Hi,
i just witnessed another game with the ponder bug. At move 55 rybka 2.3.1 with white ponders that black would play Rxf3?? expecting to be 5.xx up after winning the rook for nothing. Of course the opponent played a normal move and the game continued to a draw. The second game is from another thread and shows the same bug where rybka 2.3.1 with white at move 22 ponders a totally insane move. It expects that black will play Qa5?? loosing a queen! the opponent played a normal move and the game continued as if nothing happened. There seems to be something wrong with rybka extracting the pondermove of the previous best line. Maybe something to do with the moves being capture moves?
[Event "Wertungspartie, 5m + 0s"]
[Site "Maschinenraum"]
[Date "2007.03.07"]
[Round "?"]
[White "Constanza, Rybka 2.3.1 mp 32-b"]
[Black "Pioner"]
[Result "1/2-1/2"]
[ECO "D23"]
[WhiteElo "2705"]
[BlackElo "2633"]
[Annotator "-0.07;-0.15"]
[PlyCount "121"]
[EventDate "2007.03.07"]
[TimeControl "300"]
{HIARCS 11.1 MP UCI: 13.2 ply; 744kN/s AMD Athlon(tm) 64 X2 Dual Core
Processor 3800+ 2800MHz, (2 threads), top2.ctg,} 1. d4 {B/0 0} c6 {B/0 0} 2. c4
{B/0 0} d5 {B/0 0} 3. Nf3 {B/0 0} Nf6 {B/0 0} 4. Qb3 {B/0 0} dxc4 {-0.01/1 0}
5. Qxc4 {B/0 0} Bf5 {-0.01/1 0} 6. g3 {B/0 0} e6 {-0.01/1 0} 7. Nc3 {B/0 0}
Nbd7 {B/0 0} 8. Bg2 {B/0 0} Be7 {B/0 0} 9. O-O {B/0 0} O-O {B/0 0} 10. e3 {
B/0 0} Ne4 {B/0 0} 11. Qe2 {B/0 0} Nxc3 {B/0 0} 12. bxc3 {B/0 0} Be4 {B/0 0}
13. c4 {B/0 0} Qa5 {B/0 0} 14. Nd2 {B/0 0} Bxg2 {-0.15/13 11} 15. Kxg2 {B/0 0}
e5 {-0.17/12 11} 16. Bb2 {B/0 0} exd4 {-0.21/12 10} 17. exd4 {(Êb3) B/0 0} Rfe8
{-0.29/13 10} 18. Qd3 {B/0 0} b5 {-0.36/13 9} 19. Bc3 {-0.07/12 12} Qa4 {
(Qa3) -0.51/13 0} 20. Rfe1 {(Ëfb1) 0.00/13 15} Nb6 {(Rad8) -0.49/12 10} 21.
cxb5 {-0.02/14 32} cxb5 {-0.40/14 0} 22. Nf3 {(d5) -0.02/13 7} Qa3 {
(Rad8) -0.60/12 6} 23. Rab1 {(Êe5) -0.04/13 3} b4 {-0.53/13 12} 24. Rb3 {
0.00/16 5} Qa4 {-0.53/13 0} 25. Bd2 {-0.10/14 8} Nd5 {(Rac8) -0.37/12 0} 26.
Ne5 {0.13/13 7} Red8 {(Rac8) -0.37/12 6} 27. Rc1 {0.19/12 14} a5 {-0.31/13 0}
28. Rb2 {0.15/12 4} Bf6 {-0.18/12 0} 29. Qe4 {(f4) 0.11/14 29} Bxe5 {
(Qe8) -0.03/12 16} 30. dxe5 {0.23/14 4} Qe8 {(Qd7) 0.00/13 4} 31. Rbc2 {
(f3) 0.15/13 14} Qe6 {0.00/12 5} 32. Bg5 {(f3) 0.22/15 5} Re8 {-0.21/13 8} 33.
Rc6 {0.15/16 3} Qxe5 {-0.18/14 0} 34. Qxe5 {0.08/17 7} Rxe5 {0.02/17 0} 35.
