Game 8 
Kramnik vs. Leko

10/7/2004 - Play resumed today, with Kramnik again pushing  1e4. The players debated the Ruy Lopez, Marshall line. 14..Qh3.  First 15 moves busted out real quickly in under 10 minutes. Peter Leko from the black side again sacs a pawn with 8 .. d5. By move 13 Leko is down a pawn but has a lead in development. Kramnik's white queenside has not developed at all. Kramnik decides to try to exchange queens by playing 16 Qf1, it attacks Leko queen on h3. GM Larry Christenson says black has compensation for the pawn.

    As of move 16, the queenside pieces, rook, knight and bishop are still on their original squares. Leko is giving his 16th move a big think. Leko is already 60 minutes behind in time. Leko has spend at least 50 minutes on his 16th move. He must think of how he can use his developmental advantage. Leko said that Kramnik can look the deepest into a chess position. Leko is in a critical position, he is down a pawn, he must make use of his developmental advantage. Leko finally plays 16.. Qh5. Leko must be thinking his chances in the match are dwindling, He only has three more whites left. You might think the match has plenty of chess left. However Kramnik is one of the most difficult players to defeat. So Leko is trying to take advantage of every opportunity.  Kramnik needs to quickly develop his queenside. Leko has 53 minutes for the next 24  moves. Time could become a factor.

    Leko plays 17.. Bf5 attacking Kramnik's rook on e4. However, Kramnik is offering to  sac the exchange with move 18 f3. This move protects the rook and allows Leko to win the exchange. Kramnik is up a pawn and now materially would be down by a pawn equivalent, but he has compensation in the form of two central pawns. Kramnik has a 70 minute time advantage. Leko refuses Kramnik's gift offer and plays 18.. Nf6. Kramnik will have trouble finishing the development of his queenside. He still has a rook and bishop on the original squares.  Leko pieces are massing a kingside attack. Leko is keeping the pressure and tension up. This is a good strategy. Kramnik retreats his rook by playing 19 Re1. Leko challengers the e-file with 19.. Rae8.

   After a 10 minute think, Kramnik exchanges rooks. Leko controls the e-file. Kramnik still has not finished developing his queenside. He is playing down two pieces. All of Leko's pieces have entered the game. Leko time management may become an issue. He has an aggressive attacking position. Can he use his 25 minutes left to break Kramnik? That's the issue at hand. Leko has 23 minutes for 20 moves. Leko plays 21 Qg6 .. maybe his plan is 22 ... Bd3 followed by 23 .. Re2 .. Kramnik responds with 22 axb5.. Does Kramnik realize the danger he is in ? Kramnik is thinking about his runaway passed pawn with bxa6. But Kramnik's queen could be trapped with Leko's Re2.  23.. Re2 played. Kramnik gives up his queen with 24 Qxe2. Leko follows with 24.. Bxe2. Kramnik creates a dangerous passer with bxa6. Leko will have to sac the bishop to stop the pawn.

    This will be an incredibly unbalanced position. Kramnik pieces do not appear to be organized. I think Kramnik has miss calculated.  Not sure Kramnik can organize his pieces correctly. Leko ,down to 8 minutes, is think about 25 .. Qd3 instead of taking Kramnik's dangerous passer on a6.  25 Qd3 could lead to a dangerous mating attack by Leko.  Leko, 8 minutes left, plays 25.... Qd3.  Kramnik needs to give this dangerous position a very long think. His king is under a dangerous attack. If Kramnik can queen his pawn and prevent mate. He could win this game.  The match and the world championship could be riding on the next moves.  Kramnik giving this a huge think right now. Leko clearly will benefit from Kramnik's long think. This is the most exciting game of the match. This is a the kind of position Leko can think about his options on how to mate Kramnik. Kramnik is thinking if he has the time to push a7.  Kramnik realizes he cannot push his passed pawn and plays 26 Kf2. 

    Leko sacrifices his knight busting opens Kramnik's defenses. Kramnik's pieces are unorganized. Kramnik drops his rook. Kramnik is defeated .. he pushes his pawn to a7, however this pawn will be gobbled up by Leko's queen. Kramnik has three minor pieces, two bishops and knight against Leko's queen and bishop. This one is over. Only 31 moves have been played.  Kramnik playing on auto pilot. He plays 32 h4. Leko responds 32..g5 The play has stopped at move 32.  It's over . It's over. A great , great victory from the black pieces by Peter Leko.  

Post game:
 Leko's excerpts : I have quite mixed feelings winning this game. I was clear. I  took on Vladimir 3 times in the Anti Marshall. He tried to switch , Kramnik had so many options. It's hard to guess where to enter the marshal. Vladimir used the most principled and aggressive move to enter the Marshall. One side I am happy he entered the Marshall, other side quite sure he understood the line as well.  F3 absolutely difficult sacrifice - big surprise for me. Bigger surprise a4, very strong move, didn't expect it. Vladimir went with the spectacular queen sacrifice. Looks very strong.  But I have, in every line, special counter play. Qd3 position is winning by force in every line. Back gives mate. This was the move I saw in advance. I couldn't believe it. it works. not a draw, It's winning. Valid Kf2 offered practical chances. I took the simple win with Nxc3, because there was not too much time on the clock. I lost to Kramnik in Linares this year. I lost in a similar way. Qc2 was the decisive move I missed that day. I was quite happy to take revenge in a similar way.   20 minutes to check Qd3. Just win the game, strong position for the queen. I was not sure what to expect. Vladimir playing quickly and confident. I thought maybe he is still in preparation. Maybe even check with the machine.  I was thinking it might be preparation. but on the other hand suck positions are dangerous. Computers will give you white is wining very quickly. Exactly what Vladimir said. I have feeling some thing could go wrong.  You don't have time to check every move. Computers can mislead you, instead of helping you.  They guide you in the wrong direction. Maybe that's whap happened. 

Kramnik's excerpts: I was very disappointed to lose in such a way.  I am not quite sure if it was an analytical mistake or I confused something.  Seems to me it was an analytical mistake, but we thought white was better.  It looks very dangerous. Qd3 big surprise. Maybe I  confused things - confused some moves or lines.  Maybe I confused the wrong move order.  The match situation is worse, ok, but nothing is lost. I still have many games to go. I can win one game and tie the score and win the match. I will try to do my best.  It's very unpleasant to lose,  But it was better to lose this way. I believe this game is beautiful. It going to be remembered as one of the greats games in chess history. 
 Maybe bxa6 is wrong decision. Maybe I should stop and check and calculate the lines. But why give my opponent time. I didn't want to give Leko time to think.  He was in time trouble . Maybe it was wrong practical decision, maybe I should calm down and check al the lines.  To be honest I did not believe our assessment was better for white.  During the game no way for black , expect for some perpetual. I am quite surprised not perpetual , It's just mate. Simply bad luck for me . Good play for Peter.

Replay Game 8
Chessbase analysis
More from Chess 21
 Most Critical position move 26 Kf2 - This was the last deciding blunder according to Crafty
Critical Positions.

lg