Here's what he said about the game:

<Aronian: It didn't go well for me. I didn't expect Volodya to play that opening and went for a very sharp line without any understanding of it. Quite a rash decision, but it turned out that... Volodya knew it all, but not to the end. And somehow I managed to equalise, and then, of course, when Volodya lost a couple of tempos... Well, although the position was still drawn he, no doubt, got upset and didn't notice...

Surov: You're thinking of the moment 32...Rd3 - 33...Ra3?

A: Yes. Because after b6 it seemed to me during the game that it was nearly over. That the knight would take the pawn and that would promote. But there, it turned out, there was an excellent resource, noted by the computer - 38...Rad2! And if 39. Kb7 b3. After that he has the idea of giving a perpetual. It's quite simple, but when you blunder so badly, no doubt, he didn't notice it because of nerves. He was upset.

S: And what did you mean when you said that he was prepared, but not to the end?

A: I mean that it's clear that you usually look at the opening for Black: well, ok, this position is better for Black - well, I won't look any further. And that's correct. He was theoretically prepared, but his game didn't go so well for him today.

S: At Chesspro Sergey Zagrebelny, who commentated on the game live, noted that the move 22...Rfa8 instead of 22...Bh7 was almost winning.

A: No doubt, no doubt. I mean the position's very dangerous. I did everything rashly.>

<So far Kramnik's opening preparation has been flawless and his advantage is close to winning. The combination of passed pawn and wonderful lines for his rooks should be decisive. 22...Bh7?? Here is the guilty party. Black had to play 22...Rfa8! 23.0-0 (The bishop is untouchable since 23.Nxf5? is met with 23...b3! 24.Ne7+ Kf8 25.Kd2 (25.Nxd5 b2 26.Nc3 Rc1 27.Kd2 g5! preparing Rd8+) 25...b2 >

and this on 38...Rfd2 <38...Rfd2?? The wrong rook! 38...Rad2! 39.Nxb7 b3 40.Nc5 b2 41.Bxb2 Forced since the threat was 41.-- Rxg2+ 42.Rxg2 Rxg2+ 43.Kxg2 b1Q >

One idea is to make a decision about which Rook to play and then to choose the other one. because one always chooses the wrong Rook. Admittedly this cannot be called a true answer.

<[Kramnik]: I can’t even understand what happened, because I had the win written down, and I remembered the ideas. I needed to accurately remember everything, to calculate it to the end. To do what I always do. I had lots of time and the position was very simple.>

click for larger view

(position after 21.Nd4)

<You can win in two ways: 21…Rfc8 22. Bd1 Bh7 or 21…Ra1+ 22. Bd1 Rfa8. I made the first move from one win, and the second from the second: 21…Ra1+ 22. Bd1 Bh7. i.e. I chose the only path that didn’t win. Moreover, strangely enough, I’m not even particularly better.>

While Aronian’s description of HIS "preparation" for this game provides an absolutely delightful contrast:

<And do you want me to tell you why I won today? Instead of preparing for the game I walked around the city for two hours – ok, maybe not two, but definitely one and a half – looking for the "Collected Works of Platonov". And when I finally found it I was so happy that I wasn’t thinking anymore who I was playing or how to play. I knew that it’d all be fine.

But why walk around the city for two hours. Aren’t you staying in the Ritz-Carlton?

I am.

Then you could simply cross Tverskaya Street, go to the "Moscow" bookshop, and that’s that.

I looked there. They didn’t have it. There was some sort of junk. Some Platonov, K. While I needed, of course, A.

Andrei?

Who else.

And where did you find what you were looking for?

This suggests two questions:

1. After 21...Ra1+ 22 Bd1 how does 22...Rfa8 improve on 22..Bh7?

2. How does 21...Rfc8 22 Bd1 Bh7 improve on 21...Ra1+ 22 Bd1 Bh7?

One answer to the first question is that with a Rook on a8 instead of a N on h7, on 23 Nb3 Rb1 24 Nd2 Black has 24...Rxd1+! 25 Kxd1 Ra1+