Rc8+ {0.05/16 0} Re8 {0.05/15 0} 36. Rxe8+ {0.12/18 7} Rxe8 {0.05/18 0} 37. Rc5
{0.19/17 7} f6 {(Nc3) 0.10/17 0} 38. Rxd5 {0.14/14 1} fxg5 {0.17/16 2} 39. Rxa5
{0.11/17 1} h6 {(g4) 0.17/15 5} 40. Ra4 {(h3) 0.05/16 4} Re2 {0.12/15 5} 41.
Kf3 {(Ëxb4) 0.00/18 5} Rb2 {(Rc2) 0.11/15 7} 42. Ra7 {0.01/17 3} Kh7 {0.11/15 0
} 43. Ra6 {0.00/16 1} h5 {(Kg8) 0.10/15 5} 44. h3 {0.05/14 3} g6 {
(Rc2) 0.09/15 7} 45. Ra7+ {0.00/17 3} Kh6 {0.09/15 0} 46. Ra4 {(Ëa8) 0.00/18 2}
b3 {(Kg7) 0.04/15 6} 47. axb3 {0.00/16 0} Rxb3+ {(Rxf2+) 0.03/14 1} 48. Ke4 {
(Êðg2) 0.00/19 3} h4 {(Rb2) -0.21/14 8} 49. gxh4 {0.00/17 1} gxh4 {-0.07/13 3}
50. f3 {0.00/20 2} Kh5 {(Kg5) 0.00/12 0} 51. Rd4 {(Ëa5+) 0.00/19 3} Ra3 {
(Kg5) -0.03/13 6} 52. Rd5+ {(Ëd8) 0.00/18 2} Kh6 {(g5) 0.00/14 5} 53. Rd1 {
(Ëd4) 0.00/21 3} Rb3 {(Kg5) 0.00/14 5} 54. Kf4 {(Ëd8) 0.00/20 3} g5+ {
(Rb4+) -0.01/13 5} 55. Ke4 {0.00/15 1} Rb6 {(Rxf3) 0.00/14 3} 56. Ke5 {
(Ëd4) 0.00/16 2} Kg6 {(Rb8) -0.07/13 6} 57. Rd4 {(Ëd8) -0.08/15 3} Rf6 {
-0.28/13 4} 58. Rd3 {-0.14/19 4} Rf5+ {-0.25/15 0} 59. Ke4 {-0.14/20 3} Rf4+ {
-0.25/15 0} 60. Ke3 {(Êðe5) -0.14/15 1} Rb4 {(Rf6) -0.26/13 4} 61. Ra3 {
(Ëd4) (Lag: Av=0.57s, max=5.8s) 0.00/19 2} 1/2-1/2
In the following game rybka 2.3.1 with white on move 22 ponders a totally insane move that black will play Qa5?? hanging a queen - seeing itself with +5. Black obvioulsy does not hang its queen and the came continues normally....
[Event "Wertungspartie, 5m + 0s"]
[Site "Maschinenraum"]
[Date "2007.03.05"]
[Round "?"]
[White "Casshern 2200"]
[Black "Morula, Rybka 2.2 mp."]
[Result "1-0"]
[ECO "B00"]
[WhiteElo "1921"]
[BlackElo "2691"]
[Annotator "0.49;0.34"]
[PlyCount "109"]
[EventDate "2007.03.06"]
[TimeControl "300"]
{Rybka 2.2 mp.: 17.3 ply; 293kN/s Intel(R) Core(TM)2 CPU 6600 @ 2.
40GHz 2404MHz, (2 threads), top2.ctg,} 1. d4 {B/0 0} h6 {B/0 0} 2. e4 {B/0 0}
d6 {B/0 0} 3. Nc3 {B/0 0} g5 {0.34/13 10} 4. Bc4 {(f4) B/0 0} Nc6 {0.34/14 14}
5. h4 {0.49/13 12} g4 {0.32/15 4} 6. Nge2 {0.53/14 5} Bg7 {(Nf6) 0.30/15 36} 7.
Be3 {(0-0) B/0 0} Nf6 {0.38/15 46} 8. Ng3 {0.46/13 13} Bd7 {(a6) 0.36/15 8} 9.
