2018-09-01

CCCC rounds 1, 2 and 3.....


CCCC continues running and we are on round 3 currently. The level of play is very high as it was expected and interesting games and positions arise.
Shredder is on the lead now with 3 points out of 3 games but other engines can reach that too(Komodo, Houdini).

First 2 rounds, were not without many adventures for Leela since in the first round it was winning with black against Ivanhoe, but for some reason game crashed. It was not because of any engine's fault. Leela was clearly winning and Stockfish with 7-man tablebases show a win for black so this is a won position. Of course nothing says Leela would actually win the game.

The official CCCC rules state that:
"If a neutral event occurs due to a technical error by the CCCC system or anything outside the two engines playing, the game will be restarted and replayed in it's entirety.
The original game will be adjudicated and will count if the result of the restarted game does not match the result of the original game."

Restarted here means not from the end position the original game was adjourned, but from starting position again.
The game was restarted then and it was a draw.
But according to their above rule the result of restarted game is irrelevant and the actual result will be the adjudication decision of the original game.
They took that decision and they have decided Leela would win in the winning position so they awarded a win for Leela.

[Event "CCCC"] [Site "?"] [Date "2018.09.01"] [Round "?"] [White "Ivanhoe"] [Black "Lc0 17.11089"] [Result "*"] [ECO "B52"] [PlyCount "263"] [EventDate "2018.??.??"] 1. e4 c5 2. Nf3 d6 3. Nc3 Nf6 4. Bb5+ Bd7 5. Bxd7+ Nbxd7 6. O-O e6 7. d3 Be7 8. Bf4 O-O 9. h3 Nb8 10. e5 dxe5 11. Nxe5 Nfd7 12. Nc4 Nb6 13. Ne3 Nc6 14. Re1 Qd4 15. Bh2 Qd7 16. a4 Rad8 17. a5 Nd5 18. Nc4 f6 19. Ne4 Kh8 20. c3 e5 21. Qf3 Kg8 22. h4 Rf7 23. Qh5 h6 24. Qf3 Bf8 25. Ned2 b5 26. axb6 axb6 27. Nf1 Qe6 28. Ra6 Qc8 29. Raa1 b5 30. Na5 b4 31. Nxc6 Qxc6 32. Rec1 Qe8 33. Rab1 Rfd7 34. d4 cxd4 35. cxd4 exd4 36. Qd3 Ne7 37. Bc7 Ra8 38. Nd2 Qh5 39. Bh2 Qxh4 40. Nc4 Nd5 41. Re1 Qg4 42. Nd2 Rad8 43. Nf3 Ne7 44. Re4 Qg6 45. Rc1 Nd5 46. Nh4 Qf7 47. Nf3 Nc3 48. Ree1 Na4 49. Nd2 Nc5 50. Qh3 Rd5 51. Ra1 Qh5 52. Qxh5 Rxh5 53. b3 Nd3 54. Rf1 Rc5 55. Nc4 Rc6 56. Rfd1 Nc5 57. Nd2 Ne6 58. Nc4 d3 59. Kf1 Rcc8 60. f3 Nd4 61. Rxd3 Nxb3 62. Rxb3 Rxc4 63. Bg3 h5 64. Rb2 g5 65. Rab1 Kf7 66. Be1 Rb8 67. Ke2 Bd6 68. Kd3 Rc7 69. Ke2 Rcb7 70. Rb3 h4 71. Rd3 Rb6 72. Rdb3 Kg6 73. Kf1 Ra6 74. R3b2 Kf5 75. Bd2 Rab6 76. Rb3 Rb5 77. Rd3 Bf4 78. Bxf4 Kxf4 79. Rbb3 Ke5 80. Rb2 b3 81. g3 hxg3 82. Kg2 Kf4 83. Rd4+ Ke5 84. Rd3 Kf4 85. Rd4+ Kf5 86. Rd3 R8b6 87. Kxg3 Rb4 88. Rd5+ Kg6 89. Rd3 Rb8 90. Kf2 R8b5 91. Kg2 Kf5 92. Kg3 Rb8 93. Rd5+ Ke6 94. Rd3 R8b6 95. Re2+ Kf5 96. Rb2 Rb7 97. Rd5+ Ke6 98. Rd3 Rb8 99. Re2+ Kf5 100. Rb2 R8b5 101. Rc3 Kg6 102. Kf2 Rb6 103. Kg3 Rb8 104. Kf2 R4b6 105. Rd3 Rb4 106. Rc3 Kf5 107. Kg3 R8b7 108. Rd3 Ke6 109. Re2+ Kf7 110. Rb2 Kg7 111. Kf2 R7b6 112. Ke2 f5 113. Kf2 Kf7 114. Re3 Kf6 115. Kg2 R4b5 116. Kh3 Rb7 117. Rd3 Rb4 118. Kg3 R4b6 119. Kf2 Kf7 120. Re3 Rb4 121. Re5 Kf6 122. Re3 Rb8 123. Kg2 R8b7 124. Kf2 Kg6 125. Re6+ Kf7 126. Re5 R4b5 127. Re3 R5b6 128. Re5 Kf6 129. Re3 Rb4 130. Kg2 R7b6 131. Kh3 R4b5 132. Rd3 {Here the game stopped, not because of one engine's fault but because of being something wrong with hardware/connection of the organizers. Black is clearly winning here. Game was adjourned and result was adjudicated as a black win.} (132. Rd3 {the game would continue according to Stockfish's moves.....} Rb8 133. Kg2 f4 134. Kf1 Kg6 135. Rc3 Kf5 136. Ke1 Re8+ 137. Kf2 Re3 138. Rc8 Rd5 139. Kg2 Rdd3 140. Rf8+ Ke6 141. Kh3 Rxf3+ 142. Kg4 Rg3+ 143. Kh5 f3 144. Re8+ Kf7 145. Rb8 Rg2 $19 {0-1})




In the second game where Leela with white faced Gull, the game reached a tablebase drawn 5 piece position but Gull refused to play. So by their rules the full point was given to Leela. Their rules for that state:
"In the event that one of the participating engines crashes, stalls, or disconnects during a game, our tournament software will automatically replay the game. The replayed game will be for exhibition only and only the original game will be scored.
At the end of the tournament, the game in question will be examined by the tournament director and the Chess.com team to determine proper adjudication, as follows:
If the engine that crashed was in a winning position, the game will be adjudicated as a draw.
If the position was a clear draw, the game will be adjudicated as a win for the crashing engine's opponent.Otherwise, the game will be adjudicated as a loss for the crashing engine."


 So since this was a drawn position it has been adjudicated as a loss for the engine that crashed, Gull.
A replay game for demonstration purposes was made where it ended in a draw.



The original Leela-Gull game where Gull crashed in the final position:
[Event "CCCC 1: Rapid Rumble (15|5) Stage 1"] [Site "Chess.com"] [Date "2018.09.01"] [Round "?"] [White "Lc0 17.11089"] [Black "Gull 3.syz"] [Result "*"] [ECO "D37"] [WhiteElo "3300"] [BlackElo "3300"] [PlyCount "257"] [EventDate "2018.??.??"] [EventType "rapid"] [TimeControl "900+5"] 1. d4 {[%eval 23,13] [%emt 0:00:28]} Nf6 {[%eval -18,23]} 2. c4 {[%eval 22,15] [%emt 0:00:00]} e6 {[%eval -23,24]} 3. Nf3 {[%eval 21,23] [%emt 0:00:00]} d5 { [%eval -23,23] [%emt 0:00:28]} 4. Nc3 {[%eval 20,23] [%emt 0:00:00]} Be7 { [%eval -24,23]} 5. Bf4 {[%eval 31,21] [%emt 0:00:18]} O-O {[%eval -24,23] [%emt 0:00:00]} 6. e3 {[%eval 32,24] [%emt 0:00:08]} Nbd7 {[%eval -24,23] [%emt 0:00:00]} 7. a3 {[%eval 35,22] [%emt 0:00:08]} c5 {[%eval -18,22] [%emt 0:00:35]} 8. cxd5 {[%eval 34,25] [%emt 0:00:00]} Nxd5 {[%eval -10,21] [%emt 0: 00:13]} 9. Nxd5 {[%eval 36,30] [%emt 0:00:00]} exd5 {[%eval -10,21] [%emt 0:00: 03]} 10. dxc5 {[%eval 36,17] [%emt 0:00:00]} Nxc5 {[%eval -10,21] [%emt 0:00: 00]} 11. Be2 {[%eval 37,21] [%emt 0:00:00]} Bf6 {[%eval -6,23] [%emt 0:00:39]} 12. Be5 {[%eval 36,31] [%emt 0:00:00]} Bxe5 {[%eval -10,24] [%emt 0:00:35]} 13. Nxe5 {[%eval 38,20] [%emt 0:00:00]} Qd6 {[%eval -10,24] [%emt 0:00:00]} 14. Nf3 {[%eval 44,12] [%emt 0:00:24]} Bg4 {[%eval -16,23] [%emt 0:00:52]} 15. O-O { [%eval 65,12] [%emt 0:00:27]} a5 {[%eval -18,23] [%emt 0:00:22]} 16. h3 { [%eval 72,12] [%emt 0:00:32]} Bxf3 {[%eval -16,24] [%emt 0:00:31]} 17. Bxf3 { [%eval 84,17] [%emt 0:00:23]} Rfd8 {[%eval -16,24] [%emt 0:00:00]} 18. Rc1 { [%eval 91,17] [%emt 0:00:23]} a4 {[%eval -16,24] [%emt 0:00:00]} 19. Rc2 { [%eval 122,18] [%emt 0:00:35]} Nb3 {[%eval -11,23] [%emt 0:00:10]} 20. Qd3 { [%eval 126,17] [%emt 0:00:09]} Ra5 {[%eval -10,23] [%emt 0:00:23]} 21. Rd1 { [%eval 130,17] [%emt 0:00:19]} g6 {[%eval -8,23] [%emt 0:00:00]} 22. Rc4 { [%eval 126,18] [%emt 0:00:07]} dxc4 {[%eval 0,23] [%emt 0:00:07]} 23. Qxd6 { [%eval 137,26] [%emt 0:00:20]} Rxd6 {[%eval -8,23] [%emt 0:00:00]} 24. Rxd6 { [%eval 126,29] [%emt 0:00:07]} Rb5 {[%eval -8,23] [%emt 0:00:00]} 25. Rd7 { [%eval 113,28] [%emt 0:00:08]} Nc5 {[%eval -8,27]} 26. Rc7 {[%eval 108,42] [%emt 0:00:00]} b6 {[%eval -8,27] [%emt 0:00:00]} 27. Bd5 {[%eval 167,22] [%emt 0:00:22]} Kh8 {[%eval -28,27] [%emt 0:00:31]} 28. Bxc4 {[%eval 162,25] [%emt 0:00:50]} Rxb2 {[%eval -12,21] [%emt 0:00:06]} 29. Rxf7 {[%eval 156,26] [%emt 0:00:01]} Ne4 {[%eval -9,23] [%emt 0:00:16]} 30. Rf4 {[%eval 177,19] [%emt 0:00:33]} Nc3 {[%eval -4,22] [%emt 0:00:08]} 31. g4 {[%eval 179,32] [%emt 0:00:34]} b5 {[%eval -15,23] [%emt 0:00:09]} 32. Rf8+ {[%eval 179,29] [%emt 0:00:00]} Kg7 {[%eval -15,23] [%emt 0:00:00]} 33. Rf7+ {[%eval 180,25] [%emt 0:00:00]} Kh8 {[%eval -15,25] [%emt 0:00:11]} 34. Rf8+ {[%eval 179,25] [%emt 0:00:00]} Kg7 35. Rf7+ {[%eval 180,22] [%emt 0:00:00]} Kh8 {[%eval 0,29] [%emt 0:00:05]} 36. Be6 {[%eval 180,23] [%emt 0:00:00]} h6 {[%eval -8,26]} 37. Rc7 {[%eval 176,25] [%emt 0:00:00]} Nd1 {[%eval 0,25] [%emt 0:00:13]} 38. Rf7 { [%eval 175,24] [%emt 0:00:00]} Nc3 {[%eval 0,28] [%emt 0:00:06]} 39. Kg2 { [%eval 174,18] [%emt 0:00:00]} Nb1 {[%eval -19,23] [%emt 0:00:31]} 40. Rf6 { [%eval 174,21] [%emt 0:00:33]} Kg7 {[%eval -3,26] [%emt 0:00:18]} 41. Rf7+ { [%eval 169,17] [%emt 0:00:20]} Kh8 {[%eval 0,34] [%emt 0:00:06]} 42. Kg3 { [%eval 172,19] [%emt 0:00:10]} Nxa3 {[%eval -4,21] [%emt 0:00:07]} 43. Rf8+ { [%eval 171,18] [%emt 0:00:14]} Kg7 {[%eval 0,28] [%emt 0:00:09]} 44. Rf7+ { [%eval 148,20] [%emt 0:00:13]} Kh8 {[%eval 0,28] [%emt 0:00:00]} 45. Ra7 { [%eval 150,20] [%emt 0:00:47]} Nc2 {[%eval -17,23]} 46. g5 {[%eval 138,34] [%emt 0:00:00]} h5 {[%eval 0,23] [%emt 0:00:07]} 47. Kf4 {[%eval 137,31] [%emt 0:00:01]} Nb4 {[%eval 0,25]} 48. Kg3 {[%eval 126,35] [%emt 0:00:00]} Nc2 { [%eval 0,25] [%emt 0:00:04]} 49. Ra8+ {[%eval 144,19] [%emt 0:00:34]} Kg7 { [%eval 0,28] [%emt 0:00:14]} 50. h4 {[%eval 141,21] [%emt 0:00:06]} a3 { [%eval 0,28] [%emt 0:00:07]} 51. Kf4 {[%eval 60,20] [%emt 0:00:32]} Ne1 { [%eval 0,25] [%emt 0:00:03]} 52. Ra7+ {[%eval 40,24] [%emt 0:00:45]} Kf8 { [%eval 0,27] [%emt 0:00:08]} 53. Ke5 {[%eval 35,27] [%emt 0:00:30]} Rxf2 { [%eval 0,25] [%emt 0:00:02]} 54. Rxa3 {[%eval 32,29] [%emt 0:00:42]} Nf3+ { [%eval 0,25] [%emt 0:00:00]} 55. Kf4 {[%eval 30,28] [%emt 0:00:32]} Nd2+ { [%eval 0,27] [%emt 0:00:08]} 56. Ke5 {[%eval 29,29] [%emt 0:00:00]} Nf3+ { [%eval 0,30] [%emt 0:00:08]} 57. Kd6 {[%eval 29,28] [%emt 0:00:00]} Nxh4 { [%eval 0,30] [%emt 0:00:00]} 58. Bd5 {[%eval 29,27] [%emt 0:00:00]} Nf5+ { [%eval 0,26] [%emt 0:00:10]} 59. Ke5 {[%eval 41,18] [%emt 0:00:18]} h4 { [%eval 0,26] [%emt 0:00:09]} 60. Ra8+ {[%eval 61,20] [%emt 0:00:18]} Ke7 { [%eval 0,24] [%emt 0:00:02]} 61. Ra7+ {[%eval 64,24] [%emt 0:00:07]} Kd8 { [%eval 0,27] [%emt 0:00:01]} 62. e4 {[%eval 63,23] [%emt 0:00:17]} Ne7 { [%eval 0,27] [%emt 0:00:00]} 63. Ra8+ {[%eval 50,28] [%emt 0:00:29]} Kc7 { [%eval 0,27] [%emt 0:00:00]} 64. Ra7+ {[%eval 49,31] [%emt 0:00:01]} Kd8 { [%eval 0,27] [%emt 0:00:00]} 65. Ke6 {[%eval 49,29] [%emt 0:00:01]} Nxd5 { [%eval 0,26] [%emt 0:00:03]} 66. exd5 {[%eval 47,28] [%emt 0:00:01]} Re2+ { [%eval 0,28] [%emt 0:00:06]} 67. Kd6 {[%eval 52,22] [%emt 0:00:15]} Ke8 { [%eval 0,28] [%emt 0:00:00]} 68. Rh7 {[%eval 35,22] [%emt 0:00:10]} Rh2 { [%eval 0,28] [%emt 0:00:00]} 69. Ke6 {[%eval 73,20] [%emt 0:00:17]} Re2+ { [%eval 0,28] [%emt 0:00:00]} 70. Kd6 {[%eval 66,23] [%emt 0:00:33]} Rh2 71. Ke6 {[%eval 63,26] [%emt 0:00:01]} Re2+ 72. Kf6 {[%eval 63,25] [%emt 0:00:01]} Rd2 73. Kxg6 {[%eval 87,17] [%emt 0:00:36]} Rxd5 74. Rxh4 {[%eval 85,19] [%emt 0: 00:05]} Rd6+ 75. Kg7 {[%eval 184,16] [%emt 0:00:26]} Rd7+ 76. Kf6 {[%eval 159, 19] [%emt 0:00:14]} Rd6+ 77. Kg7 {[%eval 143,22] [%emt 0:00:05]} Rd7+ 78. Kh8 { [%eval 144,22] [%emt 0:00:08]} Rd6 {[%eval 0,63] [%emt 0:00:00]} 79. Rf4 { [%eval 139,26] [%emt 0:00:13]} b4 {[%eval 0,63] [%emt 0:00:00]} 80. Rxb4 { [%eval 129,26] [%emt 0:00:05]} Rg6 {[%eval 0,63] [%emt 0:00:00]} 81. Rb5 { [%eval 111,27] [%emt 0:00:31]} Kf7 {[%eval 0,63] [%emt 0:00:00]} 82. Rf5+ { [%eval 100,24] [%emt 0:00:18]} Ke6 83. Ra5 {[%eval 89,21] [%emt 0:00:26]} Kf7 84. Rf5+ {[%eval 84,22] [%emt 0:00:18]} Ke6 85. Rc5 {[%eval 80,20] [%emt 0:00: 09]} Kf7 86. Rc7+ {[%eval 77,20] [%emt 0:00:07]} Kf8 87. Rc8+ {[%eval 72,20] [%emt 0:00:22]} Ke7 88. Rc7+ {[%eval 75,19] [%emt 0:00:23]} Kf8 89. Rc5 { [%eval 72,19] [%emt 0:00:07]} Kf7 {[%eval 0,63] [%emt 0:00:00]} 90. Rb5 { [%eval 67,18] [%emt 0:00:22]} Rg8+ 91. Kh7 {[%eval 65,14] [%emt 0:00:00]} Rg7+ 92. Kh6 {[%eval 63,15] [%emt 0:00:11]} Rg6+ 93. Kh5 {[%eval 62,14] [%emt 0:00: 13]} Rd6 94. g6+ {[%eval 62,10] [%emt 0:00:21]} Kg8 95. Kh6 {[%eval 62,10] [%emt 0:00:08]} Rd8 {[%eval 0,63] [%emt 0:00:00]} 96. Rb7 {[%eval 62,9] [%emt 0:00:10]} Rc8 {[%eval 0,63] [%emt 0:00:00]} 97. Rg7+ {[%eval 62,9] [%emt 0:00: 09]} Kh8 {[%eval 0,63] [%emt 0:00:00]} 98. Ra7 {[%eval 61,9] [%emt 0:00:05]} Kg8 99. Ra6 {[%eval 61,8] [%emt 0:00:05]} Rd8 {[%eval 0,63] [%emt 0:00:00]} 100. Rc6 {[%eval 60,8] [%emt 0:00:05]} Rb8 {[%eval 0,63] [%emt 0:00:00]} 101. Rd6 {[%eval 60,8] [%emt 0:00:04]} Ra8 {[%eval 0,63] [%emt 0:00:00]} 102. Rb6 { [%eval 60,8] [%emt 0:00:06]} Rc8 {[%eval 0,63] [%emt 0:00:00]} 103. Rb7 { [%eval 59,8] [%emt 0:00:04]} Rd8 104. Rg7+ {[%eval 60,8] [%emt 0:00:04]} Kh8 105. Ra7 {[%eval 58,9] [%emt 0:00:05]} Kg8 {[%eval 0,63] [%emt 0:00:00]} 106. Ra6 {[%eval 58,8] [%emt 0:00:04]} Rc8 {[%eval 0,63] [%emt 0:00:00]} 107. Rb6 { [%eval 58,8] [%emt 0:00:04]} Rd8 {[%eval 0,63] [%emt 0:00:00]} 108. Re6 { [%eval 57,8] [%emt 0:00:05]} Rc8 {[%eval 0,63] [%emt 0:00:00]} 109. Re7 { [%eval 56,8] [%emt 0:00:05]} Rd8 {[%eval 0,63] [%emt 0:00:00]} 110. Rc7 { [%eval 55,8] [%emt 0:00:04]} Rb8 111. Rc6 {[%eval 56,8] [%emt 0:00:05]} Rd8 { [%eval 0,63] [%emt 0:00:00]} 112. g7 {[%eval 55,8] [%emt 0:00:04]} Rb8 113. Rd6 {[%eval 52,9] [%emt 0:00:05]} Rc8 114. Rb6 {[%eval 51,9] [%emt 0:00:05]} Rd8 115. Re6 {[%eval 50,9] [%emt 0:00:03]} Rc8 {[%eval 0,63] [%emt 0:00:00]} 116. Rd6 {[%eval 46,10] [%emt 0:00:04]} Rb8 117. Rc6 {[%eval 46,9] [%emt 0:00:05]} Rd8 118. Rb6 {[%eval 45,9] [%emt 0:00:04]} Ra8 119. Rd6 {[%eval 45,9] [%emt 0: 00:05]} Rb8 120. Re6 {[%eval 41,9] [%emt 0:00:04]} Rc8 121. Ra6 {[%eval 36,8] [%emt 0:00:04]} Rb8 122. Rd6 {[%eval 35,9] [%emt 0:00:03]} Rc8 123. Re6 { [%eval 33,9] [%emt 0:00:05]} Rb8 124. Ra6 {[%eval 32,8] [%emt 0:00:05]} Rd8 125. Rb6 {[%eval 31,9] [%emt 0:00:05]} Ra8 126. Rg6 {[%eval 28,8] [%emt 0:00: 04]} Ra1 127. Rd6 {[%eval 20,9] [%emt 0:00:04]} Rh1+ 128. Kg6 {[%eval 16,10] [%emt 0:00:03]} Rd1 {[%eval 0,63] [%emt 0:00:00]} 129. Re6 {Here Gull crashed. The game as per CCCC rules is awarded to Leela as it is a drawn position.} *



So Leela has 2 points out of 2 games but her performance is poor so far since over the board she drew 3 games out of 4 and the win was by adjudication because she had a winning position in the first game.
Till now the rate of games is 1 game per 48 minutes. So Leela will play around 18:30 UTC and then in about 8 hours, i.e 02:00 UTC, etc.

Games can be seen from the following site: CCCC

So far we have the results:


      ROUND 1
1 Andscacs vs Xiphos 0-1
2 Wasp vs Arasan 0-1
3 Booot vs Vajolet 1/2-1/2
4 Texel vs Crafty 1-0
5 Ethereal vs Stockfish 0-1
6 Shredder vs Fire 1-0
7 Fizbo vs Senpai 1/2-1/2
8 Pedone vs Fritz 0-1
9 Gull vs Nirvana 1/2-1/2
10 Nemorino vs Houdini 0-1
11 Ivanhoe vs Lco 0-1 
12 Laser vs Komodo 0-1

         ROUND 2
13 Wasp vs Andscacs 0-1
14 Vajolet vs Xiphos 1-0
15 Arasan vs Texel 1/2-1/2
16 Stockfish vs Booot 1-0
17 Crafty vs Shredder 0-1
18 Senpai vs Ethereal 1/2-1/2
19 Fire vs Pedone 1-0
20 Nirvana vs Fizbo 1/2-1/2
21 Fritz vs Nemorino 1/2-1/2
22 Lco vs Gull 1-0
23 Houdini vs Laser 1-0
24 Komodo vs Ivanhoe 1-0

         ROUND 3
25 Andscacs vs Vajolet 1-0
26 Texel vs Wasp 1/2-1/2
27 Xiphos vs Stockfish 1/2-1/2
28 Shredder vs Arasan 1-0
29 Booot vs Senpai 1-0
30 Pedone vs Crafty 1-0
31 Ethereal vs Nirvana
32 Nemorino vs Fire
33 Fizbo vs Lco
34 Laser vs Fritz
35 Gull vs Komodo
36 Ivanhoe vs Houdini

So far the classification is:

1 Shredder 3.0/3
2 Stockfish 2.5/3
3 Andscacs 2.0/3
4 Texel 2.0/3
5 Lc0 2.0/2
5 Houdini 2.0/2
7 Komodo 2.0/2
8 Xiphos 1.5/3
9 Vajolet 1.5/3
10 Booot 1.5/3
11 Arasan 1.5/3
12 Fritz 1.5/2
13 Fizbo 1.0/2
13 Fire 1.0/2
15 Senpai 1.0/3
15 Nirvana 1.0/2
17 Pedone 1.0/3
18 Wasp 0.5/3
19 Nemorino 0.5/2
20 Gull 0.5/2
20 Ethereal 0.5/2
22 Laser 0.0/2
22 Ivanhoe 0.0/2
22 Crafty 0.0/3


Next 2 rounds:

      ROUND 4
37 Texel vs Andscacs
38 Stockfish vs Vajolet
39 Wasp vs Shredder
40 Senpai vs Xiphos
41 Arasan vs Pedone
42 Nirvana vs Booot
43 Crafty vs Nemorino
44 Lco vs Ethereal
45 Fire vs Laser
46 Komodo vs Fizbo
47 Fritz vs Ivanhoe
48 Houdini vs Gull

        ROUND 5
49 Andscacs vs Stockfish
50 Shredder vs Texel
51 Vajolet vs Senpai
52 Pedone vs Wasp
53 Xiphos vs Nirvana
54 Nemorino vs Arasan
55 Booot vs Lco
56 Laser vs Crafty
57 Ethereal vs Komodo
58 Ivanhoe vs Fire
59 Fizbo vs Houdini
60 Gull vs Fritz






4 comments:

  1. It is encouraging to see so many blog post incoming. I very much enjoy them!

    ReplyDelete
  2. I hope LC0 team watches closely as the same pattern here has emerged as in Div4 of TCEC, where Leela plays beginning and middle game much too quickly, builds up a big time advantage, which is completely squandered because she spent not enough time planning and executing her initial attack. Leads to draw-ish results such as against Gull, and arguably against Ivanhoe as well, as she showed no understanding of how to convert her advantage to a win. The SF 7-man TB result is moot, bc Leela is not SF, doesn't have TB support, and didn't have time available to analyze game (as again, all that saved up time was squandered in a pointless shuffling endgame). The likely cause: Humans forgot to tune her time mgt for the case of pondering on. This is why *human* tinkering in time mgt is against the 'zero' philosophy. Time mgt is a part of chess-playing strategy, as time controls are a part of the chess rules (just as much as 50 move rule and 3-fold repetition are). Sooner or later, Leela's NN will need to take control of whatever parameters and decisions humans are currently managing with time mgt and even MCTS search (such as Cpuct settings). Hopefully that time comes sooner rather than later.

    However, that being said, I do respect (highly) the main priority of this project to replicate the A0 results. That should remain top priority. I'm just saying that the correct solution for time management is clear (and if the CCCC contest continues with the same pattern, it seems the 'human tuning' is already at 3 strikes (yer out) with Div4 and Div3 of TCEC 13, and this first round of C4), and it would be nice if perhaps some planning for it could begin.

    ReplyDelete
  3. Re the comment "The SF 7-man TB result is moot, bc Leela is not SF, doesn't have TB support, and didn't have time available to analyze game (as again, all that saved up time was squandered in a pointless shuffling endgame)."

    It is true that Leela does not show great skill in finishing off the position vs Ivanhoe. However, the position is just so won that even despite a load of shuffling Leela doid eventuially manage to win it vs SF with TBs in a couple of tests I did. It is reasonable to assume it would have won it against Ivanhoes too.

    ReplyDelete
    Replies
    1. Okay, thanks for that info, Peter. I'll take your word for it and take back my 'moot' comment, which was based on a faulty assumption. Thanks for the research and correction!

      Delete