On-line engine blitz tourney June

[D Sceviour]

[d]8/5pk1/p5p1/4p2p/7P/1p3qP1/2n5/1N3QK1 b - - 1 45

The following bug occurred in Kingslayer-Schooner. Here is the output from the debug file:

Code: Select all

5893454 <first &#58; 8 1704 0 10838  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1c3 d1e1 c3b1
5893456 <first &#58; 9 1672 0 15561  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1c3 d1e1 c3b1 e5e4
5893463 <first &#58; 10 1774 1 30375  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1a3 d1e1 a3c4 f2e4 c4e5
5893475 <first &#58; 11 1723 2 57657  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1c3 b3b2 c3d1 f2d1 g2h2 b2b1q
5893491 <first &#58; 12 1830 4 96367  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1a3 d1e1 g2f3 e5e4 f3g2 e4e3 g2f3 b3b2
5893522 <first &#58; 13 1854 7 166027  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1a3 d1e1 a3c4 f2e4 g2f3 e4g3 f3g2

This cannot be duplicated in the subsequent analysis. It would be nice to have a guess (even a bad guess) as to what to look for and why it does this. This frustrating non-reproducible type of error occurs in maybe one in 100 games. Normal asserts and light debug code show no problems arising that would make it play unbearable moves. Nor was there any memory crash or segmentation violation. I am considering opening a new log file to record non-standard events. Suggestions as to what to include might be helpful. Perhaps restrict the saved information only for the PV moves, otherwise the hard drive might fill up too quickly. For the log file maybe include:

(1) game information header
(2) generated move list to see if moves like f1f3 are even in there.
(3) position structure information for each move

[MikeB]

Here are the 9 McBrain games - click on the ellipsis ( 3 dots upper upper left for the pull down)

[pgn][Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "1"]
[White "McBrain"]
[Black "Nemeton"]
[Result "1-0"]
[WhiteElo "2375"]
[BlackElo "1633"]
[TimeControl "300+1"]
[Annotator "10. +1.87"]

1. e4 d6 2. d4 g6 3. Nf3 Bg7 4. c3 Nf6 5. Bd3 O-O 6. O-O c5 7. Be3 cxd4 8.
cxd4 Nc6 9. d5 Nxd5 10. exd5 {+1.87/26 +303} Bxb2 11. Nbd2 {+1.77/26 +302}
Nb4 12. Rb1 {+1.88/29 +291} Nxd3 13. Ne4 {+1.89/26 +289} Nxf2 14. Nxf2
{+2.08/30 +269} Bg7 15. Re1 {+1.89/31 +254} e5 16. dxe6 {+2.51/25 +250}
fxe6 17. Qd2 {+2.62/27 +241} Rf5 18. Bh6 {+3.09/25 +237} Bh8 19. Ng4
{+5.14/27 +234} Rd5 20. Qf2 {+5.16/25 +233} Qe7 21. Bg5 {+7.44/26 +228}
Rxg5 22. Nxg5 {+7.74/25 +227} Qxg5 23. Rf1 {+10.11/29 +223} Qe7 24. Nh6+
{+10.50/26 +223} Kg7 25. Qd2 {+11.32/30 +217} a6 26. Rf7+ {+15.93/25 +212}
Qxf7 27. Nxf7 {+16.51/23 +211} Kxf7 28. Qxd6 {+53.11/24 +209} Ke8 29. Rf1
{+1000.22/31 +203} Bg7 30. Qf4 {+1000.19/36 +203} Kd8 31. Rd1+
{+1000.18/34 +181} Ke8 32. Qh4 {+1000.17/37 +176} e5 33. Qd8+
{+1000.16/35 +175} Kf7 34. Rf1+ {+1000.15/33 +174} Ke6 35. Qg8+
{+1000.14/23 +175} Kd6 36. Qxg7 {+1000.13/29 +176} Bg4 37. Qxb7
{+1000.12/27 +176} Rc8 38. Qb4+ {+1000.10/26 +176} Kd5 39. Qxg4
{+1000.09/28 +177} Rc4 40. Qd7+ {+1000.08/30 +178} Kc5 41. Re1
{+1000.07/32 +178} e4 42. Rd1 {+1000.06/35 +179} Kb6 43. Qd6+
{+1000.05/42 +180} Ka7 44. Rb1 {+1000.04/54 +180} Rc1+ 45. Rxc1
{+1000.03/70 +181} h5 46. Rc7+ {+1000.02/65 +182} Kb8 47. Qd8#
{+1000.01/127 +182}
{Nemeton checkmated} 1-0


[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "2"]
[White "McBrain"]
[Black "gogobello"]
[Result "1-0"]
[WhiteElo "2376"]
[BlackElo "1874"]
[TimeControl "300+1"]
[Annotator "20. +0.24 14... -0.21"]

1. e4 e5 2. Nf3 Nf6 3. d4 Nxe4 4. Bd3 d5 5. Nxe5 Nd7 6. Nxd7 Bxd7 7. O-O
Bd6 8. c4 c6 9. cxd5 cxd5 10. Nc3 Nxc3 11. bxc3 O-O 12. Qh5 g6 13. Qxd5 Qc7
14. c4 Bxh2+ {-0.21/19 +290} 15. Kh1 Bc6 {-0.29/20 +280} 16. Qg5 Bd6
{-0.15/21 +271} 17. d5 Rae8 {-0.15/21 +263} 18. Bb2 Be5 {-0.23/22 +254} 19.
Rfb1 Bxb2 {-0.07/20 +246} 20. Rxb2 {+0.24/29 +305} Bd7 {-0.09/21 +245} 21.
a4 {+0.35/34 +269} b6 {-0.13/20 +237} 22. a5 {+0.34/33 +259} bxa5
{-0.21/21 +238} 23. Qg3 {+0.41/30 +253} Re5 {-0.20/23 +237} 24. Kg1
{+1.22/26 +247} f6 {+0.00/18 +229} 25. f4 {+1.33/30 +248} Re7
{-0.56/21 +222} 26. Qf2 {+1.62/27 +247} Rc8 {-0.78/21 +217} 27. Rab1
{+1.93/32 +225} Be8 {-0.90/21 +217} 28. Rb7 {+2.29/29 +221} Qxb7
{-1.35/23 +216} 29. Rxb7 {+4.08/24 +213} Rxb7 {-1.45/23 +214} 30. c5
{+4.33/23 +213} Bb5 {-1.94/23 +172} 31. Bxb5 {+6.40/34 +196} Rxb5
{-2.91/20 +163} 32. c6 {+7.64/31 +187} a6 {-3.13/19 +164} 33. Qe3
{+20.10/24 +174} Kg7 {-4.33/22 +165} 34. Qe6 {+28.53/21 +168} Rbb8
{-5.03/22 +157} 35. d6 {+128.37/29 +162} Re8 {-7.34/19 +149} 36. Qd7+
{+128.38/25 +162} Kh6 {-9.85/24 +142} 37. c7 {+128.40/27 +163} Rec8
{-10.10/24 +135} 38. cxb8=R {+128.42/29 +147} Rxb8 {-10.87/21 +128} 39. Qc7
{+128.43/29 +147} Rh8 {-11.33/23 +122} 40. d7 {+128.44/21 +148} f5
{-10.82/20 +117} 41. Qxa5 {+128.45/27 +149} Ra8 {-11.41/20 +111} 42. d8=Q
{+128.46/24 +149} Rxd8 {-11.83/18 +105} 43. Qxd8 {+128.47/31 +150} Kg7
{-150.00/17 +101} 44. g4 {+128.48/33 +151} fxg4 {-150.00/35 +96} 45. f5
{+128.49/33 +152} gxf5 {-150.00/48 +91} 46. Qg5+ {+128.50/42 +152} Kh8
{-150.00/51 +87} 47. Qxf5 {+128.50/47 +153} a5 {-150.00/64 +87} 48. Qf8#
{+1000.01/127 +154}
{gogobello checkmated} 1-0


[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "3"]
[White "Raptor"]
[Black "McBrain"]
[Result "1/2-1/2"]
[WhiteElo "1977"]
[BlackElo "2378"]
[TimeControl "300+1"]
[Annotator "17. +0.42 16... +0.41"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O 8.
c3 d5 9. exd5 Nxd5 10. Nxe5 Nxe5 11. Rxe5 c6 12. d4 Bd6 13. Re1 Qh4 14. g3
Qh3 15. Be3 Bg4 16. Qd3 Rae8 {+0.41/24 +310} 17. Nd2 {+0.42/25 +251} Qh5
18. Ne4 {+0.75/24 +233} Bf5 {-0.34/30 +306} 19. Bd2 Rxe4 {+0.10/24 +307}
20. Rxe4 {+0.06/26 +201} Qg6 {-0.34/35 +304} 21. Rae1 {+0.00/27 +191} Nf6
{-0.34/33 +294} 22. f3 {+0.00/28 +191} Bxg3 {-0.34/33 +293} 23. hxg3
{+0.00/28 +183} Qxg3+ {-0.34/36 +294} 24. Kh1 {+0.00/30 +175} Bg4
{-0.34/35 +282} 25. Bxf7+ {+0.20/24 +172} Rxf7 {-0.34/36 +270} 26. Rxg4
{+0.15/27 +173} Nxg4 {-0.34/36 +253} 27. Re8+ {+0.08/27 +174} Rf8
{-0.34/33 +254} 28. Rxf8+ {+0.08/27 +156} Kxf8 {-0.34/37 +255} 29. Qf5+
{+0.00/25 +150} Nf6 {-0.34/38 +233} 30. Bg5 {+0.00/25 +151} Kf7
{-0.34/34 +218} 31. Bxf6 gxf6 {+0.42/21 +219} 32. Qxh7+ {+0.34/21 +143} Ke6
{-0.34/36 +215} 33. Qe4+ {+0.34/23 +144} Kd6 {-0.34/33 +214} 34. Qe2
{+0.27/22 +141} a5 {-0.34/33 +206} 35. Qe4 {+0.00/24 +123} a4
{-0.34/34 +198} 36. Qf5 {+0.00/22 +119} Qg5 {-0.34/36 +189} 37. Qxg5
{+0.36/24 +117} fxg5 {+0.00/49 +185} 38. Kg2 {+0.03/30 +117} Kd5
{+0.00/54 +170} 39. Kg3 {+0.01/33 +117} Kc4 {+0.00/59 +168} 40. f4
{+0.01/33 +115} gxf4+ {+0.00/59 +158} 41. Kxf4 {+0.01/36 +116} b4
{+0.00/71 +157} 42. cxb4 {+0.00/31 +111} Kxd4 {+0.00/127 +158} 43. a3 Kc4
{+0.00/37 +159} 44. Ke5 c5 {+0.00/39 +160} 45. b5 Kxb5 {+0.00/43 +160} 46.
Kd5 {+0.00/47 +95} Kb6 {+0.00/1 +161} 47. Ke4 {+0.00/49 +90} Kc6
{+0.00/52 +153} 48. Kd3 {+0.00/50 +86} Kb5 {+0.00/52 +143} 49. Kc3
{+0.00/53 +87} Kc6 {+0.00/54 +137} 50. Kc4 {+0.00/55 +82} Kb6
{+0.00/58 +126} 51. Kc3 {+0.00/58 +83} Kb5 {+0.00/58 +115} 52. Kd3
{+0.00/55 +73} Kc6 {+0.00/63 +116} 53. Kc4 {+0.00/54 +65} Kb6
{+0.00/62 +85} 54. Kc3 {+0.00/58 +66} Ka5 {+0.00/64 +53} 55. Kc4 Kb6
{+0.00/62 +53}
{Game drawn by repetition} 1/2-1/2

