Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / A position where Rybka is better
- - By kaustubh (**) [in] Date 2007-06-30 08:48
If you look at this game between Topalov vs Anand
[Event "Corus"]
[Site "Wijk aan Zee NED"]
[Date "2007.01.21"]
[EventDate "2007.01.13"]
[Round "8"]
[Result "1-0"]
[White "Veselin Topalov"]
[Black "Viswanathan Anand"]
[ECO "E15"]
[WhiteElo "2783"]
[BlackElo "2779"]
[PlyCount "69"]

1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Ba6 5. b3 Bb4+ 6. Bd2 Be7
7. Bg2 c6 8. Bc3 d5 9. Ne5 Nfd7 10. Nxd7 Nxd7 11. Nd2 O-O
12. O-O Nf6 13. e4 dxe4 14. a4 Nd5 15. cxd5 Bxf1 16. d6 Bxg2
17. dxe7 Qxe7 18. Kxg2 f5 19. b4 Qd7 20. Qe2 Qd5 21. f3 exf3+
22. Nxf3 h6 23. Re1 Rfe8 24. Qc2 Rad8 25. Bd2 Qd7 26. Kf2 Rc8
27. Bf4 Qd5 28. Re5 Qd7 29. h4 Ra8 30. Bd2 Rac8 31. Qc4 Kh7
32. Bc3 Qd6 33. Ne1 b5 34. Qc5 Qd8 35. Nd3 1-0

Anand sacrifices his knight at 14. a4 Nd5 it opens diagonal for his bishop which gets white rook. But after some moves Anand is left with only two rook and queen , and Topalov has queen ,knight, bishop and one rook .It was a difficult position therefore Anand resigned. So i think Nd5 is not good move.

I tried to analysis using fritz 10, shredder 10  , every engine thinks that Nd5 is good move , but only Rybka understands this position better and does not consider Nd5 as good move , it proves that Rybka's evaluation is better .

r2q1rk1/p3bppp/bpp1pn2/8/P1PPp3/1PB3P1/3N1PBP/R2Q1RK1 b - a3 0 1


Analysis by Fritz 10:

14...Nd5 15.Nxe4
  µ  (-0.77)   Depth: 4/14   00:00:00  4kN
14...Nd5 15.Nxe4 f5 16.cxd5 Bxf1 17.Qxf1 fxe4 18.dxc6
  µ  (-0.77)   Depth: 5/16   00:00:00  5kN
14...Nd5 15.Nxe4 f5 16.cxd5 Bxf1 17.Qxf1 fxe4 18.dxc6
  µ  (-0.77)   Depth: 6/18   00:00:00  9kN
14...Nd5 15.Nxe4 f5 16.cxd5 Bxf1 17.Qxf1 fxe4
  µ  (-0.89)   Depth: 7/21   00:00:00  17kN
14...Nd5 15.Nxe4
  ³  (-0.61)   Depth: 8/20   00:00:00  27kN
14...Nd5 15.Nxe4 f5 16.cxd5 Bxf1
  ³  (-0.60)   Depth: 8/20   00:00:00  40kN
14...Nd5 15.Nxe4
  ³  (-0.32)   Depth: 9/21   00:00:00  70kN
14...Nd5
  =  (0.10)   Depth: 9/29   00:00:02  495kN
14...Nd5 15.Nxe4 Nxc3 16.Nxc3 Qd7 17.Qg4 Bb7 18.Rfd1 a5 19.d5 Bd6 20.Be4
  =  (0.10)   Depth: 10/21   00:00:08  626kN
14...Nd5 15.Nxe4 f5 16.cxd5 Bxf1 17.Qxf1 fxe4 18.dxc6 Qd5 19.Qe1 Rac8 20.Bxe4 Qxb3 21.f4 Bf6
  =  (0.06)   Depth: 11/27   00:00:10  912kN
14...Nd5 15.Nxe4 Nxc3 16.Nxc3 Qd7 17.Qg4 Bf6 18.Ne2 Bb7 19.Rad1 Rfd8 20.Be4 a5
  =  (-0.16)   Depth: 12/29   00:00:12  1410kN
14...Nd5 15.Nxe4 Nxc3 16.Nxc3 Qd7 17.Qg4 Bb7 18.Rfd1 Rfd8 19.Qh5 Bf6 20.Ne2 c5 21.Bxb7
  =  (-0.02)   Depth: 13/31   00:00:22  3422kN
14...Nd5 15.Nxe4 Nxc3 16.Nxc3 Qd7 17.Qg4 Bb7 18.Rfd1 Rad8 19.Rac1 a5 20.Qh5
  =  (-0.09)   Depth: 14/32   00:00:45  8386kN