O-O {0.54/13 11} a6 {(e6) 0.38/14 0} 10. a4 {0.56/13 26} e6 {0.38/15 0} 11. Re1
{0.54/13 10} O-O {(Qe7) 0.39/14 14} 12. h5 {0.68/11 10} e5 {0.42/15 0} 13. d5 {
0.75/14 9} Nd4 {0.47/16 0} 14. Qd2 {0.75/13 13} Rc8 {0.49/15 0} 15. Nce2 {
(Ted1) 0.69/13 13} c5 {0.47/14 7} 16. a5 {(dxc6) 0.69/13 2} Kh7 {0.44/13 11}
17. c3 {0.68/13 1} Ne8 {(Qe7) 0.43/13 5} 18. Qd3 {0.75/12 8} Qc7 {0.47/15 0}
19. Bd2 {0.80/13 5} Nxe2+ {0.48/14 3} 20. Qxe2 {0.75/14 25} Bf6 {0.45/16 0} 21.
f4 {0.75/13 7} gxf3 {(Ng7) 0.33/16 0} 22. Qxf3 {0.68/11 0} Qd8 {
(Qxa5) 0.39/16 5} 23. Nf5 {0.51/12 6} Bg5 {0.37/15 0} 24. Bxg5 {
(Tad1) 0.46/14 6} Qxg5 {0.34/14 1} 25. Rf1 {0.50/12 0} Rg8 {(Bxf5) 0.31/14 2}
26. Be2 {0.46/13 5} Rc7 {(Bxf5) 0.32/15 0} 27. Rad1 {(Tf2) 0.37/13 7} f6 {
0.19/13 3} 28. b3 {(Tb1) 0.34/13 2} Bxf5 {0.19/14 1} 29. Qxf5+ {0.30/14 2} Qxf5
{0.14/17 0} 30. Rxf5 {0.29/16 1} Rcg7 {(Rg3) 0.13/18 2} 31. g4 {0.34/15 6}
Rxg4+ {(b6) 0.23/20 0} 32. Bxg4 {0.35/14 0} Rxg4+ {0.20/17 0} 33. Kf2 {
0.51/16 1} Rxe4 {0.19/17 0} 34. Rb1 {0.58/17 3} Rg4 {(Rh4) 0.32/18 0} 35. b4 {
0.62/17 4} cxb4 {0.38/18 1} 36. Rxb4 {0.59/17 0} Rxb4 {0.58/20 3} 37. cxb4 {
0.78/22 2} Kg7 {0.54/22 0} 38. Rf3 {0.73/21 4} Nc7 {0.77/21 2} 39. Rg3+ {
1.01/18 2} Kf7 {0.83/21 1} 40. Rg6 {1.07/18 0} Nxd5 {0.87/19 1} 41. Rxh6 {
1.52/18 1} Kg7 {(Kg8) 0.97/18 1} 42. Rg6+ {1.61/16 1} Kh8 {0.96/20 3} 43. h6 {
(b5) 1.59/19 2} Nxb4 {1.06/20 3} 44. Rxf6 {2.56/19 4} Nc6 {(d5) 1.86/20 6} 45.
Kf3 {(Txd6) 2.56/17 2} d5 {(Nxa5) 2.15/18 4} 46. Kg4 {2.94/19 2} e4 {2.76/19 0}
47. Kg5 {(Te6) 4.19/19 1} Kg8 {(e3) 3.69/21 3} 48. Kh5 {(Te6) 4.84/18 2} e3 {
(d4) 4.00/17 3} 49. Re6 {5.23/19 6} Kh7 {5.58/20 0} 50. Rxe3 {5.45/16 0} d4 {
5.77/19 2} 51. Re6 {5.64/17 3} d3 {#21/21 9} 52. Rd6 {6.50/18 1} Ne5 {
14.50/18 0} 53. Rd5 {6.50/16 1} d2 {(Nc6) #18/18 0} 54. Rxd2 {6.69/13 1} Kg8 {
#17/16 0} 55. Rd8+ {
(Td5) Morula,Rybka 2.2 mp. gibt auf (Lag: Av=0.22s, max=0.6s) 7.31/17 1} 1-0
i just witnessed another game with the ponder bug. At move 55 rybka 2.3.1 with white ponders that black would play Rxf3?? expecting to be 5.xx up after winning the rook for nothing. Of course the opponent played a normal move and the game continued to a draw. The second game is from another thread and shows the same bug where rybka 2.3.1 with white at move 22 ponders a totally insane move. It expects that black will play Qa5?? loosing a queen! the opponent played a normal move and the game continued as if nothing happened. There seems to be something wrong with rybka extracting the pondermove of the previous best line. Maybe something to do with the moves being capture moves?
