27 January 2013

Tata Steel Chess, Final Round

Yesterday's win over Hikaru Nakamura assured Magnus Carlsen of first place. When Levon Aronian agreed to a draw with Anish Giri, Carlsen was assured that he would not be sharing first. What remained to determine today was whether Carlson would break Garry Kasparov's 1999 record of 10/13 in this event. Second place, too, remains to be determined.

The games today began at noon Central European Time. As that is 3:00 am in my part of the world, I cherished no illusions about watching these games while in progress. I told my wife, when we wake up, I will check the results. At 7:00 am, I checked the games. Carlsen found himself in some danger with the King's Indian Defense, but Anish Giri's could not find the way. In the end, Giri's king was under attack but had sufficient defense. Magnus Carlsen tied Kasparov's record of 10/13.

Viswanathan Anand got a playable endgame against Wang Hao, but was outplayed and lost. Levon Aronian needs a draw for clear second. Fabiano Caruana is ahead the exchange against Aronian, and both have passed pawns. Aronian has two pawns to Caruana's one.

After watching the game to its conclusion from the point illustrated above, I began to go back through the game from the beginning and make some notes.

Caruana,Fabiano (2781) - Aronian,Levon (2802) [C67]
75th Tata Steel Chess Group A Wijk aan Zee, 27.01.2013

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0–0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8+ Kxd8

White to move

This position in the Berlin Wall has become frequent when Black wants a draw. Indeed, the percentage of draws is quite high. But as this game and others have demonstrated, both sides have chances to play for an advantage.


9.Nc3 is vastly more popular. 9.h3 appears to have been introduced into play by Women's World Champion Nona Gaprindashvili in 1978. The move had been suggested as interesting in Dragoljub Minic's annotations to Gulko -- Bisguier, Sombor 1974 (see Chess Informant 18/264). It may be no more than a move order nuance, but it has been growing in popularity, especially in the past three years.

Caruana has played 9.Nc3 in his previous game, but Aronian has played 9.h3 against Vladimir Kramnik.

Reference game:

Aronian,Levon (2820) - Kramnik,Vladimir (2801) [C67]
Zuerich m Zuerich (4), 25.04.2012

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0–0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8+ Kxd8 9.h3 Ke8 10.Nc3 h5 11.Bg5 Be6 12.b3 Be7 13.Rad1 h4 14.Rfe1 Rd8 15.Rxd8+ Kxd8 16.Ne4 b6 17.Bf4 Kc8 18.Neg5 Bxg5 19.Bxg5 Bd5 20.Nh2 c5 21.Rd1 Bc6 22.c3 a5 23.Ng4 Bd7 24.f3 a4 25.Kf2 ½–½

9...Bd7 10.Nc3 Kc8

10...h6 is more common.


White scores an astonishing 73% after 11.g4, but drawing conclusions from a mere fifteen games would be foolishness. Clearly, 11.g4 is an aggressive try for action on the kingside, but Black's tenth move demonstrates thta his king seeks sanctuary on b7.

11...b6 12.Bb2

Black to move


12...h6 could have transposed into a more frequent position.

13.Rad1 Nh4

This position had been played twice before, and White won both games. Presumably, Aronian has done some preparation and developed some ideas.

14.Nd4 h5

White to move

This position is unique in the annals of chess history. Did Aronian prepare it? There are imbalances that allow both sides to struggle for advantage.

15.Kh2 Kb7 16.g3 Ng6 17.Rfe1 c5 18.Nf3 Bf5 19.Rd2 Rhd8 20.Rde2 Nf8 21.Ne4 Ne6

White to move

It appears to me that both players have sought to improve their pieces. My chess engine, Stockfish 2.3, considers this position as favoring Black.

22.Bc1 a5 23.Neg5 Bxg5 24.Bxg5 Rd5 25.Be3 a4

White to move


The players are in a complicated positional struggle. What alterations of the pawn structure might be beneficial? What minor piece exchanges are worth pursuing?

My chess engine prefers 26.Ng5 and evaluates White's position as deteriorating after the move played by Caruana. Do engines understand this sort of position?

The text move allows Aronian to create a material imbalance, giving up a rook for a minor piece and two pawns.

26...Rxe5 27.f4 Rxe3 28.Rxe3 Bxc2 29.bxa4 Rxa4 30.a3 Nd4 31.Nf3 Bf5 32.Rc1 Be6 33.Nd2 c4 34.g4 hxg4 35.hxg4

Black to move

Black would seem to have a clear advantage. His queenside pawns are compact and mobile. White's a-pawn may be difficult to defend.

35...Nb3 36.Rc2 Nxd2 37.Rxd2 Bxg4 38.Rg2 Bh5 39.Rxg7 Ra5 40.Kg3 Bg6 41.Rc3 Rc5 42.Rg8 b5 43.Re8 Rd5 44.Kf2 Bd3 45.Ke3 c5 46.Rc1

Black to move


My engine favors the immediate 46...b4 here. It's evaluation suggests that here Black is letting the advantage slip away. In the final press conference, Magnus Carlsen said that he got the most from each game throughout the tournament. Aronian, who finished 1 1/2 points behind, with a score that has often been good enough to win the event, had two draws in games where at one point he had a decisive advantage.

Had he been able to solve over the board problems that Anand and his team prepared for the World Chess Championship match with Gelfand, but that was used in Wijk aan Zee instead, Aronian might have avoided his one loss. Carlsen called that loss the best game of the event.

47.Re7+ Kb6 48.Rxf7 Rh3+ 49.Kf2 b4 50.Re1 Rh2+ 51.Kf3 Rh6 52.axb4 cxb4 53.f5 b3 54.Rf8 Kc5 55.f6 Kd4 56.Kf4 Kc3 57.Ke5

Black to move 

It was from this position that I began watching the game more or less continuously. Both players, it seems to me, have problems to solve. Neither faces problems that strike me as insurmountable. The key positions, I now realize, occurred while I slept. Aronian managed to create an apparently winning position, and then let the win slip from his grasp.

57...Rh5+ 58.Ke6 Kd2 59.Rg1 Rh6 60.Rd8 b2 61.Ke7 Rh7+ 62.f7 c3 63.Rg2+ Ke3 64.Rb8 Bc4

White to move

Looking at this position on the Tata Steel Chess website during the game, I told my beloved wife that Caruana could trade a rook for two pawns, losing his pawn in the process. The result would be a theoretically drawn endgame of rook and bishop against rook.

Almost as soon as I subjected my poor wife to my patzer commentary, the website showed that three rapid moves had been played, and my fantasy position appeared on the board.

65.Rbxb2 Rxf7+ 66.Kd6 cxb2 67.Rxb2 Kd4

White to move

Inasmuch as I thought it was well-known that rook and bishop vs. rook is drawn, I cannot comprehend why an additional 36 moves were played. It may be because the theoretical draw was not without checkmate threats, and Caruana appeared to be playing with approximately three minutes on his clock, with thirty seconds added per move. Perhaps Aronian wanted to test his opponent's tactics in blitz mode. Perhaps Aronian was frustrated at having let the win slip away.

68.Rd2+ Bd3 69.Rd1 Rf6+ 70.Ke7 Ra6 71.Kd7 Ke4 72.Ke7 Bb5 73.Rc1 Kd5 74.Rd1+ Ke5 75.Re1+ Kf5 76.Re3 Bc6 77.Re1 Be4 78.Rf1+ Ke5 79.Re1 Ra7+ 80.Ke8 Kd5 81.Kf8 Bf5 82.Re7 Ra6 83.Re1 Be4 84.Re2 Re6 85.Ra2 Ke5 86.Ra7 Rb6 87.Re7+ Kf4 88.Ra7 Bd5 89.Rc7 Ke5 90.Ke7 Rh6 91.Kd8 Bc6 92.Re7+ Kd5 93.Rg7 Re6 94.Kc7 Kc5 95.Rg5+ Bd5 96.Rg7 Ra6 97.Kd8 Kd6 98.Rd7+ Ke5 99.Rc7 Ra5 100.Re7+ Be6 101.Rb7 Rc5 102.Ke7 Rc1 103.Ra7 Bc8 104.Ra5+ ½–½

I knew that rook and bishop against rook is a theoretical draw while watching the game, but in fact there are many positions that are not drawn. None of these positions occurred in this game, but a few were threatened. Moreover, the weaker side's defense can prove difficult. In practice, these ending frequently are won. I expect that I will offer another post on this ending in the coming days.

1 comment:

  1. Yes, Rook vs Rook and Bishop is a theoretical draw, but almost all grandmasters play it out because drawing is apparently not trivial. Especially if, as you note, the defender doesn't have much time. There aren't easily remembered set-ups (like Philidor's position in R + P vs R endings) to help the defender navigate.

    I think Kasparov even played out a Rook and Knight vs Rook ending once versus Judit Polgar and managed to win when Polgar slipped?