[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "4"]
[White "McBrain"]
[Black "ArasanX"]
[Result "1-0"]
[WhiteElo "2363"]
[BlackElo "2062"]
[TimeControl "300+1"]
[Annotator "28. +1.43"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
c3 O-O 9. h3 Re8 10. Ng5 Rf8 11. Nf3 Na5 12. Bc2 c5 13. d4 cxd4 14. cxd4
Qc7 15. Nc3 Bb7 16. d5 Rac8 17. Bd3 Nd7 18. Ne2 f5 19. Ng3 f4 20. Nf5 Bd8
21. b3 g6 22. Nh6+ Kg7 23. Ng4 h5 24. Ngh2 Qb6 25. Bd2 Nc5 26. Bc2 Nd7 27.
Rc1 b4 28. Rf1 {+1.43/28 +312} Ba8 29. Kh1 {+2.46/23 +303} Nb7 30. g3
{+2.49/28 +253} Be7 31. gxf4 {+2.98/28 +239} exf4 32. Rg1 {+3.10/30 +240}
Nd8 33. Bb1 {+3.30/30 +222} Rxc1 34. Qxc1 {+3.50/29 +209} Kh7 35. Bxf4
{+3.71/29 +201} Qc5 36. Qd2 {+3.70/28 +199} Qc3 37. Rc1 {+3.89/30 +196}
Qxd2 38. Bxd2 {+3.76/28 +197} Nc5 39. Bxb4 {+3.83/31 +194} Nf7 40. Bd2
{+4.03/30 +171} a5 41. Kg2 {+4.10/30 +171} Bd8 42. Be3 {+4.11/29 +163} Bb6
43. Nf1 {+4.29/29 +153} a4 44. b4 {+4.62/30 +147} Nd7 45. Bxb6
{+4.69/26 +147} Nxb6 46. Ne3 {+4.81/29 +148} Kg8 47. Rc7 {+5.03/26 +144}
Rb8 48. Nd4 {+5.00/26 +144} Rb7 49. Rc1 {+5.13/27 +143} Kf8 50. f4
{+5.46/26 +135} Rb8 51. Bd3 {+5.58/26 +134} Kg7 52. h4 {+5.90/28 +130} Kh7
53. a3 {+6.64/25 +126} Kh8 54. Rc7 {+6.97/26 +122} Rb7 55. Rxb7
{+7.33/26 +114} Bxb7 56. Nc4 {+7.28/25 +115} Nxc4 57. Bxc4 {+7.95/29 +108}
Kg8 58. Bb5 {+8.18/28 +107}
{ArasanX resigns} 1-0

[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "5"]
[White "Texel"]
[Black "McBrain"]
[Result "0-1"]
[WhiteElo "2222"]
[BlackElo "2369"]
[TimeControl "300+1"]
[Annotator "11. +0.19 12... +0.00"]

1. d4 d5 2. c4 e6 3. Nc3 c6 4. Nf3 Nf6 5. cxd5 exd5 6. Bf4 Be7 7. h3 Bf5 8.
e3 Nbd7 9. Nh4 Be4 10. Be2 h6 11. Nf3 {+0.19/22 +296} O-O 12. Rc1
{+0.13/22 +284} Bh7 {+0.00/28 +302} 13. O-O {+0.17/23 +273} Nb6
{-0.01/30 +303} 14. Ne5 {+0.17/21 +261} Bd6 {+0.06/28 +297} 15. Bd3
{+0.15/22 +255} Bxd3 {+0.07/31 +286} 16. Nxd3 {+0.19/25 +242} Bxf4
{+0.05/33 +286} 17. Nxf4 {+0.20/26 +232} Nc4 {+0.08/32 +287} 18. b3
{+0.18/26 +220} Nd6 {+0.02/31 +276} 19. Nd3 {+0.16/25 +221} Qe7
{+0.04/31 +269} 20. Qf3 {+0.19/25 +211} Nfe4 {+0.07/30 +258} 21. Qf4
{+0.17/24 +201} g5 {+0.09/30 +230} 22. Qg4 {+0.37/23 +193} Nxc3
{+0.09/32 +231} 23. Rxc3 {+0.44/24 +184} Kg7 {+0.06/30 +229} 24. a4
{+0.44/23 +180} Qe4 {+0.00/34 +197} 25. Qe2 {+0.41/23 +172} Rae8
{+0.08/29 +190} 26. Nc5 {+0.32/21 +164} Qg6 {+0.02/35 +191} 27. Nd7
{+0.31/23 +157} Rh8 {+0.00/30 +184} 28. Ne5 {+0.25/24 +158} Qf5
{+0.16/31 +175} 29. Rc2 {+0.20/23 +159} h5 {+0.16/37 +168} 30. a5
{+0.15/21 +156} a6 {+0.16/38 +129} 31. Re1 {+0.13/23 +156} Ne4
{+0.24/31 +118} 32. b4 {+0.09/22 +149} Re7 {+0.24/31 +111} 33. Rec1
{+0.00/23 +142} f6 {+0.49/26 +108} 34. Nd3 {-0.27/21 +137} g4
{+0.54/30 +106} 35. Nf4 {-0.27/23 +133} g3 {+0.39/32 +106} 36. fxg3
{-0.22/24 +122} Nxg3 {+0.40/35 +107} 37. Qe1 {-0.20/24 +117} h4
{+0.08/33 +80} 38. Rf2 {-0.27/27 +118} Qg5 {+0.34/29 +74} 39. Rf3
{-0.27/25 +111} Qh6 {+0.48/25 +70} 40. Kh2 {+0.00/21 +106} Rhe8
{+0.31/29 +61} 41. b5 {+0.00/23 +101} axb5 {+0.70/25 +59} 42. a6
{+0.00/24 +99} f5 {+0.55/26 +58} 43. Qa5 {-0.52/21 +84} Qd6 {+1.08/23 +56}
44. a7 {-0.79/21 +83} Ra8 {+1.03/30 +50} 45. Rc2 {-0.94/22 +76} b4
{+1.17/29 +46} 46. Qb6 {-0.95/23 +72} Qd8 {+0.95/28 +46} 47. Qc5
{-1.13/22 +69} b3 {+1.35/29 +44} 48. Rb2 {-1.26/22 +69} Ne4 {+1.18/24 +44}
49. Qc1 {-1.37/21 +66} Rxa7 {+1.11/31 +35} 50. Rxb3 {-1.32/23 +67} Kg8
{+1.33/27 +33} 51. Qe1 {-1.32/21 +64} Rf7 {+1.33/29 +26} 52. Rb2
{-1.48/21 +57} Qg5 {+1.25/31 +27} 53. Kg1 {-1.47/22 +54} Re7 {+1.47/28 +26}
54. Rb6 {-1.57/22 +53} Ra2 {+2.56/30 +25} 55. Qc1 {-2.89/21 +50} Nd2
{+4.26/26 +24} 56. Rf2 {-3.06/21 +49} Qg3 {+4.72/28 +23} 57. Rb2
{-3.26/22 +46} Nb3 {+4.95/31 +24} 58. Nxd5 {-3.27/22 +44} Nxc1
{+5.22/26 +23} 59. Nxe7+ {-3.15/23 +45} Kh8 {+4.97/22 +24} 60. Rxa2
{-3.32/22 +42} Nxa2 {+5.89/26 +20} 61. Nxf5 {-3.43/22 +43} Qg8
{+5.86/19 +21} 62. Rf4 {-3.55/21 +42} Qb3 {+6.13/24 +20} 63. Rxh4+
{-3.67/20 +41} Kg8 {+6.25/23 +20} 64. Nh6+ {-3.71/20 +38} Kf8
{+6.56/25 +18} 65. Rf4+ {-4.02/20 +39} Ke8 {+6.90/24 +15} 66. Nf5
{-4.07/21 +34} Nc3 {+7.54/23 +13} 67. Ng3 {-4.71/20 +35} b5 {+8.60/20 +12}
68. Kh2 {-5.37/18 +33} b4 {+10.08/23 +12} 69. h4 {-5.95/18 +31} Qb1
{+12.27/22 +13} 70. d5 {-6.48/17 +29} cxd5 {+14.12/21 +12} 71. Rf1
{-8.42/18 +27} Qd3 {+19.92/21 +9} 72. Rf4 {-9.95/20 +28} b3 {+23.20/17 +8}
73. Rb4 {-11.58/19 +26} Qc2 {+128.39/22 +9} 74. Rb8+ Kd7 {-15.86/19 +9} 75.
h5 b2 {-17.73/17 +10} 76. Rb7+ {-1008.88/16 +17} Kc6 {+128.44/16 +11} 77.
Rb4 {-1004.87/16 +16} b1=Q {+128.45/19 +12} 78. Rxb1 {-1000.19/18 +15} Qxb1
{+128.46/26 +12} 79. h6 {-1000.13/19 +15} Qg6 {+128.47/22 +13} 80. h7
{-1000.12/20 +14} Qxh7+ {+128.48/24 +14} 81. Kg1 {-1000.10/19 +14} Ne2+
{+128.49/25 +15} 82. Nxe2 {-1000.20/18 +13} d4 {+128.50/24 +15} 83. Nxd4+
{-1000.16/24 +12} Kd5 {+128.51/19 +16} 84. Kf2 {-1000.14/23 +12} Ke4
{+128.51/18 +17} 85. Nf3 {-1000.13/27 +11} Qh6 {+128.51/19 +17} 86. Nd2+
{-1000.13/27 +11} Kd3 {+128.51/18 +18} 87. Nf3 {-1000.12/25 +12} Qxe3+
{+1000.21/17 +19} 88. Kg3 {-1000.11/23 +12} Ke4 {+1000.17/26 +19} 89. Kg4
{-1000.10/21 +13} Qe2 {+128.51/14 +20} 90. Kg3 Qe3 {-1000.10/21 +21} 91.
Kg4 {-1000.10/21 +15} Qf4+ {+1000.15/1 +22} 92. Kh3 {-1000.09/19 +13} Kf5
{+1000.09/29 +22} 93. Nh4+ {-1000.08/17 +14} Kg5 {+1000.08/1 +23} 94. Nf3+
{-1000.07/15 +15} Kh5 {+1000.07/1 +24} 95. g4+ {-1000.06/13 +16} Qxg4+
{+1000.06/47 +25} 96. Kh2 {-1000.05/11 +17} Qxf3 {+1000.05/60 +25} 97. Kg1
{-1000.04/9 +16} Kg4 {+1000.04/81 +26} 98. Kh2 {-1000.03/7 +15} Qe2+
{+1000.03/127 +27} 99. Kg1 {-1000.02/5 +16} Kg3 {+1000.02/127 +27} 100. Kh1
{-1000.01/3 +17} Qd1# {+1000.01/127 +28}
{Texel checkmated} 0-1


