Game 9: Anand-Kramnik: Live blogging with regular updates
3:30 Post-Press Conference Update:
Anand: "I must have done something wrong in the opening." He realized this after 16.f4 Qg7. He thought/expected 18...gxf4 instead of 18...cxd4, but acknowledged that 18...cxd4 19.Qxc4 a5 was also strong. He believed that 20.Kh1 was the only move. Later in the game, he didn't think Kramnik's 33...f5 was precise, though it wasn't clear how Kramnik could have broken through. Once he played 38.Rd7 he was fine, but then 38...a4 threw him off. On move 40 he considered 40.Rd1 b2 41.Kg2 Rc1 42.Rdd7, but with so little time on his clock he didn't want to do this. He saw the bishop and two pawns for rook and pawn ending, and since he had that once before, in a game with van Wely, he went for that. It's a draw - White puts his bishop on the long diagonal and there are too many squares for Black to chase it off. In the game, Kramnik's ...h5-h4 set a nice trap: if White plays 43.Kf2, then 43...Ra8 suddenly creates some danger. After 43.Rc6, however, it's just a draw.
Kramnik: He finally felt he had some chances and played well in general. Anand's Qc2, Rd1 plan fairly new but isn't considered very good on account of ...Nh5, but he decided he was tired of "hour and a half lessons" and time trouble going into Anand's preparation, so he decided to play "something else". [By this he meant anything else, as long as it's reasonable and not the main line of Anand's prep.] Later in the game, he agreed that Anand's 20.Kh1 was the only move, and after that he couldn't find anything better than what he did. Later, he didn't understand why White gave up the "b-pawn". [I think he's referring to 23.a4 and the sequence culminating in 26...Rxb5.]
His explanation of the unfortunate 35...Qc7 was very interesting and all too human. He was told after the game by someone that 35...Rg8 was "just winning", but the move he had been considering instead was the rather unfortunate 35...f4. After thinking about it for a couple of minutes(!), he realized that it allowed 36.Qh7#, and thus 36...Qc7: "at least I won't get mated".
After 37.Bc4, a4 was the last chance. As for 41...h5, it had a couple of small traps, of which he mentioned the one Anand did with 43.Kf2 Ra8, when it is "getting hot" for White. In generaly, he was happy with his play, but was "not lucky enough to win the game."
Q & A:
Among the more interesting questions, Anand was asked if his errors were partly due to the pressure of the situation. Anand said it was "better not to go there". Maybe later, but "definitely today I didn't play very well." Someone asked Kramnik if he's playing better, as the last two games he's succeeded in getting "his" positions. Kramnik replied that it's not about style but the quality of positions. He doesn't think he's necessarily playing better, but is getting better positions. It's easier to play when one isn't in a difficult position with less time. At any rate, he affirmed that he will fight to the end.
And with that, this long post concludes. Later today I should have the "regular" post with game annotations, and tomorrow morning we'll have live-ish coverage of game 10.
2:55
45...Rxc4 Draw agreed
That makes the match score 6-3, Anand, and he now needs just one draw in the last three games to retain his title. Stay tuned for the post-press conference update!
2:50
After half an hour, Kramnik came up with 41...h5, perhaps hoping to improve the situation on the kingside before recouping the piece. Whatever the story, it doesn't avoid a hopelessly drawn finale. Anand played 42.Kg2, getting the king off the back rank, and the continuation has been 42...h4 43.Rc6 hxg3 44.hxg3 Rg8 45.Rxe6. Lots of other approaches drew too - 45.Rxb3 Rgxg3+ 46.Kf2 Rxb3 47.Rxb3 Rxb3 48.Rxe6 would be one extremely comfortable choice - but this is certainly good enough.
Times, not that they really matter at this point, are 44 minutes for Anand, 25 and counting for Kramnik.
2:30
Kramnik is thinking hard, trying to find some way of drawing blood from a stone, but it doesn't seem there's any way to do it. We're still in the same position after 41.Rb2, and as long as White doesn't fall for the trick mentioned in the last update (41...Rc2 42.Rxc2 bxc2 43.Bf1? Rb2!!-+) it should be a routine draw (either by 43.Bb5 Rxb5 44.Rxc2 or 43.Bxe6 Rb1+ 44.Kg2 c1Q 45.Rxc1 Rxc1 46.Bxf5).