Analysis by Shredder 10:

14...Qd6
  =  (-0.08)   Depth: 1/4   00:00:00
14...Qd6
  =  (-0.08)   Depth: 1/4   00:00:00
14...Qd6 15.Nxe4
  =  (-0.08)   Depth: 2/3   00:00:00
14...Qd6 15.Nxe4 Nxe4
  =  (0.07)   Depth: 3/8   00:00:00
14...Qc7
  =  (0.06)   Depth: 3/8   00:00:01
14...Qc7 15.Qe2 Nd5
  =  (0.06)   Depth: 3/10   00:00:01
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 3/10   00:00:01
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 3/10   00:00:01
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 4/4   00:00:01
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 5/5   00:00:02
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 6/6   00:00:02
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 7/9   00:00:02  17kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 8/8   00:00:03  30kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 9/9   00:00:05  39kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 10/10   00:00:06  61kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 11/11   00:00:06  159kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 12/16   00:00:08  290kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 13/13   00:00:10  651kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 14/16   00:00:12  902kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 15/16   00:00:16  1527kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 16/16   00:00:23  2630kN
14...Nd5 15.Nxe4
  =  (0.02)   Depth: 17/17   00:00:35  4468kN

Analysis by Rybka 2.3 LK mp 32-bit :

14...Qc7 15.Nxe4 Bb7 16.Qd3 a5 17.Bd2
  ²  (0.26)   Depth: 2   00:00:00
14...e3 15.fxe3 Qd7 16.Nf3 Ng4 17.Ne5 Nxe5 18.dxe5 Qc7
  =  (0.19)   Depth: 2   00:00:00
14...e3 15.fxe3 Qd7 16.Nf3 Ng4 17.Ne5 Nxe5 18.dxe5 Qc7
  =  (0.21)   Depth: 3   00:00:00
14...e3 15.fxe3 Qd7 16.Nf3 Ng4 17.Ne5 Nxe5 18.dxe5 Qc7
  =  (0.25)   Depth: 4   00:00:00
14...e3 15.fxe3 Qd7 16.Nf3 Ng4 17.Ne5 Nxe5 18.dxe5 Qc7
  ²  (0.29)   Depth: 5   00:00:00
14...Qc7 15.Nxe4 Bb7 16.Qd3 a5 17.Bd2
  =  (0.25)   Depth: 5   00:00:00
14...Qc7 15.Nxe4 Bb7 16.Qd3 a5 17.Bd2
  ²  (0.26)   Depth: 6   00:00:00  2kN
14...Qc7 15.Nxe4 Bb7 16.Qd3 a5 17.Bd2
  =  (0.25)   Depth: 7   00:00:00  5kN
14...Qc7 15.Nxe4 Bb7 16.Qd3 a5 17.Bd2
  ²  (0.26)   Depth: 8   00:00:00  7kN
14...Qc7 15.Nxe4 h6 16.Re1 Rfd8 17.Ra2
  ²  (0.30)   Depth: 9   00:00:05  96kN
14...e3 15.fxe3 Qd6 16.Qf3 Rac8 17.Rf2
  ²  (0.26)   Depth: 9   00:00:06  115kN
14...Bb7 15.Nxe4 a5 16.Qe2 Qc7 17.Bf3
  =  (0.25)   Depth: 9   00:00:07  117kN
14...Bb7 15.Nxe4 a5 16.Qe2 Qc7 17.Bf3 Nxe4 18.Bxe4 Bf6
  ²  (0.26)   Depth: 10   00:00:07  143kN
14...Bb7 15.Nxe4 a5 16.Qe2 Qc7 17.Rfd1 Rfd8 18.Ra2
  ²  (0.26)   Depth: 11   00:00:11  195kN
14...Bb7 15.Nxe4 a5 16.Qe2 Qc7 17.Rfd1 Rfd8 18.Ra2 g6
  ²  (0.27)   Depth: 12   00:00:19  332kN
14...Qc7 15.Nxe4 Nd5 16.Bd2 Nb4 17.Bf4 Qd7 18.Qe2 Rfe8
  =  (0.23)   Depth: 12   00:00:41  738kN
Parent - - By Uri Blass (*****) [il] Date 2007-06-30 15:11
Why do you use an old version of rybka?
Rybka2.3LK is not the latest version and latest version likes Anand's move.

