-Friedman 72
Bagirov-Kraidman 57 Chekhov-Karsa 129
Bagirov-Veingold 64 Chemin-Georgiev 176
Balashov-Korchnoi 119 Chiburdanidze-Galliamova 92
Balashov-Tiviakov 30 Chiburdanidze-Marié 190
Balashov-Ulibin 94 Cruz-Seirawan 19
Balashov-Vaganian 145 Cruz-Seirawan 33
Ballon-Mikhalchishin 184 Cuartas-Zuidema 97
Bareev-Faragé 139 Cvitan 104
Barle-Mikhalchishin 89 Dammes-Sosonko 105
Barlov-Abramovié 45 Dan-Pytel 168
Barlov-Schfissler 78 Dao Thien Hai-Ivanchuk 120
Barlov-Seirawan 128 Dautov-Alterman 65
Bayer-Polasek 104 Deék-Horvéth 32
Beliavsky-Azmaiparashvili 95 Degraeve-Hansen 26
Beliavsky-Dolmatov 163 Dokhoian-Shirov 177
Beliavsky-Gelfand 131 Drasko-Vratonjié 13
Beliavsky-Geller 170 Dreev-Anastasian 26
Beliavsky-Hodgson 61 Dreev-Beliavsky 104
Beliavsky-Kupreichik 174 Ehlvest-Polugaevsky 65
Beliavsky-Mikhalchishin 132 Ehlvest—Rausis 136
Beliavsky-Miles 154 Ehlvest—Shirov 10
Beliavsky-Neverov 143 Ehlvest-Topalov 190
206 Index

EIiskases-Levenfish 81 Hfibner-Polgér 89
Emma-Riemersma 56 Hfibner-Spassky 116
Estrin-Gusev 31 Hulak-Beliavsky 173
Euwe-Alekhine 59 Huzman-Mikhalchishin 61
Evreinov 26 Ilincié-Abramovié 52
Faragé-Csom 172 Illivitsky-Taimanov 84
Fedotov-Arkhipov 34 Ilyin-Zhenevsky-Botvinnik 27
Fercec-Cvitan 131 Ioseliani-Onischuk 107
Fercec-Mikhalchishin 83 Ivanchuk-Eingom 51
Filipescu-Citron 27 Ivanchuk-Kasimdzhanov 20
Filipov-Kopatsny 90 Ivanchuk-Kramnik 164
Finkel-Mikhalchishin 20 Ivanchuk-Lautier 88
Fischer-Bisguier 23 Ivanov-Sagalchik 30
Fischer-Geller 94 Janvarev-Schcherbakov 25
Fischer-Letelier 21 Kamsky-Karpov 59
Fischer-Larsen 23 Kamsky-Cvitan 144
Flohr-Vidmar 18 Kan-Keres 115
Formanek-Mikhalchishin 170 Kamer-Renter 97
Gausel-Agdestein 142 Karpov-Hort 145
Gelfand 105 Karpov-Kasparov
Gelfand-Lautier 131 Karpov-Kasparov 157
Gelfand-Lemer 108 Karpov-Timman 185
Gelfand-Shirov 135 Karpov-Yudasin 103
Genba—Irzhanov 143 Karpov-Yusupov 117
Georgiev-Khalifman 114 Kasparov~Andersson 127
Gligorié—Fischer 22 Kasparov-Short 103
Godena-Lalié 44 Keller-Mikenas 66
Greenfeld-Golod 11 Keres-Aronin 183
Gretarson-Magerramov 133 Keres-Mikenas 107
Grfinberg-Brunner 85 Keres-Szabé 166
Gufeld-Grigorian 39 Kirov-Kosié 102
Gufeld-Minev 183 Klovan-Elistratov 29
Guliev-Tukmakov 25 Kochiev-Lemer 43
Gulko-Gurevich 102 Konopka—Shcherbakov 67
Gulko-Sveshnikov 144 Konstantinopolsky~Fridman 107
Gurevich-Andersson 134 Korchnoi-Kengis 99
Gurevich-Rechlis 101 Korchnoi-Ljubojevié 151
Guseinov-Beliavsky 70 Korchnoi-Pinter 125
Gutman—Tseitlin 105 Korolev-Palm 185
Havsky-Yuferov 27 Kovacevié-Tosié 114
Hector-Speelman Kozakov-Asparuhov
Heinig-Liebert 33 Kozlov-Mikhalchishin 77
Hellers-Eingom 12 Kozul-Mikhalchishin 87
Herrera-Vasquez 57 Kozul-Nikolié 68
Hertneck-Narciso 151 Kramnik-Beliavsky 79
Hoch 28 Kramnik-Lautier 17
Hai-Mikhalchishin 182 Krasenkov-Iskusnik 87
Holmov-Hasin 104 Kremenetsky—Razuvaev 37
Holmov-Timoschenko 74 Krumpachnik-Maksimenko 190
Horvéth- Sherzer 177 Krumpachnik-Polak
Hiibner-Ftéénik 127 Kudrin-Cebalo 188
 Index 207

