World Chess Championship (Game 7)

[Anil]

[Event "World Chess Championship 2008"]
[Site "Bonn"]
[Date "2008.10.23"]
[Round "7"]
[White "Anand, Viswanathan"]
[Black "Kramnik, Vladimir"]
[Result "*"]
[WhiteElo "2783"]
[BlackElo "2772"]
[PlyCount "64"]
[EventDate "2008.10.23"]
[EventType "tourn"]
[EventCountry "GER"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 dxc4 5. a4 Bf5 6. e3 e6 7. Bxc4 Bb4 8. O-O
Nbd7 9. Qe2 Bg6 10. e4 O-O 11. Bd3 Bh5 12. e5 Nd5 13. Nxd5 cxd5 14. Qe3 Re8 15.
Ne1 Bg6 16. Bxg6 hxg6 17. Nd3 Qb6 18. Nxb4 Qxb4 19. b3 Rac8 20. Ba3 Qc3 21.
Rac1 Qxe3 22. fxe3 f6 23. Bd6 g5 24. h3 Kf7 25. Kf2 Kg6 26. Ke2 fxe5 27. dxe5
b6 28. b4 Rc4 29. Rxc4 dxc4 30. Rc1 Rc8 31. g4 a5 32. b5 c3 *

Will there be a fortress?

If there is a fortress being formed, then no engine can see it at present. All the scores/evals would be useless.

[AdminX]

[Event "World Chess Championship 2008"]
[Site "Bonn"]
[Date "2008.10.23"]
[Round "7"]
[White "Anand, Viswanathan"]
[Black "Kramnik, Vladimir"]
[Result "*"]
[WhiteElo "2783"]
[BlackElo "2772"]
[PlyCount "64"]
[EventDate "2008.10.23"]
[EventType "tourn"]
[EventCountry "GER"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 dxc4 5. a4 Bf5 6. e3 e6 7. Bxc4 Bb4 8. O-O
Nbd7 9. Qe2 Bg6 10. e4 O-O 11. Bd3 Bh5 12. e5 Nd5 13. Nxd5 cxd5 14. Qe3 Re8 15.
Ne1 Bg6 16. Bxg6 hxg6 17. Nd3 Qb6 18. Nxb4 Qxb4 19. b3 Rac8 20. Ba3 Qc3 21.
Rac1 Qxe3 22. fxe3 f6 23. Bd6 g5 24. h3 Kf7 25. Kf2 Kg6 26. Ke2 fxe5 27. dxe5
b6 28. b4 Rc4 29. Rxc4 dxc4 30. Rc1 Rc8 31. g4 a5 32. b5 c3 *

Will there be a fortress?

If there is a fortress being formed, then no engine can see it at present. All the scores/evals would be useless.

Chess engines have a very hard time with fortress understanding. But look at the pawn structure and you can see the fortress idea.

[Jack Lad]

You may be right - it could still be a draw especially if the rooks go.

[Anil]

<snipped>
Will there be a fortress?

If there is a fortress being formed, then no engine can see it at present. All the scores/evals would be useless.

Chess engines have a very hard time with fortress understanding.

If a person can create an engine that can understand (or just create a solid fortress) then no engine can win against it. We will see a lot of draw games then ... even against the mighty Rybka 3.

[AdminX]

[Event "World Chess Championship 2008"]
[Site "Bonn"]
[Date "2008.10.23"]
[Round "7"]
[White "Anand, Viswanathan"]
[Black "Kramnik, Vladimir"]
[Result "*"]
[WhiteElo "2783"]
[BlackElo "2772"]
[PlyCount "65"]
[EventDate "2008.10.23"]
[EventType "tourn"]
[EventCountry "GER"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 dxc4 5. a4 Bf5 6. e3 e6 7. Bxc4 Bb4 8. O-O
Nbd7 9. Qe2 Bg6 10. e4 O-O 11. Bd3 Bh5 12. e5 Nd5 13. Nxd5 cxd5 14. Qe3 Re8 15.
Ne1 Bg6 16. Bxg6 hxg6 17. Nd3 Qb6 18. Nxb4 Qxb4 19. b3 Rac8 20. Ba3 Qc3 21.
Rac1 Qxe3 22. fxe3 f6 23. Bd6 g5 24. h3 Kf7 25. Kf2 Kg6 26. Ke2 fxe5 27. dxe5
b6 28. b4 Rc4 29. Rxc4 dxc4 30. Rc1 Rc8 31. g4 a5 32. b5 c3 33. Rc2 *