[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "6"]
[White "McBrain"]
[Black "WaDuuttie"]
[Result "1/2-1/2"]
[WhiteElo "2380"]
[BlackElo "2037"]
[TimeControl "300+1"]
[Annotator "28. +0.14 22... -0.54"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Be3 e5 7. Nf3 Be7 8.
Bc4 O-O 9. O-O Be6 10. Bxe6 fxe6 11. Na4 Ng4 12. Qd3 b5 13. Nb6 Ra7 14. Nd5
Nxe3 15. Qxe3 Rb7 16. Qd2 Qd7 17. Nxe7+ Qxe7 18. Rfd1 Rb6 19. a4 bxa4 20.
Rxa4 Qb7 21. c3 Rc8 22. Ne1 Nd7 {-0.54/17 +306} 23. Nd3 Nc5 {-0.66/17 +294}
24. Nxc5 Rxc5 {-0.66/18 +283} 25. b4 Rc4 {-0.70/18 +269} 26. f3 Rbc6
{-0.48/19 +253} 27. Ra3 Kf7 {-0.52/20 +244} 28. Rf1 {+0.14/34 +311} h6
{-0.40/20 +200} 29. Rfa1 {+0.14/44 +312} Qc8 {-0.22/19 +179} 30. h3
{+0.14/46 +301} Ke7 {-0.40/18 +176} 31. Kh2 {+0.14/35 +286} Rxc3
{-0.50/20 +173} 32. Rxc3 {+0.27/30 +277} Rxc3 {-0.46/21 +165} 33. Rxa6
{+0.06/35 +278} Qxa6 {-0.52/19 +160} 34. Qxc3 {+0.07/38 +268} Qb6
{-0.56/19 +157} 35. h4 {+0.20/36 +248} Kd7 {-0.40/17 +153} 36. Qe1
{+0.11/38 +249} Kc6 {-0.40/16 +149} 37. Kh3 {+0.00/41 +235} Qb5
{-0.52/16 +147} 38. Qd2 {+0.00/33 +236} Qa4 {-0.50/17 +145} 39. h5
{+0.00/42 +230} Kc7 {-0.28/16 +142} 40. Qe2 {+0.00/34 +225} Kb6
{-0.16/17 +140} 41. Qc4 {+0.00/36 +216} Qe8 {+0.41/19 +138} 42. Kg4
{+0.00/38 +199} Qe7 {+0.41/21 +136} 43. Kh3 {+0.00/39 +199} Qf7
{+0.41/22 +132} 44. Kg4 {+0.00/41 +195} Qf4+ {+0.41/23 +127} 45. Kh3
{+0.00/41 +196} Qf7 {+0.41/25 +122} 46. Kg4 {+0.00/40 +196} Qf4+
{+0.41/26 +119} 47. Kh3 {+0.00/44 +197} Qf7 {+0.41/26 +117}
{Game drawn by repetition} 1/2-1/2