Kupreichik—Mikhalchishin 14 Miles-Kindermann 167

Kupreichik-Zheliandinov 102 Miles-Van der Sterren 160
Kuzmin-Petrosian 19 Milov-Pelletier 142
Kuzmin-Petrosian 30 Mitrofanov 110
Larsen 107 Mokry-Pribyl 73
Larsen-Browne 76 Morovic-Agdestein 105
Larsen-Uhlmann 33 Morovic-Yusupov 154
Lastin-Cvitan 144 Morozevich-Balashov 129
Lasker-Levenfish 102 Mortensen-Plueg 29
Lasker-Marshall 180 Munios-Salazar 107
Lautier-Yermolinsky 103 Natapov-Schuravlov
Lein-Suetin 107 Neverov-Dreev 124
Leké-Akopian 101 Nezhmctdinov-Luik 123
Lemer-Dorfman 67 Nezhmctdinov-Romanovsky l 2 1
Lesiége-Ivanov 101 Nikolaevsky-Gufeld 43
Levenfish-Lisitsin 102 Nikolié-Ftéénik 63
Levy-Peev 108 Nikolié-Movsesian 83
Lilienthal-Smyslov 77 Nikolié-Pofiisch 138
Ljuboj evié-Ivanchuk 116 Nimzowitsch-Chigorin 27
Lombardy-Fischer 23 Novikov-Beliavsky 64
Lopyshnoi-Dreev 108 Novikov-Kaidanov 162
Lputian—Tukmakov 122 Novikov-Lalié 58
Maciej a-Grabarczak 8 Olafsson-Tal 106
Madsen-Hansen 60 Oll-Benjamin 18
Magerramov-Kohlmeyer 139 Ostenstad-Kuzmin 11
Manukovsky-Pustovalov 109 Osterman-Mikhalchishin 190
Marié-McNab 86 Owen-Morphy 78
Marié-Zaitseva 52 Padevsky-Latinov 26
Markovié-Ivanovié 9 Panno-Donner 137
Markowski-Gdanski 189 Papendieck-Geiger 29
Marosi-Yudasin 186 Parma-Gligorié 189
Marshall-Alekhine 183 Pelletier-Arencibia 147
Martinovié-Mikhalchishin 179 Pelletier-Rozentalis 82
Martinovié-Yudasin 114 Felling-Miles 137
Matlak-Tseshkovsky 13 Permiakov-Petkevich 103
Matulovic-Uitumen 49 Petrosian-Larsen 31
Mednis—Gurevich 71 Petrosian-Rashkovsky 175
Mestel-Chekhov 33 Petrosian-Tal 169
Mikhailov-Volchok 141 Piampuu-Lancava 104
Mikhalchishin-Azmaiparashvili 1 10 Piket-Sokolov 106
Mikhalohishin-Beliavsky 119 Plaskett-Rowson 10
Mikhalchishin-Cvetkovié 187 Polgér-Smyslov 27
Mikhalchishin-Eslon 66 Polnareva-Akhsharumova 45
Mikhalchishin-Gufeld 180 Polugaevsky-Bronstein 189
Mikhalchishin—Kavcié 189 Polugaevsky-Parma 73
Mikhalchishin-Khmelnitsky 71 Ponomariov-Plaskett 169
Mikhalchishin-Nestorovich 182 Portisch-Kramnik 118
Mikhalchishin—Stangel 88 Portisch-Pietzsch 92
Mikhalchishin—Savchenko 151 Pofiisch—Ribli 112
Mikhalchiishin-Sveshnikov 150 Psakhis-Bé'msch 115
Miles-Adorjan 165 Rausis-Faragé 137
208 Index

RazuvaeV—Kirov Taimanov-Chekhov
RazuvaeV—Ostojic Taimanov-Gligoric
Renet— 01 afsson Tal—Balashov
Reti—Bogoljubow Tal— Suetin
Ribli—Hertneck Tavadian—Tseshkovsky
Ribli—Mikh alchi shin Tiets-Forsberg
Ricardi—Sorin Tikhomirova—Morozova
Rittner—Bernstein Timman—Ree
Rom anishin—Markowski Timman—Salov
Rom anishin—Nunn Timman—Van Wely
Romanishin—Rodriguez Toothill—Heemsoth
Romanovsky— Stahlberg Topalov-Beliavsky
RossolimO—Fischer Topalov-Kasparov
RozentaliS—Christiansen Torre—Portisch
Rublevsky— Sh ariazdinov Tosic-Gyimesi
Rufen acht—Orseth Trabattoni—Barlov
Ryumin—Ilyin—Zhenevsky Uhlmann—Robatsch
Sajtar—Benko Ulibin—Nevednichi
SakaeV—Novikov Vaganian—Portisch
Sale—Mikhalchi shin Vaganian—Schlosser
Salov-Kam sky Vaganian—Smirin
Salov-Khalifman Valvo- Levitt
Salov-Malaniuk Van der Doel—Klovan
Salov-Yudasin Van der Sterren—Douven
Salwe Van der Wiel—Larsen
Schandorff—Speelman Van Laatum—Mikhalchishin
Schlechter—Lasker Van Wely—Adams
Schmittdiel—Mikh alchishin Vasiukov-Timoschenko
Sermek—Hulak Vaulin—Voikhovsky
Shirov-Kramnik Vladimirov-Novopashin
Shirov- Lautier Vladimirov-Rashkovsky
Shirov-Morozevich Vujala—Smith
Shirov- Timm an Vukic-Pietzsch
Shirov-Van Wely Vukic-Vujosevic
Short— Kasparov Vukovic-Eingorn
Short— Top alov Vyzhmanavin—Chiburdanidze
Smagin—Mikhalchishin Vyzhmanavin—Lerner
Sm agin—Naumkin Ward—Baburin
Smy slov- Epishin Wells-Hector
Sokolov-Khalifman Wirthensohn—Mikhalchishin
Solozhenkin—Anchesi Yates-Alekhine
Sorin—Alterm an Yermolinsky—Ivanov
Speelman—Saltaev Yermolinsky—Seirawan
Stein—Bobotsov Yudasin—Kramnik
Suba—Chiburdanidze Yudasin—Osnos
Sulipa—Gricak YurtaeV—Temirbaev
Sveshnikov-Sokolov Yusupov-Gerusel
Svidler—Lobron Zaitsev-Hiibner
Sz abo-Keres Zaichik—Sorin
Szabo-Penrose Zinar
Szelaig—Pinski Zotkin—Kudrin