Crazyhouse tournaments and rating list

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Crazyhouse tournaments and rating list

Post by Ferdy »

hgm wrote: BTW, I just uploaded CrazyWa 0.0.9. This now also implement repetition detection. I also fixed a bug in the LMR, that it sometimes reduced directly into QS, because the limiting of the reduction did not always work. But fixig the bug made it weaker, so I figured I should let the LMR kick in at lower depth already. This version does that, and although it is too early to have a statistically significant conclusion, this put it in the lead in my 40/1 test matches against both Imortal 1.0 and Sunsetter 7g after 25 games each. (While before the bug fix it was doing at best 36% aganist Sunsetter 7g.)

This version still clears the hash table before every move. I wanted to do that during development to make errors reproducable. (Unfortunately this did not help, though.) It does crash a lot less often than previous versions; I located one cause of crashes (the randomization during the first 5 moves), and fixed another suspect (execution of the check-detection code after null move).
Running Blitz tour, I got this in the engine's dir, dump.txt

Code: Select all

crash: g7g3 (ranKey = 419672283)
 {626207} N@g7 {8a620707} f8g7 {60606} P@d7 {87060601} e8d7 {50005} N@b8 {9b050501} h8b8 {40004} g4g7 {8a040401} d7c6 {30003} B@e8 {9e030301} b8e8 {20002} c1f4 {10101} e8c8 {1} g7c7 {86010101} c8c7 23 21 00 00 00 23 3f 00 ff ff ff    ff fc ff ff ff ff ff ff ff ff ff
 20 20 00 00 00 20 20 20 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 00 20 00 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 00 00 00 22 00 00 ff ff ff    fd fc fd ff fe ff ff ff ff ff ff
 00 00 44 00 00 42 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 5f 00 40 40 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 40 40 43 00 00 00 00 40 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 00 27 00 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 ff ff ff ff ff ff ff ff ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 ff ff ff ff ff ff ff ff ff ff ff    f0 f0 f0 f0 f0 f0 f0 f0 f0 f0 f0
 ff ff ff ff ff ff ff ff ff ff ff    ff ff ff ff e9 ff ff ff ff ff ff


crash: b7a7 (ranKey = 673996001)
 {626212} b5c5 {611111} P@b7 {101010} c6b7 {f0f0f} P@d7 {870f0f0f} e8d7 {e0e0e} a7b7 {850e0e0e} P@c7 {d0d0d} B@e8 {9e0d0d0b} h8e8 {c0a0c} b7b8n {b0b00} a1a1 00 00 00 00 00 23 3f 00 ff ff ff    ff ff ff fe ff ff ff ff ff ff ff
 20 00 00 21 21 20 20 20 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 20 20 20 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 00 20 00 00 00 00 ff ff ff    fe fe fd ff fe ff ff ff ff ff ff
 00 00 43 40 00 41 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 00 20 40 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 40 5f 22 40 40 40 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 35 43 00 43 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 ff ff ff ff ff ff ff ff ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 ff ff ff ff ff ff ff ff ff ff ff    f0 f0 f0 f0 f0 f0 f0 f0 f0 f0 f0
 ff ff ff ff ff ff ff ff ff ff ff    ff ff ff ff cc ff ff ff ff ff ff


crash: bad hash move         b6e575d6: e4f4
 (ranKey = 699585017)
 {626214} f3g5 {611312} P@h2 {1d131212} g1h2 {111111} g4h4 {49111111} h2g1 {101010} e6f5 {f0f0f} e4f5 {e0e0c} P@a4 {d0b0b} e2a6 {a0a0a} b7a6 {90909} e1e7 {88090908} e8e7 {80707} c3a4 {60606} h4f2 {1b060605} g1f2 {50404} d7f5 23 00 00 00 00 00 00 00 ff ff ff    ff fb ff fe ff fe ff ff ff ff ff
 20 20 20 24 00 3f 20 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 21 00 00 00 00 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 21 00 00 00 00 22 00 00 ff ff ff    fd ff fe fe ff ff ff ff ff ff ff
 00 00 00 00 00 42 21 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 40 00 41 40 00 00 00 00 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 00 00 40 00 5f 40 40 40 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 43 00 00 00 00 00 00 43 ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 ff ff ff ff ff ff ff ff ff ff ff    ff ff ff ff ff ff ff ff ff ff ff
 ff ff ff ff ff ff ff ff ff ff ff    f0 f0 f0 f0 f0 f0 f0 f0 f0 f0 f0
 ff ff ff ff ff ff ff ff ff ff ff    ff ff ff ff dc ff ff ff ff ff ff
The games:

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "1"]
[White "TJchess 1.3 64bit"]
[Black "CrazyWa 0.0.9 32bit"]
[Result "1-0"]
[PlyCount "67"]
[Termination "abandoned"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. d4 {book} d5 {book} 2. Bf4 {book} Bf5 {book} 3. e3 {book} e6 {book}
4. Nd2 {book} Nd7 {-0.29/11 8.4s} 5. Bd3 {+0.29/12 4.0s} Bxd3 {+0.18/12 5.0s}
6. cxd3 {+0.35/12 5.6s} Bb4 {+0.04/10 6.1s} 7. Rc1 {+0.16/10 3.4s}
B@a5 {+0.44/10 6.9s} 8. B@c3 {+0.33/12 3.6s} Bxc3 {+1.93/12 3.3s}
9. bxc3 {+0.33/12 6.8s} B@b2 {+1.74/11 3.9s} 10. Ne2 {+0.69/11 5.8s}
Bxc1 {+1.59/11 7.4s} 11. Qxc1 {+0.82/11 5.6s} Ne7 {+1.66/9 7.1s}
12. Qb2 {+0.85/11 7.5s} Ng6 {+1.82/8 4.0s} 13. B@a4 {+0.87/10 4.8s}
Bb6 {+1.79/8 5.0s} 14. B@a3 {+1.06/10 4.1s} R@a5 {+2.02/9 8.7s}
15. Bxd7+ {+0.57/11 7.0s} Qxd7 {+2.45/10 8.5s} 16. Bb4 {+0.58/11 9.1s}
Ra4 {+1.98/9 6.6s} 17. N@c5 {+0.68/10 6.8s} Bxc5 {+3.03/10 5.8s}
18. Bxc5 {+0.48/11 5.3s} B@b5 {+2.98/8 6.7s} 19. Bxc7 {+1.73/10 4.3s}
Nh4 {+0.53/9 14s} 20. O-O {+2.56/10 6.4s} N@a6 {+0.44/9 5.2s}
21. B@c2 {+3.06/10 5.3s} Nxc7 {-1.16/10 3.4s} 22. Bxa4 {+4.90/11 7.5s}
Bxa4 {-1.61/10 5.3s} 23. R@e7+ {+5.61/11 9.3s} Qxe7 {-1.83/10 12s}
24. Bxe7 {+4.81/11 25s} Nf5 {-1.40/7 3.2s} 25. Qxb7 {+5.57/8 1.2s}
R@c8 {-2.85/7 4.3s} 26. P@d6 {+6.03/8 2.6s} B@c6 {-2.81/8 4.8s}
27. Qb4 {+5.98/9 3.9s} Na6 {-3.38/9 5.4s} 28. Qa3 {+6.08/9 3.1s}
Rab8 {-2.44/8 6.8s} 29. Q@a5 {+5.99/9 3.6s} Bab5 {-4.03/9 3.8s}
30. Qxa6 {+7.37/10 5.6s} Bxa6 {-4.99/9 6.7s} 31. N@c7+ {+7.21/10 4.0s}
Kd7 {-7.04/10 7.3s} 32. Nxa6 {+7.05/10 4.3s} Rb5 {-7.05/8 4.3s}
33. Nc5+ {+7.68/10 11s} Ke8 {-8.33/10 26s}
34. Qxa7 {+8.91/10 10s, Black disconnects} 1-0