[d]2r5/3n2p1/1p1Bp1k1/pP2P1p1/P5P1/2p1P2P/2R1K3/8 b - - 0 33

[sorcerers_apprentice]

OK.

How about +1.34?

33. Kd3 Kh6 34. Rxc3 Rxc3+ 35. Kxc3 Rybka3_8cpu (0:01.13)+1.34|d25

And I think Anand's best chances are always when he plays for a win.

It does indeed look grimmer for Kramnik after just a couple of moves. Don't get me wrong: I'm rooting for Anand, but I'd hate seeing him squander such a lead !!

Anyway, let's just wait and see.. Ah, if only Anand won again today ... !! It would start to take on 'Fischerian' proportions!

Cheers!

[Anil]

You may be right - it could still be a draw especially if the rooks go.

But, in this game if the Rooks are exchanged, then there are 2 targets (pawns on b6 and a6) for the dark-squared black-bishop.

OK, it seems the fortress holds good if Bishop-Knight and Rook-Rook are taken off the board.

[AdminX]

[Event "World Chess Championship 2008"]
[Site "Bonn"]
[Date "2008.10.23"]
[Round "7"]
[White "Anand, Viswanathan"]
[Black "Kramnik, Vladimir"]
[Result "*"]
[WhiteElo "2783"]
[BlackElo "2772"]
[PlyCount "66"]
[EventDate "2008.10.23"]
[EventType "tourn"]
[EventCountry "GER"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 dxc4 5. a4 Bf5 6. e3 e6 7. Bxc4 Bb4 8. O-O
Nbd7 9. Qe2 Bg6 10. e4 O-O 11. Bd3 Bh5 12. e5 Nd5 13. Nxd5 cxd5 14. Qe3 Re8 15.
Ne1 Bg6 16. Bxg6 hxg6 17. Nd3 Qb6 18. Nxb4 Qxb4 19. b3 Rac8 20. Ba3 Qc3 21.
Rac1 Qxe3 22. fxe3 f6 23. Bd6 g5 24. h3 Kf7 25. Kf2 Kg6 26. Ke2 fxe5 27. dxe5
b6 28. b4 Rc4 29. Rxc4 dxc4 30. Rc1 Rc8 31. g4 a5 32. b5 c3 33. Rc2 Kf7 *

[d]2r5/3n1kp1/1p1Bp3/pP2P1p1/P5P1/2p1P2P/2R1K3/8 w - - 0 34

[Anil]

How about trying: e4 then Ke3 and then Kd4. Only if black rook doesn't come on c4 before Kd4.

The score even after some moves and at depth 30 is +1.34 for white. This is surely an indication of fortress for an engine to understand.

34. Kd3 Rd8 35. Kxc3 Rc8+ 36. Kd2 Rxc2+ 37. Kxc2 Kg6 38. Kb3 Kh7 39. Kc4 Kg6 40. Kc3 Rybka3_8cpu (0:03.33)+1.34|d30

[Albert Silver]

You may be right - it could still be a draw especially if the rooks go.

But, in this game if the Rooks are exchanged, then there are 2 targets (pawns on b6 and a6) for the dark-squared black-bishop.

OK, it seems the fortress holds good if Bishop-Knight and Rook-Rook are taken off the board.

Not really. Exchange the rooks and what does White do if Black does nothing?