Candidates Round 3 – What went wrong for Svidler?

Topalov – Aronian, 1/2-1/2
Andreikin – Karjakin, 1/2-1/2
Mamedyarov – Anand, 0-1
Svidler – Kramnik, 1/2-1/2

So Anand is leading after three rounds with two excellent wins and essentially flawless play. I am not sure many would want to see him as the challenger after the match in November, but as Grischuk said: the player that wins the candidates is not the same many as when he started the tournament. Still, there is a long way to go and anything can happen.

Topalov challenged Aronian in the Anti-Marshall, but at one point he did not play the most challenging continuation:

After 28. Qxh7 Qd5! Black quickly equalised. A draw was reached after: 29. f4 d3 30. f5 d2 31. Qh8+ Kf7 32. Qh5+ Kf8 33.
Qh8+ Kf7 34. Qh5+ Kf8 35. Qh8+ 1/2-1/2

More challenging was 28. Qxa5 Bb7 (28… Qb7? 29. Qf5+ and 30. Qxh7 is good for White.) 29. Qxb4+ c5 30. Qb6 Qd5 31. f3, though Black can still hold here.

The following  variation seems surprisingly forced: 31…Nf4 32. c4 dxc3 33. Rxe8+ Kxe8 34. Re1+ Kd7 35. bxc3 Nxg2 36. Rb1 Bc6 37. Qa7+ Ke6 38. Qxg7 Qxf3 39. Rf1 Qe3+ 40. Kh2 Qe5+ 41. Qxe5+ Kxe5 42. Kg3 h5 43. h4 Nxh4 44. Kxh4 Bxa4

45. Kxh5 Ke4 46. Kg5 Kd3 47. Rf3+ Kc2 48. Kf4 Bb5 49. Ke5 Bd3 and Black draws.

Svidler was completely winning in a very eventful game. It has been annotated on ChessBase and will be so at other places too; but I still think I can add a bit…

1. c4 c5 2. Nf3 Nf6 3. Nc3 Nc6 4. g3 g6 5. d4 cxd4 6. Nxd4 Bg7 7. Bg2 O-O 8. O-O Nxd4 9. Qxd4 d6 10. Qd3 Be6 11. Bd2 Qc8 12. b3 Bh3 13. Rac1 Bxg2 14. Kxg2

Kramnik played very fast to get here, but then slowed down and had nothing
prepared. Svidler outplayed him with with simple means and by move 25 Black’s
position was appauling.
14…Qc6+ 15. f3 e6 16. Rfd1 Rad8 17. Bf4!
Preventing …Nd7.
17…Rd7 18. Qe3 b6 19. Rd3 Rc8 20. Qd2 Ne8 21. e4 a6 22. e5 h6?!
Played to prevent Bg5 ideas, but actually this makes matters even worse.
23. h4 Rcd8 24. Rd1 b5 25. c5! Qxc5
25… d5 26. b4 and White is in complete control. The knight either goes to d4 or to d6.
26. Ne4 Qb6

This is the first big moment in the game. Actually, for Svidler this is the biggest moment in the game. Like Carlsen in the 10th match game in Chennai, he was too eager to take on d6.

27. Nxd6?

27. Bxh6! would have won the game it seems. The main idea is that h4-h5 comes quickly in most lines.

a) 27… d5 28. Bxg7 Kxg7 29. Qg5 and Nf6. Black is busted.

b) 28. h5 Nf6

(28… Rc8 loses very elegantly to: 29. hxg6 fxg6 30. Qg5 Kf7

31. Qxe5!! dxe5 32. Rxd7+ Kg8 33. Re7! and Black is checkmated.)

29. Nxf6+ Bxf6 30. Bg5! Bxg5 31. Qxg5 Black loses material and/or is mated down the h-file.

Both players were down to just over 10 minutes around here.

27… Bf8! 28. h5 Nxd6 29. exd6 g5 30. Be5 Rc8?! (30… Bg7!) 31. Rc1 Rxc1 32. Qxc1 Qb7 33. g4 b4 34. Qc4 Bg7 35. Bg3 Qb5?!

36. Be1?

Svidler believes his opponent and misses that after 36. Qxb5 axb5 37. f4! gxf4 38. Bxf4 he has the dual threat of Bd2 or Re3-e4, in both cases winning the b4-pawn. Even without it White is doing great, but Svidler maybe have an incorrect inclination towards keeping the queens on the board?

36… Qe5 37. Bg3 Qe2+

No repetition; no second chance!

38. Bf2 Bf8 39. Qxa6

39. Rd4!? is better, but Black should hold.

39… e5?

A strange blunder, showing that Kramnik is not as stable as those of us who considered him the favourite might think. 39… Qe5! would have won the d6-pawn and equalised.

40. Qc4 Qxa2

Svidler had a lot of time to calculate the variations accurately; but missed a stunning defence.

41. Qc6?

As this does not work, he should have relied on 41. Qxb4 Qa6 42. Rd5 Rxd6 43. Rxe5 when he would have stayed a pawn up with reasonable chances.

Kramnik now plays only moves. As they have to be found one by one, they are actually not that difficult, though the final moment is remarkable.

41… e4 42. fxe4 Qe2 43. Rf3 Rxd6 44. Qe8 f6 45. e5

Black is seemingly lost; but the pawn on e5 is pinned and Black has a marvelous defence.

45…f5! 46. gxf5 Rf6!!

Surprisingly White cannot make progress. He really needed f5-f6 to break through…

47. Kg3 Qe4 48. Bc5 Qe1+ 49. Bf2 Qe4 50. Bc5 Qe1+ 51. Bf2 1/2-1/2

10 thoughts on “Candidates Round 3 – What went wrong for Svidler?”

  1. Amazing game and what a spectacle in the press conference! I must say I am a BIG fan of Kramnik when it comes to press conferences because he’s capable of explaining chess in a way that gives me the feeling I actually understand what chess is about and could play a game like him (which of course I can’t). Same goes for Svidler for his ‘explanatory prose’ style. And then you get these two combined – wow.

    “So basically I missed my chances somewhere around move 35… because…of over-excitement basically. (Not without reason).” – Svidler nicely sums it up before being asked about the position after move 26.

  2. What I like most about the press conference is that we can see how meaningless it is to talk about a mistake on move 41. The players were nowhere near seeing it only two moves earlier! And there were 100s of lines that the computer will immediately disregard, which the players had to work out themselves; something we at times can forget.

  3. @Jacob Aagaard
    Exactly, I totally agree. This is chess at its finest: a battle between two titans, on their own. No need for computers here, they will not make you learn anything more than what these two show (and indeed, probably less, because they do not explain but simply ignore ‘losing’ lines).

  4. GM Aagaard, thank you for the analysis and the link to the press conference (which is roughly @ 5hrs 30 mins in). What is fascinating are the identities and differences between such a game and the games of ordinary players. The differences are the most glaring and notable—the skill, the speed, the genius and accuracy. The commonality is when they speak of things not seen, or erring because of “excitement,” or looking at a line and not remembering why you thought it was bad (when the post-mortem shows it to be good). This “human all too human” fallibility makes it possible to relate to a game played at a level most of us will never get near.
    Yes, computers hide the mess and the work. “Facts” and “objectivities” are made only by selection (which involves a lot of ignoring) and interpretation (guiding ideas, strategies, etc). The philosophical idealists were right!

  5. @Mark Moorman
    I am sure you can see that I looked at the games before actually checking out the post mortems.

    Tomorrow I will try to remember to put the timings for the post mortems in the post.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top