Chess variant tournament: Shatranj

hgm · Post by **hgm** » Tue Mar 05, 2013 6:55 am

Capturing all one's opponent's pieces apart from the king (baring the king) was a win, unless your opponent could capture your last piece on his or her next move, then in most parts of the Islamic world it was a draw, but in Medina it was a win.[2]

WinBoard and ShaMax use the standard rule (i.e. KK is draw). The implementation can be simple by delaying the claim one ply (which thus has become mandatory, or WB would forfeit you for a false claim!), to await the recapture:

If the side to move is not bare, and the opponent is, the side to move has won. The engine can treat such bareness as an in-check condition, awarding the same score as when it would have a capture on the King.

KFEKF is generally a draw, if the defending Ferz can find shelter with its King. Which, in ~15% of the positions where it has the move it can not.

enhorning · Post by **enhorning** » Wed Mar 06, 2013 1:50 am

Half-way standings, after 3 cycles out of 6:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Tiy Neb Sha Sja Pul HaC Che Dab
 1. Tiyaga v1.0               ### 1=1 1=1 111 =11 111 111 111
                              ### ==1 =1= 111 1=1 111 111 111   90%  38.0 (780.0, 670.0)

 2. NebiyuChess_1.43          0=0 ### 1== 111 111 111 111 111
                              ==0 ### 1== 111 111 111 111 111   85%  35.5 (795.0, 571.0)

 3. ShaMax 4.8S               0=0 0== ### 1== ==1 111 11= 11=
                              =0= 0== ### 1=1 =0= 1=1 1=1 =11   63%  26.5 (849.0, 430.0)

 4. Sjaak 524                 000 000 0== ### 101 111 =1= 111
                              000 000 0=0 ### 1=0 =11 1=1 111   50%  21.0 (882.0, 254.0)

 5. Pulsar2009-9b             =00 000 ==0 010 ### === === 11=
                              0=0 000 =1= 0=1 ### ==0 ==1 ==1   40%  17.0 (906.0, 266.5)

 6. HaChu 0.5beta             000 000 000 000 === ### =1= 111
                              000 000 0=0 =00 ==1 ### =00 111   31%  13.0 (930.0, 141.5)

 7. ChessV_WinBoard           000 000 00= =0= === =0= ### 011
                              000 000 0=0 0=0 ==0 =11 ### =1=   30%  12.5 (933.0, 164.0)

 8. Dabbaba 6.52 JA           000 000 00= 000 00= 000 100 ###
                              000 000 =00 000 ==0 000 =0= ###   11%   4.5 (981.0,  77.0)

Unfortunately, progress is a bit slower now, as I need some cpus for other stuff during the week, so only running 2 games concurrently.

Also, Nebiyu just had its first loss against somebody else than Tiyaga - well done, Sjaak! Looking at the game, I am not sure if this was again due to the lack of knowledge about bare Kings, or if the ending was lost anyhow:

[Event "Computer Chess Game"]
[Site "ENHORNING"]
[Date "2013.03.05"]
[Round "43"]
[White "Sjaak 524"]
[Black "NebiyuChess_1.43"]
[Result "1-0"]
[TimeControl "40/1200"]
[Variant "shatranj"]
[FEN "rnbkqbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBKQBNR w 0 1"]
[SetUp "1"]

1. Nf3 Be6 {-0.05/22 29} 2. Nc3 {+0.05/15 17} g6 {-0.15/22 26} 3. d3
{+0.12/16 18} d6 {-0.05/23 24} 4. d4 {+0.06/16 32} Nf6 {+0.00/24 29} 5. d5
{+0.04/15 20} Bc8 {+0.15/24 29} 6. e3 {+0.04/15 27} c6 {+0.15/25 29} 7.
dxc6 {+0.00/16 32} bxc6 {+0.15/24 29} 8. e4 {-0.02/15 23} Nbd7
{+0.15/24 29} 9. Be3 {-0.03/14 25} Be6 {+0.20/22 25} 10. Bd3 {-0.04/14 18}
Rb8 {+0.10/22 29} 11. b3 {-0.07/14 21} Rb6 {+0.10/20 29} 12. h3
{+0.04/13 36} d5 {+0.15/23 27} 13. Na4 {+0.00/14 34} dxe4 {+0.25/23 29} 14.
Nxb6 {+0.04/16 43} axb6 {+0.25/25 23} 15. Nd4 {+0.11/16 22} Ne5
{+0.30/24 30} 16. Bb1 {+0.28/15 27} Bd6 {+0.20/21 30} 17. Ne2 {+0.35/15 25}
g5 {+0.20/21 22} 18. a3 {+0.41/15 18} Rg8 {+0.20/21 24} 19. a4
{+0.34/15 31} Qd7 {+0.05/21 30} 20. a5 {+0.59/15 20} b5 {+0.00/22 28} 21.
a6 {+0.68/15 35} h6 {-0.10/21 31} 22. a7 {+1.16/14 21} Rg6 {-0.10/19 31}
23. Bc5 {+1.39/13 24} h5 {-0.05/20 31} 24. Qd2 {+1.51/13 24} Bc4
{-0.30/18 30} 25. bxc4 {+1.63/14 38} Nxc4 {-0.40/20 25} 26. Qe3
{+1.69/14 38} e6 {-0.30/21 32} 27. a8=Q {+1.98/14 23} g4 {-0.35/20 32} 28.
h4 {+1.89/14 24} e5 {-0.40/20 31} 29. Kc1 {+2.07/14 32} Qe6 {-0.40/22 32}
30. Ng3 {+1.92/15 54} Kc7 {-0.45/23 32} 31. Ra6 {+2.04/14 29} Bf4
{-0.55/20 29} 32. Re1 {+2.17/12 40} Qd5 {-0.50/19 33} 33. Rd1 {+2.23/12 50}
Kd6 {-0.40/22 27} 34. Ba3 {+1.82/13 23} b4 {-0.30/22 29} 35. Ra4
{+1.60/14 53} Nxa3 {-0.30/21 32} 36. Rxb4 {+1.49/14 51} Nb5 {-0.35/23 32}
37. Rb2 {+1.43/12 27} Rg8 {-0.10/23 33} 38. Qb7 {+0.92/13 25} Kc7
{-0.10/23 31} 39. Qxc6 {+0.58/14 37} Qxc6 {-0.15/23 39} 40. Ne2
{+0.43/14 39} Bd6 {-0.15/24 51} 41. Rf1 {+0.46/14 24} Na3 {+0.05/22 28} 42.
Ra2 {+0.43/15 34} Nc4 {+0.00/24 23} 43. Ra4 {+0.61/15 31} Nb6 {+0.00/24 29}
44. Ra7+ {+0.60/15 15} Qb7 {+0.00/23 29} 45. Ra5 {+0.54/16 31} Nbd7
{+0.00/22 22} 46. Rd1 {+0.56/16 18} Re8 {+0.00/22 29} 47. Ng3 {+0.62/14 31}
Re6 {+0.00/23 22} 48. c3 {+0.68/14 18} Kb6 {+0.00/24 25} 49. Ra3
{+0.68/14 25} Kc6 {+0.00/22 30} 50. Nf5 {+0.68/14 29} Nb6 {+0.00/21 30} 51.
Ra5 {+0.75/12 17} Nc4 {+0.00/24 30} 52. Ra4 {+0.82/15 25} Nb6 {+0.00/24 30}
53. Ra7 {+0.83/15 22} Nfd7 {+0.00/21 30} 54. Ra2 {+0.80/13 30} Nf6
{-0.05/21 23} 55. Kc2 {+0.75/12 22} Nfd7 {-0.15/22 30} 56. Ng7
{+0.80/13 18} Rh6 {-0.20/23 23} 57. Ra5 {+0.78/14 35} Nc4 {-0.15/25 31} 58.
Nf5 {+0.81/15 18} Rf6 {-0.10/25 31} 59. Ne7+ {+0.81/14 21} Kb6
{+0.00/25 31} 60. Ra4 {+0.72/14 40} Kb5 {+0.00/26 24} 61. Ra7 {+0.66/13 30}
Nc5 {+0.00/26 31} 62. Nd5 {+0.66/14 32} Rf5 {+0.00/26 31} 63. Ra2
{+1.13/13 20} Bf4 {+0.00/25 29} 64. Nc7+ {+0.99/13 25} Kc6 {-0.15/26 32}
65. Ne8 {+1.13/15 43} g3 {+0.00/25 32} 66. Ng7 {+0.97/15 54} Rf6
{-0.30/25 32} 67. Nxh5 {+0.97/15 36} Rg6 {-0.50/25 32} 68. Qd4
{+1.03/14 22} gxf2 {-0.45/24 32} 69. Qxc5 {+0.95/14 24} Ne3+ {-0.40/25 31}
70. Kc1 {+0.94/16 28} Nxd1 {-0.70/24 25} 71. Kxd1 {+1.06/16 40} Rxg2
{-0.85/24 26} 72. Ra5 {+1.23/17 34} f1=Q {-0.85/25 32} 73. Ke1
{+1.22/17 32} Qe2 {-1.10/25 35} 74. Qd4 {+1.21/17 42} f6 {-1.05/25 35} 75.
Ra2 {+1.39/15 26} Qf3 {-0.80/28 35} 76. Rxg2 {+1.75/18 25} Qxg2
{-0.75/28 8} 77. Qe3 {+1.72/20 25} f5 {-0.70/31 43} 78. Ng7 {+1.74/21 34}
Qf3 {-0.90/29 34} 79. Nxf5 {+1.75/22 52} Kd5 {-0.90/31 49} 80. Kd1
{+1.74/21 44} Qc6 {-0.85/29 54} 81. Qf2 {+1.76/20 19} Kc4 {-0.90/29 24} 82.
Kc2 {+1.77/21 31} Kd5 {-1.05/29 21} 83. Ne3+ {+1.80/21 21} Ke6
{-1.00/31 29} 84. Kb3 {+1.80/21 32} Kd6 {-1.10/30 26} 85. h5 {+1.82/20 23}
Qd7 {-0.80/29 29} 86. Qg3 {+1.80/18 19} Ke6 {-0.95/29 29} 87. c4
{+1.77/20 32} Kf6 {-0.90/29 25} 88. Qh4 {+1.85/20 23} Qe6 {-0.90/29 29} 89.
c5 {+1.84/21 33} Qd7 {-1.40/28 29} 90. Kb4 {+1.90/20 20} Qc6 {-1.55/30 29}
91. Ka5 {+1.96/22 33} Qd7 {-1.60/31 29} 92. Kb6 {+2.04/19 45} Qe6
{-1.70/29 23} 93. c6 {+2.07/20 20} Qe2 {-1.75/30 23} 94. c7 {+2.65/19 17}
Qd7 {-1.90/30 23} 95. c8=Q {+2.65/19 19} Qxc8 {-1.85/34 30} 96. Kc7
{+2.65/22 18} Qf3 {-1.90/35 28} 97. Kxc8 {+2.65/23 27} Ke7 {-1.90/33 26}
98. Qg5 {+2.65/23 34} Qe2 {-1.90/33 31} 99. Kc7 {+2.65/21 23} Qf3
{-1.90/34 31} 100. Kc6 {+2.65/21 14} Ke6 {-1.90/36 31} 101. Kb7
{+2.65/23 37} Qe2 {-1.90/34 31} 102. Kc7 {+2.85/23 55} Qf3 {-1.90/34 31}
103. Kd8 {+2.65/22 36} Kd6 {-3.05/31 26} 104. Ke8 {+2.65/21 52} Ke6
{-3.35/32 32} 105. Kf8 {+3.35/21 36} Qg4 {-3.05/32 32} 106. h6
{+3.24/21 23} Bxh6+ {-3.50/32 26} 107. Qxh6 {+3.35/22 23} Qf5 {-3.45/34 26}
108. Kg7 {+160.00/18 4} Qg4 {-5.05/33 27} 109. Nxg4 {+159.98/16 0.4} Kf5
{-5.05/32 33} 110. Ne3+ {+159.94/15 0.4} Kf4 {-5.10/31 30} 111. Nd5+
{+160.00/7 0.1} Kf3 {-5.05/33 34} 112. Kf6 {+159.98/14 0.8} e3
{-5.10/36 27} 113. Nxe3 {+160.00/3 0.1} Kxe3 {-3.50/38 31} 114. Kxe5
{+160.00/8 0.1} Kf3 {-3.55/40 31}
{Xboard adjudication: Bare king} 1-0

enhorning · Post by **enhorning** » Thu Mar 07, 2013 7:50 am

Standings after 4 cycles out of 6:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Tiya Nebi ShaM Sjaa Puls HaCh Ches Dabb
 1. Tiyaga v1.0               #### 1=1= 1=1= 1111 =111 1111 1111 1111
                              #### =11= 1=11 1111 =111 1111 1111 1111   92%  51.5 (1380.0, 1214.8)

 2. NebiyuChess_1.43          0=0= #### 1==1 1111 1111 1111 1111 1111
                              =00= #### ==11 1101 1111 1111 11=1 1111   83%  46.5 (1420.0, 1004.5)

 3. ShaMax 4.8S               0=0= 0==0 #### 1==1 ==1= 1111 11=1 11==
                              0=00 ==00 #### =100 0==1 =1=1 =111 1111   60%  33.5 (1524.0, 685.3)

 4. Sjaak 524                 0000 0000 0==0 #### 1011 111= =1=1 1111
                              0000 0010 =011 #### =011 1111 =110 1111   55%  31.0 (1544.0, 549.5)

 5. Pulsar2009-9b             =000 0000 ==0= 0100 #### ==== ==== 11==
                              =000 0000 1==0 =100 #### =0== =11= =111   39%  22.0 (1616.0, 425.0)

 6. HaChu 0.5beta             0000 0000 0000 000= ==== #### =1== 1111
                              0000 0000 =0=0 0000 =1== #### 00== 1111   31%  17.5 (1652.0, 247.5)

 7. ChessV_WinBoard           0000 0000 00=0 =0=0 ==== =0== #### 011=
                              0000 00=0 =000 =001 =00= 11== #### 1=1=   30%  17.0 (1656.0, 306.5)

 8. Dabbaba 6.52 JA           0000 0000 00== 0000 00== 0000 100= ####
                              0000 0000 0000 0000 =000 0000 0=0= ####    9%   5.0 (1752.0, 109.0)

Ferdy · Post by **Ferdy** » Thu Mar 07, 2013 7:13 pm

enhorning wrote:By the way, any suggestions on what variant to run next? As usual, there will be a 2-3 weeks break after this tournament before I run the next one.

Have a look on the capablanca or perhaps the gothic chess variants. There are additional pieces of chancellor = combined rook and knight and an archbishop with a combined moves of knight and bishop. The board is larger now at 10x8, isolated attack on certain areas contributes to interesting plays of the new pieces.

enhorning · Post by **enhorning** » Fri Mar 08, 2013 11:24 pm

Standings after 5 cycles out of 6:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Tiyag Nebiy ShaMa Sjaak Pulsa HaChu Chess Dabba
 1. Tiyaga v1.0               ##### 1=1== 1=1=1 11111 =111= 11111 11111 11111
                              ##### 11=1= =1111 11111 11111 11111 11111 11111   93%  65.0 (2150.0, 1913.8)

 2. NebiyuChess_1.43          0=0== ##### 1==1= 11111 11111 11111 11111 11111
                              00=0= ##### =1111 10111 1111= 11111 1=111 11111   84%  58.5 (2215.0, 1581.3)

 3. ShaMax 4.8S               0=0=0 0==0= ##### 1==1= ==1=0 1111= 11=1= 11==1
                              =0000 =0000 ##### 100== ==11= 1=11= 1111= 11111   59%  41.0 (2390.0, 1009.0)

 4. Sjaak 524                 00000 00000 0==0= ##### 1011= 111=1 =1=1= 11111
                              00000 01000 011== ##### 0111= 11111 11011 11111   56%  39.5 (2405.0, 854.3)

 5. Pulsar2009-9b             =000= 00000 ==0=1 0100= ##### ===== ===== 11===
                              00000 0000= ==00= 1000= ##### 0==1= 11==1 11111   41%  28.5 (2515.0, 663.3)

 6. HaChu 0.5beta             00000 00000 0000= 000=0 ===== ##### =1==0 11111
                              00000 00000 0=00= 00000 1==0= ##### 0==10 11111   31%  21.5 (2585.0, 368.3)

 7. ChessV_WinBoard           00000 00000 00=0= =0=0= ===== =0==1 ##### 011=1
                              00000 0=000 0000= 00100 00==0 1==01 ##### =1=11   30%  21.0 (2590.0, 445.0)

 8. Dabbaba 6.52 JA           00000 00000 00==0 00000 00=== 00000 100=0 #####
                              00000 00000 00000 00000 00000 00000 =0=00 #####    7%   5.0 (2750.0, 136.3)

Sjaak has caught up a lot to ShaMax in the last couple of cycles... in the standings after 3 cycles, there was 5.5 points separating them - now, the gap is down to 1.5 points.

hgm · Post by **hgm** » Sat Mar 09, 2013 8:37 am

Indeed, it will be a close finish! I consider that still a pretty good performance of ShaMax. Probably its lack of proper Pawn-structure evaluation hurts it much less in this game, where promotion hardly increases a Pawn in value, than in other Chess variants. King safety is probably also less important.

But even without a large 'eval gap', it does pretty well, considering that it is virtual without move sorting. (Only the hash move is duplicated in front of the list, if it was not a castling or double-push (which you also don't have in Shatranj!

), and in any case it starts with the moves of the piece that could do the hash move.) I am thinking about equiping Pair-o-Max with true move sorting, using MVV/LVA, no longer searching the hash move twice, and implementing a killer heuristic (perhaps only with one killer, the null-move killer).

enhorning · Post by **enhorning** » Sun Mar 10, 2013 12:03 am

So, when I woke up this morning, all games except one had been finished. In an unfortunate twist of events, it was the game Sjaak - ShaMax, and in the results before this final game where that they were both on 48 points.

Unfortunately, Sjaak seems to have become stuck (in a disadvantageous position) - its clock had run out by many hours, and its process was still running, but no output in the engine window, and no move:

[Event "Computer Chess Game"]
[Site "ENHORNING"]
[Date "2013.03.09"]
[Round "80"]
[White "Sjaak 524"]
[Black "ShaMax 4.8S"]
[Result "*"]
[TimeControl "40/1200"]
[Variant "shatranj"]
[FEN "rnbkqbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBKQBNR w 0 1"]
[SetUp "1"]

1. e3 Be6 {+0.21/12 23} 2. e4 {-0.03/16 18} Bd6 {+0.13/12 23} 3. Bd3
{-0.01/18 31} f6 {+0.11/12 19} 4. Be3 {+0.00/17 27} Nc6 {-0.03/14 28} 5.
Ne2 {+0.01/17 31} b6 {-0.08/13 27} 6. f3 {+0.03/16 20} Ne5 {-0.02/13 19} 7.
h3 {+0.07/15 32} Bc4 {-0.06/12 19} 8. Nec3 {+0.12/14 18} c6 {+0.04/13 18}
9. Qf2 {+0.21/16 32} e6 {-0.14/13 49} 10. Qg3 {+0.30/15 27} Ne7
{-0.24/12 24} 11. b3 {+0.29/15 44} Ba6 {-0.26/12 24} 12. Ne2 {+0.20/14 24}
Bc4 {-0.10/13 16} 13. bxc4 {+0.54/15 19} Nxc4 {-0.13/13 17} 14. a3
{+0.53/14 25} Nb2+ {+0.02/12 35} 15. Ke1 {+0.58/15 18} Nc4 {-0.01/13 1:25}
16. a4 {+0.66/14 19} g6 {-0.13/12 40} 17. Na3 {+0.87/13 21} Nxa3
{-0.08/15 27} 18. Rxa3 {+0.80/15 29} h6 {+0.03/13 27} 19. f4 {+0.93/13 18}
e5 {+0.14/14 23} 20. fxe5 {+1.10/16 38} fxe5 {-0.04/14 23} 21. a5
{+1.08/15 35} Rb8 {-0.13/13 14} 22. Rf1 {+1.27/13 24} Bb4 {-0.16/13 20} 23.
axb6 {+1.41/14 33} axb6 {-0.04/14 22} 24. Ra6 {+1.38/12 19} Bxd2
{-0.04/13 18} 25. Kxd2 {+1.53/14 37} Nc8 {-0.22/13 18} 26. Rf6
{+1.41/13 40} Rg8 {-0.12/14 20} 27. Nc3 {+1.24/14 31} Ne7 {-0.11/14 24} 28.
Ra2 {+1.36/13 38} Rg7 {-0.20/13 17} 29. Rf1 {+1.21/14 30} g5 {-0.21/13 40}
30. Rfa1 {+1.34/13 21} g4 {-0.21/14 17} 31. h4 {+1.37/13 25} Qf7
{-0.24/13 29} 32. Rf1 {+1.36/14 37} Qe8 {-0.39/13 19} 33. Ra4 {+1.31/14 42}
Rg6 {-0.28/13 16} 34. h5 {+1.46/14 28} Rd6 {-0.29/13 22} 35. Raa1
{+1.51/13 33} b5 {-0.23/12 16} 36. Rf8 {+1.25/13 29} Rd4 {-0.08/12 18} 37.
Rf6 {+1.26/14 27} Nc8 {-0.10/13 17} 38. Rxh6 {+1.17/14 35} Nd6
{+0.10/14 23} 39. Kc1 {+1.20/15 59} Nxe4 {-0.20/15 19} 40. Ne2
{+1.06/16 48} Rd6 {-0.25/16 33} 41. Rh8 {+1.06/15 31} Re6 {-0.22/13 18} 42.
Rg8 {+1.19/14 19} Nf6 {-0.23/15 24} 43. Rg5 {+1.12/14 25} Nd5 {-0.24/14 31}
44. Kd2 {+1.21/13 16} Nf6 {-0.01/15 33} 45. Ke1 {+1.20/13 17} Qf7
{-0.25/13 24} 46. Ra7 {+1.26/13 28} Re8 {-0.26/13 22} 47. Rf5 {+1.35/13 17}
Re6 {-0.25/14 20} 48. Kd1 {+1.30/13 33} Qe8 {-0.25/12 20} 49. Kd2
{+1.22/13 47} Ne4+ {+0.05/13 23} 50. Ke1 {+1.23/14 32} Rc8 {-0.22/13 20}
51. Kd1 {+1.24/14 19} Rb8 {-0.22/14 23} 52. Qf2 {+1.18/13 37} Rf6
{-0.20/14 28} 53. Ng3 {+1.18/14 19} Rxf5 {-0.33/15 22} 54. Bxf5
{+1.11/16 35} Nxg3 {-0.18/18 33} 55. Qxg3 {+0.92/16 54} Ke7 {-0.11/17 54}
56. h6 {+0.75/14 29} Kf6 {-0.12/18 20} 57. Bxd7 {+0.49/17 32} Rd8
{-0.19/18 27} 58. Ke2 {+0.53/18 17} Qxd7 {-0.11/17 28} 59. Qh4
{+0.47/17 31} Qe6 {-0.18/18 32} 60. Qg5+ {+0.49/15 33} Kf5 {-0.19/19 46}
61. Ke1 {+0.47/15 24} Rf8 {-0.23/16 32} 62. Re7 {+0.44/16 21} Rd8
{-0.25/17 32} 63. Ke2 {+0.51/15 21} Rh8 {-0.21/17 21} 64. h7 {+0.56/18 22}
Qd5 {-0.26/18 22} 65. Rg7 {+0.56/19 36} Ke4 {-0.28/19 1:10} 66. Qf6
{+0.60/16 27} Kf5 {-0.27/17 26} 67. Qe7 {+0.56/17 25} Kf4 {-0.26/17 23} 68.
Qd6 {+0.61/17 25} e4 {-0.23/18 19} 69. Kf2 {+0.48/16 32} g3+ {-0.24/20 23}
70. Ke2 {+0.60/18 34} Kf5 {-0.18/20 24} 71. Qe7 {+0.57/17 23} b4
{-0.19/18 49} 72. Bc5 {+0.59/18 24} Ke5 {-0.21/20 29} 73. Ke3 {+0.59/19 26}
Qe6 {-0.19/21 32} 74. Kd2 {+0.55/18 46} Qd5 {-0.19/19 21} 75. Be3
{+0.55/17 31} Qe6 {-0.19/20 30} 76. Kd1 {+0.55/17 27} Qd5 {-0.22/19 25} 77.
Kc1 {+0.55/19 34} Qc4 {-0.27/19 48} 78. Kd2 {+0.55/19 48} Ke6 {-0.20/17 38}
79. Kd1 {+0.55/18 53} Qd5 {-0.17/19 23} 80. c3 {+0.25/17 1:18} bxc3
{+0.06/20 40} 81. Bc5 {+0.30/17 41} Ke5 {+0.83/20 29} 82. Ba3 {-0.42/15 27}
Qc4 {+0.80/18 25} 83. Bc1 {-0.56/15 30} Qd3 {+0.97/20 27} 84. Be3
{-1.10/16 17} Re8 {+1.21/19 28} 85. Rxg3 {-1.02/15 33} Rxe7 {+1.75/20 36}
86. Rh3 {-1.54/17 24} Rg7 {+1.78/20 26} 87. g3 {-1.67/17 32} Kf5
{+1.82/20 37} 88. h8=Q {-1.73/17 18} Rf7 {+1.98/20 48} 89. Rh4
{-1.83/18 33} Ke5 {+2.03/21 31} 90. Ke1 {-1.77/19 22} Rf3 {+2.04/21 25} 91.
Bc1 {-1.80/18 30} Rxg3 {+1.94/21 31} 92. Kf2 {-1.80/19 31} Rg6
{+1.98/20 33} 93. Rh7 {-1.78/15 17} Rf6+ {+1.98/20 48} 94. Kg2
{-1.77/15 16} c2 {+1.91/18 43} 95. Re7+ {-1.77/15 24} Re6 {+2.03/19 30} 96.
Rd7 {-1.80/19 32} Rg6+ {+2.03/19 42} 97. Kf2 {-1.95/19 33} Rh6
{+2.04/19 33} 98. Qg7 {-1.92/19 32} Rh2+ {+2.70/21 28} 99. Kg3
{-2.19/17 28} Rh1 {+2.72/20 19} 100. Be3 {-2.64/16 33} Re1 {+2.86/19 31}
101. Bc5 {-2.66/17 33} Ke6 {+2.94/17 21} 102. Ra7 {-2.72/16 33} Kd6
{+3.01/19 19} 103. Ba3 {-2.81/17 27} Ra1 {+2.91/20 32} 104. Qf8
{-2.85/18 33} c1=Q {+2.94/20 28} 105. Qe7+ {-2.85/18 25} Ke6 {+3.36/20 18}
106. Rc7 {-2.84/18 26} Qd2 {+3.40/21 41} 107. Bc5 {-2.85/19 29} Kd5
{+3.46/20 37} 108. Ba7 {-2.85/19 25} e3 {+3.36/19 41} 109. Rd7+
{-2.85/16 23} Ke6 {+3.51/19 18} 110. Rc7 {-2.85/16 21} Ra5 {+3.52/19 29}
111. Qf8 {-2.89/17 37} Kd6 {+3.50/18 19} 112. Rb7 {-3.04/15 45} e2
{+3.53/18 20} 113. Qe7+ {-3.05/13 30} Ke6 {+3.64/19 29} 114. Qf8
{-3.15/15 49} e1=Q {+3.62/18 31} 115. Rc7 {-3.05/17 51} Kd6 {+3.54/18 31}
116. Rf7 {-3.29/15 44} Qe3 {+3.56/18 20} 117. Kh4 {-3.38/14 26} Q1d2
{+3.53/17 26} 118. Kg3 {-3.37/13 26} Qe4 {+3.52/18 25} 119. Kg4
{-3.38/15 24} Qc3 {+3.51/18 28} 120. Rb7 {-3.40/14 28} Ke5 {+3.51/18 22}
121. Re7+ {-3.05/14 22} Kd5 {+3.47/20 20} 122. Kg5 {-3.05/14 18} Ra2
{+3.49/18 19} 123. Kg6 {-3.32/14 43} Rg2+ {+3.51/19 24} 124. Kf7
{-3.36/15 46} Rf2+ {+3.51/20 32} 125. Kg8 {-3.24/15 35} Qcd4 {+3.49/18 30}
126. Qg7 {-3.25/15 37} Qe5 {+3.47/17 20} 127. Re8 {-3.52/15 19} Q3d4
{+3.49/17 25} 128. Rd8+ {-3.55/16 18} Ke6 {+3.49/19 32} 129. Rb8
{-3.55/17 30} Rf3 {+3.50/17 24} 130. Rb7 {-3.55/16 30} Rf1 {+3.50/20 42}
131. Qh6 {-3.57/17 30} Rh1 {+3.47/18 21} 132. Qg7 {-3.55/16 18} Re1
{+3.47/20 35} 133. Kf8 {-3.51/17 21} Rf1+ {+3.49/20 26} 134. Ke8
{-3.52/16 21} Rf2 {+3.50/20 20} 135. Qf8 {-3.53/16 25} Ra2 {+3.48/19 26}
136. Re7+ {-3.52/16 26} Kd5 {+3.47/20 37} 137. Rd7+ {-3.56/17 32} Kc4
{+3.42/19 32} 138. Rb7 {-3.55/15 32} Rf2 {+3.44/18 26} 139. Qe7
{-3.49/16 22} Kd5 {+3.52/20 22} 140. Rc7 {-3.52/16 33} Rh2 {+3.49/20 22}
141. Kd7 {-3.50/17 33} Rh6 {+3.54/20 25} 142. Kc8 {-3.41/16 28} Rh7
{+3.49/19 27} 143. Kb7 {-2.71/17 34} Rg7 {+3.50/20 32} 144. Rd7+
{-2.67/18 17} Ke6 {+4.19/23 35} 145. Kxc6 {-2.67/18 18} Rxe7
{+5.39/23 1:07}
*

Its engine output window:
15 -5.16 1.7M 0:03.42 146. Rxe7 Kxe7 147. Kc7 Qd5 148. Kb8 Kd7 149. Kb7 Qf4 150. Kb6 Kc8 151. Ka6 Kc7 152. Kb5 Kb7 153. Ka5
14 -2.19 363117 0:00.64 146. Rxe7 Kxe7 147. Kb6 Qd5 148. Kb5 Qf4 149. Ka4 Ke8 150. Kb4 Kf8 151. Kb3 Kf7 152. Kb4 Kg8
13 -2.19 264284 0:00.46 146. Rxe7 Kxe7 147. Kb6 Qd5 148. Kb5 Qf4 149. Ka4 Ke8 150. Kb4 Kf8
12 -2.19 138208 0:00.25 146. Rxe7 Kxe7 147. Kb6 Qd5 148. Kb5 Qf4 149. Ka4 Kd7 150. Ka5 Ke8 151. Kb4 Kf8
11 -2.17 68963 0:00.11 146. Rxe7 Kxe7 147. Kc7 Qd5 148. Kb6 Qc4 149. Kc7 Qd6 150. Kc6 Qd3 151. Kb6 Qe2
10 -2.19 50417 0:00.08 146. Rxe7 Kxe7 147. Kc7 Qd5 148. Kb6 Qc4 149. Kc7 Qf4 150. Kc6 Qd3
9 -2.10 23603 0:00.03 146. Rxe7 Kxe7 147. Kb6 Qd5 148. Ka6 Qd6 149. Ka5 Kf6 150. Ka6
8 -2.19 13268 0:00.02 146. Rxe7 Kxe7 147. Kb6 Qd5 148. Kc7 Qf4 149. Kb6 Qg5
7 -2.10 7523 0:00.01 146. Rxe7 Kxe7 147. Kb7 Qd5 148. Ka8 Qf4 149. Kb7
6 -2.20 1543 0:00.00 146. Rxe7 Kxe7 147. Kc7 Qf4 148. Kc8 Qd5
5 -2.10 872 0:00.00 146. Rxe7 Kxe7 147. Kb5 Qd5 148. Kb6
4 -2.19 393 0:00.00 146. Rxe7 Kxe7 147. Kb5 Qf4
3 -2.10 122 0:00.00 146. Rxe7 Kxe7 147. Kb7
2 -2.16 34 0:00.00 146. Rxe7 Kxe7

... so, not sure what it is doing, over 10 hours later, still using cpu but not making a move.

How unfortunate that the race between Sjaak and ShaMax has to end up with an adjudication against Sjaak... though at least it was behind on the board as well!

enhorning · Post by **enhorning** » Sun Mar 10, 2013 12:15 am

Final standings:

Code: Select all

Cross table, sorted by score percentage, Buchholz, SB

                              Tiyaga Nebiyu ShaMax Sjaak  Pulsar ChessV HaChu  Dabbab
 1. Tiyaga v1.0               ###### 1=1==1 1=1=1= 111111 =111=1 111111 111111 111111
                              ###### 1=1==1 111111 111111 111111 111111 111111 111111   93%  78.5 (3090.0, 2772.0)

 2. NebiyuChess_1.43          0=0==0 ###### 1==1== 111111 111111 111111 111111 111111
                              0=0==0 ###### 11111= 011111 111==1 =11111 111111 111111   83%  70.0 (3192.0, 2267.5)

 3. ShaMax 4.8S               0=0=0= 0==0== ###### 1==1=1 ==1=0= 11=1=1 1111=1 11==11
                              000000 00000= ###### 00===1 =11=1= 111=== =11=1= 111111   58%  49.0 (3444.0, 1439.5)

 4. Sjaak 524                 000000 000000 0==0=0 ###### 1011=0 =1=1=1 111=11 111111
                              000000 100000 11===0 ###### 111=11 101111 111111 111111   57%  48.0 (3456.0, 1220.0)

 5. Pulsar2009-9b             =000=0 000000 ==0=1= 0100=1 ###### =====1 =====0 11===1
                              000000 000==0 =00=0= 000=00 ###### 1==1=1 ==1==1 111111   41%  34.5 (3618.0, 936.0)

 6. ChessV_WinBoard           000000 000000 00=0=0 =0=0=0 =====0 ###### =0==11 011=1=
                              000000 =00000 000=== 010000 0==0=0 ###### ==011= 1=1111   31%  26.0 (3720.0, 638.0)

 7. HaChu 0.5beta             000000 000000 0000=0 000=00 =====1 =1==00 ###### 111111
                              000000 000000 =00=0= 000000 ==0==0 ==100= ###### 111111   30%  25.0 (3732.0, 501.8)

 8. Dabbaba 6.52 JA           000000 000000 00==00 000000 00===0 100=0= 000000 ######
                              000000 000000 000000 000000 000000 0=0000 000000 ######    6%   5.0 (3972.0, 165.8)

Congratulations to Tiyaga for a dominating performance - no losses, and only 11 draws out of 84 games!
Feeding this into bayeselo gives these ratings and LOS matrix:

Code: Select all

ResultSet-EloRating>ratings
Rank Name               Elo    +    - games score oppo. draws
   1 Tiyaga v1.0        401  103  103    84   93%   -57   13%
   2 NebiyuChess_1.43   284   84   84    84   83%   -41   17%
   3 ShaMax 4.8S         60   70   70    84   58%    -9   38%
   4 Sjaak 524           54   72   72    84   57%    -8   14%
   5 Pulsar2009-9b      -77   71   71    84   41%    11   39%
   6 ChessV_WinBoard   -152   73   73    84   31%    22   31%
   7 HaChu 0.5beta     -162   73   73    84   30%    23   24%
   8 Dabbaba 6.52 JA   -408  102  102    84    6%    58   10%
ResultSet-EloRating>los
                  Ti Ne Sh Sj Pu Ch Ha Da
Tiyaga v1.0          99 99 99100100100100
NebiyuChess_1.43   0    99 99100100100100
ShaMax 4.8S        0  0    56 99 99 99100
Sjaak 524          0  0 43    99 99 99100
Pulsar2009-9b      0  0  0  0    98 99 99
ChessV_WinBoard    0  0  0  0  1    61 99
HaChu 0.5beta      0  0  0  0  0 38    99
Dabbaba 6.52 JA    0  0  0  0  0  0  0

... which again highlights just how close Sjaak and ShaMax are to each other in strength.

I also thought it was interesting that Sjaak ended up with so many fewer draws than the programs above and below it - Pulsar and ShaMax were the two most draw-inclined programs in the tournament, whereas Sjaak did not have that many draws at all:

ShaMax 32 draws
Sjaak 12 draws
Pulsar 33 draws

As usual, if anybody wants the pgn-file with the 336 games, let me know and I'll be happy to email it to you.

Finally, next tournament (in 2-3 weeks time) will be Capablanca, as suggested, and for that tournament, I will probably use different starting setups instead of different opening moves, as a lot of different starting setups have been proposed for that set of pieces and the 10x8 board.

After that, I will most likely return to another variant with low-powered promotion for the following tournament - Makruk or Thai chess. As my understanding is that Winboard does not support the endgame counting rule, the tournament will not use these, to save me from having to manually score games (which I had to do a fair bit in this tournament, as a couple of the programs claimed the win immediately after baring the enemy King, which WB rejects as false).

hgm · Post by **hgm** » Sun Mar 10, 2013 8:34 am

I think that game should be an easy win for Black. KFFKE is generally won if there is a Ferz on the color of the E, so trading R is fatal, and the only way to avoid the trade will lose white the Elephant. With so many Ferzes to shelter behind, and the white King already pressed in a corner, this should be an easy win. (KRFKR would be draw, and I never did any 6-men EGT, but I would expect KRFFKR to be already won, let alone with 3 (if not all on the same color)

Sjaak was still planing to make the fatal trade as far as it shows its search, but perhaps the fact that the loss came within the horizon, and the need to change to a variation where the Rooks stay on the board at high depth were what caused it to get stuck.

Thanks for the tourney! I would like to have the PGN.

If you want I can implement the end-game counting rule in WB. But as the engines most likely won't know it, this might not make much sense.

enhorning · Post by **enhorning** » Sun Mar 10, 2013 8:40 am

hgm wrote:If you want I can implement the end-game counting rule in WB. But as the engines most likely won't know it, this might not make much sense.

Unless there are engines that are aware of it (are there any?), then I'd rather play without it... as in some end-games, the trick becomes trading off the losing side's last piece first when you can mate within the move limit.

Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj

Re: Chess variant tournament: Shatranj