[Event "Wertungspartie, 5m + 0s"]
[Site "Maschinenraum"]
[Date "2007.03.07"]
[Round "?"]
[White "Constanza, Rybka 2.3.1 mp 32-b"]
[Black "Pioner"]
[Result "1/2-1/2"]
[ECO "D23"]
[WhiteElo "2705"]
[BlackElo "2633"]
[Annotator "-0.07;-0.15"]
[PlyCount "121"]
[EventDate "2007.03.07"]
[TimeControl "300"]
{HIARCS 11.1 MP UCI: 13.2 ply; 744kN/s AMD Athlon(tm) 64 X2 Dual Core
Processor 3800+ 2800MHz, (2 threads), top2.ctg,} 1. d4 {B/0 0} c6 {B/0 0} 2. c4
{B/0 0} d5 {B/0 0} 3. Nf3 {B/0 0} Nf6 {B/0 0} 4. Qb3 {B/0 0} dxc4 {-0.01/1 0}
5. Qxc4 {B/0 0} Bf5 {-0.01/1 0} 6. g3 {B/0 0} e6 {-0.01/1 0} 7. Nc3 {B/0 0}
Nbd7 {B/0 0} 8. Bg2 {B/0 0} Be7 {B/0 0} 9. O-O {B/0 0} O-O {B/0 0} 10. e3 {
B/0 0} Ne4 {B/0 0} 11. Qe2 {B/0 0} Nxc3 {B/0 0} 12. bxc3 {B/0 0} Be4 {B/0 0}
13. c4 {B/0 0} Qa5 {B/0 0} 14. Nd2 {B/0 0} Bxg2 {-0.15/13 11} 15. Kxg2 {B/0 0}
e5 {-0.17/12 11} 16. Bb2 {B/0 0} exd4 {-0.21/12 10} 17. exd4 {(Êb3) B/0 0} Rfe8
{-0.29/13 10} 18. Qd3 {B/0 0} b5 {-0.36/13 9} 19. Bc3 {-0.07/12 12} Qa4 {
(Qa3) -0.51/13 0} 20. Rfe1 {(Ëfb1) 0.00/13 15} Nb6 {(Rad8) -0.49/12 10} 21.
cxb5 {-0.02/14 32} cxb5 {-0.40/14 0} 22. Nf3 {(d5) -0.02/13 7} Qa3 {
(Rad8) -0.60/12 6} 23. Rab1 {(Êe5) -0.04/13 3} b4 {-0.53/13 12} 24. Rb3 {
0.00/16 5} Qa4 {-0.53/13 0} 25. Bd2 {-0.10/14 8} Nd5 {(Rac8) -0.37/12 0} 26.
Ne5 {0.13/13 7} Red8 {(Rac8) -0.37/12 6} 27. Rc1 {0.19/12 14} a5 {-0.31/13 0}
28. Rb2 {0.15/12 4} Bf6 {-0.18/12 0} 29. Qe4 {(f4) 0.11/14 29} Bxe5 {
(Qe8) -0.03/12 16} 30. dxe5 {0.23/14 4} Qe8 {(Qd7) 0.00/13 4} 31. Rbc2 {
(f3) 0.15/13 14} Qe6 {0.00/12 5} 32. Bg5 {(f3) 0.22/15 5} Re8 {-0.21/13 8} 33.
Rc6 {0.15/16 3} Qxe5 {-0.18/14 0} 34. Qxe5 {0.08/17 7} Rxe5 {0.02/17 0} 35.
Rc8+ {0.05/16 0} Re8 {0.05/15 0} 36. Rxe8+ {0.12/18 7} Rxe8 {0.05/18 0} 37. Rc5
{0.19/17 7} f6 {(Nc3) 0.10/17 0} 38. Rxd5 {0.14/14 1} fxg5 {0.17/16 2} 39. Rxa5
{0.11/17 1} h6 {(g4) 0.17/15 5} 40. Ra4 {(h3) 0.05/16 4} Re2 {0.12/15 5} 41.
Kf3 {(Ëxb4) 0.00/18 5} Rb2 {(Rc2) 0.11/15 7} 42. Ra7 {0.01/17 3} Kh7 {0.11/15 0
} 43. Ra6 {0.00/16 1} h5 {(Kg8) 0.10/15 5} 44. h3 {0.05/14 3} g6 {
(Rc2) 0.09/15 7} 45. Ra7+ {0.00/17 3} Kh6 {0.09/15 0} 46. Ra4 {(Ëa8) 0.00/18 2}
b3 {(Kg7) 0.04/15 6} 47. axb3 {0.00/16 0} Rxb3+ {(Rxf2+) 0.03/14 1} 48. Ke4 {
(Êðg2) 0.00/19 3} h4 {(Rb2) -0.21/14 8} 49. gxh4 {0.00/17 1} gxh4 {-0.07/13 3}
50. f3 {0.00/20 2} Kh5 {(Kg5) 0.00/12 0} 51. Rd4 {(Ëa5+) 0.00/19 3} Ra3 {
(Kg5) -0.03/13 6} 52. Rd5+ {(Ëd8) 0.00/18 2} Kh6 {(g5) 0.00/14 5} 53. Rd1 {
(Ëd4) 0.00/21 3} Rb3 {(Kg5) 0.00/14 5} 54. Kf4 {(Ëd8) 0.00/20 3} g5+ {
(Rb4+) -0.01/13 5} 55. Ke4 {0.00/15 1} Rb6 {(Rxf3) 0.00/14 3} 56. Ke5 {
(Ëd4) 0.00/16 2} Kg6 {(Rb8) -0.07/13 6} 57. Rd4 {(Ëd8) -0.08/15 3} Rf6 {
-0.28/13 4} 58. Rd3 {-0.14/19 4} Rf5+ {-0.25/15 0} 59. Ke4 {-0.14/20 3} Rf4+ {
-0.25/15 0} 60. Ke3 {(Êðe5) -0.14/15 1} Rb4 {(Rf6) -0.26/13 4} 61. Ra3 {
(Ëd4) (Lag: Av=0.57s, max=5.8s) 0.00/19 2} 1/2-1/2
In the following game rybka 2.3.1 with white on move 22 ponders a totally insane move that black will play Qa5?? hanging a queen - seeing itself with +5. Black obvioulsy does not hang its queen and the came continues normally....
[Event "Wertungspartie, 5m + 0s"]
[Site "Maschinenraum"]
[Date "2007.03.05"]
[Round "?"]
[White "Casshern 2200"]
[Black "Morula, Rybka 2.2 mp."]
[Result "1-0"]
[ECO "B00"]
[WhiteElo "1921"]
[BlackElo "2691"]
[Annotator "0.49;0.34"]
[PlyCount "109"]
[EventDate "2007.03.06"]
[TimeControl "300"]
{Rybka 2.2 mp.: 17.3 ply; 293kN/s Intel(R) Core(TM)2 CPU 6600 @ 2.
40GHz 2404MHz, (2 threads), top2.ctg,} 1. d4 {B/0 0} h6 {B/0 0} 2. e4 {B/0 0}
d6 {B/0 0} 3. Nc3 {B/0 0} g5 {0.34/13 10} 4. Bc4 {(f4) B/0 0} Nc6 {0.34/14 14}
5. h4 {0.49/13 12} g4 {0.32/15 4} 6. Nge2 {0.53/14 5} Bg7 {(Nf6) 0.30/15 36} 7.
Be3 {(0-0) B/0 0} Nf6 {0.38/15 46} 8. Ng3 {0.46/13 13} Bd7 {(a6) 0.36/15 8} 9.
O-O {0.54/13 11} a6 {(e6) 0.38/14 0} 10. a4 {0.56/13 26} e6 {0.38/15 0} 11. Re1
{0.54/13 10} O-O {(Qe7) 0.39/14 14} 12. h5 {0.68/11 10} e5 {0.42/15 0} 13. d5 {
0.75/14 9} Nd4 {0.47/16 0} 14. Qd2 {0.75/13 13} Rc8 {0.49/15 0} 15. Nce2 {
(Ted1) 0.69/13 13} c5 {0.47/14 7} 16. a5 {(dxc6) 0.69/13 2} Kh7 {0.44/13 11}
17. c3 {0.68/13 1} Ne8 {(Qe7) 0.43/13 5} 18. Qd3 {0.75/12 8} Qc7 {0.47/15 0}
19. Bd2 {0.80/13 5} Nxe2+ {0.48/14 3} 20. Qxe2 {0.75/14 25} Bf6 {0.45/16 0} 21.
f4 {0.75/13 7} gxf3 {(Ng7) 0.33/16 0} 22. Qxf3 {0.68/11 0} Qd8 {
(Qxa5) 0.39/16 5} 23. Nf5 {0.51/12 6} Bg5 {0.37/15 0} 24. Bxg5 {
(Tad1) 0.46/14 6} Qxg5 {0.34/14 1} 25. Rf1 {0.50/12 0} Rg8 {(Bxf5) 0.31/14 2}
26. Be2 {0.46/13 5} Rc7 {(Bxf5) 0.32/15 0} 27. Rad1 {(Tf2) 0.37/13 7} f6 {
0.19/13 3} 28. b3 {(Tb1) 0.34/13 2} Bxf5 {0.19/14 1} 29. Qxf5+ {0.30/14 2} Qxf5
{0.14/17 0} 30. Rxf5 {0.29/16 1} Rcg7 {(Rg3) 0.13/18 2} 31. g4 {0.34/15 6}
Rxg4+ {(b6) 0.23/20 0} 32. Bxg4 {0.35/14 0} Rxg4+ {0.20/17 0} 33. Kf2 {
0.51/16 1} Rxe4 {0.19/17 0} 34. Rb1 {0.58/17 3} Rg4 {(Rh4) 0.32/18 0} 35. b4 {
0.62/17 4} cxb4 {0.38/18 1} 36. Rxb4 {0.59/17 0} Rxb4 {0.58/20 3} 37. cxb4 {
0.78/22 2} Kg7 {0.54/22 0} 38. Rf3 {0.73/21 4} Nc7 {0.77/21 2} 39. Rg3+ {
1.01/18 2} Kf7 {0.83/21 1} 40. Rg6 {1.07/18 0} Nxd5 {0.87/19 1} 41. Rxh6 {
1.52/18 1} Kg7 {(Kg8) 0.97/18 1} 42. Rg6+ {1.61/16 1} Kh8 {0.96/20 3} 43. h6 {
(b5) 1.59/19 2} Nxb4 {1.06/20 3} 44. Rxf6 {2.56/19 4} Nc6 {(d5) 1.86/20 6} 45.
Kf3 {(Txd6) 2.56/17 2} d5 {(Nxa5) 2.15/18 4} 46. Kg4 {2.94/19 2} e4 {2.76/19 0}
47. Kg5 {(Te6) 4.19/19 1} Kg8 {(e3) 3.69/21 3} 48. Kh5 {(Te6) 4.84/18 2} e3 {
(d4) 4.00/17 3} 49. Re6 {5.23/19 6} Kh7 {5.58/20 0} 50. Rxe3 {5.45/16 0} d4 {
5.77/19 2} 51. Re6 {5.64/17 3} d3 {#21/21 9} 52. Rd6 {6.50/18 1} Ne5 {
14.50/18 0} 53. Rd5 {6.50/16 1} d2 {(Nc6) #18/18 0} 54. Rxd2 {6.69/13 1} Kg8 {
#17/16 0} 55. Rd8+ {
(Td5) Morula,Rybka 2.2 mp. gibt auf (Lag: Av=0.22s, max=0.6s) 7.31/17 1} 1-0
It seems it always happens if rybkas main line on the move before is only one move. Then as the main line doesnt contain a second move rybka chooses a seemingly random move to ponder. It is strange that this happens quite often now - there are games where this happens 5 - 10 times. Not always are the mispondered moves as bad as in the above examples but they almost always drop the eval by a considerable amount. This strange behaviour was not present in rybka 2.2n2. And if it is it only occured extremely rarly when compared to rybka 2.3.1.
hope this helps....
hope this helps....
Thanks. I'm aware of this issue now and have a number of examples.
Vas
Vas
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill
