Wijk R03: A highly intstructive endgame struggle

1/15/2008 – All games of round three drawn – but one was a tense 85-move battle from which there is a lot to learn. Watching the tournament on Playchess.com Russian politician Garry Kasparov tried to explain its complexities – without the use of the computer. Our commentator Mihail Marin did have tablebases and attempts to do just that, in an endgame training session you cannot afford to miss.

ChessBase 14 Download ChessBase 14 Download

Everyone uses ChessBase, from the World Champion to the amateur next door. Start your personal success story with ChessBase 14 and enjoy your chess even more!

Along with the ChessBase 14 program you can access the Live Database of 8 million games, and receive three months of free ChesssBase Account Premium membership and all of our online apps! Have a look today!


Wijk aan Zee 2008

GM Mihail Marin at work in a tournament

The following express commentary was provided by Romanian grandmaster Mihail Marin, who is the author of a number of very popular ChessBase training CDs and articles for ChessBase Magazine. GM Marin will study the games of the World Championship tournament in much greater detail and provide the full results of his analysis in the next issue of ChessBase Magazine.

Round three commentary by GM Mihail Marin

A peaceful round: all the games ended in draws, some of them rather quickly. In three cases the presence of opposite coloured bishops allowed predicting the final result dozens of moves before the end of the games...

And yet, we had a star game, in which a very interesting theoretical ending was played, arousing the enthusiasm of none other than Garry Kasparov. I have annotated the final part of Radjabov-van Wely extensively, aiming to reveal all its essential aspects.

The actors of the Monday night drama: Teimour Radjabov, Azerbaijan, (white) and...

Loek van Wely, many times Dutch champion (black)

Radjabov,T (2735) - Van Wely,L (2681) [D43]
Corus A Wijk aan Zee NED (3), 14.01.2008 [Mihail Marin]

1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.Nf3 e6 5.Bg5 h6 6.Bh4 dxc4 7.e4 g5 8.Bg3 b5 9.Be2 Bb7 10.0-0 Nbd7 11.Ne5 Bg7 12.Nxd7 Nxd7 13.Bd6 a6 14.Bh5 Bf8 15.Bxf8 Rxf8 16.d5 cxd5 17.exd5 Nf6 18.dxe6 Qxd1 19.Bxf7+ Ke7 20.Raxd1 Rfd8 21.Ne2 Be4 22.f4 b4 23.fxg5 hxg5 24.Ng3 Rxd1 25.Rxd1 Bh7 26.Nh5 c3 27.Nxf6 Kxf6 28.bxc3 bxc3 29.Rc1 c2 30.Kf2 Rd8 31.Ke2 Rd4 32.h3 Be4 33.Bh5 Kxe6 34.g3 Ke5 35.h4 gxh4 36.gxh4

Black has been pressing for a long time, but now he allows his opponent an unexpected transposition to a theoretical endgame. 36...Rd5? Garry Kasparov pointed out that van Wely had missed an "/portals/all/_for_legal_reasons.jpg" win with 36...Bf5. John Nunn, who had all the necessary tablebases installed, confirmed that it was "elementary". 37.Bf3! Rd1 38.Bxe4 Rxc1 39.Kd2 Rg1 40.Bxc2 Kd4 41.Bd3 a5 42.a4 Rh1 43.Bb5 Rh2+ 44.Kd1 Ke3 45.Kc1 Rxh4

This is an important theoretical position. The evaluation and the recommended methods of play have varied along the decades. Of course, things have been settled from theoretic point of view after the creation of the six men tablebases, but in many cases one does understand very little about the endgame by learning that a cople of moves draw, while all the others lose in 38 moves or less.

In the subsequent analysis I have mainly relied on the lines provided by Averbakh in his five-tomes endgame manual (the best ever published, in my humble opinion). It is remarkable that his analysis does not contain "real" mistakes. Although the tablebases sometimes recommend lines that win in one or two moves earlier (36 instead of 38, for example), I have stuck to the human lines, because they are easier to explain as a part of the same general plan. (In fact, there is a lot to be explained, because Averbakh's lines have huge instructive value, but only if you draw the rules yourself, because the verbal explanations are rather restrained.) Computer lines sometimes rely on casual tactics, which are nice, but are more difficult to remember if you get this endgame in an over-the-board game.

Among other things, chess is supposed to be an exciting, educative and entertaining game, or, to put it briefly, just fun. By watching the "cold" recommendations of the tablebases, one risks getting a feeling of false strength, similar to that of driving a BMW at 200 kms per hour, which has nothing to do with one's physical condition. It tends to generate exaggerated criticism of top GMs, who sometimes overlook "such an obvious win". The notion of "fun" is basically related with the pure human effort in difficult and complex situations; the current position is complex enough, as we shall convince ourselves soon.

Here is a small episode that illustrates quite well what I mean. When this endgame was reached, none other than Garry Kasparov became enthusistiacally involved in the evaluation of the position, mentioning the crucial game Salwe-Rubinstein (see below) and also two of his own games (against Pinter and Jussupow, both played in 1993; more about these games later). The great champion tried to remember how or if at all it could be won, what were the best chances, etc. rather than simply connecting to the tablebases. A human effort from a great player who loves chess passionately...

Let us return to our analysis now. To start with, the diagrammed position is drawn.

Black's only winning chance consists of sacrificing the exchange for the a4-pawn, but he cannot achieve this under favourable circumstances. Things would be different if the pair of pawns would be placed on any other file, but it is known that rook pawns offer additional saving chances in pawn endings. In order to reach the desired result, White has to avoid three (!) main situations.

  1. His king has to avoid being pushed onto the b-file or further.
  2. His king has to avoid being pushed onto the fifth rank or further.
  3. His king has to avoid being trapped in the immediate neighbourhood of the pawn!

The first two points are easy to explain. In case of an exchange sacrifice, the king has to be in time to reach the b1-square. However, the latter aspect is far from obvious and, for many decades, remained unnoticed by theoreticians.

Let us move on now and take contact with the mentioned situations as they arise in some of the sidelines and with the methods of achieving (or, on the contrary, avoiding) certain situations.

46.Kc2 Rh7 47.Ba6 Rc7+

48.Kd1! The only good move! 48.Kb2? would lead to the situation corresponding to point (3), which was met (with reversed wings) in a famous game Salwe-Rubinstein, Prague 1908, quoted in most ending books. Rubinstein managed to win the game, but only by a... mistaken method! He chased the king away from the dangerous zone and eventually succeeded in pushing it until the fifth rank, after which the win was relatively easy.

In 1954, Baranov discovered that Black can avoid being pushed too far and that the logical result would have been a draw. His conclusions are valid up to this moment, but only for the case when the defending king escapes the zone surrounding the pawns. Later, the renown endgame expert Maizelis (who was one of Averbakh's collaborators for the first edition of his books, published in the late '50s), discovered the dangers facing the king in the neighbourhood of the pawns. Lysytzin's famous endgame manual, first published in 1956, only mentions Baranov's conclusions, without showing how Rubinstein could have won... without the help of his opponent. Anyway, after 48...Kd2 49.Kb3 Rc6 50.Bb5 Rc3+ 51.Kb2 we reach a position from Maizelis' analysis.

Analysis diagram

Black's pieces are not optimally placed yet. In order to win, he has to carry out a major regrouping. His king should reach b4, in order to restrict the bishop's mobility, but the enemy king should not be allowed to escape from the dangerous zone at the same time! 51...Rh3 52.Bd7 Rh4 53.Be8 Rb4+ 54.Ka3 Kc3 Black has strengthened his domination, but the b4-square is not available yet. 55.Bf7 Rb1

Analysis diagram

56.Ba2!? White stubbornly keeps his king on a3, keeping b4 defended, but his bishop will not be able to return to the b5-e8 diagonal under favourable circumstances. The voluntary retreat 56.Ka2 leads to a crucial position after 56...Rc1 57.Be8 Kb4 58.Kb2 Rc5

Analysis diagram

Black has completed his regrouping and White will get in successive zugzwang positions. It is typical for such endings that the rook restricts both enemy pieces at the same time. The next step is to push the enemy king on the back rank. 59.Bb5 Rc8 Apart from king retreats, which is precisely what Black aims to provoke, White has only one move: 60.Bd7, but this places the bishop on a vulnerable square, allowing the rook to switch to lateral attacks, without letting the king escape from the corner! This will be an important element in several phases of the line actually played in the game Radjabov-Van Wely. 60...Rd8! 61.Bb5 Rd2+ 62.Kc1 Kc3 Black has made further progress and now the same method as on the previous step ensures him the win. 63.Kb1 Kb3 64.Kc1 Rd8

Analysis diagram

White is in zugzwang again. The only move that maintains the pawn defended and avoids mate in one is 65.Bc6 but this loses the bishop to 65...Rc8.

Let us go back to the position after 56.Ba2. 56...Rc1 (Avoiding the trap 56...Ra1?? with complete domination and... draw by stalemate!) 57.Bb3 Ra1+ 58.Ba2 Kd4. The king retreats in order to enable lateral attacks of the rook. With the a4-pawn vulnerable now, White cannot save the game. 59.Kb2 Re1 60.Bf7 (A desperate attempt to return with the bishop on the b5-e8 diagonal. 60.Bb1 leaves the c4-square undefended, allowing a relatively easy win after 60...Kc4 61.Bf5 Kb4 62.Bd7 Rd1 63.Bb5 Rd2+ and so on, like in the variation starting with 56.Ka2 above. 60.Bb3 keeps the c4-square defended, but deprives the own king of the b3-square and places the bishop on a vulnerable square. 60...Kc5! 61.Kc3 Re3+! With the bishop on a favourable square, the white king could escape now to the right wing, but here this is impossible. 62.Kb2 Kb4 63.Bd1 Rg3 64.Bc2 Rg2 65.Kc1 Kc3 and it is all over.) 60...Re7 61.Bh5 (Or 61.Bg8 Kc5 62.Kc3 Re3+ 63.Kb2 Kb4 with a familiar position.) 61...Kc4 62.Bf3 Kb4 63.Bc6

Analysis diagram

Finally, the bishop has returned on the "good" diagonal, but Black has carried out his essential regrouping in the meanwhile. 63...Rc7 64.Be8 Rc5 65.Bb5 Rc8 and we have reached a familiar zugzwang position.

I will return now to the aforementioned games of Kasparov. In both cases, the situation was slightly different: the defending part had an advantage of space (here, this would mean that pawns are on a5 and a6). If the king is free, the attacker has to push it one file further than in our current game, namely untill the opposite edge of the board.

In Kasparov's game against Pinter, the Hungarian GM managed to avoid this and achieved a draw. Of more interest is the Jussupov game. Just like in the lines above, the king was trapped in the surroundings of the pawns. Apparently, this should be less dangerous, because the king has more space available, but on the other hand the bishop's "good" diagonal is shorter! Curiously, the evaluation remains the same: the side with a rook should win by successive zugzwang positions. For a while, Kasparov conducted his attack in the best way, but then let the enemy king escape, with inevitable draw...

Back to the game in Wijk: 48...Rc6 49.Bb5 Rc3 50.Ba6 Ra3 51.Bb5 Ra2

52.Kc1! We can notice an important element. Being a long-ranged piece, the bishop, too, can act on both wings. Here, the control of the d3-square is as important as the defence of the pawn. For instance, the careless 52.Bc6? would lose to 52...Rd2+! 53.Kc1 (53.Ke1 allows 53...Rc2 with a double (and deadly) threat.) 53...Kd3 followed by ...Kc3.; However, 52.Be8 is playable, too. After 52...Rd2+ 53.Ke1! the bishop is safe on e8, because the interposition of the black king on the e-file prevents the double attack ...Re2+. 52...Kd4 53.Kd1 Kc3 54.Ke1 Rd2 55.Ba6 Rd4 56.Bb5 Kc2

Finally, Black switches plans. He will try to push the enemy king as far as possible. 57.Ke2 Re4+ 58.Kf3 Re7 59.Kf2 Kc3 60.Kf3 Kd4 61.Ba6 Re3+

62.Kf2! One important defensive method is the diagonal opposition. In case of the careless 62.Kf4? Black achieves the situation from point 1) with 62...Re6 63.Bb5 Rf6+ 64.Kg3 when 64...Kc3 followed by the exchange sacrifice on a4 ensures an elementary win. 62...Ke4 63.Bb5 Kf4 64.Ba6 Re4 65.Bb5 Re6 66.Bc4 Re4 67.Bb5 Rd4 68.Ke2 Rd6 69.Bd3 Rh6 70.Bb5 Ke4 71.Bd3+ Kd4 72.Bb5 Rh2+ 73.Kf3 Ra2 74.Bd7 Ra3+ 75.Kf2 Kd3 76.Kf3 Kd2+

77.Kf4! The king has to approach the dangerous 5th rank in order to maintain the diagonal opposition. 77.Kf2? loses to 77...Rd3 78.Bb5 Re3 with zugzwang. If the bishop moves, it will be attacked by the rook, followed by a check on the c-file, while 79.Kf1 Rf3+ immediately drives the king into the forbidden zone. 77...Re3

78.Bb5! Again, the control of the d3-square is essential. After a careless move such as 78.Bc6? Black establishes a frontal opposition with 78...Kd3 79.Bb5+ Kd4 , forcing the king to go upper than he would have wished. 80.Kf5 (80.Bd7 is met by the familiar manoeuvre 80...Re7 81.Bb5 Rf7+) 80...Re4 81.Bc6 Re5+! 82.Kf6 (Or 82.Kf4 Re6!) 82...Re3 83.Kf5 and now Black starts his decisive attack on the opposite wing with 83...Kc5 84.Bb5 Kb4 85.Kf4 Ra3 86.Ke4 Rxa4 87.Bxa4 Kxa4 88.Kd3 Kb3 winning. 78...Re7 79.Kf3 Kc3 80.Ba6 Kd4 81.Bb5 Re6

82.Kf2! Radjabov deals correctly with the last dangerous moment for White in this game. 82.Bd7? loses to 82...Re3+ 83.Kf2 Kd3 and so on. 82...Re5 83.Kf3 Kc5 84.Kf2 Kb6 85.Bd3 Kc5 and, having convinced himself that Radjabov knows "/portals/all/_for_legal_reasons.jpg" about this endgame, Van Wely resigned himself to the inevitable. 1/2-1/2.
  • Click to replay the above analysis and the three quoted games. Note that in our JavaScript replay you can click on the notation on the right and the board on the left will follow the moves.


Discussion and Feedback Join the public discussion or submit your feedback to the editors


Rules for reader comments


Not registered yet? Register