[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "7"]
[White "Baron"]
[Black "McBrain"]
[Result "0-1"]
[WhiteElo "1908"]
[BlackElo "2366"]
[TimeControl "300+1"]
[Annotator "13... +0.25"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. d3 Bc5 5. c3 O-O 6. Nbd2 d6 7. h3 a6 8.
Bxc6 bxc6 9. O-O Bb6 10. Nc4 h6 11. Be3 Re8 12. Re1 Bxe3 13. Nxe3 Nh5
{+0.25/27 +298} 14. Qc2 Qf6 {+0.43/24 +295} 15. d4 Bxh3 {+0.69/26 +290} 16.
dxe5 dxe5 {+0.78/27 +285} 17. Qa4 Rab8 {+0.94/26 +281} 18. b3 Re6
{+1.67/26 +277} 19. Qxa6 Bxg2 {+2.07/32 +270} 20. Kxg2 Qg6+ {+2.03/32 +271}
21. Kf1 Nf4 {+3.08/29 +266} 22. Nh2 Qxe4 {+6.27/25 +260} 23. f3 Qg6
{+6.51/26 +260} 24. Qa7 Rbe8 {+7.84/27 +254} 25. Rad1 Qg3 {+11.45/30 +246}
26. Rd2 e4 {+12.02/28 +245} 27. Nd1 exf3 {+55.03/36 +239} 28. Rxe6 Rxe6
{+64.97/37 +240} 29. Qf2 Qh3+ {+128.25/33 +239} 30. Kg1 Rg6+
{+1000.60/37 +240} 31. Kh1 Rg2 {+1000.15/33 +230} 32. Qxg2 fxg2+
{+1000.14/33 +224} 33. Kg1 Qg3 {+1000.13/33 +223} 34. Nf2 Qxc3
{+1000.12/35 +224} 35. Rd1 h5 {+1000.11/32 +225} 36. a4 Kh7
{+1000.10/25 +225} 37. a5 Qe3 {+1000.09/29 +226} 38. b4 f5
{+1000.07/29 +227} 39. Nf3 Qxf3 {+1000.06/40 +227} 40. Re1 h4
{+1000.05/46 +228} 41. b5 Qg3 {+1000.04/51 +229} 42. bxc6 h3
{+1000.03/83 +230} 43. Ng4 fxg4 {+1000.02/127 +230} 44. a6 h2#
{+1000.01/127 +231}
{Baron checkmated} 0-1


[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "8"]
[White "McBrain"]
[Black "Chiron"]
[Result "1-0"]
[WhiteElo "2369"]
[BlackElo "2187"]
[TimeControl "300+1"]
[Annotator "17. +0.28"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. O-O Nxe4 5. Re1 Nd6 6. Nxe5 Be7 7. Bf1
Nxe5 8. Rxe5 O-O 9. d4 Bf6 10. Re1 Re8 11. Bf4 Rxe1 12. Qxe1 Ne8 13. c3 d5
14. Nd2 Bf5 15. Qe2 c6 16. Re1 a5 17. g4 {+0.28/27 +305} Be6 18. h3
{+0.29/28 +293} a4 19. a3 {+0.27/29 +292} h6 20. Qe3 {+0.25/31 +285} Qb6
21. Rb1 {+0.31/32 +280} Be7 22. Bd3 {+0.31/28 +269} Nf6 23. Bc2
{+0.38/29 +231} Qb5 24. Be5 {+0.35/25 +227} Nd7 25. Bf4 {+0.20/31 +186} Nf6
26. Qf3 {+0.25/30 +180} Re8 27. Qd1 {+0.38/30 +170} Ra8 28. Kg2
{+0.43/28 +153} c5 29. Nf3 {+0.46/30 +147} c4 30. Ne5 {+0.40/32 +147} Bd6
31. Qc1 {+0.42/31 +148} Nd7 32. Nf3 {+0.37/34 +148} Bxf4 33. Qxf4
{+0.59/26 +142} Re8 34. h4 {+0.48/31 +120} Nf8 35. g5 {+0.64/28 +120} h5
36. Qd6 {+0.65/32 +120} g6 37. Ne5 {+0.50/42 +84} Nd7 38. f4 {+0.43/39 +79}
Nxe5 39. fxe5 {+0.43/45 +67} b6 40. Qc7 {+0.64/37 +65} Bd7 41. Kf2
{+0.64/45 +61} Kg7 42. Ke2 {+0.64/39 +58} Bc8 43. Kd2 {+0.64/40 +54} Qd7
44. Qxd7 {+0.64/44 +55} Bxd7 45. Re1 {+0.64/52 +53} b5 46. Rf1
{+0.64/54 +53} Bh3 47. Re1 {+0.64/55 +51} Bg4 48. Rf1 {+0.64/51 +45} Bh3
49. Rf6 {+0.64/54 +37} Be6 50. Ke3 {+0.64/55 +38} Re7 51. Rf3
{+0.64/54 +35} Bg4 52. Rg3 {+0.64/62 +34} Be6 53. Kd2 {+0.64/63 +31} Bg4
54. Ke3 {+0.64/64 +29} Be6 55. Rg1 {+0.64/63 +22} Bh3 56. Rh1
{+0.64/58 +20} Bg4 57. Rf1 {+0.64/61 +21} Be6 58. Bb1 {+0.64/55 +19} Bh3
59. Rf6 {+0.64/64 +18} Be6 60. Bc2 {+0.64/65 +19} Re8 61. Rf1
{+0.64/65 +20} Re7 62. Kf4 {+0.64/60 +20} Re8 63. Kg3 {+0.64/53 +19} Re7
64. Kf4 {+0.64/64 +9} Re8 65. Ra1 {+0.64/57 +5} Bg4 66. Rh1 {+0.64/60 +4}
Be2 67. Rc1 {+0.64/48 +4} Kf8 68. Re1 {+0.64/54 +4} Bg4 69. Kg3
{+0.64/54 +4} Be6 70. Bd1 {+0.64/53 +3} Kg8 71. Rh1 {+0.64/51 +2} Bf5 72.
Rf1 {+0.64/44 +1} Be6 73. Kf4 {+0.64/50 +1} Bh3 74. Rf2 {+0.64/43 +1} Bc8
75. Kg3 {+0.64/41 +1} Rd8 76. Rg2 {+0.64/39 +1} Kg7 77. Kf4 {+0.64/38 +1}
Kg8 78. Bc2 {+0.64/36 +1} Bd7 79. Ke3 {+0.64/40 +1} Be6 80. Kd2
{+0.64/52 +1} Bc8 81. Re2 {+0.64/43 +1} Bd7 82. Rf2 {+0.64/44 +1} Be6 83.
Rf1 {+0.64/51 +1} Rd7 84. Ke2 {+0.64/45 +1} Re7 85. Kf3 {+0.64/51 +2} Bh3
86. Re1 {+0.64/48 +2} Bc8 87. Ke3 {+0.64/48 +1} Bh3 88. Rg1 {+0.64/44 +1}
Be6 89. Bd1 {+0.39/18 +1} Kf8 90. Bf3 {+0.31/30 +1} Rd7 91. Bg2
{+0.00/34 +1} Bf5 92. Rf1 {+0.00/39 +2} Ke7 93. Rxf5 {+0.00/40 +3} gxf5 94.
Bh3 {+0.00/39 +3} b4 95. cxb4 {+0.43/28 +3} f4+ 96. Kd2 {+0.00/34 +3} Rd8
97. Bg2 {+1.42/25 +3} Ke6 98. Ke2 {+1.05/26 +4} Rb8 99. Bf3 {+0.08/33 +5}
Rh8 100. Ke1 {+1.83/28 +5} Kf5 101. Bxd5 {+3.50/26 +5} Rf8 102. b5
{+6.26/22 +5} Kg4 103. g6 {+7.29/24 +6} fxg6 104. Bc6 {+7.98/25 +6} Kg3
105. b6 {+9.62/27 +7} f3 106. b7 {+10.45/24 +8} Kg2 107. e6 {+11.03/27 +8}
g5 108. e7 {+12.03/28 +9} Re8 109. Kd2 {+12.53/24 +10} Rxe7 110. b8=Q
{+15.69/23 +8} Re6 111. Bxf3+ {+128.40/29 +8} Kxf3 112. Qb7+
{+128.41/31 +9} Kg3 113. Qf7 {+128.42/33 +10} Re4 114. hxg5
{+128.44/36 +10} Rg4 115. g6 {+1000.12/37 +8} Rxd4+ 116. Ke3
{+1000.11/38 +9} Rd3+ 117. Ke4 {+1000.10/45 +10} Kg2 118. g7
{+1000.09/46 +11} Rg3 119. g8=Q {+1000.08/46 +11} Rxg8 120. Qxg8+
{+1000.07/50 +11} Kf2 121. Qxc4 {+1000.06/69 +11} h4 122. Qc2+
{+1000.05/92 +12} Kg3 123. Ke3 {+1000.04/127 +12} Kg4 124. Qg6+
{+1000.03/127 +13} Kh3 125. Kf2 {+1000.02/127 +14} Kh2 126. Qg2#
{+1000.01/127 +14}
{Chiron checkmated} 1-0


[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2017.06.17"]
[Round "9"]
[White "Goldbar"]
[Black "McBrain"]
[Result "0-1"]
[WhiteElo "1931"]
[BlackElo "2378"]
[TimeControl "300+1"]
[Annotator "19. +0.27 18... +0.12"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. O-O Nf6 5. d3 d6 6. c3 O-O 7. h3 a6 8.
Re1 h6 9. a4 a5 10. Nbd2 Be6 11. b3 d5 12. exd5 Nxd5 13. Qc2 Re8 14. Ne4
Bf8 15. Ng3 Nb6 16. Bb5 Bd5 17. Nh2 Nd7 18. Rb1 Ne7 {+0.12/15 +312} 19. Ng4
{+0.27/15 +295} Ng6 {-0.09/27 +307} 20. Bd2 {+0.23/15 +285} c6
{+0.11/27 +299} 21. Bc4 {+0.12/16 +286} Nb6 {+0.14/30 +279} 22. Be3
{+0.13/14 +279} Qc7 {+0.08/29 +276} 23. Nh5 {+0.13/15 +276} Kh8
{+0.13/28 +261} 24. Bxb6 {+0.09/15 +266} Qxb6 {+0.12/32 +247} 25. Ne3
{+0.00/16 +266} Qd8 {+0.23/28 +237} 26. Nxd5 {-0.01/16 +261} cxd5
{+0.52/26 +227} 27. Bb5 {-0.02/19 +261} Re7 {+0.63/27 +219} 28. d4
{-0.11/16 +257} e4 {+0.68/29 +212} 29. c4 {-0.01/15 +254} f5
{+0.80/29 +199} 30. c5 {-0.17/16 +238} Rf7 {+1.15/30 +184} 31. g3
{-0.44/17 +236} f4 {+2.12/25 +180} 32. Qe2 {-1.11/18 +216} f3
{+2.95/26 +175} 33. Qd2 {-0.60/18 +193} Be7 {+3.66/29 +169} 34. b4
{-0.58/17 +190} Rf5 {+3.96/25 +164} 35. g4 {-0.68/17 +187} Rxh5
{+8.05/27 +159} 36. gxh5 {-1.51/17 +167} Bg5 {+9.78/30 +149} 37. Qc2
{-3.00/17 +167} Qc8 {+12.79/30 +142} 38. Rxe4 {-6.02/18 +158} Qxh3
{+15.82/31 +143} 39. Bf1 {-6.02/16 +155} Qxh5 {+16.81/30 +142} 40. Bc4
{-6.09/16 +145} dxc4 {+1000.08/39 +139} 41. Qxc4 {-99.86/16 +130} Bf4
{+1000.07/54 +133} 42. Rxf4 {-99.88/31 +131} Nxf4 {+1000.06/57 +133} 43.
Qf1 {-99.90/60 +131} Re8 {+1000.05/72 +120} 44. bxa5 {-99.92/62 +132} Ne2+
{+1000.04/76 +116} 45. Qxe2 {-99.94/62 +133} Qg6+ {+1000.03/103 +115} 46.
Kf1 {-99.96/62 +134} Qxb1+ {+1000.02/127 +115} 47. Qd1 {-99.98/62 +135}
Qxd1# {+1000.01/127 +116}
{Goldbar checkmated} 0-1
[/pgn]

By far the most interesting position and move was in game 8 playing against the co-leader at the time, Chiron. The position was a dead draw, and from all appearances , it looked like it was going to be a draw. The last pawn move /capture was 45. ...b5, so at move 93, in just a two more moves , it would be draw by 50 Move rule. McBrain "respect" setting ( similar to "contempt" but opposite sign) was a minus 10.

[d]8/3rkp2/6p1/1p1pPbPp/p1pP3P/P1P1K3/1P4B1/5R2 w - - 94 93

93. Rxf5 {+0.00/40 +3}

My thought was ok - it was about to be drawn, but even if McBrain plays Rxf6 , the game is still drawn but it will last a few more moves. Both engines at this point only had few seconds on the clock with a 1 second increment game - so time may have been a factor on what transpire below. McBrain had 3 seconds on the clock after playing Rxf5.

The crucial position is here:

[d]7r/5p2/4k3/3pP1Pp/pPpP1p1P/P4B2/1P6/4K3 b - - 9 100

The game is still drawn, but black plays Kf5?? and the game is over.

Code: Select all

38	+8.70 	9.20G	4:36.71	Bxd5 Kg6 Be4+ Kg7 Kd2 Re8 Kc3 Kf8 Bf3 Re6 b5 Ke8 Kb4 Kd8 Kxa4 Re8 Kb4 Kc8 a4 Rd8 Kxc4 Kd7 a5 Rc8+ Kd3 Rc1 a6 Ra1 Bxh5 Kc7 Bxf7 Rd1+ Ke4 f3 Ke3 f2 Kxf2 Rxd4 g6 Kb6 Kf3 Kxb5 g7 Rd8 h5 Kxa6 Bc4+ Ka7 
 38	+8.25!	6.46G	3:15.25	Bxd5! 
 38	+7.91?	6.43G	3:14.36	Bxd5 Rf8? 
 38	+8.18!	5.74G	2:54.42	Bxd5! 
 38	+8.15!	5.46G	2:46.22	Bxd5! 
 38	+8.00!	5.37G	2:43.39	Bxd5! 
 38	+7.80?	5.36G	2:43.11	Bxd5 Kg6? 
 38	+7.97?	5.33G	2:42.28	Bxd5 Rf8? 
 37	+8.10 	5.27G	2:40.59	Bxd5 Kg6 Be4+ Kg7 Kd2 Re8 Kc3 Kf8 Bf3 Rd8 d5 Rb8 Be4 Ke7 Bh1 Re8 Bf3 Kd7 Kd4 Rh8 Kxc4 Re8 Kd4 Rc8 Kd3 Ke7 Ke4 Rc4+ Kf5 Rd4 d6+ Kd7 Bxh5 f3 Bxf3 Rxh4 Bd1 Ke8 Bxa4+ Kf8 Bc6 Rh2 b5 Rf2+ Ke4 Rc2 Kd5 Rf2 a4 Rd2+ Ke4 Re2+ Kd4 Rd2+ Kc4 
 37	+8.08!	4.55G	2:18.81	Bxd5! 

So this is a good example of why using a 'respect' or 'contempt' setting may be worthwhile to use on occasion as an engine may be able to swindle a won game from a drawn game.

[JVMerlino]

Thanks for the tournament, HG!

[Ferdy]

[d]8/5pk1/p5p1/4p2p/7P/1p3qP1/2n5/1N3QK1 b - - 1 45

The following bug occurred in Kingslayer-Schooner. Here is the output from the debug file:

Code: Select all

5893454 <first &#58; 8 1704 0 10838  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1c3 d1e1 c3b1
5893456 <first &#58; 9 1672 0 15561  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1c3 d1e1 c3b1 e5e4
5893463 <first &#58; 10 1774 1 30375  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1a3 d1e1 a3c4 f2e4 c4e5
5893475 <first &#58; 11 1723 2 57657  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1c3 b3b2 c3d1 f2d1 g2h2 b2b1q
5893491 <first &#58; 12 1830 4 96367  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1a3 d1e1 g2f3 e5e4 f3g2 e4e3 g2f3 b3b2
5893522 <first &#58; 13 1854 7 166027  c2e3 f1f2 f3d1 g1h2 e3g4 h2g2 g4f2 b1a3 d1e1 a3c4 f2e4 g2f3 e4g3 f3g2

This cannot be duplicated in the subsequent analysis. It would be nice to have a guess (even a bad guess) as to what to look for and why it does this. This frustrating non-reproducible type of error occurs in maybe one in 100 games. Normal asserts and light debug code show no problems arising that would make it play unbearable moves. Nor was there any memory crash or segmentation violation. I am considering opening a new log file to record non-standard events. Suggestions as to what to include might be helpful. Perhaps restrict the saved information only for the PV moves, otherwise the hard drive might fill up too quickly. For the log file maybe include:

(1) game information header
(2) generated move list to see if moves like f1f3 are even in there.
(3) position structure information for each move

Try to check your move generator.

Here is Deuterium's perft 6, check if we are the same.

Code: Select all

position fen 8/5pk1/p5p1/4p2p/7P/1p3qP1/2n5/1N3QK1 b - - 1 45
perft 6

 f7f6 - 6038102
 f7f5 - 5870846
 a6a5 - 6800911
 g6g5 - 8461722
 e5e4 - 4758657
 b3b2 - 7529826
 c2a1 - 5238486
 c2e1 - 5021410
 c2a3 - 6213117
 c2e3 - 6554472
 c2b4 - 6654452
 c2d4 - 6536826
 f3d1 - 2724456
 f3f1 - 60915
 f3h1 - 395077
 f3e2 - 4060598
 f3f2 - 616091
 f3g2 - 304818
 f3c3 - 9865670
 f3d3 - 8038534
 f3e3 - 1501743
 f3g3 - 395346
 f3e4 - 9319274
 f3f4 - 9321048
 f3g4 - 10253920
 f3d5 - 10556869
 f3f5 - 8296371
 f3c6 - 11757092
 f3f6 - 8102058
 f3b7 - 9918438
 f3a8 - 9716139
 g7f8 - 6040871
 g7g8 - 6156224
 g7h8 - 5788826
 g7h7 - 6237128
 g7f6 - 3219632
 g7h6 - 5085252

 Perft 6 Summary
 Nodes: 223411217

[hgm]

This cannot be duplicated in the subsequent analysis.

First priority would be to make sure the engine will always be completely deterministic: clear hash table(s), killers and history before every move, play with ponder off. And if you use explicit randomization, remember the random seed, so that you can set it to the same value. Then just play games until you see a gross blunder. If such blunders happen every 100 games, that should not take too long.

In absence of that, what I do when a problem doesn't reproduce from the position it occurred in, it to start a few moves earlier, and limit the depth of those searches to the depth they had in the original game. If the engine picks the same moves, this goes a long way towards reproducing the contents of the various tables at the start of the erroneous search, with as a consequence that the original error re-appears.

[D Sceviour]

Here is Deuterium's perft 6, check if we are the same.

Code: Select all

position fen 8/5pk1/p5p1/4p2p/7P/1p3qP1/2n5/1N3QK1 b - - 1 45
perft 6

 f7f6 - 6038102
 f7f5 - 5870846
 a6a5 - 6800911
 g6g5 - 8461722
 e5e4 - 4758657
 b3b2 - 7529826
 c2a1 - 5238486
 c2e1 - 5021410
 c2a3 - 6213117
 c2e3 - 6554472
 c2b4 - 6654452
 c2d4 - 6536826
 f3d1 - 2724456
 f3f1 - 60915
 f3h1 - 395077
 f3e2 - 4060598
 f3f2 - 616091
 f3g2 - 304818
 f3c3 - 9865670
 f3d3 - 8038534
 f3e3 - 1501743
 f3g3 - 395346
 f3e4 - 9319274
 f3f4 - 9321048
 f3g4 - 10253920
 f3d5 - 10556869
 f3f5 - 8296371
 f3c6 - 11757092
 f3f6 - 8102058
 f3b7 - 9918438
 f3a8 - 9716139
 g7f8 - 6040871
 g7g8 - 6156224
 g7h8 - 5788826
 g7h7 - 6237128
 g7f6 - 3219632
 g7h6 - 5085252

 Perft 6 Summary
 Nodes: 223411217

Thank you. First instincts are that there might be something wrong with the move generator.

Schooner does not have a Perft test (although it does have a Perf test). Instead, it uses a full width search function called FWS in the console "help". It predates the invention of Perft from a time long ago. It used to allow an exact discovery of all move values without bounds, but FWS has been damaged somewhat with selective searching techniques. It might be time to update with Perft to compare with other programs.

Here is a FWS 6 output. The high branching factor is an indication that some selective searching has been turned off. There is nothing wrong with the move generator here:

Code: Select all

Search Iteration =  6
PV [6] 991  f3-g3 f1-g2 g3-h4 g2-d2 h4-g3 g1-h1 c2-e3 d2-b2 g3-e1
Branching factor = 5.087933957525186 avg =  8.424032221745771
Total Researches = 48

Nodes =             2343727
Total Nodes =       2917055

PV [0] 991  f3-g3 f1-g2 g3-h4 g2-d2 h4-g3 g1-h1 c2-e3 d2-b2 g3-e1
move f3g3
elap: 1.614596816260018     limit: 7.492307692307692 seconds
Total Nodes=2917056         nps:1806677

#Move Value
1 f3-g3  991              2 f7-f5  977  
3 f3-f4  968              4 f3-d3  968  
5 f3-g4  968              6 f3-c3  968  
7 f3-e2  968              8 f3-d1  968  
9 c2-e1  968              10 f3-f6  968  
11 f3-e4  968             12 g7-h7  968  
13 g7-h6  968             14 c2-b4  968  
15 g7-g8  968             16 g7-f8  968  
17 g7-h8  968             18 g7-f6  968  
19 c2-a1  968             20 f3-f5  968  
21 f3-d5  968             22 f3-c6  968  
23 f3-b7  968             24 f3-a8  968  
25 e5-e4  964             26 a6-a5  953  
27 f7-f6  950             28 f3-h1  942  
29 c2-a3  941             30 b3-b2  938  
31 f3-e3  930             32 f3-f1  882  
33 g6-g5  849             34 c2-d4  737  
35 f3-g2  712             36 c2-e3  667  
37 f3-f2 -122  

[D Sceviour]

In absence of that, what I do when a problem doesn't reproduce from the position it occurred in, it to start a few moves earlier, and limit the depth of those searches to the depth they had in the original game. If the engine picks the same moves, this goes a long way towards reproducing the contents of the various tables at the start of the erroneous search, with as a consequence that the original error re-appears.

Starting from a position before the fault is a good method. However, nothing useful occurred in this game by re-tracing.

Another method is to create a parallel log file with excess information to recover the complete thinking process. The problem is to decide how much information to save, and in what format the information should be saved. I do not want to program a large database style binary.

[AlvaroBegue]

If you haven't implemented perft in Schooner, just implement perft in Schooner. It's a tiny function, and completely trivial to write compared to alpha-beta search.

[D Sceviour]

If you haven't implemented perft in Schooner, just implement perft in Schooner. It's a tiny function, and completely trivial to write compared to alpha-beta search.

This might be more difficult than it sounds. Perft is best for comparing clones. Crafty uses a complete move list, and only tests for legality after making a move. Schooner does some legality testing for king moves during the move generation process. Robolitto actually prunes move during the move generation process with a SEE value. I have yet to figure out what Stockfish is doing.

[AlvaroBegue]

If you haven't implemented perft in Schooner, just implement perft in Schooner. It's a tiny function, and completely trivial to write compared to alpha-beta search.

This might be more difficult than it sounds. Perft is best for comparing clones. Crafty uses a complete move list, and only tests for legality after making a move. Schooner does some legality testing for king moves during the move generation process. Robolitto actually prunes move during the move generation process with a SEE value. I have yet to figure out what Stockfish is doing.

Perft is perfectly well defined and is independent of the kind of details of implementation you describe.

You don't count pseudo-legal moves in Perft. Whether you do that by actually not generating those moves or by checking the legality of the position after playing them, that's completely up to you. Whether you write out a complete list of moves or you generate them in stages, that's completely up to you.

Do you not have a mechanism to iterate over the legal moves in Schooner?