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "4"]
[White "CrazyWa 0.0.9 32bit"]
[Black "Imortal 2.9001 32bit"]
[Result "0-1"]
[PlyCount "34"]
[Termination "abandoned"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} c5 {book} 2. Nf3 {book} d6 {book} 3. Bc4 {book} e6 {book}
4. Nc3 {book} Nc6 {-0.42/13 6.3s} 5. d4 {+1.32/10 3.9s} cxd4 {-0.58/11 9.1s}
6. Nxd4 {+1.37/9 5.5s} Bd7 {-0.53/11 6.2s} 7. Ndb5 {+1.58/9 6.1s}
P@c7 {-0.62/11 6.2s} 8. Bf4 {+1.36/8 4.4s} a6 {-0.50/11 8.3s}
9. Na3 {+1.23/9 6.1s} Qf6 {-0.02/11 8.0s} 10. Qd2 {+1.63/8 3.9s}
Qg6 {-0.27/12 6.0s} 11. O-O {+1.47/8 3.7s} Nf6 {+0.06/11 5.5s}
12. Rfe1 {+1.63/9 4.9s} Ng4 {+0.17/11 6.9s} 13. P@f5 {+2.26/9 3.8s}
Qh5 {-1.95/12 7.1s} 14. h3 {+3.98/10 9.5s} Be7 {-3.52/10 5.5s}
15. hxg4 {+4.67/9 6.3s} Qxg4 {-4.12/12 6.6s} 16. Be2 {+5.48/9 9.1s}
Qh4 {-4.54/12 6.4s} 17. N@f3 {+5.72/9 3.8s}
Qg4 {-3.97/12 6.2s, White disconnects} 0-1
Top
Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Windows dump description

Post by Ferdy »

Running Blitz tour, I got this in the engine's dir, dump.txt
Windows dump.

Image
Top
Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

CCVA RR Blitz Tour Nr 26

Post by Ferdy »

Tour Nr 26 finished, single RR, 32 games per pairing.

Code: Select all

Summary:

   # PLAYER                          :  RATING   +/-   GAMES  PTS(%)
   1 Stockfish zh 02Nov2016 64bit    :  2992.3  79.9     224    92.0
   2 Imortal 2.9001 32bit            :  2781.5  59.2     224    76.6
   3 TJchess 1.3 64bit               :  2663.1  54.1     224    65.6
   4 Sunsetter 9 32bit               :  2638.1  54.9     224    63.2
   5 CrazyWa 0.0.9 32bit             :  2347.7  53.6     224    33.0
   6 Sjeng 11.2 32bit                :  2300.0  57.7     224    28.1
   7 NebiyuAlien 1.45a 32bit         :  2243.5  61.1     224    22.5
   8 Sjaak II 1.3.1a 64bit           :  2205.3  61.0     224    19.0

White advantage = 50.84 +/- 16.08
Draw rate (equal opponents) = 1.02 % +/- 0.41

Sjeng is set at 2300

Code: Select all

Head to head statistics:

 1) Stockfish zh 02Nov2016 64bit         2992.3 :    224 (+205,=2,-17),  92.0 %

    vs.                                         :  games (   +, =,  -),   (%) :    Diff,    SD, CFS (%)
    Imortal 2.9001 32bit                        :     32 (  24, 1,  7),  76.6 :  +210.8,  50.8,  100.0
    TJchess 1.3 64bit                           :     32 (  28, 0,  4),  87.5 :  +329.2,  49.9,  100.0
    Sunsetter 9 32bit                           :     32 (  26, 1,  5),  82.8 :  +354.2,  51.5,  100.0
    CrazyWa 0.0.9 32bit                         :     32 (  32, 0,  0), 100.0 :  +644.6,  54.7,  100.0
    Sjeng 11.2 32bit                            :     32 (  32, 0,  0), 100.0 :  +692.3,  57.5,  100.0
    NebiyuAlien 1.45a 32bit                     :     32 (  32, 0,  0), 100.0 :  +748.8,  58.6,  100.0
    Sjaak II 1.3.1a 64bit                       :     32 (  31, 0,  1),  96.9 :  +787.0,  59.3,  100.0

 2) Imortal 2.9001 32bit                 2781.5 :    224 (+170,=3,-51),  76.6 %

    vs.                                         :  games (   +, =,  -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (   7, 1, 24),  23.4 :  -210.8,  50.8,    0.0
    TJchess 1.3 64bit                           :     32 (  21, 0, 11),  65.6 :  +118.4,  41.2,   99.8
    Sunsetter 9 32bit                           :     32 (  24, 0,  8),  75.0 :  +143.4,  40.5,  100.0
    CrazyWa 0.0.9 32bit                         :     32 (  29, 0,  3),  90.6 :  +433.8,  45.9,  100.0
    Sjeng 11.2 32bit                            :     32 (  28, 1,  3),  89.1 :  +481.5,  46.6,  100.0
    NebiyuAlien 1.45a 32bit                     :     32 (  31, 1,  0),  98.4 :  +538.0,  48.7,  100.0
    Sjaak II 1.3.1a 64bit                       :     32 (  30, 0,  2),  93.8 :  +576.2,  49.0,  100.0

