Chess variant tournaments

[enhorning]

Looking forward for Tadyak's performance

It's been doing very well in my smaller test tournaments, and looks to be one of the favourites for coming out on the top!

[enhorning]

Okay, the tournament has started!

Single-core on i-970, 512 MB hash, no ponder, 40 moves in 20 minutes.

Second game saw Joker mount an impressive attack against Lime:

[Event "Computer Chess Game"]
[Site "ENHORNING"]
[Date "2012.12.16"]
[Round "2"]
[White "Joker KM"]
[Black "Lime KM 621_ja"]
[Result "1-0"]
[TimeControl "40/1200"]
[Variant "knightmate"]
[Annotator "1. +0.03 1... +0.06"]

1. d4 {+0.03/14} d5 {+0.06/11 19} 2. e3 {+0.02/14 23} Bd7 {+0.06/11 14} 3.
c4 {+0.09/14 10} c6 {+0.19/12 18} 4. Mc2 {+0.08/14 33} Mc7 {+0.14/11 12} 5.
f3 {+0.11/13 15} g6 {+0.03/11 17} 6. Mf2 {+0.09/13 30} Mg7 {+0.11/11 20} 7.
Mc3 {+0.00/13 12} Mf6 {+0.06/11 11} 8. Be2 {+0.00/14 57} Bg7 {+0.10/12 30}
9. e4 {+0.04/14 29} dxe4 {+0.09/12 11} 10. fxe4 {+0.02/14 28} e5
{+0.09/10 33} 11. Kg1 {+0.05/13 38} exd4 {+0.20/12 26} 12. Mxd4
{+0.07/14 44} Me5 {+0.24/11 20} 13. Mxe5 {+0.02/14 56} Bxe5 {+0.37/11 24}
14. Mg3 {+0.07/14 37} Qe7 {+0.53/10 12} 15. Qd3 {+0.11/13 25} Rd8
{+0.51/10 35} 16. Rf2 {+0.15/13 52} Bg4 {+0.63/10 15} 17. Qc2 {+0.01/13 24}
Bxe2 {+0.64/12 27} 18. Qxe2 {+0.14/14 32} Bxg3 {+0.66/12 24} 19. hxg3
{+0.00/14 21} Kg8 {+0.67/11 15} 20. e5 {+0.00/13 19} Rfe8 {+0.76/10 38} 21.
Bf4 {+0.08/12 17} Qc5 {+0.73/11 23} 22. Qf3 {+0.14/13 31} Rd7 {+0.77/11 16}
23. b3 {+0.14/13 22} a5 {+0.75/11 17} 24. Rf1 {+0.15/13 20} a4
{+0.70/10 15} 25. g4 {+0.41/13 31} axb3 {+0.60/11 37} 26. axb3
{+0.88/14 49} Md8 {+0.32/11 21} 27. g5 {+1.15/14 15} Qf8 {+0.04/11 45} 28.
Bh2 {+2.24/15 26} Qc5 {+0.77/11 46} 29. Qf6 {+2.98/16 46} Qd4 {+0.93/13 32}
30. e6 {+3.02/16 15} Qxf6 {-2.26/13 22} 31. Rxf6 {+8.96/18 14} fxe6
{-2.41/13 49} 32. Bd6 {+100.10/19 22} Me7 {-99.75/11 30} 33. Bxe7
{+100.09/17 0.1} Rd2 {-99.75/10 31} 34. Bd6 {+100.08/15 2.4} Rxg2+
{-12.37/11 55} 35. Kh3 {+100.07/13 0.1} Rxg5 {-99.80/11 56} 36. Kxg5
{+100.06/1 0.1} Rd8 {-99.88/12 40} 37. Rf7 {+100.05/9 0.1} Rd7
{-99.92/12 33} 38. Rxd7 {+100.04/7 0.1} Kh6 {-99.92/12 1:09} 39. Rxh7+
{+100.03/2 0.1} Kg4 {-99.94/11 48} 40. Rf3 {+100.02/3 0.1} e5
{-99.98/15 47} 41. Rh4# {+100.01/1 0.1}
{Xboard adjudication: Checkmate} 1-0

Position after black's move 27 - an impressive array of White's heavy pieces about to be unleashed along the F file:

[d]3krqn1/1p1r1p1p/2p3p1/4P1P1/2P2B2/1P3Q2/5RP1/5RN1 w - - 1 28
(Knights are royal)

2 games finished, 154 games left to go. I am leaving Nebiyu's (due to technical issues with that program) and Dabbaba's (due to possibly an improved version coming) games for last.

[JBNielsen]

Thanks for adding Dabbaba.

I hope to see a nightrider-tournament some day, by the way...

Dabbaba uses ca 20MB for hashtables (512.000 elements). I have no plans to increase that right now.

As Graham has just started a new tournament where Dabbaba participates, I am trying to get a slightly improved Dabbaba 6.80 ready to release.

Any idea when this new version might get released? If it's fairly soon, I could hold off on Dabbaba's games and play them last, to get the new version participating.

As far as Nightrider tournament - it is on my list of possibilities, but for now, fairly low down, as it is not a version I have personally played.

I've been playing a couple of smaller and quicker test tournaments, while getting used to Winboard, and trying to come to grips with various technical issues... going to start the real tournament tonight, I think.

I'm also leaning towards increasing the time controls - to 40 moves in 20 minutes (30 seconds per move average) instead of 40 moves in 400 seconds / 10 seconds per move average.

I will not release a new version of Dabbaba yet: http://talkchess.com/forum/viewtopic.ph ... 62&t=46405

best, jens

[JBNielsen]

As far as Nightrider tournament - it is on my list of possibilities, but for now, fairly low down, as it is not a version I have personally played.

Here is some inspiration for Nightrider chess!:

http://talkchess.com/forum/viewtopic.php?t=21930

best, jens

[enhorning]

Hmm, Winboard adjucated this as a false draw claim?

[Event "Computer Chess Game"]
[Site "ENHORNING"]
[Date "2012.12.17"]
[Round "8"]
[White "Faile KM 144_ja"]
[Black "Lime KM 621_ja"]
[Result "1-0"]
[TimeControl "40/1200"]
[Variant "knightmate"]
[Annotator "1. +0.08 1... -0.10"]

1. e4 {+0.08/11} c6 {-0.10/11 21} 2. d4 {+0.08/11 29} Mc7 {-0.06/11 14} 3.
Be2 {+0.08/10 29} d5 {-0.08/11 30} 4. exd5 {+0.19/11 29} cxd5 {-0.13/12 18}
5. f4 {+0.26/11 24} g6 {+0.02/12 12} 6. Mf2 {+0.14/10 24} Mg7 {+0.03/11 22}
7. Kg1 {+0.12/11 35} Mf6 {+0.25/11 16} 8. Be3 {+0.11/11 27} Bg7
{+0.24/12 25} 9. Re1 {+0.07/10 32} Md6 {+0.25/12 28} 10. b3 {+0.10/10 30}
Bf5 {+0.28/11 13} 11. c4 {+0.13/10 29} Kg8 {+0.29/10 32} 12. Mc2
{+0.13/10 29} Rc8 {+0.21/10 13} 13. Mf3 {+0.13/10 29} Bd7 {+0.21/10 26} 14.
Bd3 {+0.14/10 26} Rc7 {+0.26/10 14} 15. c5 {+0.19/12 27} Mc6 {+0.29/11 24}
16. b4 {+0.21/11 32} Qa8 {+0.25/11 32} 17. a4 {+0.29/11 20} Bf5
{+0.15/11 15} 18. b5 {+0.41/13 36} Bxd3 {-0.12/11 36} 19. Qxd3
{+0.35/12 31} Md7 {-0.24/13 36} 20. a5 {+0.38/11 28} Rb8 {-0.24/11 36} 21.
a6 {+0.46/11 23} b6 {-0.38/13 16} 22. c6 {+0.44/13 34} Md6 {-0.40/14 21}
23. Ra2 {+0.44/13 31} e6 {-0.40/12 38} 24. Mc3 {+0.42/12 30} Mf5
{-0.36/12 18} 25. Rae2 {+0.42/11 29} Rbc8 {-0.38/11 18} 26. Qd2
{+0.42/11 29} Me4 {-0.33/14 33} 27. Bf2 {+0.42/11 22} Mxf3 {-0.34/13 18}
28. Kxf3 {+0.59/11 28} Bf6 {-0.40/12 17} 29. Kg1 {+0.61/13 34} Be7
{-0.44/13 26} 30. Qe3 {+0.61/12 31} Qb8 {-0.54/12 26} 31. Qg3 {+0.64/12 30}
Mxc6 {-0.54/12 31} 32. bxc6 {+0.67/11 29} Rxc6 {-0.54/11 19} 33. Mb3
{+0.66/11 24} b5 {-0.60/12 26} 34. Ra2 {+0.70/11 31} Rc1 {-0.56/11 23} 35.
Be3 {+0.71/11 31} Rxe1+ {-0.78/12 24} 36. Qxe1 {+0.75/10 30} Qb6
{-0.78/12 27} 37. Mb4 {+0.93/11 29} Bxb4 {-0.73/13 1:11} 38. Qxb4
{+0.92/12 29} Qc6 {-0.73/11 1:34} 39. Rb2 {+0.97/11 29} Qxa6
{-0.42/13 1:13} 40. Qxb5 {+0.37/11 20} Qxb5 {-0.29/14 1:58} 41. Rxb5
{+0.27/11 35} Rc2 {-0.69/13 29} 42. Rb1 {+0.21/11 31} Rxg2+ {-0.24/12 29}
43. Kf3 {+0.15/11 27} Re2 {-0.20/13 29} 44. Re1 {+0.19/11 31} Rxe1
{-0.09/13 10} 45. Kxe1 {+0.12/15 26} Ke7 {-0.11/18 16} 46. Kf3
{+0.09/15 31} Kf5 {-0.09/16 30} 47. Bd2 {+0.21/15 29} Kd6 {-0.45/17 11} 48.
Ke5 {+0.40/13 22} Kb5 {-0.70/18 26} 49. Kc6 {+0.86/14 34} a6 {-1.04/17 11}
50. Bb4 {+1.01/14 31} Kc7 {-1.19/20 23} 51. h4 {+1.11/14 30} Kb5
{-0.61/19 23} 52. Bc5 {+1.01/14 29} Kc3 {-0.63/19 22} 53. Kd8 {+1.11/14 28}
a5 {-0.63/17 33} 54. Kxf7 {+1.07/14 30} a4 {-0.90/16 33} 55. Kd6
{+1.13/15 29} Ke2 {-0.95/15 33} 56. Ke8 {+0.90/15 58} Kxf4 {-1.19/17 17}
57. Kg7 {+0.87/16 29} Kg2 {-1.25/17 33} 58. Kxe6 {+1.22/16 25} Kxh4
{-1.76/18 27} 59. Kf4 {+1.61/15 30} a3 {-1.76/16 34} 60. Bxa3 {+1.98/15 27}
Kf5 {-1.77/20 34} 61. Ke6 {+1.99/15 32} h6 {-1.30/21 14} 62. Kf4
{+2.28/16 30} g5+ {-1.30/19 35} 63. Kxd5 {+2.42/15 29} Kxd4 {-2.07/21 24}
64. Kf6 {+2.39/15 29} h5 {-1.24/21 14} 65. Kxh5 {+3.42/15 20} Ke2
{-1.24/17 30} 66. Kf6 {+3.42/19 28} g4 {+0.00/32 6} 67. Kxg4 {+3.42/21 28}
{False draw claim: ''} 1-0

[d]8/8/8/8/6N1/B7/4n3/8 b - - 0 67

As far as I know Knightmate chess, that looks like a draw. Unless somebody can give me good reasons why not to, I will overrule Winboard and change the result to be a draw.

[enhorning]

One third of the games have been played: 52/156. Cross-table so far:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Ta Jo Ge Fa Ja Sa Da Li Fa Va Sj Fi
 1. Tadyak 1.1                ##          11       11    11 11 11  100%  10.0 ( 22.0,  22.0)
 2. Joker KM                     ##                01 11    11      83%   5.0 ( 14.0,  11.0)
 3. Gerbil KM 0.2_ja                ##          10    11 01 11      75%   6.0 ( 21.0,  14.5)
 4. Faile KM 144_ja                    ## 00    11 0=       11 11   65%   6.5 ( 27.0,  12.5)
 5. Jabba 1.0 KM              00       11 ## 10                11   62%   5.0 ( 45.0,  19.0)
 6. SamChess KM 1.2                       01 ##    =1 =0 11 01      60%   6.0 ( 28.0,  16.0)
 7. Dabbaba 652_ja                  01 00       ##    1=       11   56%   4.5 ( 31.0,  10.5)
 8. Lime KM 621_ja            00 10    1=    =0    ##               37%   3.0 ( 55.0,  17.8)
 9. Fairy-Max 4.8Rt              00 00       =1 0=    ## 10         30%   3.0 ( 47.0,  13.3)
10. Vanilla Chess KM 26g_ja   00    10       00       01 ##         25%   2.0 ( 50.0,   9.0)
11. Sjaak r470                00 00 00 00    10             ##      10%   1.0 ( 67.0,   6.0)
12. Fimbulwinter KM 50_ja     00       00 00    00             ##    0%   0.0 ( 52.0,   0.0)

[enhorning]

Two thirds of the games have been played: 104 / 156. Cross-table so far:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Ta Ne Jo Li Sa Ge Fa Ja Sj Va Fa Da Fi
 1. Tadyak 1.1                ##    11 11    10    11 11 11 11 11 11   94%  17.0 (123.0, 113.5)
 2. NebiyuChess_1.43             ##             11          11 01      83%   5.0 ( 44.0,  37.5)
 3. Joker KM                  00    ## 01 10 =1       11    11 11      68%   9.5 (126.0,  68.3)
 4. Lime KM 621_ja            00    10 ## =0    1= 11          11 11   64%   9.0 (127.0,  63.3)
 5. SamChess KM 1.2                 01 =1 ## 10    01 01 11 =0 11      62%  10.0 (121.0,  75.0)
 6. Gerbil KM 0.2_ja          01    =0    01 ## 10    11 01 11 10      59%   9.5 (141.0,  76.8)
 7. Faile KM 144_ja              00    0=    01 ## 00 11 11 01 11 11   58%  10.5 (114.0,  58.0)
 8. Jabba 1.0 KM              00       00 10    11 ## 00 11    11 11   56%   9.0 (132.0,  58.0)
 9. Sjaak r470                00    00    10 00 00 11 ##       01 11   37%   6.0 (146.0,  36.5)
10. Vanilla Chess KM 26g_ja   00          00 10 00 00    ## 01 11 11   37%   6.0 (137.0,  29.5)
11. Fairy-Max 4.8Rt           00 00 00    =1 00 10       10 ## 0= 10   28%   5.0 (150.0,  35.8)
12. Dabbaba 652_ja            00 10 00 00 00 01 00 00 10 00 1= ## 11   27%   6.5 (195.0,  30.0)
13. Fimbulwinter KM 50_ja     00       00       00 00 00 00 01 00 ##    6%   1.0 (138.0,   5.0)

[enhorning]

It is finished.

Congratulations to Tadyak for an outstanding performance, winning 23 games out of 24, with a 5 points margin down to the next program.

Following Tadyak, there was a group of 7 programs from 12-18 points, with no big gaps. Then, a group of 3 programs at 5.5-6.5 points, and last Fimbulwinter, which did at least manage to score a win.

Final cross-table:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Ta Li Ne Jo Ge Fa Sa Sj Ja Da Va Fa Fi
 1. Tadyak 1.1                ## 11 11 11 10 11 11 11 11 11 11 11 11   96%  23.0 (266.0, 252.0)
 2. Lime KM 621_ja            00 ## 11 10 11 1= =0 10 11 11 11 11 11   75%  18.0 (276.0, 176.8)
 3. NebiyuChess_1.43          00 00 ## =1 01 11 11 01 11 01 11 11 11   69%  16.5 (279.0, 156.5)
 4. Joker KM                  00 01 =0 ## =1 == 10 11 01 11 11 11 11   67%  16.0 (280.0, 146.8)
 5. Gerbil KM 0.2_ja          01 00 10 =0 ## 10 01 11 1= 10 01 11 11   58%  14.0 (284.0, 140.5)
 6. Faile KM 144_ja           00 0= 00 == 01 ## =1 11 00 11 11 01 11   54%  13.0 (286.0, 114.3)
 7. SamChess KM 1.2           00 =1 00 01 10 =0 ## 01 01 11 11 =0 11   52%  12.5 (287.0, 117.3)
 8. Sjaak r470                00 01 10 00 00 00 10 ## 11 01 11 11 11   50%  12.0 (288.0, 102.5)
 9. Jabba 1.0 KM              00 00 00 10 0= 11 10 00 ## 11 11 =1 11   50%  12.0 (288.0,  96.8)
10. Dabbaba 652_ja            00 00 10 00 01 00 00 10 00 ## 00 1= 11   27%   6.5 (299.0,  52.8)
11. Vanilla Chess KM 26g_ja   00 00 00 00 10 00 00 00 00 11 ## 01 11   25%   6.0 (300.0,  34.5)
12. Fairy-Max 4.8Rt           00 00 00 00 00 10 =1 00 =0 0= 10 ## 10   23%   5.5 (301.0,  48.0)
13. Fimbulwinter KM 50_ja     00 00 00 00 00 00 00 00 00 00 00 01 ##    4%   1.0 (310.0,   5.5)

Feeding this into bayeselo gives these results:

Code: Select all

Ratings:
Rank Name                      Elo    +    - games score oppo. draws
   1 Tadyak 1.1                437  268  268    24   96%   -36    0%
   2 Lime KM 621_ja            194  187  187    24   75%   -16    8%
   3 NebiyuChess_1.43          152  184  184    24   69%   -13    4%
   4 Joker KM                  127  172  172    24   67%   -11   17%
   5 Gerbil KM 0.2_ja           66  177  177    24   58%    -5    8%
   6 Faile KM 144_ja            44  169  169    24   54%    -4   17%
   7 SamChess KM 1.2            19  174  174    24   52%    -2   13%
   8 Sjaak r470                 -1  181  181    24   50%     0    0%
   9 Jabba 1.0 KM               -5  174  174    24   50%     0    8%
  10 Dabbaba 652_ja           -188  191  191    24   27%    16    4%
  11 Vanilla Chess KM 26g_ja  -202  195  195    24   25%    17    0%
  12 Fairy-Max 4.8Rt          -206  190  190    24   23%    17   13%
  13 Fimbulwinter KM 50_ja    -436  270  270    24    4%    36    0%

LOS Matrix:
                         Ta Li Ne Jo Ge Fa Sa Sj Ja Da Va Fa Fi
Tadyak 1.1                  97 99 99 99 99 99 99 99 99 99 99 99
Lime KM 621_ja            2    66 76 91 94 96 97 98 99 99 99 99
NebiyuChess_1.43          0 33    60 81 87 92 94 95 99 99 99 99
Joker KM                  0 23 39    75 82 88 91 92 99 99 99 99
Gerbil KM 0.2_ja          0  8 18 24    59 69 76 78 99 99 99 99
Faile KM 144_ja           0  5 12 17 40    61 68 70 99 99 99 99
SamChess KM 1.2           0  3  7 11 30 38    58 60 98 98 99 99
Sjaak r470                0  2  5  8 23 31 41    51 97 97 98 99
Jabba 1.0 KM              0  1  4  7 21 29 39 48    97 97 98 99
Dabbaba 652_ja            0  0  0  0  0  0  1  2  2    55 57 97
Vanilla Chess KM 26g_ja   0  0  0  0  0  0  1  2  2 44    51 97
Fairy-Max 4.8Rt           0  0  0  0  0  0  0  1  1 42 48    96
Fimbulwinter KM 50_ja     0  0  0  0  0  0  0  0  0  2  2  3

Despite some technical hick-ups here and there, it was fun to watch the games. Positions where both sides thought they were up by a pawn or more abounded - presumably due to evaluation functions for variants not being as well explored as for Fide Chess.

Next up is a Crazyhouse tournament, probably a couple of weeks from now. Preliminary participants (assuming I can get them all up and running):
Pulsar
Sjeng
Imortal
TJChess

I'll start a new thread for that tournament, once the time draws closer.

If anybody wants the PGN of these games, email me (enhorning@gmail.com) or send me a PM. (Or is there a way to post attachments to this forum? I don't see one... that is yet another feature lacking.)

[Ferdy]

Thanks for running the tour, alright Tadyak dominates

Next up is a Crazyhouse tournament, probably a couple of weeks from now. Preliminary participants (assuming I can get them all up and running):
Pulsar
Sjeng
Imortal
TJChess

Additional zh engines that you might want to include.
1. Sunsetter 7e
http://sourceforge.net/projects/sunsett ... etter-w23/

2. KKFChess v2.6.6
http://freespace.virgin.net/andrew.fankk/

3. Perhaps Nebiyu can also play.

Be aware that Sjeng has a bug, sample is

Code: Select all

Xboard: Forfeit due to invalid move: q@h2

The q should have been a Q if this is a drop move.

[enhorning]

3. Perhaps Nebiyu can also play.

Hmm, in Winboard, Crazyhouse is not listed as one of the supported variant for Nebiyu. If it can indeed play Crazyhouse, what do I need to do to make it so?

I went through the pgn-file from the tournament and looked at openings statistics. The following lines were played in at least 5 games each:

1. d4 d5 2. e3
1. d4 d5 2. f4
1. d4 d5 2. e4 dxe4 3. Bf4
1. d4 d5 2. Qd3
1. d4 e6 2. e4 d5
1. d4 f5
1. e4 d5 2. exd5 Qxd5
1. e4 g6
1. e4 c5
1. e3 d5
1. c4
1. b3
1. g3

The longest duplicated opening was 14 ply: 1. e4 d5 2. exd5 Qxd5 3. c4 Qa5 4. Mc2 Bd7 5. b3 c5 6. Bb2 Mc7 7. Qf3 Kc8.

1.d4 d5 was by far the most common 2-ply opening, played in 37 games out of 156. Second most popular was 1. e4 d5, which was played in 13 games.