Table of Contents

CopyRight Information
A Study Guide

Practical examples of typical simplified positions

Basic Theoretical Positions
Theoretical Summary
Test Yourself
Piece Coordination
Active King
Active Rook
Choosing direction for the king
Power of connected passed pawns
Evaluating Consequences of Exchanges
Passed pawns must be pushed

About the Author

Symbols and Abbreviations Used in the Book

Swipe left for next chapter

2
 Rook Endgames - Study Guide

By Roman Jiganchine

The right of Roman Jiganchine to be identified as the author of this work has been asserted in
accordance with the Copyrights, Designs and Patents Act 1988. All rights reserved. No part of this
publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means
(electronic, electrostatic, magnetic tape, photocopying, recording or otherwise), without prior
permission from the author.

Jiganchine, Roman. Rook Endgames - Study Guide. Electronic Edition.

For questions, comments, or more information on this book, contact: Roman Jiganchine at
roman.jiganchine@outlook.com.

Copyright © 2015 by Roman Jiganchine.

3
A Study Guide

This ebook is a follow up to my endgame ebooks Bishop endgames - Do the opposites Attract? and
Rook and Knight Endgames - Strategy and Tactics. The idea behind this series is the approach that
Nimzowitsch called «a radioactive method» - selecting a rather narrow topic, and by learning a lot
about it, improve one’s overall understanding of chess.

This book is dedicated to rook endgames, but the goal is to provide a more broad illustration for how a
chess player should go about studying any type of endgame. The steps that I found effective and that a
lot of other chess players and coaches recommend are as follows:
1. Identify and understand most common general principles guiding a specific type of endgames
2. Find a few important theoretical positions that should be well studied and memorized
3. Study your own games related to this type of endgames

Elaborating on the last point, the path between short term memory and long term memorization lies via
the part of the brain responsible for emotions, so we best remember ideas that stem from our own
tournament practice. For that reason your own games should be the trigger for studying endgame
theory, with the steps above serving as a blueprint. This book will walk you through these steps for
endgames with rook and connected two pawns vs. rook and one pawn where the pawns on
different flanks. In the summer of 2000, I had two endgames with this material balance against
national masters in the same tournament, which led me to study these endgames in great detail, turning
into a great learning experience. Reviewing the analysis many years later with the use of the latest
computer software - has also produced new discoveries.

4
Practical examples of typical simplified positions

After looking at the role of piece activity - let's look at a portion of theoretical rook endgame positions -
with rook and two passed pawns vs. rook and one passed pawn. In my game we have seen that even
though I was up a pawn, I had to worry a lot about the passed pawn. This was happening mainly
because my king was badly placed. In fact, I had to give up one of my pawns to bring him to the correct
defensive position, where he would block the pawn. In the next example Black failed to do so:
Bernstein, O. - Prins, L.
London-A, 1946

White's turn 53.Rg6+! Now Black is in danger as his king is cut off and the rook (because of its own
pawns) has trouble fighting against the 'e' pawn. 53...Kh5 54.Rg1! Ra4? A natural move, but a decisive
mistake.
Black should have kept the rook behind the 'e' pawn 54...Rb3+ 55.Kf4 Rb2 56.e5 Re2=
55.e5 c5 56.e6!

5
56...Ra8
56...Ra6 57.Kf4 and now 57...Rxe6? loses because of 58.Kf5+-
57.Kf4 c4 58.e7! c3
58...Re8 59.Kf5 same idea! 59...Kh4 60.Kf6 b4 61.Rc1 c3 62.Rb1+-
59.Kf5 Kh4 60.Rc1 b4 61.Rb1

I guess Black would prefer if his king was away from the board at all - so many times its bad position
allows White to gain tempi. 61...Rc8
61...Rb8 62.Rxb4+ Rxb4 63.e8=Q+-

6
62.Rxb4+ Kg3
62...Kh5 63.Rc4! Rxc4 64.e8=Q+ Kh4 65.Qh8+ Kg3 66.Qg8+ Kf2 67.Qxc4+-
63.Rb3 Kf2 64.Rxc3 Rxc3 65.e8=Q An impressive play by White! 1-0

Rudakovsky - Tolush
USSR Championship, 1945

Again, the white king does not take part in the action. As shown by Levenfish and Smyslov, Black
could have saved the game by correct play: 1...Kg7 2.Rb6 Kh7 3.h5 Rc2 This is a typical strategy -
sacrificing the pawn, but achieving the drawn Rook + two pawns Vs. Rook endgame. 4.Rxb3 Rc4
5.Rg3 Kh6 6.Kg2 Kg5=

7
This is a well-known theoretical position
In the three previous examples we have seen the same material: rook and two pawns vs. rook and one
pawn, where the stronger side has two connected passed pawns. Strangely enough, the side with two
pawns did not win in any of these games. Is this logical? What does theory say about these endgames?
Well, theory's conclusion is that the side with 2 pawns usually wins. Once such a general guideline is
established in our heads, we can begin to study particular positions where this principle is true, and also
those that are exceptions to the general rule. Those, who have read David Bronstein's brilliant book
about the 1953 Candidates tournament, might remember the following game:
Stahlberg, G. - Petrosian, T.
Zuerich ct, 1953

8
This is an example of a "normal" position of this type of endings; Black wins without any problems:
60.Kf2 g3+ 61.Kf3 Rb3+ 62.Ke2 Rb2+ 63.Kf1 Rf2+ 64.Kg1

64...Rd2 The white pawn is safely stopped, and Black intends to simply continue with Kg4, h5-h4,
Kh3, etc. 0-1
So now, after we have seen several endgames of the same type, we are somewhat interested in being
able to evaluate them, but so far we only know a few general concepts about them. What is the next
step?

Knowing the principles is a very important thing, however, sometimes it is just necessary to learn some
precise positions. The number of these positions should be limited; after all we need to save the
memory cells in our brains for memorizing openings. But these few positions that we are going to
remember have to be studied thoroughly and understood completely. Some of these positions have
deserved to be called critical. First, let me make a simple analogy with a position that everyone is very
well familiar with:
Pawn endgame

9
We all remember that if here it is White's turn, then the result is a draw; with Black to move, he loses.
This critical position is very useful when we evaluate more complicated positions. As soon as we reach
it in our calculations, we can immediately predict the result, without having to make the analysis 2
moves longer.

10
Basic Theoretical Positions

We have seen a very impressive game Bernstein - Prins. However we probably should not memorize
the exact position from this game - it was fairly complicated and we are not likely to have the same
position in our game. But the next position is indeed critical; it could be a good idea to remember it.
Theoretical position 1a

Everything depends on whose turn is it. With Black to move he achieves a draw: 1...a2 2.Ra7+ Kg6
3.h5+ Kh6 4.Ra8 Kg7=

11
Here White cannot advance without losing the 'g3' pawn. 5.Kg5
5.Kf4? Rf1+
5...Rg1 6.Ra7+ Kg8 7.Rxa2 Rxg3+ 8.Kh6 with a draw 1/2-1/2

Theoretical position 1b

If it is White's turn, however, he can win: 1.Kh5 a2 2.g4 Kh7 3.g5 Kg7 4.Ra7+ Kf8 5.g6 Kg8

6.Kh6+-
From this example one can make a conclusion: When the pawns of the stronger side have reached the

12
fourth rank, the position is won. Otherwise it is a draw. This rule, however, has two exceptions; one of
them is this particular position.
Theoretical position 2

As one can see, the configuration of the white pawns here is slightly different and this allows White to
win: 1.g5 Kg7 2.Kh5 Rh1 3.Ra7+ Kf8 4.Rxa2 Rxh3+ 5.Kg6

Here too, Black wins one of the pawns, but the resulting situation is much more favourable for White.
He wins, as the black king is going to find himself on the 'long side'. 5...Rg3 6.Ra8+ Ke7 7.Rg8+-
followed by Kh7.

13
The second exception is the case when one of White's pawns is on the second rank.
Theoretical position 3
Pawn on g2

This is a very hopeless situation for Black. 1.h4 Kg6 2.Kg4 Kf6 3.Ra6+ Kf7 4.Kg5

and now if 4...Rg1 then White simply takes the black pawn, defending the 'g' pawn, and remains 2
pawns up. 5.Rxa2 1-0
Note that in these theoretical positions the black rook is very passive, as it is placed in front of the
pawn. What would happen if the rook on the last diagram had been on b2, instead of a1? As the

14
following example shows, White is still winning:
Vucic, M. - Gurevich, D.
St.Martin op, 1993

Compare it to theoretical position Number Two. Again the White pawns are on h3 and g4 and this
gives him a win. A GM could not save this position, as the 'g' pawn advances before the 'h' pawn,
defending the king from horizontal checks. 60.Ra5 Rh2 61.Ra6+ Kf7 62.g5 Rc2 63.Kh5 Kg7 64.Ra7+
Kf8 65.h4 Rh2 66.g6

66...Kg8 67.Kg5 Kf8 68.h5 Rg2+ 69.Kh6

15
1-0
However, sometimes it is possible to save the game if the rook defends the pawn from the second rank.
Tarrasch, S. - Chigorin, M.
St. Petersburg match, 1893

This example (played just a century earlier than the game Vucic-Gurevich) shows that the question of
"where to keep the rook" is not a simple one. Even Mikhail Chigorin, one of the strongest players of the
19th century was not able to find the correct answer: 40.Kh2 Kg6 41.Ra5 a3! 42.g3 Rc2+ 43.Kh3

16
43...Ra2? This loses.
43...a2! would have saved the game, as shown in analysis by I. Maizelis: 44.h5+ Kf6 45.Kh4 (45.g4
Rc5 46.Rxa2 Kg5=) 45...Rh2+ 46.Kg4 Rb2 47.Ra6+ Kg7 48.Kg5 (48.Ra5 gives nothing, as after
48...Rb4+ 49.Kg5?? would be a disaster in view of 49...Rb5+ 50.Rxb5 a1=Qµ This is in fact one of the
major defensive ideas for Black if his rook is defending the pawn from the second rank.) 48...Rb5+
49.Kh4 Rb2 50.g4 Kf7! 51.h6

51...Rb6= Again, this typical idea of sacrificing the 'a' pawn allows Black to defend successfully.
44.Kg4 Ra1 45.Ra6+ Kf7 46.Kg5 a2

17
It took Black too much time (3 moves) to get to the 'defensive position', with the rook on a1 and pawn
on a2. Given that White advances with gain of tempi (rook checks), Black is clearly not on time. 47.g4!
As we have seen, in such positions when the pawns have time to reach the fourth rank, it usually gives
a win. 47...Ke7 48.Ra7+ Ke8 49.h5 Kf8 50.h6 Rb1 51.Rxa2+-

With two extra pawns, White wins rather easily. 51...Kg8 52.Ra8+ Kh7 53.Ra7+ Kh8 54.Rf7 Rb5+
55.Kg6 Rb6+ 56.Rf6 Rb8 57.g5 Ra8 58.Rf5 Kg8 59.Rd5 Rb8 60.Kh5 Ra8 61.g6 Kh8 62.Rf5 Rg8
63.g7+ 1-0

Charbonneau, P. - Jiganchine, R.
CYCC U18, 2000.07.06

18
After a lot of suffering in the inferior endgame, I finally got to this position with about half an hour left
on my clock for the rest of the game (no increment) and still hoping to make a draw. However, Black
has to solve a very important problem - where to put the rook. I understood the significance of the
moment and spent quite a lot of time here. And yet I still went wrong. The position is almost identical
to the one that Chigorin had against Tarrasch. My mistake was also very similar to Chigorin's:
47...Rf1?? A mistake that decided the game. What was I thinking about? I have to confess, I simply
forgot about the fact that the 'b' pawn was still on the second rank... I remembered that the pawns need
to reach the fourth rank for a win, and it seemed to me that they wouldn't have time to do so...
47...h3! was correct 48.Rh5 Rf3 49.a4 Kc6

19
Now White may try slightly different ideas, but none of them seems to work: 50.b3 This leads to a
theoretical draw, but there is nothing better. (50.b4 Rg3 and the white king is simply cut off. 51.a5 Rf3
52.b5+ Kb7 53.Rh7+ Kb8 54.a6 Rf5 55.Rb7+ Ka8=; 50.a5 Kb7 51.Rh6 Rf5 52.b4 Rf3 with a position
very similar to the one in the previous line. 53.b5 Rf5 54.Rh7+ Kb8=) 50...Rf2+ 51.Ka3 h2! This
position is an exact mirror of Tarrasch-Chigorin, Match (9), 1893. (51...Rh2? , Chigorin's move, loses.)
52.a5 Rg2 53.Ka4 Ra2+ 54.Kb4 Rg2 55.Rh6+ Kb7 56.Kb5 Rg5+ 57.Ka4 Rg2 58.b4 Kc7 59.a6
(59.Rh7+ Kc6 60.b5+ Kc5 61.a6 Ra2+ 62.Kb3 Ra1 63.Rxh2 Kxb5=) 59...Rg6=47...Rf3 is also a
possible move order 48.Rh5 h3 transposingAnother method for a draw was to bring the king to the
kingside, making it difficult for White to keep his rook behind the 'h' pawn: 47...Ke6! Ke7 was also
possible. 48.b4 (48.Rh5 Kf6 49.b4 Kg6 50.Rh8 Kg7=) 48...Kf6 49.Rd5 h3 50.Rh5 Rf3! 51.a4 Rf2+
52.Kb3 Rf3+ 53.Kc4

53...Rf4+= forcing a perpetual because White must avoid an exchange of rooks: 54.Kb5? Rf5+
55.Rxf5+ Kxf5 56.a5 h2-+
48.Rh5 Rh1
It was now too late to put the rook on the side: 48...Rf4 49.Kb3! Rf3+ 50.Ka4 h3 51.Rh6 Rf2 52.b4 h2
53.Ka5+- and White wins since his 'b' pawn will cover the king from black rook's checks.
49.a4 h3

20
50.Kb3 Only now I realized with horror that White's "b" pawn is on the second rank. Such positions
are lost because White can simply advance the pawn and the king. When the black rook leaves h1, the h
pawn (assuming it reaches h2) is going to be lost without any compensation. This is exactly what we
saw in theoretical position # 3. 50...Kc6 51.Kb4 Rh2
I have to lose two tempi because otherwise White just leaves his pawn on b2, where it serves as a
shield: 51...h2 52.a5 Kc7 53.Kb5 Rb1 54.Rxh2+- defending b2
52.b3 Rh1 53.Rh6+ Kc7
53...Kd7 54.Kb5 h2 55.a5+-
54.Kb5 h2 This is now almost an exact mirror of theoretical position #3. 55.b4

21
I finally get what I want - the 'b' pawn has advanced, but the pawns have reached the fourth rank -
another indicator that the position is lost 55...Kd7 56.a5 Kc7 57.Rh7+ Kd6 58.a6 Kd5 59.a7 Ra1
60.Kb6 1-0
A couple of years later I ended up on the White side of the same opening variation, and unsurprisingly
another complex rook endgame ensued. By then I had already reviewed some of the theoretical ideas
covered in this book, and the home preparation had gained me a lot of extra time of the clock.
However, as the game showed, it is still hard to avoid mistakes in these sharp positions where a single
imprecise move can shift the evaluation.
Jiganchine, Roman - Sokourinski, Sergei [B13]
Czt, 2002.03.28

1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Nc6 6.Nf3 Bg4 7.cxd5 Nxd5 8.Qb3 Bxf3 9.gxf3 e6
10.Qxb7 Nxd4 11.Bb5+ Nxb5 12.Qc6+ Ke7 13.Qxb5 Qd7 14.Nxd5+ Qxd5 15.Bg5+ f6 16.Qxd5
exd5 17.Be3 Ke6 18.O-O-O Bb4 19.a3 Rhc8+ 20.Kb1 Bc5 21.Rhe1 Bxe3 22.Rxe3+ Kd6 23.Rde1
Rc4 24.Re7N
24.Re6+ Kc5 25.Ra6 is another continuation, but my intention during preparation was to avoid any
risks: destroying Black's kingside pawns was consistent with my plan. 25...Kb5 26.Rd6 Kc5 27.Rd7
Rh4 28.Rc1+ Kb6 29.Rdc7 Re8 30.R1c6+ Kb5 31.Ka2 a5 32.Rc3 Kb6 33.R7c6+ Kb7 34.R6c5 Re6
35.Rxd5 Rxh2 36.Rd7+ Kb6 37.Rxg7 h5 38.Rb3+ Kc6 39.Ra7 Re5 40.Ra6+ Charbonneau,P-
Jiganchine,R/CYCC 2000/1-0 (60)
24...Rh4 25.R1e6+ Kc5 26.Rxg7 Rxh2 27.Rxf6 Rxf2 28.Rxh7 Rb8 29.Rc7+ Kd4 30.Rc2 Rxc2
31.Kxc2

22
I had this position on my computer the morning before the game. 31...Ke3 seemed like a natural try to
advance the d pawn and attack on f3, so I checked this move with a chess engine. The machine's
response was rather convincing: 31...Ke3?!
31...a5! 32.Ra6 Rc8+ 33.Kd2 Rb8 34.Kc2 Rc8+ 35.Kd2 with a draw was how the game Belikov-
Dreev, 1992 ended
32.Re6+ A useful check that forces the black king to commit. 32...Kxf3?! A natural move, but a
strategic mistake, that is not entirely obvious.
32...Kd4 33.f4 Rh8 would have given Black excellent drawing chances.
33.b4!

23
Black finally regained a pawn, but the cost is too high: his king is cut-off, and White threatens to
transfer the king to d4. 33...Kf4 34.Kc3!? I was purposefully keeping the b3 square protected.
However, this was unnecessary, because the a5 break followed by check on b3 is not dangerous.
34.Kd3 a5 35.bxa5 Rb3+? 36.Kd4 Rxa3 37.Ra6!+-
34...Kf5 35.Ra6 Ke4
35...Rb7 36.Kd4 Rd7 is too passive, the Black king is stuck on the kingside.
36.Rxa7 Usually, these positions with 2 pawns against 1 are winning when the Black king tries to
support the lone passer: White simply sacrifices the rook for the pawn, and then opponent's king is too
far away to stop White's two pawns. Sacrificing a rook back does not help because White has two
pawns. But here the 'd' pawn is far advanced, and moreover, the 'd' pawn is very close to the 'a' and 'b'
pawns: the walk back for the king will be shorter than normally. 36...Re8? Black prepares to advance
the 'd' pawn and wants to cover the king against White's rook checks. But this move leads to an
objectively lost position.
However instead the pawn should have been pushed immediately, while the Black rook belonged on
the 'd' file - behind the passed pawn! 36...d4+! 37.Kb3 but 37...Rd8!= is a draw according to tablebases.
(and not 37...d3? 38.Re7+ Kd4 39.Rd7+! (39.a4 Rd8!=) 39...Ke3 40.a4+-)
37.Kb3
37.Rd7!+- was objectively winning, taking advantage of opponent's last move and putting the rook
behind passed pawn.
37...Rd8!? This could have brought Black a draw because of my mistake. But analysis shows that his
position is objectively lost, since he had lost the crucial tempo. 38.Re7+ Kd3 39.b5 d4 40.a4 Kd2
41.b6 d3

24
42.Rh7?! This move goes against the very purpose of playing Re7 before - making the Black king
block his own pawn. Now things become way too complicated.
The easiest win that I missed was: 42.b7! Passed pawns must be pushed! 42...Kd1 (42...Rb8 43.a5 Kc1
44.a6 d2 45.Rc7+ Kb1 46.Rd7+-; 42...Kc1 43.Rc7+ Kb1 44.Rc8+-) 43.a5 d2 and now simply 44.Rh7
Ke2 45.Rh2+ Ke3 46.Rxd2 Kxd2 47.a6 Rb8 48.Kc4+-
42...Ke3 43.Rh1 d2

44.Kb4? The best square for the king is on c4 where it both pushes away the Black king, and does not
fall into a skewer along b file. However before the promotion of the Black pawn - going to c4 is bad,
because of Rd8-c8-Rc1. Therefore it was better to start with pushing pawns, rather than king. Again,

25
the simple principle of 'Passed pawns must be pushed' serves better than complex calculation or
analysis.
44.a5! d1=Q+ 45.Rxd1 Rxd1 46.Kc4 Here the king goes directly to c4 while in game it had to go to b4
first. c4 is a better square to shoulder off the Black king. 46...Rb1 47.Kc5 Ke4 48.Kc6 Rc1+ 49.Kb7
Kd5 50.a6 Kc5 51.Ka7! Kb5 52.b7 Rc7 53.Ka8 Kxa6 54.b8=Q+- White is just on time.44.Kc4?
Rc8+! 45.Kb5 Rc1 46.Rh3+ Ke4 47.Rh4+ Kd5 48.Rh5+ Ke4=44.Rd1 Ke2 45.Rxd2+ Kxd2 (45...Rxd2
46.b7! Rd1 47.Kb2 Rd8 48.a5+-) 46.a5+- was also winning.
44...d1=Q Now the position is objectively drawn again. 45.Rxd1 Rxd1 According to Fine, the rook
should be always behind the advancing pawns, just as in this position. Here Black achieves a draw.
46.Kc5 Rc1+
Another way to a draw was 46...Ke4! 47.b7 Rb1 48.Kc6 Rc1+ 49.Kb6 Kd5 50.a5 Rb1+ 51.Kc7 Rc1+
52.Kb8 Kc6 53.a6 Rb1! 54.Ka8 Kc7 55.a7 Rxb7= stalemate
47.Kd5 Rb1 48.a5

48...Rb5+?? only detailed analysis with an engine shows that this is where Black made a crucial error.
My opponent played this only because he did not see (well, neither did I) that 48...Kd3! could hold the
position for Black: keeping the rook placed ideally - behind the most advanced pawn. 49.Kc6 Rc1+!
50.Kb7 The only way to support the advance of pawns is to block one of them with the king (50.Kd7
Rb1) 50...Kc4 51.a6 Kb5 52.a7 Ra1!= A typical strategy: the black king attacks the backward pawn
and Black is happy to sacrifice the rook for the 'a' pawn.
49.Kc6 Rxa5 50.b7 Ra6+ 51.Kc5 Ra5+ 52.Kc4 Ra4+ 53.Kc3 Ra3+ 54.Kb2

26
54...Rd3 55.b8=Q and the win was easy, but mainly because I had more time on the clock than my
opponent. These endings can be difficult to win, as even Peter Svidler failed to win it against Boris
Gelfand in the 2002 FIDE Knock-out. 55...Rd2+ 56.Kc3 Rd3+ 57.Kc4 Rd4+ 58.Kc5 Rd3 59.Qe5+
Kd2 60.Kc4 Rf3 61.Qg5+ Ke2 62.Kd4 Kf2
62...Rd3+ 63.Ke4 Rb3 was apparently most resilient.
63.Ke4 Rg3 64.Qd2+ Kf1 65.Kf4 Rg2 66.Qc1+ Ke2 67.Qc2+ Kf1 68.Qc1+ Kf2 69.Qd1 Rh2
70.Qd2+ Kg1 71.Qe1+ Kg2 72.Kg4 Rh1 73.Qe2+ Kg1 74.Kg3 The game was interesting because of
pre-game preparation up to 33rd move, but careful analysis shows that we exchanged quite a few
mistakes in the deep endgame, many of those were based on the violation of the simple principles of
rook endgames. 1-0

27
Theoretical Summary

We have seen a relatively large number of examples of positions rook and two pawns vs. rook and one
pawn, where a stronger side has two connected passed pawns. As the reader has seen, these positions
tend to be rather complicated. (In fact, even knowing the ‘critical positions’ may by itself not be
enough, a player needs to have at least some basic knowledge of other sections of theory on the rook
endgames; without this bulk of knowledge some of the critical positions may be useless). However, it is
still possible to draw some conclusions about these positions:
1. If with the black rook on a1 and pawn on a2, White has a pawn on the second rank, he wins. Black
should avoid such a situation.
2. If with the black rook on a1 and pawn on a2, White’s pawns have reached the fourth rank, he wins.
3. White should try to keep his pawns in such a way that they would defend his king from checks by the
rook. A configuration h3-g4 seems to be very favourable, no matter where the White rook is - on the
second rank, or on a1.
Of course, these are just rough guidelines, every particular position has to be looked at with a lot of
attention. Note that these principles mainly tell where to keep the rook. But the question of where to
keep the king is also very crucial. I suggest you try to make some conclusions about this by yourself,
when answering the training questions in the next section:

28
 Show in Quiz Mode

Test Yourself

Let's apply the knowledge of principles and exact positions we saw above to look for correct solutions
to some practical positions:

Piece Coordination
Schlechter, Carl - Perlis, Julius [D11]
Karlsbad, 1911.08.22

White to move

Show/Hide Solution

Solution: 41.e5! Rd2+

41...fxe5 42.Ke6+- and the black pawn serves as a shield for the white king.41...Rxe5+ 42.Kc6 Ke8
43.d7+ Ke7 44.Ra8+-
42.Ke6 Kc8 43.d7+ Kb8 44.Ra6 Kc7

29
45.Rd6
Black resigned in view of 45.Rd6 Rxd6+ 46.exd6+ Kd8 47.b5 h3 48.b6 h2 49.b7 h1=Q 50.b8=Q#
1-0

Steinitz - Lasker
World Championship Match, 1894

Question: How should Black defend?

30
 Show/Hide Solution

Answer: the game continued 1...Ra4+ As shown by Levenfish and Smyslov, Black could defend
successfully after
1...Rf8!? 2.Re5 g4 3.Rg5 g3 4.Rxg3 Rf4+ 5.Kd3 Rf5!=
2.Kd3 Ra3+?
2...Ra8 would still give a draw, but Lasker misses his chance...
3.Ke4 g4 4.Kf5 Ra8 5.e7 Re8 6.Kf6 g3

7.Kf7 With the white king supporting the pawns, the win is trivial. 7...Kd7 8.d6 g2 9.Rg1 1-0

Active King
Jiganchine, R. - Cabanas, F.
Christmas Open, 1999.12.12

31
Question: would taking the pawn on h7 be a good decision?

Show/Hide Solution

Answer: No, activating the king is more important. 28.Kc2! Even from the first glance at the position it
becomes clear that Black's king is much more active than White's. Black's rook is, on the other hand
quite passive for now. With this manoeuvre White simultaneously activates his own king and prevents
Black's rook from becoming hyperactive.
28.Rxh7? would allow Black to activate his rook and return the material with interest. 28...Re8 29.Rg7
Re1+ 30.Kc2 Re2+ 31.Kb3 Rxf2 32.Rxg6+ Kc5 33.Rg3 f4 34.Rh3 Kd4³
28...Kc5
28...Ra7 29.Rxa7 Kxa7 30.Kd3 Kb6 31.Kd4 Kc6 32.f4 is winning for White, as he will eventually
create outside passed pawn.
29.Rxh7 Kd4 30.Rh4+ Ke5

32
White is up a pawn and eventually managed to convert it into a win. 1-0

Active Rook
The major rule of rook endgames is that you should keep your rook active. When you look at a given
position - usually it is obvious whether or not your rook is active. However, sensing the moment and
finding the tactical opportunity for activating the rook actually requires a bit of experience and
judgement. An important corollary of the above rule is that you should also try to keep your opponent’s
rook passive. What does it mean for a rook to be active?
1. it attacks opponent’s pawns and protects its own
2. it can attack or cut off opponent’s king
3. it has freedom for manoeuvre, so zugzwang is never a problem
4. if there are vital open files or ranks - it controls one of them
Tarrasch, Siegbert - Rubinstein, Akiba
San Sebastian, 1911

33
Black is down a pawn, but Rubinstein managed to create sufficient counterplay. Black to move

Show/Hide Solution

Solution: 32...Rd8! 33.Ra6 Rd2! 34.Rxb6+ another pawn falls, but... 34...Kg5

Now it becomes clear that all Black's pieces and pawns (except for the one on h7) are going to join the
attack against the white monarch. This certainly compensates for the lost material. 35.Ke1 Rc2 36.Rb5
Kg4 37.h3+ Kxh3 38.Rxf5 Rxb2 39.Rf4 Rxa2 40.Rxe4 h5 41.c4 Kg2 42.Rf4 Rc2

34
As material is further reduced, achieving a draw here is not a problem for Black. 43.Rh4 Kf3 44.Kd1
Rxf2 45.c5 Ke3 46.Rxh5 Kd4 1/2-1/2

Nathani - Jiganchine, Roman [C36]

BC Junior 2001, 2001.11.12

Black to move

Show/Hide Solution

35
Solution: 27...Rb7? I decided to play it 'safe', keeping as many pawns on the board as possible, and
gave up the advantage by misplacing my rook; this leads to a very passive position of the rook.
27...Rb5!? 28.Rxa7 Rxe5 29.Kg3 Rc5 30.Ra2 Ke7

The white rook is passive on a2; Black is up a pawn, and has a more active rook - this is a lot likely to
be winning than the position that I got in the game 31.Kf4 Ke6 32.Ke3 f5 33.Ra7 Rc3+ 34.Kd4 Rg3
35.Ra6 Rxg2 36.Rxc6+ Kd7 37.Rc5 g6µ
28.Ra6 Rc7

36
It would have been hard to make my rook any more passive than it currently is on c7.
28...Re7 29.Rxc6 Rxe5 30.Ra6 Re7 31.g4 Rc7 32.Kg3 Ke7 33.Kf4 is hard to evaluate
29.Kg3 Ke7 30.Kf4 Ke6 31.c4

White supported the 'e5' pawn, and while Black's rook is passive, he can't make much progress with his
king either. 31...f6 Because my pieces are tied up defending the weaknesses, I have to force another
transformation, which helps White to get rid of his weak pawn 32.exf6 Kxf6 33.c5 Ke6 34.Ke4 Kd7
35.Ke5 Kc8 36.Kd6

36...Rf7 I am sure my opponent was a bit surprised here. This is fairly unusual - capturing a pawn with
either king or rook instantly loses, because in the pawn endgame Black will have a distant passed pawn.

37
37.Ra3
37.Rxc6+ Kb7-+37.Kxc6 Rf6+ 38.Kb5 Rxa6 39.Kxa6 Kb8 40.g4 g5 41.Kb5 Kc7 42.Ka6 Kc6
43.Kxa7 Kxc5-+
37...Rf6+ 38.Ke7 Kb7 39.Rb3+ Kc7

40.Rg3?! This complicates things unnecessarily.

Almost immediate draw was achieved by 40.Ra3 Kb7 41.Rb3+ Kc7= (41...Ka6 42.g4 Rg6 (42...Ka5
43.g5 Rf5 44.Kd6 Ka4 45.Rb7²) 43.Kd7 Ka5 44.Rb7 Ka6 45.Rb3 Black's king and rook are both
badly placed, so he does not really have much of a chance to play for a win here.)
40...Rg6 41.Ra3 Rxg2 This allows a pawn exchange, which is something Black should avoid. Now my
king is also cut off.
Unfortunately it seems that keeping a7 pawn was not winning either 41...Kb7 42.g4 Rg5 43.Kf7 h5
44.gxh5 Rxh5 45.Kxg7 Rxc5

38
46.h4= Disconnected pawns are not well supported by Black's king and rook, so White can hold this
position 46...Rd5 47.Rh3 c5 48.h5 Kc6 49.h6 c4 50.h7 Rd8 51.h8=Q Rxh8 52.Rxh8 Kd5 53.Kf6 c3
54.Ra8 Kd4 55.Rxa7 c2 56.Rc7 Kd3=
42.Rxa7+ Kb8 43.Rd7 Kc8 44.Rd8+ Kb7

45.Rd7+? This check throws away a comfortable draw since now my king has time to come to b5 and
attack the white pawn.
45.Rd3 Ka6 46.Rb3 With the king cut off along the 'b' file, there is little hope to win for Black. 46...Rg6
(46...Ka5 47.Kd7 Rg6 48.Rb6=) 47.Kd7=
45...Ka6 46.Kd6 Rg6+ 47.Ke5 Rh6

39
48.Rd6? This allows Black to keep two connected pawns on the kingside, which as we know is deadly.
More resilient was 48.Rxg7 Rxh3 49.Kd6 Rh6+ 50.Kc7 Kb5 51.Rg5 still with some drawing chances.
48...Rh5+ 49.Kd4 Kb5 50.Rd7 Rh4+ 51.Ke5 g6 52.Rd3 Kxc5 This game is a comedy of errors, as I
transformed a position with an extra pawn into one in which I buried my rook in a very passive square
where it could barely move. My opponent had many opportunities for a draw, mostly by cutting off my
king from the action, but he missed them, allowed my king to get back to action, and Black won. 0-1

Choosing direction for the king

Dreev, A. - Ehlvest, J.
Tallin, 1986

40
As we have seen in the previous examples, the defending side has to decide where to place his rook.
Now the big dilemma is "Where does the black king go?" Question: Should Black bring his king to the
kingside to support his own passed pawn, or should he retreat to the queenside, to block the opponent's
passed pawns?

Show/Hide Solution

Answer: 1...Ke5? was played in the game.

However, 1...Kc5! was the right plan 2.b4+ (2.Rc4+ Kd5 3.Ke3 g4 4.Kf2 Rb8 5.Rc3 Kd4 6.Rg3
Kc5=) 2...Kb6=
2.b4 Kf5? It is still not too late to come back with
2...Kd5
3.Rg1 g4?
3...Ke6=
4.Kc4 g3 5.Kd5 Rd8+ 6.Kc6 Rc8+ 7.Kb7 Rg8

41
8.a5 The white pawns are supported by the king, and cannot be stopped, while White himself has no
problem with giving up his rook for the g pawn. 8...Rg7+ 9.Kb6 Ke5 10.a6 Rg6+ 11.Kc5 1-0

Reshevsky, S. - Alekhine, A.
AVRO, 1938

Question: It is Black's turn and once again, the monarch is at the crossroads. Moving the king to e6 is
natural, but what theoretical position are we going to achieve then - lost or drawn?