Times: Anand 58 minutes, Kramnik 33 minutes and counting.
2:05
The time control has been made, and it seems the position is slightly more interesting than it seemed at first glance. White grabbed the piece with 39.Rxc7 - probably an inaccuracy, as 39.Ra1! draws without allowing the tricks in the game - and after 39...axb3 40.Rf2 Rb8 the time control was made. Anand had two minutes left at the end, Kramnik one.
Anand just played 41.Rb2, and now there's a trap: 41...Rc2 42.Rxc2 bxc2 43.Bf1, expecting 43...Rb1 44.Rxc2 Rxf1+ 45.Kg2 with a probable draw, is refuted by the brilliant 43...Rb2!! White's pieces are stuck, and Black will play ...e5, bring his king over and up, and win. However, with 43.Bb5! White survives: 43...Rxb5 44.Rxc2 and although Black can push after 44...Kg7, it should still be a draw.
1:55
It's not official, but it might as well be: the position must be a draw now. After 35.Qb7? Kramnik played the incomprehensible 35...Qc7? (?? given the match situation), Anand happily traded queens - 36.Qxc7 Bxc7 and now the position is really a simple draw: opposite-colored bishops, active pieces and the e6 and f5 pawns as targets leave Black without any winning chances. If he were up in the match, the safety-first 35...Qc7 would be fine, but not here - especially since both 35...Bc7 and 35...Rg8 gave Black really dangerous attacking chances.
Anyway, the game has continued 37.Bc4 Re8 38.Rd7 a4 (yes, the bishop on c7 is hanging but no, it doesn't change the outcome of the game), and the game will be drawn unless some sort of nuclear-sized meltdown occurs.
1:45
I think Kramnik's edge is dissipating, a point confirmed by Seirawan, the engines, and especially Kramnik's body language. Here are the moves since last time: 30.Qf3 Rc5 31.Bd3 Rc3 32.g3 Kh8 33.Qb7 f5 34.Qb6. Both players have 10 minutes left, and now 34...Qe5, with the idea of ...Rg8 and blowing up g3 would be nice, if it didn't fail to 35.Bxf5! Uh oh, he has played 34...Qe5, and now we'll see if Anand finds it. If he does, then Kramnik's winning chances are gone, kaput, over.
Nope! Anand played 35.Qb7, and now he might be lost: 35...Bc7 and Black's attacking prospects are terrific. Well, that's time pressure for you: the players are down to 8 minutes each, and I'm going to post this.
1:25:
An important development: Kramnik has cashed in his chips, winning a pawn with 25...Bxb5 26.axb5 Rxb5. The drawback of this idea is that opposite-colored bishops arise, and this at least seems to favor White in two ways. First, of course, opposite-colored bishop endings tend to be pretty drawish. Second, in a middlegame, opposite-colored bishops tend to favor the side with chances for an attack, and it seems at least superficially that Black's king is looser than White's.
Yet after 27.Be4 Bc3! 28.Bc2 Be5 the momentum is very much in Black's favor. White is not active, and after the further moves 29.Qf2 Bb8 it's clear that Black has the attacking chances for the moment, not White. Kramnik is looking much more lively and confident at the board, so this could be a tough day for Anand.
1:15:
The last few moves have been eventful: after a long think - 24 minutes - Kramnik played 20...Rac8! There followed 21.Qxd4 gxf4 22.Bf3 Ba6 23.a4 Rc5 24.Qxf4 Rxe5, and Kramnik's edge seems a solid one. White's queenside majority is completely stymied and Black has the bishop pair, so it's not an easy or pleasant position for White to hold. Anand just now played 25.b3, reasonably getting the pawn away from the Qg7's eyeballing.
Times: Both players have 22 minutes left to complete their 40th moves.
12:50
A new move and new news. The new move was the wholly unexpected 20.Kh1, getting out of the way of any a7-g1 pins after something xd4 Rac8 and ...Bc5. It also has the idea of playing Bf3 and meeting ...Bxf3 with gxf3 and Rg1. Anand now has 27 minutes left for 20 moves in a pretty sharp position, so Kramnik is in his best position of the match.