 3) TJchess 1.3 64bit                    2663.1 :    224 (+147,=0,-77),  65.6 %

    vs.                                         :  games (   +, =,  -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (   4, 0, 28),  12.5 :  -329.2,  49.9,    0.0
    Imortal 2.9001 32bit                        :     32 (  11, 0, 21),  34.4 :  -118.4,  41.2,    0.2
    Sunsetter 9 32bit                           :     32 (  17, 0, 15),  53.1 :   +25.0,  40.3,   73.3
    CrazyWa 0.0.9 32bit                         :     32 (  26, 0,  6),  81.3 :  +315.4,  41.9,  100.0
    Sjeng 11.2 32bit                            :     32 (  28, 0,  4),  87.5 :  +363.1,  44.0,  100.0
    NebiyuAlien 1.45a 32bit                     :     32 (  30, 0,  2),  93.8 :  +419.6,  46.6,  100.0
    Sjaak II 1.3.1a 64bit                       :     32 (  31, 0,  1),  96.9 :  +457.8,  46.0,  100.0

 4) Sunsetter 9 32bit                    2638.1 :    224 (+140,=3,-81),  63.2 %

    vs.                                         :  games (   +, =,  -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (   5, 1, 26),  17.2 :  -354.2,  51.5,    0.0
    Imortal 2.9001 32bit                        :     32 (   8, 0, 24),  25.0 :  -143.4,  40.5,    0.0
    TJchess 1.3 64bit                           :     32 (  15, 0, 17),  46.9 :   -25.0,  40.3,   26.7
    CrazyWa 0.0.9 32bit                         :     32 (  26, 0,  6),  81.3 :  +290.4,  41.4,  100.0
    Sjeng 11.2 32bit                            :     32 (  27, 1,  4),  85.9 :  +338.1,  44.3,  100.0
    NebiyuAlien 1.45a 32bit                     :     32 (  31, 0,  1),  96.9 :  +394.5,  47.1,  100.0
    Sjaak II 1.3.1a 64bit                       :     32 (  28, 1,  3),  89.1 :  +432.8,  46.2,  100.0

 5) CrazyWa 0.0.9 32bit                  2347.7 :    224 (+74,=0,-150),  33.0 %

    vs.                                         :  games (  +, =,   -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (  0, 0,  32),   0.0 :  -644.6,  54.7,    0.0
    Imortal 2.9001 32bit                        :     32 (  3, 0,  29),   9.4 :  -433.8,  45.9,    0.0
    TJchess 1.3 64bit                           :     32 (  6, 0,  26),  18.8 :  -315.4,  41.9,    0.0
    Sunsetter 9 32bit                           :     32 (  6, 0,  26),  18.8 :  -290.4,  41.4,    0.0
    Sjeng 11.2 32bit                            :     32 ( 18, 0,  14),  56.3 :   +47.7,  40.8,   87.9
    NebiyuAlien 1.45a 32bit                     :     32 ( 19, 0,  13),  59.4 :  +104.2,  41.2,   99.4
    Sjaak II 1.3.1a 64bit                       :     32 ( 22, 0,  10),  68.8 :  +142.4,  41.6,  100.0

 6) Sjeng 11.2 32bit                     2300.0 :    224 (+62,=2,-160),  28.1 %

    vs.                                         :  games (  +, =,   -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (  0, 0,  32),   0.0 :  -692.3,  57.5,    0.0
    Imortal 2.9001 32bit                        :     32 (  3, 1,  28),  10.9 :  -481.5,  46.6,    0.0
    TJchess 1.3 64bit                           :     32 (  4, 0,  28),  12.5 :  -363.1,  44.0,    0.0
    Sunsetter 9 32bit                           :     32 (  4, 1,  27),  14.1 :  -338.1,  44.3,    0.0
    CrazyWa 0.0.9 32bit                         :     32 ( 14, 0,  18),  43.8 :   -47.7,  40.8,   12.1
    NebiyuAlien 1.45a 32bit                     :     32 ( 19, 0,  13),  59.4 :   +56.5,  42.2,   91.0
    Sjaak II 1.3.1a 64bit                       :     32 ( 18, 0,  14),  56.3 :   +94.7,  42.1,   98.8

 7) NebiyuAlien 1.45a 32bit              2243.5 :    224 (+50,=1,-173),  22.5 %

    vs.                                         :  games (  +, =,   -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (  0, 0,  32),   0.0 :  -748.8,  58.6,    0.0
    Imortal 2.9001 32bit                        :     32 (  0, 1,  31),   1.6 :  -538.0,  48.7,    0.0
    TJchess 1.3 64bit                           :     32 (  2, 0,  30),   6.3 :  -419.6,  46.6,    0.0
    Sunsetter 9 32bit                           :     32 (  1, 0,  31),   3.1 :  -394.5,  47.1,    0.0
    CrazyWa 0.0.9 32bit                         :     32 ( 13, 0,  19),  40.6 :  -104.2,  41.2,    0.6
    Sjeng 11.2 32bit                            :     32 ( 13, 0,  19),  40.6 :   -56.5,  42.2,    9.0
    Sjaak II 1.3.1a 64bit                       :     32 ( 21, 0,  11),  65.6 :   +38.3,  42.1,   81.8

 8) Sjaak II 1.3.1a 64bit                2205.3 :    224 (+42,=1,-181),  19.0 %

    vs.                                         :  games (  +, =,   -),   (%) :    Diff,    SD, CFS (%)
    Stockfish zh 02Nov2016 64bit                :     32 (  1, 0,  31),   3.1 :  -787.0,  59.3,    0.0
    Imortal 2.9001 32bit                        :     32 (  2, 0,  30),   6.3 :  -576.2,  49.0,    0.0
    TJchess 1.3 64bit                           :     32 (  1, 0,  31),   3.1 :  -457.8,  46.0,    0.0
    Sunsetter 9 32bit                           :     32 (  3, 1,  28),  10.9 :  -432.8,  46.2,    0.0
    CrazyWa 0.0.9 32bit                         :     32 ( 10, 0,  22),  31.3 :  -142.4,  41.6,    0.0
    Sjeng 11.2 32bit                            :     32 ( 14, 0,  18),  43.8 :   -94.7,  42.1,    1.2
    NebiyuAlien 1.45a 32bit                     :     32 ( 11, 0,  21),  34.4 :   -38.3,  42.1,   18.2

