Rook and Bishop vs.

Rook Check out these

bestselling titles from
USCFSales.com:
Rook and bishop vs. rook endings can be exceptionally difficult if the
defender’s king is restricted to the edge of the board. With the king in a
less dangerous position, the defense is easier – if you know the main
ideas: “Cochrane’s position” and “the seventh rank defense.” In
Dvoretsky’s Endgame Manual, we examined the instructive game
Timman – Lutz (Wijk aan Zee 1995), in which both defensive methods
were used. Nevertheless, the defender finally strayed off the proper
course, turning the game into a won Philidor position. However, Timman
had to take the draw in accordance with the “fifty-move rule.”
The The ChessCafe
Puzzle Book (CD)
In practice, players in this situation – nearly always suffering from a
Instructor severe time shortage – rarely manage to avoid committing fatal errors,
by Karsten Müller

altering the evaluation of the position. I could present a number of

instances, some from the careers of leading grandmasters. I hope this
Mark Dvoretsky article helps reinforce your understanding.

The first endgame we will examine was played over a decade ago. I use it
to draw your attention to one significant theoretical detail, which I myself
only found out about very recently. The other two examples are taken
from recent games played by strong grandmasters.
A Practical Guide
to Rook Endgames (CD)
San Segundo – Beliavsky by Nikolay Minev
Madrid 1997
Read an excerpt here.
One of the defensive systems is called
“the seventh-rank defense”: the rook
stays within one square of the king.

White had already maintained the

Play through and download position for a number of moves by this
the games from defensive method, and should have
ChessCafe.com in the continued in the same style: 108.Kf2!.
DGT Game Viewer. After 108…Rh2+ 109.Ke1, Black’s rook
is attacked, and thus he has no time to
bring his king closer. And the attempt to
The Complete
play for zugzwang by 108…Bd3 doesn’t work, because White has a
DGT Product Line stalemate defense: 109.Ke1! Ke3 110.Re2+!. This short variation makes Victor Bologan:
clear why this defensive method is only effective from the second Selected Games
(seventh) rank, or from the knight’s file. 1985-2004
by Victor Bologan
Instead of 109…Ke3, Black could try 109…Re3+!?, hoping for 110.
Kf2?! Rf3+!. Now the king has to retreat to g1 or g2, and the defensive Read an excerpt here.
redoubt on the second rank breaks down, making the defense more
difficult. Such a position did indeed occur, in our third example:
Rychagov – Grischuk.

A safer defense is 110.Kd1! Rf3 111.Kc1 (or 111.Rh2 Ke3 112.Kc1)

111…Ke3 112.Rh2; for instance, 112…Kd4 113.Kd2, returning to the
second-rank defense on a different set of squares. And if 112…Rf1+ 113.
Kb2 Kd4, then White could set up the “Cochrane position” by 114.Rh4+
Be4 115.Kb3 Rb1+ 116.Ka4 (but not 116.Ka3? Kc3!–+).

The fourth-rank pin restricts the mobility

of Black’s king. On 116…Kd3, there
 follows 117.Ka5! Bd5 118.Rb4, freeing
the king from the edge of the board. A
symmetrical variation would be 116…
Kd5 117.Ka3! Bd3 118.Rb4. 116…Re1
117.Kb4 is useless; and if Black plays
the waiting 116…Rb8, White can wait
too: 117.Rg4.

By the way, the position after 116…Kd5

(if you rotate the board on its horizontal axis, reversing the upper and
lower halves) also occurred in the game Iljushin – Inarkiev, Moscow
2008.

Here, instead of 117.Ka6!, White played

the inaccurate 117.Rg5?!, to which his
opponent replied 117…Bc7+! 118.Ka6
Kc4. The defense was now more
complicated, since the king could no
longer leave the edge of the board.
Twelve moves later, Iljushin committed
the decisive error and lost.

After this theoretical excursion, let us

return to our game where, instead of the
approved defensive method, Pablo San Segundo played a losing move:

108.Rd8? Rh2+ 109.Kd1 Ke3 110.Rb8??

Unbelievable! White overlooks a mate in one, 110…Rh1# – and

Beliavsky doesn’t see it. On the other hand, as we shall see, 110.Kc1 Rc2
+ 111.Kd1 Rc3 (but not 111…Rb2 112.Rc8 =) would not have saved him
either. After the text, the game is also lost.

110…Rd2+? 111.Kc1 Rc2+ 112.Kd1 Rc3

113.Rb2?

Loses quicker. In the opinion of the

annotators – Hecht in ChessBase,
Beliavsky and Mikhalchishin in their
book, Modern Endgame Practice –
White is saved after 113.Re8. This isn’t
true: 113…Rc7 114.Re5 Rh7 115.Kc1
Rb7–+.

113…Bf3+ 0-1

Bologan – Rublevsky
Poikovsky 2007

The last capture occurred on move 76.

Black has held out for more than half the
required 50 moves, and hasn’t very far
to go before the “final” 126th move.
Rublevsky has successfully employed
the “seventh-rank defense,” and had he
played 103…Rb7 here, could have
continued with it even longer. But he
allowed his opponent to change the
course of the battle.

103…Kd8?! 104.Kf6 Rc6 105.Ra7 Rd6 106.Ke5 Rc6 107.Kf6 Rd6 108.
Rb7 Rc6 109.Rh7 Rd6 110.Ke5 Rc6 111.Rd7+ Kc8 112.Kd5 Rc7 113.
Rd6+ Kb7 114.Rd8 Kb6 115.Bd7 Ka7?! (115…Rc5+ 116.Kd6 Rc7
would have been simpler) 116.Bc6 Kb6! (only move!) 117.Rb8+ Ka7
118.Rb1
 The only way to hold the draw in this
position was 118…Rh7(g7)! 119.Kc5
Rh3(g3)! – but not 119…Rh2? 120.Be4!
+–. The exact reason why Black would
have to bring the rook precisely to the
third rank would have been difficult to
see, even with much more time available
to the players. In our next example, a
similar situation occurs, but turned 90
degrees to one side. There, the key factor
becomes control of the c-file.

118…Re7? 119.Kc5 Re5+

Now 119…Re3 is useless, in view of 120.Ra1+ Kb8 121.Kd6, as the rook

is placed too close to the king, and is therefore unable to deliver a check
from the side.

120.Bd5 Rh5?

After 120…Re7, White’s remaining six moves would not be enough for
mate or the win of the rook: 121.Rh1 Rc7+ 122.Bc6 Kb8 123.Kd6 Rg7
124.Rh8+ Ka7 125.Kc5 Ka6 126.Rh1 Rg5+ 127.Bd5 Ka7 128.Rh8+–
(Black would have had to resign, if this had been White’s 125th move).

121.Rb7+! Ka6 122.Rg7!

122.Rb8 Rh7 123.Rb3 Ka7 124.Ra3+ Kb8 125.Ra8+ Kc7 126.Ra7+ wins
also. But White’s last move would be the 50th move without a capture,
and the rook only gets taken on the following move – so it would be a
draw! Viorel Bologan accurately counted the moves, and saw that in the
other line, he would win in time (one move before the deadline, in fact).

122…Re5 123.Rg1 Ka7 124.Rg8 1-0

We begin our study of the following example long before the material
imbalance under consideration appears – there are a number of interesting
and instructive moments before it that we don’t want to miss.

Rychagov – Grischuk
Russian Championship, Moscow 2007

In-between moves are a powerful

tactical weapon, but they can be hard to
handle. Frequently, in-between moves
can be so subtle that players either fail to
notice them altogether or underestimate
them. That’s what happened in this game.

At first glance, everything’s clear: Black

brings his rook to e7, winning the pinned
knight. Note that he may do this either
immediately, or after a preliminary
check on e1. Why give the check? We will find the answer to that
question once we discover and calculate the tactical defense White has at
his disposal.

Black wins after 42…Re1+!! 43.Kf2 Re7 44.g4!? fg! 45.Ra5+ Kf4 46.
Nf6 g3+ (the pawn advances with check – this is why it was necessary to
bring the white king forward to f2) 47.Kf1 g2+ (47…Rf7!? is also very
strong) 48.Kf2 Re2+ 49.Kxe2 g1Q.

42…Re7? 43.g4! f4 (43…fg 44.Ra5+ Kh4 45.Nf6 =) 44.Ra5+ Kh4 45.

Nf6 Kg3 46.Nh5+ Kxg4 47.Nf6+ Kg3
 48.Ra3+?

48.Nh5+ looked right. The text is

illogical: why allow the pawn to reach
f3? In chess – a game ruled by iron logic
– every inconsequential move is soundly
punished (Rudolf Spielmann).

48…f3 49.Nh5+ Kg4

In Grischuk’s opinion, Black had a

simpler way: 49…Kh4 50.Nf4 Kg4, or 49.Ra5 Re1+ 50.Kf2 Re2+ 51.Kf1
Rb2 52.a4 Be4 (with the king on g4, this move would not be available
because of the check on f6) 53.Nf4 Kg4.

50.Ra5

White does no better after 50.Nf6+ Kg5 51.Ra6 Re1+ 52.Kf2 Re2+ 53.
Kf1 Rc2–+, or 53.Kg3 Rg2+ 54.Kh3 f2–+.

50…Re6 51.Kf2

Black might have exploited White’s

badly placed pieces by 51…Re2+! 52.
Kf1 (52.Kg1 meets the same reply) 52…
Rb2! 53.Nf6+ Kf4 54.a4.

Now, the quiet move 54…Bb7!! would

have placed White in zugzwang. The
rook cannot move along the fifth rank,
because of 55…Ba6+; on 55.Ra7 f2! 56.
Rxb7 Rxb7 57.Kxf2 Rf7! decides, or if
56.Ra5 Ba6+! 57.Rxa6 Kf3. King moves
are refuted in exactly the same way: 55.
Kg1(e1) f2+ 56.Kf1 Ba6+! And knight
checks would allow Black’s king to
invade the third rank: 55.Nh5+ Ke3 56.
Re5+ Be4–+, or 55.Nd5+ Kg3 56.Nc3 Bc8–+.

51…Rh6? 52.Rc5!

As long as the bishop remains on the a8-h1 diagonal, the rook will chase
it. As soon as the bishop leaves the diagonal, White gives up his knight
for Black’s last pawn on f3.

52…Bb7 53.Rb5 Bc6 54.Rc5 Be8 55.Rc4+ Kxh5 56.Kxf3

Now I will cease commenting on the ensuing “rook and bishop vs. rook”
endgame, until it reaches the critical position, which is important for
endgame theory.

56….Ra6 57.Kf4 Kg6 58.a4 Kf6 59.a5 Rxa5 60.Rb4 Bc6 61.Rc4 Rf5+
62.Ke3 Re5+ 63.Kd4 Rd5+ 64.Ke3 Bb5 65.Rd4 Rh5 66.Kf4 Ke6 67.
Ke4 Bc6+ 68.Kf4 Rf5+ 69.Ke3 Ke5 70.Rd3 Rh5 71.Rc3 Bd5 72.Kd2
Rh2+ 73.Ke3 Rh4 74.Kd3 Rg4 75.Ke3 Bc4 76.Kf3 Rh4 77.Ke3 Re4+
78.Kf3 Be2+ 79.Kf2 Kf4 80.Rc2 Bd3 81.Rb2 Re3 82.Rb4+ Be4 83.Rb2
Rh3

84.Re2?

An inaccuracy, making White’s task

considerably more difficult. He should
have continued either with 84.Rd2 (refer
to the endgame San Segundo –
Beliavsky), or with 84.Ke1.

84…Bd3 85.Rd2 Rf3+!

In this way, Black breaks up the

defensive redoubt on the second rank. The concluding portion of the
game took place in mutual time-pressure: the thirty seconds added for
each player per move was clearly not enough time for successfully
resolving the problems they faced.

86.Kg2

White’s defense would have been simpler after 86.Kg1!? Ke3 (87.Rf2
was threatened) 87.Rg2 Rf7 (87…Bf1 88.Rg7, intending either 89.Rf7+
or 89.Kh2) 88.Rg3+ Ke4 89.Kh2. Here, Black would have had a more
difficult time coordinating his pieces.

86…Bf1+ 87.Kg1 Ke3 (87…Kg3 88.Rg2+!) 88.Rd5 (88.Rd8 or 88.Ra2

would have been no worse) 88..Bd3 89.Rg5 Be4 90.Kh2 Kf4

We can hardly make sense of what

follows without computer assistance. We
can no longer rely on simple variations,
or logically understandable approaches –
we must now deal with the hidden
geometry of the chessboard.

According to the endgame tablebase,

there are only two moves to draw here:

a) There’s the pretty stalemate defense:

91.Rg3! Rf2+ 92.Kg1 Ra2 93.Rc3; and

b) Moving the rook to one of the “correct” files (the c- or a-file): 91.Rg7!
Rf2+ 92.Kg1 Ra2 (92…Rc2 93.Ra7!) 93.Rc7!.

91.Rg8? Rf2+ 92.Kg1 Ra2?

The only move to win was 92…Rc2!! 93.Rb8 (the rook can’t get to the a8-
square – this is why White’s 91st move was a mistake) 93…Rg2+ 94.Kf1
Rd2! 95.Rb4 Ke3 96.Rb3+ Bd3+ 97.Kg1 Kf3 98.Rb8 Bc4! (the rook
takes away the vital g8-square from the rook) 99.Rb6 Rg2+ 100.Kh1
Ra2!–+.

93.Rb8?

The rook belongs on the c-file: 93.Rc8! Rg2+ (93…Ke3 94.Rc3+) 94.Kf1
Rd2 95.Rc4 (95.Kg1) 95…Ke3 96.Rc3+ Bd3+ 97.Kg1 Kf3 98.Rc8! =,
now the vital c4-square remains under White’s control. Inhuman finesses,
indeed!

93…Bd5?

Black wins by 93…Rg2+! 94.Kf1 Rd2!.

94.Rd8 Kg3 95.Kf1 Bf3

 96.Ke1?

The final mistake. The draw was his

after 96.Re8! or 96.Rg8+ Bg4 97.Re8!.

96…Re2+ 97.Kf1 Re3! 98.Rg8+ Bg4

99.Rg7 Re8 100.Rg5 Rh8 0-1

White resigned, in view of 101.Ke1 Rd8–

+.

[ChessCafe Home Page] [Book Review] [Columnists]

[Endgame Study] [The Skittles Room] [Archives]
[Links] [Online Bookstore] [About ChessCafe.com] [Contact Us]

© 2008 CyberCafes, LLC. All Rights Reserved.

"ChessCafe.com®" is a registered trademark of Russell Enterprises, Inc.