As for the bombshell mentioned in the last update, Rogers responded to my question (probably asked by plenty of others too) that it was just the speculation of one person in the press room, though he felt that it made sense, as Kramnik has nothing left to achieve in the chess world.
Analysis time. The most natural move for Kramnik is 20...gxf4, and after 21.Bf3 to play 21...Bxf3 22.Rxf3 Qxe5 with a practically won game. The bad news for Kramnik in that line is 22.gxf3, when tactics let White keep the balance after 22...Qxe5 23.Nxd4 Rac8 24.Nc6. Trying to sidestep this with 21...Rac8 doesn't give an advantage either: White equalizes with 22.Qxd4 (22...Ba6 23.Qb6 Bxb5 24.Qxb5 Rc5 25.Qa4 Rxe5=).
It does look like Black can maintain an edge with the immediate 20...Rac8, however: 21.Qxd4 Bc5 22.Qa4 gxf4 23.Bf3 Bxf3 24.Rxf3 (here 24.gxf3 makes less sense, as g1 is covered) 24...Qxe5 25.Qxf4 Qxf4 26.Rxf4 Rfd8 and White has an unpleasant defensive chore ahead of him.
Times: Anand 27, Kramnik 44 and counting.
12:30
No new moves, but GM Ian Rogers, speaking with GM Seirawan in the Foidos English commentary room, dropped a bombshell. There's a rumor that Kramnik might retire from chess if he loses this match. I'd love to know more about this, if he's just picking up some individual's wholly unsubstantiated conjecture, if there's anything more to it than this. I'll try to find out.
Times: Anand 31 minutes and counting, Kramnik 57.
12:15
As noted at the end of the last update, Kramnik played 17...c5. Anand played the predicted 18.Nxb5, but instead of 18...gxf4 Kramnik played the better 18...cxd4. Here Anand had at least three moves to choose from - 19.Bf3, 19.Bxc4 and 19.Qxc4, and as you can get guess from the bold typeface, he chose the latter. Kramnik responded with the accurate 19...a5, and seems to have the advantage here - both on the board and on the clock: Anand has 46 minutes and counting, Kramnik 57. The position is closer to equal than winning for Black, but at least he has a little something to hope for.
12:00
The game is at another critical moment. Anand played 17.e5 after 17 minutes' thought, and now Kramnik has thought for even longer - 20 minutes so far. There's a lot to think about: 17...f5, 17...c5, 17...Bxc3 and 17...Qh7 are all candidates. Here are some lines:
(1) 17...f5 18.exf6 Qxf6 (I found a scary-looking draw after 18...Rxf6: 19.fxg5 Qxg5 20.Ne4 Rxf1+ 21.Rxf1 Qg7 22.Nf6+ Kh8 23.Nh5! Qxd4+ 24.Kh1 Rf8 25.Qg6 Rxf1+ 26.Bxf1 Bc5 27.Qe8+ Kh8 28.Qf7+ Kh8 29.Qe8+ with perpetual) 19.fxg5 Qxg5 and now 20.Qe4 or 20.Bf3 is roughly balanced.
(2) 17...c5 18.Nxb5 gxf4 19.Bf3 Bxf3 20.Rxf3 a6 21.Nd6 cxd4 22.Nxc4 and I think White has a slight edge.
(3) 17...Bxc3 18.Qxc3 (or 18.bxc3 c5 19.Bf3 Bxf3 20.Rxf3 cxd4 21.Rxd4 Rfd8 22.Qd2 Rxd4 23.Qxd4 g4 24.Rg3 with full compensation) 18...f5 19.exf6 Qxf6 20.fxg5 Qxg5 21.Bf3=.
(4) 17...Qh7 18.Ne4 Kh8 19.f5 Be7, and although White is going forward like Tolush, his advance troops may find themselves dying like Frenchmen in Russia: 20.Nc5 Bxc5 21.dxc5 Rae8 22.Bh5 exf5 23.Qc3 Kg8 24.Rd7 Bc8 25.Rxa7 Qg7 -/+ and White's position is collapsing.
Note: Kramnik has just played 17...c5, after 25 minutes' thought. Times remaining: Anand 1:04 and counting, Kramnik 1:00. Here's the current position:
11:15
A little unexpectedly, Kramnik has followed up 13...Nxe5 14.Bxe5 O-O. With the bishop on b4 rather than g7, this looks a little scary, but he does have a few things in his favor. First, now that White has castled, h4 isn't the worry it would have been. (It might get played anyway, but it's not as dangerous as it would have been.) Second, the thematic 15.f4 is well met by 15...Nd7, taking advantage of the Be5. Third, White's heavy pieces aren't able to get to the kingside very easily. Something like b3 would be nice, to free up d3 for a rook lift, but 15.b3 does involve a pawn sac (15...Bxc3 16.Qxc3 Nxe4) or surrendering the strong bishop for Black's knight (15...Bxc3 16.Bxf6). Another idea, in addition to 15.h4 and 15.b3, is 15.Bg3 with the ideas of h4 or f4. Maybe then Black can play 15...Rad8 and meet 16.f4 with 16...Bxc3 17.bxc3 c5. Anand has spent 16 minutes on this move (and counting - he's at 1:23 now), so it's clearly an important moment.
Here we go - literally just as I was mousing towards "finish", Anand played 15.Bxf6, and after 15...Qxf6 followed up with 16.f4. I had briefly thought about taking and playing 16.e5, with the idea of 17.Ne4, but thought that after 16...Qf5 White wouldn't have anything special. That's probably right, but the 16.f4 follow-up had eluded my attention. Black's options here include 16...gxf4 (not so much winning a pawn but reducing White's central pawn horde so he has enough room to breathe) and 16...Qg6 (preventing e5 for the moment and getting the queen off the f-file).
Instead, again just instants before I posted, Kramnik played 16...Qg7, allowing 17.e5 or 17.fxg5 followed by 18.e5. It looks interesting, but I'm going to post now before someone moves again.
Times: Anand 1:20 and counting, Kramnik 1:25.
[Note: Regarding a reader's question, there are no offstage monitors - the players must check back on the stage to know when it's their move.]
10:45
Watching this game on Foidos, I note that Anand spends almost every minute at the board, while Kramnik is often bounces up like a jackrabbit at every opportunity, and it often costs him a minute or more after Anand makes his move. Considering that this happens 15-20 times a game, that's a pretty considerable loss of time in every game!
On to the moves: 11...Bb4 12.Ne5 Qe7 13.O-O. Kramnik's 11th move was also somewhat rare and unusual in these Anti-Moscow Gambit lines, and his 12th move was a novelty that revealed the purpose behind his 11th move. Previously 12...Rg8 and 12...Nxe5 had been played in games with reasonably high-rated players.
The queen looks a little funny on e7. It protects the f-pawn so Black can castle long, but aside from that detail Black's queen would be happier on b6. On b6 it takes aim at White's d-pawn, which will also come under fire from one or two rooks on the d-file. More importantly, the queen on b6 shores up Black's queenside pawns, which are liable to come under assault from moves like b3, a4 and d5.
Here's the current position:
Times: Anand 1:40, Kramnik 1:37 and counting. (Incidentally, it took Kramnik four minutes to return to the board after 13.O-O, which I note not because of toiletgate style idiocy, but to point out again just how much time Kramnik is wasting with pacing.)
10:10
It turns out that due to the switch in Germany to daylight savings time, today's game started an hour later than usual, relative to those of us who haven't yet made the same switch. (Too bad for those of us didn't know in advance - the extra hour's sleep would have been nice!) The score is 5.5-2.5, and my expectation is that whatever slim hopes Kramnik may have had left after yesterday's game, and it's entirely possible that today and tomorrow the players will have a couple of short draws just to get it over with. We'll see!
Here are the opening moves: 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 c6 5.Bg5 h6 6.Bh4 dxc4 7.e4 g5 8.Bg3 b5 9.Be2 Bb7 10.Qc2 Nbd7 11.Rd1.
The first thing I have to say is this: Kramnik is going out swinging! Choosing the Semi-Slav will at least make it difficult for White to dry things up, but obviously Black is taking serious risks in this opening as well - especially when it's Anand's main opening with Black. Anand has chosen a sideline with 10.Qc2 and 11.Rd1, and while the plan isn't completely new, it is comparatively rare and Kramnik is starting to think here.
