The million dollar Brains in Bahrain match between Vladimir Kramnik and Deep Fritz takes place 2nd-22nd October 2002 (Opening Ceremony 2nd October. Playing days 4, 6, 8, 10, 13, 15, 17 & 19 October, 2002).

Report Ten: Game 6 Report
by Ebrahim Al Mannai

"Rock’n’Roll Chess": Are you there Tal? It’s me, Kramnik.

The day after suffering his first loss in Game 5, Kramnik decided to make a special appearance at the children’s simul given by GM Julian Hodgson here at the match venue. The children were thrilled to meet the world champion, and Kramnik himself seemed happy to be there. He signed autographs and answered their questions, which he said were more interesting than the ones asked by the adult journalists:

Q: Do you ever get tired of chess?

A: Not so much, but it’s like eating rich and delicious food; if you ate it everyday, you would still get tired of it.

GM Julian "Jules" Hodgson giving a simultaneous, photo courtesy of Mig Greengard

Playing a gruelling world-class match against a faceless computer program, with a million dollars at stake, could understandably make one forget the ‘human’ element of chess, and the reason why we choose to spend countless hours staring at 64 squares.

Kramnik answering questions from children. Photo Mig Greengard

After meeting the kids, he seemed moved. He was beaming.

"They helped me remember why I do this, to put it all in perspective."

‘The champ signing T-shirts’ photo courtesy of Mig Greengard

Our man was revitalized. This may partially explain his wild inspiration to play a bloody sacrificial attack against the computer’s king in round 6, producing the most gripping, hair-raising game of the match. Had he not lost, due to a "miracle" defensive move for the computer, Kramnik would have called it "the most brilliant game of my life".

Sacrificing a piece against a computer with no forced win in sight is, in general, a bad idea. As has been repeated time and again, analysis of long complex variations is the computer’s forte. When facing a sharp, double-edged position with many complex tactical variations, a good computer program will usually find the best continuation. Free of ego and psychological pressure, Deep Fritz would simply crank out the possible variations (and sub-variations) at monster speed, give each a quantitive assessment, and then make its choice accordingly- regardless of the overall result.

Of course, Kramnik knows this. So why would he take such a serious risk against Deep Fritz when he had just lost the previous game? Meeting some pre-teen fans the previous day probably did provide a boost to his psyche, but did it inspire him to invoke the spirit of the great Mikhail Tal to give him some ideas during the game?

Kramnik and Frans Morsch at the post-game press conference. Photo Mig Greengard

In a sense, it did. When asked about the brilliant-but-insane 19.Nxf7!?, he said, with a smile (almost apologetically):

"It was so beautiful I couldn’t force myself not to play it. Of course, the result is important. I recognize that this is an important match, but the more I looked at this move..."

He looked down and shook his head, "All these beautiful variations - have not seen such rich, beautiful variations in all my life. I just couldn’t force myself not to play it. "

He went on to say:

"Despite the result, I don’t feel like I fully lost. I am pleased that I played Nxf7 because of the beauty, and I would have even won the game if it wasn’t for this miracle defence, Bh4. If it wasn’t for this move, then this would have been the most brilliant game of my life."

I would claim that this is one his most brilliant games, regardless of the result.

"I played like a man," he said to Danny King, and he did.

Fritz opening specialist, Alex Kure, analyzing the game with GMs Danny King and Julian Hodgson. Photo Mig Greengard

During this wild Queen’s Indian encounter, everybody in the media center was in a frenzy. Grandmasters, chess journalists and amateurs were all playing out the different possible variations. Three different computers had Fritz examining various moves. Julian Hodgson couldn’t sit still. Going back and for The game, with all its unplayed variations, would take days to analyze correctly. However, 36 hours later, that notoriously silly seeker-of-truth, Mig Greengard (avec computer), seems to have worked out the final position- in which Kramnik resigned- as a practical draw!

As I was completing the previous paragraph, GM Christopher Lutz (Kramnik’s second) approached me with the news that Frederic Friedel (Chessbase co-founder/Computer Chess godfather) knocked on Lutz’s hotel room door last night to gently break it to him that his employer resigned in a drawn position. Fred also suggested that he convey the message to Kramnik!

Apparently, both Lutz and IM Tigran Nalbandian (Kramnik’s other second) decided to keep that information away from the boss, in fear that it might ruin his mental preparation for the penultimate game. However, they also fear that Vladimir might hear the jolting news from someone else (it’s all anybody’s talking about in the media center), and come to them for an explanation!

They plan to inform him of his wasted half-point tonight, hopefully after a game 7 win. Here’s the spine-tingling game 6:

Kramnik,V - DEEP FRITZ [E15]
Brains in Bahrain (6), 15.10.2002

1.d4 Nf6 2.c4 e6 3.Nf3 b6 Fritz decides on the Queen's Indian Defence. 4.g3 Ba6 5.b3 Bb4+ 6.Bd2 Be7 7.Bg2 c6 8.Bc3 Forget the whole 'don't-move-the-same-piece-twice' advice here. The dance of bishops is considered theory. 8...d5 9.Ne5 Nfd7 10.Nxd7 Nxd7 11.Nd2 0-0 12.0-0 Rc8 13.a4 Black's next move prepares to hit the d4-pawn with ...c5. 13...Bf6 [ if immediately 13...c5 then 14.cxd5 (14.a5 b5) 14...exd5 15.Bxd5 cxd4 16.Bxd4 Nb8 17.e4 Bxf1 18.Nxf1 Nc6] 14.e4 c5 15.exd5 (15.e5? cxd4) 15...cxd4 So far, so book. 16.Bb4 Re8 17.Ne4 Kramnik spent 42 minutes before making this move. It leads to some spectacular play. 17...exd5 18.Nd6 dxc4 Deep Fritz was anticipating this and was willing to sacrifice the exchange for a super-strong d-pawn, but Kramnik had other ideas... 19.Nxf7!?

Spectacular! Before this move was played Deep Fritz was expecting 19. Bd5! and if 19...Ne5 20. f4! with a clear advantage to White. Here's the actual Deep Fritz analysis: 19.Bd5 Nc5 [ 19...Ne5 20.f4 Nd3 21.Bxf7+ Kf8 ( 21...Kh8 22.Bxe8 ) 22.Qh5 Nxb4 23.Bxe8 ] 20.Bxf7+ Kf8 21.Bxe8 Qxd6 22.Bb5 Bxb5 23.axb5+/- With White's actual move, we're presented with the most fascinating king hunt. 19...Kxf7 20.Bd5+ Kg6 21.Qg4+ Bg5 22.Be4+ [ very interesting was 22.f4 h5! 23.Bf7+ Kxf7 24.fxg5+ Kg8 25.Qxh5 threatening g6 and then Qh7 mate, but after 25...Ne5 26.g6 Nxg6 27.Qxg6 Qd7 seems to hold for Black.] 22...Rxe4 Forced if Black wants to avoid a draw by repetition. 23.Qxe4+ Kh6 [ 23...Kf7 24.Qd5+ Kg6 25.Qe4+ with the choice between perpetual check or the text move] 24.h4 Bf6 25.Bd2+ g5 26.hxg5+ Here, the Deep one took a deep think before committing to 26...Bxg5 27.Qh4+

At the time of playing 19.Nxf7, Kramnik had intended 27.Qe6+ here, but then saw 27...Nf6 28. Qh3+ Kg6! 29. f4 Bh6 30.f5+ Kf7 and dismissed it for the text move (he was under time pressure). It turns out that the more complicated 27.Qe6+ would have resulted in better winning chances for White. Following is the actual analysis done by the commentary team, Deep Fritz, and Kramnik himself, enjoy: [ 27.Qe6+ Nf6 28.Qh3+ ( 28.f4 Bh4!!

The Move that prevented this game from being the brilliant jewel in Kramnik's crown. The bishop is sacrificed in order to clear open the file leading to White's king. This is also why White avoided Qe6+.) 28...Kg6! Kramnik showed that this move gives back the piece but wins for Black- see very bottom of this analysis. ( 28...Nh5 29.f4 Bh4! 30.bxc4 ( 30.g4 Ng3 ( 30...Kg7 31.bxc4) 31.f5+ Kg7 32.f6+ Bxf6 ( 32...Kh8 33.Qxh4 Ne2+ ( 33...Nxf1 34.Rxf1) 34.Kh2 cxb3 ( 34...c3? 35.Bf4 Qf8 ( 35...Nxf4 36.Rxf4) 36.f7 Nxf4 37.Rxf4) 35.Qh6 Qc7+ ( 35...Qf8 36.f7) 36.Bf4 Qf7) 33.Qh6+ Kh8 34.Rxf6) 30...Nxg3 31.Be1 Nxf1 32.Bxh4 Qg8+ 33.Bg5+ Kg7 34.Re1! Rc7 ( 34...Ne3 35.Qd7+ Kh8 36.Qxd4+ Qg7 37.Qxg7+ ( 37.Qxe3) 37...Kxg7 38.Rxe3 Rxc4 39.Re7+ Kg6 40.Rxa7 Rxa4 41.Bd8=) 35.Qh6+ Kh8 36.Kxf1 Qxc4+ 37.Kg1 Qf7= Hodgson) 29.f4 Bh6 30.f5+ Kf7-+ Kramnik]

27...Kg6 28.Qe4+ Kg7 Black saves the king and keeps the piece, but the game isn't over yet. 29.Bxg5 Qxg5 30.Rfe1 cxb3 31.Qxd4+ Nf6 32.a5 [ if 32.Re7+ then 32...Kg6] 32...Qd5?! With his passed b-pawn 2 squares away from promotion, it is (finally) favourable for Fritz to trade queens. 33.Qxd5 Nxd5 34.axb6 axb6 0-1

Final Position. Should White Resign?

Similar to Kasparov's second game against Deep Blue in 1997, Kramnik resigned in what was later suspected to be a drawn position, at least practically against the computer. If there exists a win for Black here, then it requires very subtle understanding of this endgame position, an understanding that the Fritz team doubts the computer to possess.

Kramnik should play: 35.Rxa6 b2 36.Ra7+ Kg6 ( 36...Kf8 Here, White is able to sacrifice a rook for both b-pawns. Leaving R+N vs. R endgame. The h-pawn seems harmless. This still might be Black's best chance! 37.Rd7 Nc3 ( 37...Rc1 transposes to mainline) 38.Rd2 b1Q 39.Rxb1 Nxb1 40.Rb2 Nc3 41.Rxb6 with very difficult winning chances for Black. Practically, it should be a draw whether played by a human or a computer.) 37.Rd7 Rc1 38.Rd6+!

The idea is, in order to avoid perpetual checks the knight is forced to block the check. 38...Nf6 ( 38...Kf7 39.Rd7+ Kf8 40.Rd8+ Kf7 41.Rd7+ Kf6 42.Rd6+ Kg7 43.Rd7+ Kg6 ...etc.) 39.Rdd1 b1Q ( 39...Rc2 40.Kg2 Ng4 41.Kf3 h5 42.Re2 draw) 40.Rxc1 Qxc1 41.Rxc1 with a forced draw!