New game
r2q1rk1/p3bppp/bpp1pn2/8/P1PPp3/1PB3P1/3N1PBP/R2Q1RK1 b - - 0 1


Analysis by Rybka 2.3.2a 32-bit :

1...Qd8-c7 2.Rf1-e1
  ²  (0.29)   Depth: 5   00:00:00
1...Qd8-d7 2.Rf1-e1
  ²  (0.27)   Depth: 5   00:00:00
1...Qd8-d7 2.Rf1-e1 e4-e3
  =  (0.25)   Depth: 6   00:00:00  6kN
1...Qd8-d7 2.Rf1-e1 e4-e3 3.Re1xe3
  =  (0.23)   Depth: 7   00:00:00  13kN
1...Qd8-d7 2.Rf1-e1 e4-e3 3.Re1xe3 h7-h5
  ²  (0.26)   Depth: 8   00:00:00  17kN
1...Qd8-d7 2.Nd2xe4 Nf6xe4 3.Bg2xe4 f7-f5 4.Be4-f3 Rf8-f6 5.Qd1-d2
  ²  (0.28)   Depth: 9   00:00:01  42kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.Nd2xf1 c6xd5 4.Nf1-e3 f7-f5 5.Ra1-c1
  =  (0.07)   Depth: 9   00:00:01  63kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.Nd2xf1 c6xd5 4.Nf1-e3 f7-f5 5.Ra1-c1 Ra8-c8
  =  (0.03)   Depth: 10   00:00:01  80kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 f7-f5 6.Qd1-e2 Ra8-c8 7.Nd2-c4
  =  (0.00)   Depth: 11   00:00:04  273kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 f7-f5 6.Nd2-c4 Qe7-d7 7.Qd1-c2 Qd7-d5
  =  (0.00)   Depth: 12   00:00:06  320kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 Rf8-d8 6.Qd1-c2 f7-f5 7.Nd2-c4 c6-c5 8.Ra1-d1
  =  (-0.02)   Depth: 13   00:00:08  472kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 Rf8-d8 6.Qd1-h5 Rd8-d5 7.Qh5-e2 f7-f5 8.Nd2-c4
  =  (0.00)   Depth: 14   00:00:12  679kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 Rf8-d8 6.Qd1-h5 Rd8-d5 7.Qh5-e2 f7-f5 8.Nd2-c4
  =  (0.00)   Depth: 15   00:00:18  1048kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 Rf8-d8 6.Qd1-e2 f7-f5 7.Nd2-c4 c6-c5 8.Ra1-d1
  =  (-0.03)   Depth: 16   00:00:39  2431kN
1...Nf6-d5 2.c4xd5 Ba6xf1 3.d5-d6 Bf1xg2 4.d6xe7 Qd8xe7 5.Kg1xg2 Rf8-d8 6.Qd1-c1 f7-f5 7.Nd2-c4 c6-c5 8.d4xc5
  =  (-0.04)   Depth: 17   00:01:05  4154kN

(,  30.06.2007)
Parent - - By kaustubh (**) [in] Date 2007-06-30 17:18
I don't understand why it is like that , maybe the new version has changed its evaluation.
Parent - By Uri Blass (*****) [il] Date 2007-06-30 18:40
Of course the new version changed the evaluation.
2.3K underevaluated pawns so they did not play Nd5

Maybe Nd5 is a mistake and maybe changing the evaluation caused rybka to make a mistake(I do not know) but one position does not prove that the change was wrong.