I'm glad Kramnik is making a fight, but have a feeling the match will end today.
Anand: "I must have done something wrong in the opening." He realized this after 16.f4 Qg7. He thought/expected 18...gxf4 instead of 18...cxd4, but acknowledged that 18...cxd4 19.Qxc4 a5 was also strong. He believed that 20.Kh1 was the only move. Later in the game, he didn't think Kramnik's 33...f5 was precise, though it wasn't clear how Kramnik could have broken through. Once he played 38.Rd7 he was fine, but then 38...a4 threw him off. On move 40 he considered 40.Rd1 b2 41.Kg2 Rc1 42.Rdd7, but with so little time on his clock he didn't want to do this. He saw the bishop and two pawns for rook and pawn ending, and since he had that once before, in a game with van Wely, he went for that. It's a draw - White puts his bishop on the long diagonal and there are too many squares for Black to chase it off. In the game, Kramnik's ...h5-h4 set a nice trap: if White plays 43.Kf2, then 43...Ra8 suddenly creates some danger. After 43.Rc6, however, it's just a draw.
Kramnik: He finally felt he had some chances and played well in general. Anand's Qc2, Rd1 plan fairly new but isn't considered very good on account of ...Nh5, but he decided he was tired of "hour and a half lessons" and time trouble going into Anand's preparation, so he decided to play "something else". [By this he meant anything else, as long as it's reasonable and not the main line of Anand's prep.] Later in the game, he agreed that Anand's 20.Kh1 was the only move, and after that he couldn't find anything better than what he did. Later, he didn't understand why White gave up the "b-pawn". [I think he's referring to 23.a4 and the sequence culminating in 26...Rxb5.]
His explanation of the unfortunate 35...Qc7 was very interesting and all too human. He was told after the game by someone that 35...Rg8 was "just winning", but the move he had been considering instead was the rather unfortunate 35...f4. After thinking about it for a couple of minutes(!), he realized that it allowed 36.Qh7#, and thus 36...Qc7: "at least I won't get mated".
After 37.Bc4, a4 was the last chance. As for 41...h5, it had a couple of small traps, of which he mentioned the one Anand did with 43.Kf2 Ra8, when it is "getting hot" for White. In generaly, he was happy with his play, but was "not lucky enough to win the game."
Q & A:
Among the more interesting questions, Anand was asked if his errors were partly due to the pressure of the situation. Anand said it was "better not to go there". Maybe later, but "definitely today I didn't play very well." Someone asked Kramnik if he's playing better, as the last two games he's succeeded in getting "his" positions. Kramnik replied that it's not about style but the quality of positions. He doesn't think he's necessarily playing better, but is getting better positions. It's easier to play when one isn't in a difficult position with less time. At any rate, he affirmed that he will fight to the end.
And with that, this long post concludes. Later today I should have the "regular" post with game annotations, and tomorrow morning we'll have live-ish coverage of game 10.
2:55
45...Rxc4 Draw agreed
That makes the match score 6-3, Anand, and he now needs just one draw in the last three games to retain his title. Stay tuned for the post-press conference update!
2:50
After half an hour, Kramnik came up with 41...h5, perhaps hoping to improve the situation on the kingside before recouping the piece. Whatever the story, it doesn't avoid a hopelessly drawn finale. Anand played 42.Kg2, getting the king off the back rank, and the continuation has been 42...h4 43.Rc6 hxg3 44.hxg3 Rg8 45.Rxe6. Lots of other approaches drew too - 45.Rxb3 Rgxg3+ 46.Kf2 Rxb3 47.Rxb3 Rxb3 48.Rxe6 would be one extremely comfortable choice - but this is certainly good enough.
Times, not that they really matter at this point, are 44 minutes for Anand, 25 and counting for Kramnik.
2:30
Kramnik is thinking hard, trying to find some way of drawing blood from a stone, but it doesn't seem there's any way to do it. We're still in the same position after 41.Rb2, and as long as White doesn't fall for the trick mentioned in the last update (41...Rc2 42.Rxc2 bxc2 43.Bf1? Rb2!!-+) it should be a routine draw (either by 43.Bb5 Rxb5 44.Rxc2 or 43.Bxe6 Rb1+ 44.Kg2 c1Q 45.Rxc1 Rxc1 46.Bxf5).
Times: Anand 58 minutes, Kramnik 33 minutes and counting.
2:05
The time control has been made, and it seems the position is slightly more interesting than it seemed at first glance. White grabbed the piece with 39.Rxc7 - probably an inaccuracy, as 39.Ra1! draws without allowing the tricks in the game - and after 39...axb3 40.Rf2 Rb8 the time control was made. Anand had two minutes left at the end, Kramnik one.
Anand just played 41.Rb2, and now there's a trap: 41...Rc2 42.Rxc2 bxc2 43.Bf1, expecting 43...Rb1 44.Rxc2 Rxf1+ 45.Kg2 with a probable draw, is refuted by the brilliant 43...Rb2!! White's pieces are stuck, and Black will play ...e5, bring his king over and up, and win. However, with 43.Bb5! White survives: 43...Rxb5 44.Rxc2 and although Black can push after 44...Kg7, it should still be a draw.
1:55
It's not official, but it might as well be: the position must be a draw now. After 35.Qb7? Kramnik played the incomprehensible 35...Qc7? (?? given the match situation), Anand happily traded queens - 36.Qxc7 Bxc7 and now the position is really a simple draw: opposite-colored bishops, active pieces and the e6 and f5 pawns as targets leave Black without any winning chances. If he were up in the match, the safety-first 35...Qc7 would be fine, but not here - especially since both 35...Bc7 and 35...Rg8 gave Black really dangerous attacking chances.
Anyway, the game has continued 37.Bc4 Re8 38.Rd7 a4 (yes, the bishop on c7 is hanging but no, it doesn't change the outcome of the game), and the game will be drawn unless some sort of nuclear-sized meltdown occurs.
1:45
I think Kramnik's edge is dissipating, a point confirmed by Seirawan, the engines, and especially Kramnik's body language. Here are the moves since last time: 30.Qf3 Rc5 31.Bd3 Rc3 32.g3 Kh8 33.Qb7 f5 34.Qb6. Both players have 10 minutes left, and now 34...Qe5, with the idea of ...Rg8 and blowing up g3 would be nice, if it didn't fail to 35.Bxf5! Uh oh, he has played 34...Qe5, and now we'll see if Anand finds it. If he does, then Kramnik's winning chances are gone, kaput, over.
Nope! Anand played 35.Qb7, and now he might be lost: 35...Bc7 and Black's attacking prospects are terrific. Well, that's time pressure for you: the players are down to 8 minutes each, and I'm going to post this.
1:25:
An important development: Kramnik has cashed in his chips, winning a pawn with 25...Bxb5 26.axb5 Rxb5. The drawback of this idea is that opposite-colored bishops arise, and this at least seems to favor White in two ways. First, of course, opposite-colored bishop endings tend to be pretty drawish. Second, in a middlegame, opposite-colored bishops tend to favor the side with chances for an attack, and it seems at least superficially that Black's king is looser than White's.
Yet after 27.Be4 Bc3! 28.Bc2 Be5 the momentum is very much in Black's favor. White is not active, and after the further moves 29.Qf2 Bb8 it's clear that Black has the attacking chances for the moment, not White. Kramnik is looking much more lively and confident at the board, so this could be a tough day for Anand.
1:15:
The last few moves have been eventful: after a long think - 24 minutes - Kramnik played 20...Rac8! There followed 21.Qxd4 gxf4 22.Bf3 Ba6 23.a4 Rc5 24.Qxf4 Rxe5, and Kramnik's edge seems a solid one. White's queenside majority is completely stymied and Black has the bishop pair, so it's not an easy or pleasant position for White to hold. Anand just now played 25.b3, reasonably getting the pawn away from the Qg7's eyeballing.
Times: Both players have 22 minutes left to complete their 40th moves.
12:50
A new move and new news. The new move was the wholly unexpected 20.Kh1, getting out of the way of any a7-g1 pins after something xd4 Rac8 and ...Bc5. It also has the idea of playing Bf3 and meeting ...Bxf3 with gxf3 and Rg1. Anand now has 27 minutes left for 20 moves in a pretty sharp position, so Kramnik is in his best position of the match.
As for the bombshell mentioned in the last update, Rogers responded to my question (probably asked by plenty of others too) that it was just the speculation of one person in the press room, though he felt that it made sense, as Kramnik has nothing left to achieve in the chess world.
Analysis time. The most natural move for Kramnik is 20...gxf4, and after 21.Bf3 to play 21...Bxf3 22.Rxf3 Qxe5 with a practically won game. The bad news for Kramnik in that line is 22.gxf3, when tactics let White keep the balance after 22...Qxe5 23.Nxd4 Rac8 24.Nc6. Trying to sidestep this with 21...Rac8 doesn't give an advantage either: White equalizes with 22.Qxd4 (22...Ba6 23.Qb6 Bxb5 24.Qxb5 Rc5 25.Qa4 Rxe5=).
It does look like Black can maintain an edge with the immediate 20...Rac8, however: 21.Qxd4 Bc5 22.Qa4 gxf4 23.Bf3 Bxf3 24.Rxf3 (here 24.gxf3 makes less sense, as g1 is covered) 24...Qxe5 25.Qxf4 Qxf4 26.Rxf4 Rfd8 and White has an unpleasant defensive chore ahead of him.
Times: Anand 27, Kramnik 44 and counting.
12:30
No new moves, but GM Ian Rogers, speaking with GM Seirawan in the Foidos English commentary room, dropped a bombshell. There's a rumor that Kramnik might retire from chess if he loses this match. I'd love to know more about this, if he's just picking up some individual's wholly unsubstantiated conjecture, if there's anything more to it than this. I'll try to find out.
Times: Anand 31 minutes and counting, Kramnik 57.
12:15
As noted at the end of the last update, Kramnik played 17...c5. Anand played the predicted 18.Nxb5, but instead of 18...gxf4 Kramnik played the better 18...cxd4. Here Anand had at least three moves to choose from - 19.Bf3, 19.Bxc4 and 19.Qxc4, and as you can get guess from the bold typeface, he chose the latter. Kramnik responded with the accurate 19...a5, and seems to have the advantage here - both on the board and on the clock: Anand has 46 minutes and counting, Kramnik 57. The position is closer to equal than winning for Black, but at least he has a little something to hope for.
12:00
The game is at another critical moment. Anand played 17.e5 after 17 minutes' thought, and now Kramnik has thought for even longer - 20 minutes so far. There's a lot to think about: 17...f5, 17...c5, 17...Bxc3 and 17...Qh7 are all candidates. Here are some lines:
(1) 17...f5 18.exf6 Qxf6 (I found a scary-looking draw after 18...Rxf6: 19.fxg5 Qxg5 20.Ne4 Rxf1+ 21.Rxf1 Qg7 22.Nf6+ Kh8 23.Nh5! Qxd4+ 24.Kh1 Rf8 25.Qg6 Rxf1+ 26.Bxf1 Bc5 27.Qe8+ Kh8 28.Qf7+ Kh8 29.Qe8+ with perpetual) 19.fxg5 Qxg5 and now 20.Qe4 or 20.Bf3 is roughly balanced.
(2) 17...c5 18.Nxb5 gxf4 19.Bf3 Bxf3 20.Rxf3 a6 21.Nd6 cxd4 22.Nxc4 and I think White has a slight edge.
(3) 17...Bxc3 18.Qxc3 (or 18.bxc3 c5 19.Bf3 Bxf3 20.Rxf3 cxd4 21.Rxd4 Rfd8 22.Qd2 Rxd4 23.Qxd4 g4 24.Rg3 with full compensation) 18...f5 19.exf6 Qxf6 20.fxg5 Qxg5 21.Bf3=.
(4) 17...Qh7 18.Ne4 Kh8 19.f5 Be7, and although White is going forward like Tolush, his advance troops may find themselves dying like Frenchmen in Russia: 20.Nc5 Bxc5 21.dxc5 Rae8 22.Bh5 exf5 23.Qc3 Kg8 24.Rd7 Bc8 25.Rxa7 Qg7 -/+ and White's position is collapsing.
Note: Kramnik has just played 17...c5, after 25 minutes' thought. Times remaining: Anand 1:04 and counting, Kramnik 1:00. Here's the current position:
11:15
A little unexpectedly, Kramnik has followed up 13...Nxe5 14.Bxe5 O-O. With the bishop on b4 rather than g7, this looks a little scary, but he does have a few things in his favor. First, now that White has castled, h4 isn't the worry it would have been. (It might get played anyway, but it's not as dangerous as it would have been.) Second, the thematic 15.f4 is well met by 15...Nd7, taking advantage of the Be5. Third, White's heavy pieces aren't able to get to the kingside very easily. Something like b3 would be nice, to free up d3 for a rook lift, but 15.b3 does involve a pawn sac (15...Bxc3 16.Qxc3 Nxe4) or surrendering the strong bishop for Black's knight (15...Bxc3 16.Bxf6). Another idea, in addition to 15.h4 and 15.b3, is 15.Bg3 with the ideas of h4 or f4. Maybe then Black can play 15...Rad8 and meet 16.f4 with 16...Bxc3 17.bxc3 c5. Anand has spent 16 minutes on this move (and counting - he's at 1:23 now), so it's clearly an important moment.
Here we go - literally just as I was mousing towards "finish", Anand played 15.Bxf6, and after 15...Qxf6 followed up with 16.f4. I had briefly thought about taking and playing 16.e5, with the idea of 17.Ne4, but thought that after 16...Qf5 White wouldn't have anything special. That's probably right, but the 16.f4 follow-up had eluded my attention. Black's options here include 16...gxf4 (not so much winning a pawn but reducing White's central pawn horde so he has enough room to breathe) and 16...Qg6 (preventing e5 for the moment and getting the queen off the f-file).
Instead, again just instants before I posted, Kramnik played 16...Qg7, allowing 17.e5 or 17.fxg5 followed by 18.e5. It looks interesting, but I'm going to post now before someone moves again.
Times: Anand 1:20 and counting, Kramnik 1:25.
[Note: Regarding a reader's question, there are no offstage monitors - the players must check back on the stage to know when it's their move.]
10:45
Watching this game on Foidos, I note that Anand spends almost every minute at the board, while Kramnik is often bounces up like a jackrabbit at every opportunity, and it often costs him a minute or more after Anand makes his move. Considering that this happens 15-20 times a game, that's a pretty considerable loss of time in every game!
On to the moves: 11...Bb4 12.Ne5 Qe7 13.O-O. Kramnik's 11th move was also somewhat rare and unusual in these Anti-Moscow Gambit lines, and his 12th move was a novelty that revealed the purpose behind his 11th move. Previously 12...Rg8 and 12...Nxe5 had been played in games with reasonably high-rated players.
The queen looks a little funny on e7. It protects the f-pawn so Black can castle long, but aside from that detail Black's queen would be happier on b6. On b6 it takes aim at White's d-pawn, which will also come under fire from one or two rooks on the d-file. More importantly, the queen on b6 shores up Black's queenside pawns, which are liable to come under assault from moves like b3, a4 and d5.
Here's the current position:
Times: Anand 1:40, Kramnik 1:37 and counting. (Incidentally, it took Kramnik four minutes to return to the board after 13.O-O, which I note not because of toiletgate style idiocy, but to point out again just how much time Kramnik is wasting with pacing.)
10:10
It turns out that due to the switch in Germany to daylight savings time, today's game started an hour later than usual, relative to those of us who haven't yet made the same switch. (Too bad for those of us didn't know in advance - the extra hour's sleep would have been nice!) The score is 5.5-2.5, and my expectation is that whatever slim hopes Kramnik may have had left after yesterday's game, and it's entirely possible that today and tomorrow the players will have a couple of short draws just to get it over with. We'll see!
Here are the opening moves: 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 c6 5.Bg5 h6 6.Bh4 dxc4 7.e4 g5 8.Bg3 b5 9.Be2 Bb7 10.Qc2 Nbd7 11.Rd1.
The first thing I have to say is this: Kramnik is going out swinging! Choosing the Semi-Slav will at least make it difficult for White to dry things up, but obviously Black is taking serious risks in this opening as well - especially when it's Anand's main opening with Black. Anand has chosen a sideline with 10.Qc2 and 11.Rd1, and while the plan isn't completely new, it is comparatively rare and Kramnik is starting to think here.
I'm glad Kramnik is making a fight, but have a feeling the match will end today.