42
 Show/Hide Solution

Answer: 57...Kc6!
Instead heading to the kingside would have lost the game: 57...Ke6? 58.Kg3 Kf6 59.h3 Kg6?
60.Kh4+- Levenfish and Smyslov. Here, similarly to Vukic-D. Gurevich, White's pawns give the king
a shelter from the black rook.
58.Kg3
58.g5 Rb5=
58...Kb6 59.Ra8 Kb5 60.h3 Kb4

61.Kf4
61.Kh4 Kb3 62.g5 Rb1 63.Kh5 a1=Q 64.Rxa1 Rxa1 65.g6 Kc4 66.g7 Rg1 67.Kh6 Kd5 68.Kh7 Ke6=
61...Rc2 Now Black's plan is not to win the rook for the pawn (the black king would then be too far
from White's pawns), but rather to transfer the rook to a5 or a3 with a win. 62.Rb8+
62.h4 Rc4+ 63.Kf5? Rc5+ 64.Ke4 Ra5-+ is the threat.
62...Kc3 63.Ra8 Kb4

43
Repetition is the logical outcome. 1/2-1/2

Power of connected passed pawns

Botvinnik, Mikhail - Fischer, Robert James [D98]
Varna ol (Men) fin-A, 1962

White to move

44
 Show/Hide Solution

Solution: 52.h5!
52.Rxg6 b4 was likely lost for White as Black would have maintained two connected passed pawns on
the queenside.
52...Ra3+ 53.Kg2 gxh5 54.Rg5+ Kd6 55.Rxb5

Two rook pawns are not very dangerous in this position, which is a draw according to latest theoretical
endgame tablebases. In practice White rather easily achieved a draw as well. 55...h4 56.f4 Kc6 57.Rb8
h3+ 58.Kh2 a5 59.f5

45
Advancing the 'f' pawn is the simplest way to secure draw, because in fact White does not need that
pawn. 59...Kc7 60.Rb5 Kd6 61.f6 Ke6 62.Rb6+ Kf7 63.Ra6 Kg6 64.Rc6 a4 65.Ra6 Kf7 66.Rc6 Rd3
67.Ra6 a3 68.Kg1 Neither side is able to achieve progress. A historic game! 1/2-1/2

Evaluating Consequences of Exchanges

The following two examples show that knowing endgames theory helps to make correct decisions in
more complicated positions.
Haukenfrers - Jiganchine, Roman [A00]
Fraser Valley Championship, 2004.01.15

46
Question: should Black exchange queens and go into the rook endgame?

Show/Hide Solution

Answer: Yes, it is a reasonable practical choice, especially in time trouble, but it has to be followed up
with precise play. 45...Qxg3+!? 46.fxg3

46...b6? A natural move, but an inaccuracy.

46...Rd2!-+ was stronger, as it would have prevented the white king from getting active, and made

47
more difficult for White to put the rook behind the pawn, costing a few precious tempi. Black is
winning here according to latest endgame tablebases. 47.Kf1 c6 48.Rc3 Kc7 49.g4 Kd6 50.g5 Ke6
51.Rg3 Kf7-+

Black has achieved the ideal setup where his king is blocking the white pawn.
47.g4?
47.Kf2 c5 48.g4 would have lead to a theoretically drawn position, as White would put the rook behind
the 'g' pawn and centralize the king.
47...Rd4!

48
48.Kf2 Rxg4-+ 0-1

Grischuk, Alexander - Mamedyarov, Shakhriyar [C77]

1st FIDE GP London 2012, 2012.09.24

Question: How should White deal with the dangerous 'b' pawn?

Show/Hide Solution

Answer: Exchanging to a rook endgame here is the correct approach. 44.Bxc7!

But not 44.Rb5?? Rb8! 45.Rxb8+ Bxb8 46.Bxb8 b2-+44.Rc5?? Bxf4+ 45.Kxf4 Rxc5-+
44...Rxc7 45.Rb5 Kg7 46.Kg4
White wins by simply advancing the pawns: 46.Kg4 Rc3 47.h4 Rc4+ 48.Kh5 Rc3 49.Rb7+ Kg8
50.g6+-
1-0

The next game is the battle between one side having three passed pawns, and the other side trying to
create counterplay by activating his rook and king.
Kostin, Andrey - Jiganchine, Roman [B11]

49
 Keres mem 31st, 2006.05.19

White's turn 38.Rd4

More forceful was 38.Rd6+ Re6 (38...Kf5 39.Rxc6 Re1 40.Rc5+ Ke6 41.Rxc4+-) 39.Rxe6+ Kxe6
40.Ke4 h5 41.a4 Kf6 42.Kd4 Kf5 43.Kc5+-
38...Re1 Black decides to give up all his queenside pawns and go for the active play.
38...Rc8 39.Rxc4 Ke6 was possibly better, but felt too depressing to play, as Black had no counterplay
whatsoever after 40.Rc5+-
39.Rxc4 Rf1+ 40.Kg2 Rd1 41.Rxc6+ Kf5 42.Rxa6

50
42...Rd2+! This forces the white king to go to the first rank, giving Black a glimmer of hope. 43.Kf1
Kxf4 44.c4 g5
44...Rd4!? 45.Rc6 Rd1+ 46.Ke2 Rh1 47.a4 Rxh3 48.b5 Ra3 was probably a better chance for Black, as
at least he has two passed pawns on the kingside, not one as it happened in the game.
45.c5 h5 46.b5 g4 47.hxg4 hxg4 48.b6 Rb2 49.Ra3 Rb1+
49...g3 50.Rb3 Rxa2 51.b7 Ra1+ 52.Ke2 g2 53.b8=Q++- wins for White because he promotes with
check.
50.Ke2 Rb2+ 51.Kd1 g3

51
White to move

Show/Hide Solution

Solution: 52.Rxg3! Kxg3 53.a4 Kf3 54.a5 Ke3

White correctly sacrificed the rook for 3 powerful passed pawns, and there is no way Black can stop
them. However some accuracy was still required. Question: which pawn must be pushed first?

Show/Hide Solution

Answer: 55.a6?? It turns out that the threat of queening the pawn on a8 is not as effective as on c8.
55.c6! Kd3 56.Kc1 and White wins since Black has no time for Kc3. 56...Kc3 57.c7+-
55...Rd2+?? Black returns the favour and misses the chance to save the game.
55...Kd3!= Would have saved the game. I thought about harassing the king with a checkmate threat,
but did not see the idea of transferring the rook to the 4th rank on move 57. 56.Kc1 Kc3 57.b7 (57.a7
Ra2 58.Kb1 Ra4=) 57...Rb4 58.a7 Rh4!!

52
Amazingly White cannot avoid a perpetual. 59.Kd1 (59.Kb1 Rb4+ 60.Ka2 Ra4+ 61.Kb1 Rb4+)
59...Kd3 60.Ke1 Ke3 61.Kf1 Kf3 62.Kg1 Rg4+ and White never has enough time to promote the
pawn - it is either check, or mate in 1 is threatened, admittedly, this idea of a rook transfer to the 4th
rank is almost impossible to spot over the board, unless one is very familiar with these ideas.
56.Kc1+- Rd5

Black is now lost again. I managed to capture two pawns, hoping to put up some resistance in queen vs.
rook endgame, but lost it rather quickly. 57.b7 Rxc5+ 58.Kb2 Rb5+ 59.Kc3 Rb6 60.Kc4 Rxa6
61.b8=Q Ra4+ 62.Kc3 Re4??
62...Rg4 would have allowed to prolong the resistance.

53
63.Qg3+ Ke2 64.Qd3+ 1-0

Passed pawns must be pushed

Passed pawns in the endgame are more powerful than in the middlegame. With few pieces left on the
board, they can be supported by their own king and tie up entire opponent’s army. As usual, it takes a
bit of experience to sense that a pawn is especially dangerous in a given situation.
Jiganchine, Roman - Black, B. [B07]
BC-Washington, 1999.12.04

Black to move

Show/Hide Solution

Solution: 41...d4! The white king is so cut off, that advancing the passed pawn would give great
winning chances - if not a clearly winning position. Having reviewed my game against Hacat, it is
much easier to appreciate the strength of the 'd' pawn in this game too.
In the game Black failed to consider all possible options, and played the natural move, defending the
pawn. 41...Kf6? 42.Rb6+ Kf5 43.Rb7 Kg6 (43...d4 here comes with a loss of a tempo, so not as strong
at all 44.Rxg7 d3 45.Rxh7 Ke6) 44.Rb6+ Kg5 45.Rb7 Kf6 and draw was agreed upon.

54
42.Rb8
Going for material with 42.Rxg7 allowed Black to immediately advance the pawn and support it with
the king with decisive result: 42...d3 43.Rg8 Kd5 44.Rd8+ Kc4 45.Rd7 Kc3 46.Rc7+ Kd4 47.Rxh7 d2
48.Rd7+ Ke3 49.h4 Ra4-+

The white king is cut off along 4th rank, and after White sacrifices the rook for 'd' pawn, his kingside
pawns will be easily stopped by Black.
42...d3 43.Rd8 d2 44.h4 h5!-+

White is in some kind of amazing zugzwang. Either his king has to leave the g file, making e5-e4 break
possible, or the rook has to go to d3, with similar issues! 45.Rd3

55
45.Kh2 e4! 46.fxe4 Ke5 47.Kg3 Kxe4-+ The black king advances to support the pawn.45.Kf2?
d1=Q+-+
45...e4!! 46.fxe4 Ra3!-+

56
About the Author

Roman Jiganchine has been a chess coach of several Canadian junior players. His students in the early
2000s won many Canadian championship youth titles in various age groups. Roman has contributed to
«Chess Life», Russian «64» magazine, and for several years had an endgame column in Canadian chess
magazine «En Passant» (later renamed into «Chess Canada»). He received his early chess education in
the Moscow Petrosian Chess school, which brought to the chess world grandmasters such as
Morozevich, Riazantsev, and the Women’s World Champion Alexandra Kosteniuk. Roman moved to
Vancouver, Canada in 1998, and represented Canada in the 2000 World Youth Under 18 Championship
in Spain.

57
Symbols and Abbreviations Used in the Book

Position Evaluation
∞ – Unclear: It is unclear who (if anyone) has an advantage.
= – Even position: White and Black have more or less equal chances.
+= – White has slightly better chances.
+/− – White has much better chances.
+− – White has a clear advantage.
=+ – Black has slightly better chances.
−/+ – Black has much better chances.
−+ – Black has a clear advantage.
Move Evaluation
?? - Blunder
? - Mistake
?! - Dubious move
!? - Interesting move
! - Good move
‼ - Brilliant move

58
Table of Contents
CopyRight Information 3
A Study Guide 4
Practical examples of typical simplified positions 5
Basic Theoretical Positions 11
Theoretical Summary 28
Test Yourself 29
Piece Coordination 2
Active King 2
Active Rook 2
Choosing direction for the king 2
Power of connected passed pawns 2
Evaluating Consequences of Exchanges 2
Passed pawns must be pushed 2
About the Author 57
Symbols and Abbreviations Used in the Book 58

59