Code: Select all

Average depth and time comparison:

 nr                       player  games  pts%   sumTime aveDep    aveTime
  1 Stockfish zh 02Nov2016 64bit    224  92.0  12:03:38  20.11  00:06:336
  2            Sunsetter 9 32bit    224  63.2  09:43:25  15.39  00:05:469
  3      NebiyuAlien 1.45a 32bit    224  22.5  09:36:54  14.96  00:06:538
  4          CrazyWa 0.0.9 32bit    224  33.0  11:25:42  10.47  00:05:675
  5         Imortal 2.9001 32bit    224  76.6  09:59:22   9.81  00:05:428
  6        Sjaak II 1.3.1a 64bit    224  19.0  09:14:58   9.72  00:05:317
  7            TJchess 1.3 64bit    224  65.6  11:31:32   9.23  00:05:762
  8             Sjeng 11.2 32bit    224  28.1  10:59:54   7.23  00:05:957

file         : ccva_tour_nr_26.pgn
event        : CCVA RR Blitz Tour Nr 26
score window : [-1000000.0, +1000000.0]
notes        : 1. games and pts% are not affected by score window.
               2. Table is sorted by aveDep in descending order.
               3. aveTime is the average time/move in m:s:ms

Code: Select all

Win probability comparison when score is winning [+3.0, +6.0]:

file         : ccva_tour_nr_26.pgn
event        : CCVA RR Blitz Tour Nr 26
score window : [+3.0, +6.0]

 nr                       player games pts(%) winProb
  1         Imortal 2.9001 32bit   224  76.56    0.92
  2            Sunsetter 9 32bit   224  63.17    0.92
  3 Stockfish zh 02Nov2016 64bit   224  91.96    0.92
  4            TJchess 1.3 64bit   224  65.62    0.84
  5             Sjeng 11.2 32bit   224  28.12    0.71
  6          CrazyWa 0.0.9 32bit   224  33.04    0.50
  7      NebiyuAlien 1.45a 32bit   224  22.54    0.44
  8        Sjaak II 1.3.1a 64bit   224  18.97    0.25
Time forfeits:

Code: Select all

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "2"]
[White "TJchess 1.3 64bit"]
[Black "Stockfish zh 02Nov2016 64bit"]
[Result "1-0"]
[PlyCount "91"]
[Termination "time forfeit"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} e5 {book} 2. Nf3 {book} Nc6 {book} 3. Bc4 {book} Bc5 {book}
4. c3 {book} Nf6 {book} 5. d3 {book} O-O {book} 6. O-O {book} d6 {book}
7. b4 {+0.21/11 2.4s} Nxb4 {+0.36/18 20s} 8. cxb4 {+0.37/11 5.8s}
Bxb4 {+0.46/18 6.8s} 9. Ng5 {+0.34/10 5.0s} P@e6 {+0.80/16 5.5s}
10. a3 {+0.32/11 11s} Ba5 {+0.96/16 4.5s} 11. Bd2 {+0.23/10 4.2s}
Bxd2 {+1.78/16 3.8s} 12. Nxd2 {+0.25/11 10.0s} d5 {+1.60/18 35s}
13. B@b4 {-0.14/10 7.0s} B@d6 {+1.81/15 3.0s} 14. exd5 {+0.16/10 6.0s}
exd5 {+2.43/15 4.0s} 15. Bxd6 {-0.22/11 16s} cxd6 {+1.78/18 32s}
16. Bxd5 {-0.22/11 11s} Nxd5 {+4.61/14 4.1s} 17. Nxh7 {-0.21/11 16s}
Kxh7 {+5.42/15 3.7s} 18. B@e4+ {-0.48/10 16s} B@g6 {+5.71/15 2.3s}
19. Bxd5 {-1.05/10 14s} N@f4 {+5.70/16 14s} 20. Qb3 {-0.76/9 5.6s}
P@b2 {+5.85/15 2.5s} 21. Rab1 {-1.62/9 2.6s} Bxd3 {+4.44/16 14s}
22. Bxb7 {-0.99/9 12s} P@e2 {+6.43/13 3.9s} 23. Bxa8 {-1.12/8 3.0s}
exf1=R+ {+7.83/13 5.5s} 24. Rxf1 {-2.29/9 9.7s} P@e2 {+6.42/14 9.8s}
25. R@h5+ {-1.85/8 7.2s} Kg6 {+7.02/14 3.1s} 26. Qxd3+ {-2.83/8 7.5s}
Nxd3 {+6.23/13 0.94s} 27. N@g3 {-2.15/8 8.2s} exf1=Q+ {+3.60/15 23s}
28. Ndxf1 {-1.71/10 19s} R@h1+ {+2.49/16 20s} 29. Nxh1 {-0.09/8 4.3s}
Kxh5 {+2.74/11 2.8s} 30. B@e2+ {+1.29/7 14s} Kh6 {-0.09/15 11s}
31. R@h5+ {+1.45/8 3.7s} Kg6 {-0.09/1 0s} 32. Bxd3+ {+3.08/9 16s}
f5 {-1.87/14 6.9s} 33. N@c6 {+2.91/8 2.3s} Qe8 {-1.42/12 0.91s}
34. P@e7 {+4.02/7 1.1s} Rg8 {-1.18/13 4.4s} 35. N@c7 {+4.68/7 1.2s}
Qf7 {-1.90/13 2.3s} 36. P@h7 {+5.35/9 4.0s} Rh8 {-7.93/11 2.0s}
37. P@b7 {+5.31/8 1.0s} Kxh5 {-4.18/11 2.0s} 38. bxc8=Q {+8.85/8 2.0s}
Rxc8 {-7.45/13 2.0s} 39. P@e6 {+9.26/7 1.4s} Qg6 {-4.15/11 0.87s}
40. B@f7 {+11.00/8 3.3s} Q@f6 {-7.53/10 3.1s} 41. e8=Q {+13.17/7 1.3s}
Rxe8 {-10.45/11 2.0s} 42. Nxe8 {+13.91/7 1.3s} Qfxf7 {-11.12/10 2.0s}
43. Nhg3+ {+14.83/7 1.7s} Kg5 {-10.81/10 2.0s} 44. P@h4+ {+17.51/8 2.7s}
Kxh4 {-12.57/11 2.0s} 45. exf7 {+17.84/8 3.4s} B@h3 {-15.65/10 2.0s}
46. Nxf5+ {+M9/6 0.94s, Black loses on time} 1-0