Uri
Parent - - By vroger007 (**) [be] Date 2007-06-30 17:50
and even with that old version, I don't think that 41 seconds (12ply) is enough
to draw any sound conclusion... I'm pretty sure the "old" rybka will also prefer
Anand's move Nd5 if you let it calculate at least 17 ply.  Even more so if you
switch your rybka to ultra-optimistic mode ;-)
Here is a 17ply analysis of the position:
Analysis by Rybka 2.3.2a 32-bit UO_C100_CT_PA_TBF:
1. = (-0.03): 1...Nd5 2.cxd5 Bxf1 3.d6 Bxg2 4.dxe7 Qxe7 5.Kxg2 Rfd8 6.Qe2 f5 7.Nc4 c5 8.Rd1
2. = (0.21): 1...Qc7 2.Re1 Rad8 3.Qc2 c5 4.dxc5 e3 5.Rxe3 bxc5 6.Nf3 Bb7 7.Ne5 Bxg2 8.Kxg2
3. ² (0.27): 1...c5 2.Nxe4 Rb8 3.Re1 Nxe4 4.Bxe4 Bf6 5.Qd3 Bxd4 6.Bxh7+ Kh8 7.Rad1 f5 8.Bxd4
4. ² (0.27): 1...Rb8 2.Nxe4 c5 3.Re1 Nxe4 4.Bxe4 Bf6 5.Qd3 Bxd4 6.Bxh7+ Kh8 7.Rad1 f5 8.Bxd4
5. ² (0.28): 1...Bb7 2.Nxe4 a5 3.Re1 Qc7 4.Qd3 Rad8 5.Rad1 Ng4 6.Nd2 g6 7.Qe2 Nf6 8.Nf3
6. ² (0.28): 1...Qc8 2.Re1 Bb7 3.Qc2 a5 4.Nxe4 Qc7 5.Rad1 Rfd8 6.Bf3 Nxe4 7.Bxe4 g6 8.Qe2
7. ² (0.30): 1...Qd7 2.Re1 Bb7 3.Nxe4 a5 4.Qd3 Nxe4 5.Bxe4 f5 6.Bf3 Bf6 7.Rad1 Rfe8 8.Rd2
8. ² (0.31): 1...Rc8 2.Nxe4 Nd5 3.Bb2 Nf6 4.Ng5 h6 5.Nf3 Bb7 6.Ne5 Nd7 7.a5 Nxe5 8.dxe5
9. ² (0.31): 1...Qd6 2.Nxe4 Nxe4 3.Bxe4 f5 4.Bf3 Bf6 5.Re1 Rfe8 6.Qe2 Rad8 7.Red1 Qd7 8.Qe3
10. ² (0.31): 1...Re8 2.Nxe4 Nxe4 3.Bxe4 Rc8 4.Qf3 Qc7 5.Qe3 f5 6.Bf3 Bf6 7.Rfe1 Rcd8 8.Ra2
11. ² (0.31): 1...h6 2.Nxe4 Qc7 3.Qc2 Rfd8 4.Rfe1 Rab8 5.Nd2 Bb7 6.a5 b5 7.a6 b4 8.Bb2
12. ² (0.33): 1...g6 2.Nxe4 Rc8 3.Re1 Nxe4 4.Bxe4 Bf6 5.Qd3 Qd6 6.Red1 Rfd8 7.Qf3 Bg7 8.Qe3
13. ² (0.35): 1...Qb8 2.Nxe4 Qc7 3.Qc2 Rfd8 4.Rfd1 Nxe4 5.Qxe4 Rac8 6.Qe3 Bf6 7.Be4 Rd7 8.Ra2
14. ² (0.37): 1...e3 2.fxe3 Qd7 3.Nf3 Ng4 4.Ne5 Nxe5 5.dxe5 Qc7 6.Be4 Rad8 7.Qh5 g6 8.Qf3
15. ² (0.38): 1...Qe8 2.Re1 Qd7 3.Nxe4 Rfd8 4.Qc2 Qc7 5.Ng5 h6 6.Nf3 Nd7 7.Re3 Rac8 8.Re2
16. ² (0.38): 1...Ne8 2.Nxe4 Nd6 3.Nxd6 Qxd6 4.Qe2 Rad8 5.b4 Qc7 6.Rfb1 Bb7 7.a5 b5 8.a6
