SuperPatzer77

Perhaps his mistake in his previous game against Carlsen had an effect on his play. To borrow the words of Bent Larsen (speaking of his win against Bronstein in his book of his games) some games are worth more than one point.

Kinghunt: An interesting line we could have seen is 72. Bd5 Kxh3 73. Ke2 Kxg4 74. Kxd2, leading to the following position:

click for larger view

We are in tablebase territory now, so it's possible to do definitive analysis which I think helps establish just how tricky it is to convert this. First of all, the position is in fact objectively won, but there are several only moves black would have to find.

First only move: <74...Kg3!> Okay, not that hard to find. But the tempting 74...h5?? fails to win. Only 74...Kg3 preserves the win.

So, you found and played 74...Kg3. White responds with <75. Ke3>. We have another only move here. Can you find it? The move is <75...h5!>.

White responds to 75...h5 with <76. Ke4>. And guess what? Yep, yet another only move. <76...h4!> is critical.

White meets 76...h4 with <77. Kf5>. Another only move: <77...h3!>.

White responds with 78. Kxf6. Only move <78...Kf4!.> And only from here is it relatively straightforward.

<anandrulez: GM Nunn from official site : Carlsen was pleased with the win but said that Kramnik had to play very carelessly to lose this endgame. He pinpointed 59.Bb7 as the final error, because after 59...Kd2 White is in zugzwang. "If he had played 59. Bc8! I don't see how I can win. I would have had to bring my king back and play for ...d4, but I don't think it is enough. I have plenty of time to try, though." Earlier in the game, Carlsen described 16...b5 as "a trap. He missed that if 19.Nxd7 I have this very nice move 19...Bb7!!. Then after 22.Qxb5 he offered a draw, but I have played Kramnik often enough to know what that means - I must play on. "I was disappointed to see 24.Qe3 - I didn't see any way to win after that. But he really played the endgame carelessly. I think he could have played 44.Ke6 - after I get ...g5+ I at least have winning chances.">

Eyal: <He [Carlsen] pinpointed 59.Bb7 as the final error, because after 59...Kd2 White is in zugzwang. "If he had played 59. Bc8! I don't see how I can win.">

The point is that after 59...Kd2:

click for larger view

White has to keep his bishop on the long diagonal to keep attacking d5, otherwise ...Nd1 either wins the e3 pawn (similarly in case of 60.Kf2, when it comes with check) or achieves a passer that would cost White his bishop after 60.e4 d4. But after 60.Bc6 (or Ba8, for that matter) Ke1 he can't put the bishop on a6, to stop the invasion of the black king to f1.

However, if the bishop really must shuttle between a6 & c8 as long as the black king is within one move range from d2, it doesn’t look as if White could hold on much longer even after 59.Bc8 before Black induces the desirable zugzwang – e.g. 59....Kc1 60.Ba6 Kc2 61.Bc8 Kd1 62.Ba6 Ke1 and White has to remove the bishop from the a6-f1 diagonal and allow the invasion to f1.

60. ...Nd1+ will no longer lead to loss of the e3 pawn because of Ke2.

Yeah, but 60...Kd2! is again winning in such a case (the Nd1 threat becomes decisive)

...Kramnik also made a simple tactical blunder by taking a poisoned a-pawn, and tried offering a draw in a bad position after Qb5!

(and their Kts in slightly bigger ones)

This suggests the paradox that the "careful" move 44 h3 which keeps the Black Knight our of g4 is an example of careless play.