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "6"]
[White "Stockfish zh 02Nov2016 64bit"]
[Black "Sunsetter 9 32bit"]
[Result "0-1"]
[PlyCount "92"]
[Termination "time forfeit"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} e6 {book} 2. d4 {book} Nc6 {book} 3. Nf3 {book} Nf6 {book}
4. e5 {book} Nd5 {-0.49/21 7.5s} 5. c4 {+0.72/17 15s} Bb4+ {-0.50/20 7.2s}
6. Bd2 {+0.50/18 19s} Bxd2+ {-0.54/18 7.0s} 7. Nbxd2 {+0.95/15 2.4s}
Nde7 {-0.59/19 6.8s} 8. B@e4 {+1.06/17 22s} d5 {-0.25/20 6.6s}
9. exd6 {+1.21/15 3.0s} Qxd6 {-0.23/18 6.5s} 10. Bxc6+ {+1.55/15 1.7s}
Nxc6 {-0.36/18 6.3s} 11. N@e4 {+1.19/17 13s} Qd8 {+0.09/18 6.1s}
12. d5 {+1.72/18 21s} exd5 {+0.78/18 5.9s} 13. cxd5 {+1.58/16 0.70s}
Qxd5 {+0.75/16 5.8s} 14. P@f6 {+1.74/18 6.8s} B@f8 {+0.74/15 5.6s}
15. Bc4 {+1.22/17 7.4s} Qf5 {+0.45/15 5.5s} 16. Qb3 {+1.51/16 11s}
P@e6 {+0.18/16 5.3s} 17. P@d5 {+2.30/14 2.5s} B@b4 {+0.78/16 5.2s}
18. O-O {+2.70/14 1.3s} P@a4 {+0.06/17 7.6s} 19. Qxa4 {+3.02/18 12s}
exd5 {+0.06/18 11s} 20. Nd4 {+3.02/16 0.82s} dxc4 {-0.42/16 10s}
21. Nxf5 {+2.54/17 8.9s} Bxf5 {-0.60/14 4.2s} 22. fxg7 {+2.48/18 4.3s}
Bxg7 {+0.90/13 4.1s} 23. P@f6 {+2.64/17 3.1s} Bgf8 {+0.02/13 4.0s}
24. Q@e5+ {+3.48/15 5.5s} B@e6 {-0.36/15 3.9s} 25. Qxc7 {+4.14/15 4.3s}
Bxe4 {-0.48/14 3.9s} 26. Nxe4 {+4.42/15 2.7s} Rc8 {-1.43/13 13s}
27. Qxb7 {+5.64/15 6.8s} N@a5 {-1.91/13 5.0s} 28. Qxc8+ {+6.66/15 7.3s}
Bxc8 {-1.67/13 3.2s} 29. R@e5+ {+5.46/14 12s} Be6 {-2.19/15 7.1s}
30. Rxa5 {+8.10/12 2.6s} Q@d7 {-3.41/14 10.0s} 31. N@g7+ {+9.46/14 13s}
Kd8 {-3.69/14 6.0s} 32. Nxe6+ {+10.00/14 3.5s} fxe6 {-3.75/14 2.5s}
33. P@g7 {+10.83/13 2.1s} Bxg7 {-3.74/13 3.7s} 34. fxg7 {+10.42/13 0.99s}
N@e2+ {-4.37/12 2.4s} 35. Kh1 {+10.42/1 0s} Re8 {-5.77/14 8.0s}
36. B@g5+ {+14.98/12 3.1s} Kc7 {-6.15/13 7.2s} 37. Rxa7+ {+17.36/13 20s}
P@b7 {-9.09/13 4.3s} 38. Nf6 {+19.19/12 2.5s} N@b6 {-11.62/12 6.2s}
39. Nxd7 {+21.45/12 1.3s} Nxa4 {-9.46/12 3.8s} 40. Rxa4 {+22.06/13 2.5s}
Q@c8 {-11.45/12 5.4s} 41. B@b6+ {+M15/12 3.2s} Kxd7 {-15.23/12 1.5s}
42. N@f6+ {+M13/16 10s} Kd6 {-M7/11 2.5s} 43. Q@c7+ {+M11/17 2.4s} Qxc7
44. Bxc7+ {+M9/17 0.84s} Kxc7 {-M5/8 0.11s} 45. Nxe8+ {+M7/17 2.6s}
Kb6 {-M3/6 0.006s} 46. B@c7+ {+M5/17 4.5s}
Kb5 {-M1/4 0.013s, White loses on time} 0-1
Abandoned games:

Code: Select all

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "1"]
[White "TJchess 1.3 64bit"]
[Black "CrazyWa 0.0.9 32bit"]
[Result "1-0"]
[PlyCount "67"]
[Termination "abandoned"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. d4 {book} d5 {book} 2. Bf4 {book} Bf5 {book} 3. e3 {book} e6 {book}
4. Nd2 {book} Nd7 {-0.29/11 8.4s} 5. Bd3 {+0.29/12 4.0s} Bxd3 {+0.18/12 5.0s}
6. cxd3 {+0.35/12 5.6s} Bb4 {+0.04/10 6.1s} 7. Rc1 {+0.16/10 3.4s}
B@a5 {+0.44/10 6.9s} 8. B@c3 {+0.33/12 3.6s} Bxc3 {+1.93/12 3.3s}
9. bxc3 {+0.33/12 6.8s} B@b2 {+1.74/11 3.9s} 10. Ne2 {+0.69/11 5.8s}
Bxc1 {+1.59/11 7.4s} 11. Qxc1 {+0.82/11 5.6s} Ne7 {+1.66/9 7.1s}
12. Qb2 {+0.85/11 7.5s} Ng6 {+1.82/8 4.0s} 13. B@a4 {+0.87/10 4.8s}
Bb6 {+1.79/8 5.0s} 14. B@a3 {+1.06/10 4.1s} R@a5 {+2.02/9 8.7s}
15. Bxd7+ {+0.57/11 7.0s} Qxd7 {+2.45/10 8.5s} 16. Bb4 {+0.58/11 9.1s}
Ra4 {+1.98/9 6.6s} 17. N@c5 {+0.68/10 6.8s} Bxc5 {+3.03/10 5.8s}
18. Bxc5 {+0.48/11 5.3s} B@b5 {+2.98/8 6.7s} 19. Bxc7 {+1.73/10 4.3s}
Nh4 {+0.53/9 14s} 20. O-O {+2.56/10 6.4s} N@a6 {+0.44/9 5.2s}
21. B@c2 {+3.06/10 5.3s} Nxc7 {-1.16/10 3.4s} 22. Bxa4 {+4.90/11 7.5s}
Bxa4 {-1.61/10 5.3s} 23. R@e7+ {+5.61/11 9.3s} Qxe7 {-1.83/10 12s}
24. Bxe7 {+4.81/11 25s} Nf5 {-1.40/7 3.2s} 25. Qxb7 {+5.57/8 1.2s}
R@c8 {-2.85/7 4.3s} 26. P@d6 {+6.03/8 2.6s} B@c6 {-2.81/8 4.8s}
27. Qb4 {+5.98/9 3.9s} Na6 {-3.38/9 5.4s} 28. Qa3 {+6.08/9 3.1s}
Rab8 {-2.44/8 6.8s} 29. Q@a5 {+5.99/9 3.6s} Bab5 {-4.03/9 3.8s}
30. Qxa6 {+7.37/10 5.6s} Bxa6 {-4.99/9 6.7s} 31. N@c7+ {+7.21/10 4.0s}
Kd7 {-7.04/10 7.3s} 32. Nxa6 {+7.05/10 4.3s} Rb5 {-7.05/8 4.3s}
33. Nc5+ {+7.68/10 11s} Ke8 {-8.33/10 26s}
34. Qxa7 {+8.91/10 10s, Black disconnects} 1-0

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "2"]
[White "Sunsetter 9 32bit"]
[Black "Sjaak II 1.3.1a 64bit"]
[Result "1-0"]
[PlyCount "6"]
[Termination "abandoned"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} e5 {book} 2. Nf3 {book} Nc6 {book} 3. Bc4 {book}
Bc5 {book, Black disconnects} 1-0

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "4"]
[White "CrazyWa 0.0.9 32bit"]
[Black "Imortal 2.9001 32bit"]
[Result "0-1"]
[PlyCount "34"]
[Termination "abandoned"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} c5 {book} 2. Nf3 {book} d6 {book} 3. Bc4 {book} e6 {book}
4. Nc3 {book} Nc6 {-0.42/13 6.3s} 5. d4 {+1.32/10 3.9s} cxd4 {-0.58/11 9.1s}
6. Nxd4 {+1.37/9 5.5s} Bd7 {-0.53/11 6.2s} 7. Ndb5 {+1.58/9 6.1s}
P@c7 {-0.62/11 6.2s} 8. Bf4 {+1.36/8 4.4s} a6 {-0.50/11 8.3s}
9. Na3 {+1.23/9 6.1s} Qf6 {-0.02/11 8.0s} 10. Qd2 {+1.63/8 3.9s}
Qg6 {-0.27/12 6.0s} 11. O-O {+1.47/8 3.7s} Nf6 {+0.06/11 5.5s}
12. Rfe1 {+1.63/9 4.9s} Ng4 {+0.17/11 6.9s} 13. P@f5 {+2.26/9 3.8s}
Qh5 {-1.95/12 7.1s} 14. h3 {+3.98/10 9.5s} Be7 {-3.52/10 5.5s}
15. hxg4 {+4.67/9 6.3s} Qxg4 {-4.12/12 6.6s} 16. Be2 {+5.48/9 9.1s}
Qh4 {-4.54/12 6.4s} 17. N@f3 {+5.72/9 3.8s}
Qg4 {-3.97/12 6.2s, White disconnects} 0-1

[Event "CCVA RR Blitz Tour Nr 26"]
[Site "i7-2600K"]
[Date "2016.11.04"]
[Round "5"]
[White "Sunsetter 9 32bit"]
[Black "CrazyWa 0.0.9 32bit"]
[Result "1-0"]
[PlyCount "51"]
[Termination "abandoned"]
[TimeControl "180+2"]
[Variant "crazyhouse"]

1. e4 {book} e6 {book} 2. Nc3 {book} d5 {book} 3. exd5 {book} exd5 {book}
4. Nf3 {book} a6 {-0.31/9 9.0s} 5. d4 {+0.59/19 7.2s} Bb4 {+0.29/9 4.3s}
6. Bg5 {+0.57/17 7.3s} Ne7 {+0.13/8 5.2s} 7. P@f6 {+0.33/17 11s}
gxf6 {+0.04/10 4.2s} 8. Bxf6 {+0.13/18 6.8s} Rg8 {+0.30/9 5.7s}
9. Rb1 {+0.09/17 9.8s} Rg6 {+0.95/9 7.8s} 10. Be5 {-0.02/17 6.3s}
Bg4 {+1.27/9 4.8s} 11. Be2 {+0.05/16 6.1s} P@a3 {+1.44/8 4.9s}
12. P@d2 {+0.13/19 8.9s} Bxc3 {+1.37/9 3.4s} 13. dxc3 {+0.06/18 5.7s}
N@e4 {+1.61/9 7.8s} 14. B@d2 {+0.06/17 5.5s} axb2 {+3.36/10 6.7s}
15. Rxb2 {+0.25/17 5.4s} Bxf3 {+3.91/10 7.5s} 16. Bxf3 {+0.05/18 12s}
Nxd2 {+3.96/8 3.3s} 17. Qxd2 {-0.51/15 7.3s} N@c4 {+4.63/8 6.3s}
18. N@g7+ {-0.83/14 4.6s} Kf8 {+4.25/10 9.4s} 19. B@h6 {-1.31/15 4.5s}
Nxd2 {+4.76/9 8.9s} 20. Ne6+ {-0.28/16 4.4s} Ke8 {+5.39/9 8.6s}
21. Nxd8 {-0.36/15 4.3s} Q@c1+ {+5.39/7 7.6s} 22. Bd1 {-0.43/15 4.3s}
Rxh6 {+4.82/7 3.7s} 23. P@e6 {-0.71/13 4.1s} fxe6 {+5.49/7 4.8s}
24. Rxb7 {+0.05/13 6.1s} B@h5 {+1.58/7 7.7s} 25. P@f7+ {+1.03/15 5.8s}
Bxf7 {+1.93/8 3.3s} 26. Rxc7 {+1.72/14 3.7s, Black disconnects} 1-0
Top
Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Blitz Rating 2016-11-04

Post by Ferdy »

Blitz rating all versions:

Code: Select all

   # PLAYER                          :  RATING   +/-   GAMES  PTS(%)
   1 Stockfish zh 02Nov2016 64bit    :  3052.0  87.9     224      92
   2 Imortal 2.9001 32bit            :  2837.7  32.9    1339      76
   3 TJchess 1.3 64bit               :  2765.7  31.6    1369      66
   4 Sunsetter 9 32bit               :  2655.6  32.7    1179      49
   5 Imortal 2.0 32bit               :  2650.8  48.3     311      66
   6 TJchess 1.1 64bit               :  2573.3  65.1     168      55
   7 Sunsetter 7g 64bit              :  2530.0  47.2     296      52
   8 Imortal 1.0 32bit               :  2508.1  80.5     110      41
   9 CrazyWa 0.0.8 32bit             :  2423.7  61.1     192      51
  10 CrazyWa 0.0.9 32bit             :  2388.3  59.4     224      33
  11 CrazyWa 0.0.7 32bit             :  2315.3  60.1     216      49
  12 Sjeng 11.2 32bit                :  2300.0  25.9    1128      45
  13 NebiyuAlien 1.45a 32bit         :  2284.2  25.3    1090      40
  14 Sjaak II 1.3.1a 64bit           :  2267.8  25.8    1195      40
  15 CrazyWa 0.0.6 32bit             :  2262.4  55.1     192      63
  16 CrazyWa 0.0.4 32bit             :  2237.2  43.1     336      42
  17 CrazyWa 0.0.1 32bit             :  2216.8  66.6     129      56
  18 Pulsar 2009 9b 32bit            :  2111.1  37.2     523      37
  19 NebiyuAlien 1.45 32bit          :  2060.0  55.4     204      37
  20 CrazyWa 0.0.0 32bit             :  2002.9  82.3      96      32
  21 TSCP zh 1.1 32bit               :  1996.6  38.9     503      25
  22 KKFChess 2.6.7 beta 32bit       :  1962.6  41.9     524      22

White advantage = 59.65 +/- 5.75
Draw rate (equal opponents) = 1.31 % +/- 0.16
Using Ordo, Sjeng 11.2 32bit is set to 2300
Site for details:
https://sites.google.com/site/zhassocia ... list/blitz

Download games here.
https://sites.google.com/site/zhassociation/download
Top
Carlos777
Posts: 1730
Joined: Sun Dec 13, 2009 6:09 pm

Re: Blitz Rating 2016-11-04

Post by Carlos777 »

15'+3" on phenom II and equivalent on i5-2400

Code: Select all

Stockfish 021116                    :    198 (+170,=2,-26),  86.4 %

    vs.                             :  games (   +, =,  -),   (%) :     Diff
    Imortal 2.9001                  :     66 (  54, 0, 12),  81.8 :  +263.31
    Sunsetter 9                     :     66 (  57, 2,  7),  87.9 :  +376.07
    TJchess 1.3                     :     66 (  59, 0,  7),  89.4 :  +381.35
http://www.mediafire.com/file/w5uv6i9bd ... 0211gt.pgn
Top
User avatar
hgm
Posts: 27790
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: CCVA RR Blitz Tour Nr 26

Post by hgm »

I think I finally traced the cause of these illusive crashes. They never reproduced, but at some point I noticed that some of them occurred when their was an invalid piece on the board (which was printed when an assert triggered). For some reason undefined pieces all move like an Archbishop, but did not have their capture codes set. So they would dodge the in-check test, and then later capture the King during move generation. Which would abort move generation without proper termination of the move list, so that during search the loop over moves would go beyond its end, and encounter all kinds of garbage there, triggering the invalid-move assert. This was all understandable, and actually worked as designed.

The question was of course how the invalid pieces got on the board. The irreproducibility of the crashes (despite hash and history clearing before each move) was probably caused by a previous move already corrupting the board, so that when I set up the position (with an uncorrupted board) everything would work fine. By adding an assert for invalid pieces, I could catch the problem in the search where it occurred, which made the problem reproducible. So I could finally debug it.

Turned out it was all due to the killer moves: the routine for testing if killers were legal let castlings slip through, as it did not realize that only Kings can castle. Since I implemented castling in a bit of a tricky way, as a promotion move promoting a King to a Rook, this cause other pieces to which it was applied to promote to something else too, often an invalid piece.

After fixing that there were still crashes, not due to asserts but of the Windows type (where a popup informs you the application is malfunctioning). These turned out to be caused by an already promoted piece being subject to a promotion move. Promoting black promoted pieces led to piece codes so far out of range that looking up their value would cause a segfault. This cannot be due to killer verification, as promotions should not be able to become killers. (They are sorted with the captures.) So they must have originated as faulty hash moves from a key collision. I still must make the program resistent to that.
Top
User avatar
Guenther
Posts: 4605
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: for CrazyWa 0.09

Post by Guenther »

I think the opening and early middlegame of CrazyWa could be improved a lot, with potential to the top 4 in ZH.
(oh wait it is already in top 5 according to the blitzlist...)
Still it could win this game from -3.41 (own eval - TSCP even showed 7.15 at one point),
but until around move 15 or so it looked like a catastrophe.

I am not really a specialist for ZH, but I played some dozens of games a month ago at Lichess for fun and read a tiny bit about its general theory recently.

Yet I think 4. Na3?? (not only because it does nothing for center and attack, but also because it weakens one of the spots we are told to
defend carefully: b2/b7 g2/g7) is quite bad and retreating a developed piece with 5. Nd2? gives already big advantage to Black.
Note that 7. Qc1 already tries to defend the hole at b2 and a few more careless moves later White normally was burned down until TSCP went out of steam.

Nice tactics at the end though - it seems in attacking it is much better than in developing and defending.

This is just meant as positive criticism of course and CrazyWa is just week(s?) old and it already gained a lot until now.
I am quite sure that HGM will pull out more rabbits from his hat ;-)

Code: Select all

[Event "RWBC"]
[Site "CAPPUCCINO"]
[Date "2016.11.08"]
[Round "-"]
[White "CrazyWa 0.0.9"]
[Black "TSCP zh 1.1"]
[Result "1-0"]
[TimeControl "40/240"]
[Variant "crazyhouse"]
[Annotator "1. +0.15   1... -0.25"]

1. Nf3 {+0.15/11} d5 {-0.25/8 8} 2. d3 {+0.25/11 6} Nf6 {+0.18/7 8} 3. Bg5
{+0.27/10 4} Nc6 {+0.27/7 7} 4. Na3 {+0.20/9 4} e5 {+0.54/6 7} 5. Nd2
{-0.37/10 15} Bxa3 {+1.35/7 7} 6. bxa3 {-0.01/10 5} O-O {+1.25/6 7} 7. Qc1
{-0.12/10 8} N@e6 {+1.35/6 6} 8. Bxf6 {-0.01/11 8} Qxf6 {+1.35/5 6} 9. e3
{+0.04/9 6} Bd7 {+1.38/5 6} 10. B@f3 {+0.15/9 8} e4 {+3.09/6 6} 11. dxe4
{+0.69/10 4} B@b2 {+3.44/6 6} 12. exd5 {-0.01/10 9} Bxc1 {+6.44/6 5} 13.
Rxc1 {+0.26/9 9} Q@b2 {+6.40/5 5} 14. N@e2 {+0.49/7 4} Qxa3 {+7.15/5 5} 15.
dxc6 {-0.36/8 4} Bxc6 {+6.56/5 5} 16. Bxc6 {-1.95/8 17} P@c3 {+5.85/5 5}
17. Rd1 {-3.41/7 22} cxd2+ {+9.14/5 5} 18. Rxd2 {-3.10/7 4} bxc6
{+9.14/4 4} 19. B@c3 {-3.06/7 4} Qh4 {+8.38/4 4} 20. N@f5 {-2.75/7 6} Qe4
{+8.38/5 4} 21. B@d3 {-1.88/8 6} Qd5 {+7.64/5 4} 22. P@b2 {+0.16/8 3} Qac5
{+3.05/5 4} 23. P@d4 {+0.16/9 6} Qxf5 {-1.12/5 4} 24. dxc5 {+0.17/8 2.6}
Qxc5 {-1.17/5 4} 25. P@f5 {+0.92/8 6} Ng5 {-1.47/5 3} 26. h4 {+0.35/8 3}
N@f3+ {+0.40/4 3} 27. gxf3 {+2.41/9 6} Nxf3+ {-1.55/5 3} 28. Kd1
{+2.28/8 5} P@e5 {-1.55/4 3} 29. f6 {+6.15/7 2.7} gxf6 {-5.28/4 3} 30.
Bxh7+ {+11.83/6 4} Kxh7 {-5.35/5 3} 31. N@g5+ {+18.46/6 7} Nxg5
{-6.90/5 2.9} 32. hxg5+ {+18.71/5 4} P@h3 {-6.90/4 2.8} 33. Rxh3+
{+1000.04/10 2.4} P@h4 {-99.94/4 0.4} 34. Q@h6+ {+1000.03/87 0.8} Kg8
{-99.96/3} 35. P@h7+ {+1000.02/7 0.3} Kh8 {-99.98/2} 36. P@g7#
{+1000.01/2 0.3}
{Xboard adjudication: Checkmate} 1-0
BTW it would be wonderful if our pgn viewer at CCC could be extended for also showing ZH games?
Top
User avatar
hgm
Posts: 27790
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: for CrazyWa 0.09

Post by hgm »

Indeed, but as long as we have an unwilling TCadmin not much can be achieved.

As for CrazyWa: the first 5 moves are rather heavily randomized, by adding a random score of up to 31 cP to each move in the root. This will often produce quite poor moves in this phase. But I need it, to create enough game diversity in my tests, when there are only 2 or 3 opponents of suitable strength. I could start from opening lines, but that would not test its opening ability at all. Now at least the moves it thinks best should still be played more often, so that some of its opening abilities will leak through in the results.

I am happy I finally solved the problem of the crashes / asserts. In the last few thousand games I had no real crashes and only 3 assert exits. One where a King capture occurred after null move, which must have been a hash-key collision, as null move is not done when in check, but whether it it is an in-check position is recorded in the hash table. The other two were for a trial version that extends a check evasion that is an only-legal-move by 2 ply, which exceeded the maximum allowed depth of 80 ply. This is a risky practice, as it could go on 'forever', and that it only did so twice in 600 games is not bad at all. In the final version I will simply score the position as a draw when maximum depth is reached, but during development I wanted to know how often it happens, to judge if it would be a problem that could affect strength. Same for the King capture.
Top
User avatar
Guenther
Posts: 4605
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: Crazyhouse tournaments and rating list

Post by Guenther »

Ferdy wrote:
TonyJH wrote:Fyi, there is another engine that can play crazyhouse, called Feuerstein. The link used to be http://people.freenet.de/FritzSchach/ but it's a dead link now. I think it was open source (in Pascal).

Its handle on FICS is "Thukydides" with a crazyhouse rating of 2263.
I got Feuerstein 0.4.6.1 32bit from Ola. I apply inbetween to it because it does not say feature variants="crazyhouse", and Winboard 4.8.0b and Cute Chess do not like that. It looks fine so far now, but I need more tests.
Meanwhile I have set up a complete folder with ZH programs and an extra
Winboard also only for ZH (for my convenience and to have a separate ini file)

Now with the updated WB2UCI it seems I need a different Inbetween ini file for
Feuerstein? At least with the same setup as for SF-ZH UCI it says
crazyhouse is not supported for Feuerstein.
OTH if it really leaks memory as you told in another post I will probably let it out?
Top
User avatar
hgm
Posts: 27790
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: Crazyhouse tournaments and rating list

Post by hgm »

In principle the option

-firstFeatures {variants="crazyhouse"}

on the engine line should solve this, and make WinBoard believe that the engine sent feature variants="crazyhouse" after all feature commands it really sent.
Top

Return to “Computer Chess Club: General Topics”