17. ² (0.40): 1...Bc8 2.Nxe4 Qc7 3.Qf3 Bd7 4.Qd3 Nxe4 5.Bxe4 f5 6.Bf3 a5 7.Rad1 Bf6 8.Rfe1
18. ² (0.40): 1...Nd7 2.Nxe4 Qc7 3.Qh5 h6 4.Rfe1 Nf6 5.Qf3 Rfc8 6.Re2 Nxe4 7.Qxe4 Qd6 8.Qe3
19. ² (0.43): 1...Kh8 2.Nxe4 Nxe4 3.Bxe4 Qd7 4.Re1 f5 5.Bf3 Bf6 6.Qe2 Rfe8 7.Bh5 g6 8.Bf3
20. ² (0.47): 1...Bd6 2.Nxe4 Nxe4 3.Bxe4 Qd7 4.Re1 f5 5.Bf3 Rad8 6.Qd3 Rf7 7.Ra2 Re8 8.Bh5
21. ² (0.54): 1...g5 2.Re1 Qd7 3.Nxe4 Nxe4 4.Rxe4 f5 5.Re3 Bf6 6.Qh5 Rae8 7.Rae1 Bb7 8.Qe2
22. ² (0.61): 1...h5 2.Nxe4 Qc7 3.Nxf6+ Bxf6 4.Qxh5 Rad8 5.Rad1 c5 6.Qg4 Bb7 7.Bxb7 Qxb7 8.Rfe1
23. ± (1.17): 1...e5 2.dxe5 Nd5 3.cxd5 cxd5 4.Re1 Rc8 5.Bd4 f5 6.Rc1 Bc5 7.Bxc5 bxc5 8.Bf1
24. +- (1.60): 1...Ng4 2.Qxg4 f5 3.Qe2 Bb7 4.Rfd1 Bf6 5.a5 Rf7 6.b4 g6 7.axb6 axb6 8.Rxa8
25. +- (1.66): 1...b5 2.axb5 Bb7 3.Nxe4 Nxe4 4.Bxe4 Qd7 5.Qg4 cxb5 6.d5 g6 7.dxe6 Qc7 8.Bxb7
26. +- (1.98): 1...Bxc4 2.bxc4 Rc8 3.Qe2 c5 4.dxc5 Rxc5 5.Nxe4 Nxe4 6.Bxe4 Qc8 7.Bb4 Rc7 8.Bxe7
27. +- (2.07): 1...Nh5 2.Qxh5 f5 3.Nxe4 fxe4 4.Bxe4 g6 5.Qe5 Kf7 6.Bxc6 Bf6 7.Qe3 Rc8 8.d5
28. +- (2.32): 1...Bb5 2.axb5 cxb5 3.Nxe4 Rc8 4.Rxa7 bxc4 5.bxc4 Rxc4 6.Qb3 Rc7 7.Ra8 Rc8 8.Rxc8
29. +- (2.32): 1...Ba3 2.Rxa3 Nd5 3.cxd5 Bxf1 4.Nxf1 cxd5 5.Ne3 a5 6.Ra1 Rc8 7.Rc1 f5 8.f4
30. +- (2.75): 1...Bc5 2.dxc5 e3 3.fxe3 Qd3 4.Bxf6 Qxe3+ 5.Rf2 gxf6 6.Qf3 Qxf3 7.Bxf3 Rac8 8.cxb6
31. +- (2.91): 1...Bb4 2.Bxb4 Re8 3.Nxe4 Nxe4 4.Bxe4 Qc7 5.Re1 f5 6.Bf3 Rad8 7.a5 b5 8.cxb5
32. +- (4.77): 1...Qd5 2.cxd5 Nxd5 3.Bb2 Bxf1 4.Nxf1 f5 5.Qe2 Rfe8 6.Ne3 Rac8 7.Nxd5 exd5 8.a5
33. +- (6.15): 1...Qxd4 2.Bxd4 Rfd8 3.Bc3 Rac8 4.Qe1 Nd5 5.cxd5 cxd5 6.Bd4 Bxf1 7.Nxf1 Bc5 8.Bxc5
Roger
Parent - By kaustubh (**) [in] Date 2007-07-18 05:18
Is Nd5 a good move ?
Up Topic Rybka Support & Discussion / Rybka Discussion / A position where Rybka is better

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill