On-line engine blitz tourney April

[Maarten Claessens]

Thank you for joining WaDuuttie. I seem to have missed a lot of fun and exitement, like this game against Rookie:

[pgn]
[Event "ICS rated blitz match"]
[Site "winboard.nl"]
[Date "2015.04.25"]
[Round "7"]
[White "WaDuuttie"]
[Black "Rookie"]
[Result "1/2-1/2"]
[WhiteElo "1996"]
[BlackElo "1890"]
[TimeControl "300+1"]

1. c4 e5 2. g3 d5 3. cxd5 c6 4. dxc6 {+0.00/16 18} Nxc6 5. Nc3
{+0.16/17 26} Nf6 6. Bg2 {+0.08/17 13} Bc5 7. d3 {+0.10/17 23} O-O 8. Nf3
{+0.16/16 11} h6 9. a3 {+0.30/16 12} a5 10. O-O {+0.30/16 1.9} Qe7 11. Nd2
{+0.30/15 5} Be6 12. Nc4 {+0.24/15 15} Rfd8 13. Qa4 {+0.30/17 5} Bd7 14.
Be3 {+0.40/16 1.1} Nd4 15. Qd1 {+0.44/17 20} a4 16. Bxb7 {+0.98/15 8} Rab8
17. Bg2 {+0.92/15 4} Nb3 18. Bxc5 {+1.12/16 3} Qxc5 19. Rb1 {+0.98/19 0.2}
Be6 20. Nd2 {+1.10/20 5} Qa7 21. Nxb3 {+1.06/19 4} Bxb3 22. Qe1 {+1.06/20}
Qd4 23. Rc1 {+1.02/19 8} Be6 24. Rc2 {+1.10/18 5} Bb3 25. Rd2 {+1.34/16 4}
h5 26. e3 {+1.28/16 2.9} Qb6 27. Qe2 {+1.42/15 1.8} g6 28. h3 {+1.52/14 3}
Rbc8 29. f4 {+1.72/16 2.9} exf4 30. Rxf4 {+1.92/14 2.5} Kg7 31. d4
{+2.06/13 2.7} h4 32. gxh4 {+2.46/15 12} Qd6 33. Qf2 {+2.66/15 2.6} Nh5 34.
Rg4 {+2.74/17 5} Rxc3 35. bxc3 {+3.18/17 4} Qxa3 36. Bf3 {+2.66/17 6} Nf6
37. Rf4 {+2.64/16 1.4} Qc1+ 38. Kg2 {+2.18/15 4} Rd6 39. e4 {+2.02/17 8} a3
40. e5 {+2.04/16 1.4} a2 41. Rxf6 {+2.26/18 1.4} a1=Q 42. Rxd6 {+1.74/19 6}
Qaxc3 43. Rc6 {+1.76/18 2.8} Bc4 44. Rd1 {+1.68/19 4} Q1b2 45. Rd6
{+1.54/20} Qxf2+ 46. Kxf2 {+1.36/21 1.5} Qc2+ 47. Kg3 {+1.30/19 1.9} Be2
48. Re1 {+1.46/20 2.1} Bxf3 49. Kxf3 {+0.64/23} Qf5+ 50. Ke3 {+0.68/22 3}
Qxh3+ 51. Kd2 {+0.26/22 6} Qh2+ 52. Kd3 {+1.58/18 1.7} Qxh4 53. Rb1
{+1.58/16 1.6} Qh3+ 54. Kc4 {+1.10/19 1.6} Qc8+ 55. Kd5 {+0.78/18 7} g5 56.
Rdb6 {+0.66/17 8} Qa8+ 57. Rb7 {+1.20/16 2.6} Qd8+ 58. Ke4 {+1.86/18 3} Qc8
59. R1b6 {+1.38/19 2.0} Qc2+ 60. Kd5 {+1.36/20 0.2} Qa2+ 61. Rb3
{+1.08/19 4} Kg6 62. Kd6 {+1.68/15 2.5} Qd2 63. d5 {+1.34/15 3} g4 64. Rb2
{+1.26/16 11} Qe3 65. R7b3 {+1.26/14 0.9} Qe1 66. Rb1 {+0.80/16 7} Qa5 67.
Rb8 {+0.78/16 1.5} Kf5 68. R1b5 {+1.44/14 1.1} Qa3+ 69. Kc6 {+0.36/16 2.6}
Qc1+ 70. Rc5 {+0.92/15 4} Qh1 71. Kd7 {+1.60/14 1.5} Qa1 72. Rcb5
{+1.90/16} Kxe5 73. Ke7 {+1.56/16 0.6} f5 74. d6+ {+0.88/16 2.7} Kf4 75. d7
{+0.76/16 2.4} Qg7+ 76. Kd6 {+0.10/18 0.6} Qf6+ 77. Kc7 {+0.08/20 1.0} Qc3+
78. Kb7 {+0.04/18 1.2} Qg7 79. Rd5 {+0.04/17} g3 80. Kc6 {+0.04/17} Qc3+
81. Rc5 {+3.30/14 1.2} Qf6+ 82. Kc7 {+1.62/15 2.2} Qd4 83. Rbb5
{+0.26/15 1.6} g2 84. Rxf5+ {+0.26/15 0.1} Ke4 85. Rg5 {+0.24/17 0.6} g1=Q
86. Rxg1 {+0.04/19 0.7} Qa7+ 87. Kc6 {+0.00/18 1.1} Qa8+ 88. Rb7
{+0.00/20 0.9} Qa6+ 89. Kc5 {+0.00/21 0.8} Qa5+ 90. Kc4 {+0.00/22 0.1} Qd5+
91. Kc3 {+0.00/23 1.3} Qc6+ 92. Kd2 {+0.00/24 1.1} Qh6+ 93. Kd1
{+0.18/18 1.0} Qd6+ 94. Ke1 {+0.22/16 0.7} Qe5 95. d8=R {+0.22/17 0.2} Qa1+
96. Ke2 {+0.20/19 1.2} Qa6+ 97. Kd2 {+0.20/20 1.2} Qxb7 98. Re8+
{+0.24/16 1.0} Kd5 99. Rg4 {+0.22/15 0.9} Qb2+ 100. Ke3 {+0.20/17 0.9} Qb3+
101. Kf2 {+0.20/20} Qb6+ 102. Re3 {+0.22/18 0.5} Qf6+ 103. Rf3
{+0.22/18 0.8} Qe5 104. Rd3+ {+0.22/18 0.7} Kc5 105. Re3 {+0.22/20 0.6} Qd5
106. Ke2 {+0.24/17 0.7} Qf5 107. Rc3+ {+0.24/19 0.3} Kb5 108. Rcc4
{+0.24/20 0.1} Qf7 109. Rcf4 {+0.24/20 1.7} Qa2+ 110. Kf3 {+0.24/18 0.6}
Qb3+ 111. Kf2 {+0.24/19 0.2} Qc2+ 112. Ke3 {+0.24/20 0.1} Qc5+ 113. Rd4
{+0.24/20 1.9} Ka5 114. Ke4 {+0.24/18 0.8} Qe7+ 115. Kf5 {+0.24/19 0.2} Kb6
116. Rg6+ {+0.24/17 1.0} Kc7 117. Rf4 {+0.26/14 0.6} Qe3 118. Rf6
{+0.26/16 2.4} Qg3 119. Ke6 {+0.30/15 0.8} Qb3+ 120. Ke7 {+0.30/20 0.7}
Qe3+ 121. Kf7 {+0.30/23 0.9} Qb3+ 122. Re6 {+0.30/17 0.6} Kd8 123. Kf6
{+0.26/18 0.9} Qc3+ 124. Re5 {+0.26/19 0.8} Kc7 125. Ke6 {+0.26/17 0.7} Kc6
126. Rff5 {+0.28/18 0.8} Qd4 127. Ra5 {+0.30/18 0.7} Qc4+ 128. Ke7
{+0.26/19 1.0} Qb4+ 129. Kf7 {+0.26/21 1.0} Qc4+ 130. Kf8 {+0.26/22 1.8}
Qb4+ 131. Ke8 {+0.26/20 0.6} Qb8+ 132. Kf7 {+0.22/18 0.6} Qb3+ 133. Kf6
{+0.08/21 1.0} Kc7 134. Ra7+ {+0.26/16 1.4} Kb8 135. Rf7 {+0.30/17 0.7} Qg3
136. Rg5 {+0.30/18 1.4} Qc3+ 137. Kg6 {+0.30/21 0.2} Qd4 138. Rb5+
{+0.22/18 1.2} Kc8 139. Rbf5 {+0.10/18 2.7} Qe4 140. Kh5 {+0.00/18 0.7}
Qh1+ 141. Kg5 {+0.00/20 0.2} Qg2+ 142. Kf4 {+0.00/21 0.1} Qh2+ 143. Ke4
{+0.00/20 0.6} Qh1+ 144. Ke5 {+0.00/20 0.7} Qa1+ 145. Kd6 {+0.00/22 0.8}
Qd4+ 146. Ke6 {+0.00/23} Qe4+ 147. Re5 {+0.00/22 0.9} Qc6+ 148. Ke7
{+0.00/22 0.8}
{Game drawn by the 50 move rule} 1/2-1/2
[/pgn]

At one point WaDuuttie was 3 pawns up, but it was not enough to win:
[d]2rr4/5pk1/1q3np1/8/p2P1R1P/PbN1P2P/1P1RQ1B1/6K1 b - - 0 32
Surely WaDuuttie missed some better continuations and Rookie defended well !

[mvk]

Thank you for joining WaDuuttie. I seem to have missed a lot of fun and exitement, like this game against Rookie:

At one point WaDuuttie was 3 pawns up, but it was not enough to win:
[d]2rr4/5pk1/1q3np1/8/p2P1R1P/PbN1P2P/1P1RQ1B1/6K1 b - - 0 32
Surely WaDuuttie missed some better continuations and Rookie defended well !

This was a nice game to watch. Rookie was down by just -0.5 here, so there was quite an evaluation difference. It dropped to -1.5 later on, and at that point I thought it was lost:

[D] 8/5pk1/3r1np1/4P3/3P1R1P/pbP2B1P/3R1QK1/2q5 b - - 0 40

But then it goes for the promotion and for a while black has two queens. The second isn't worth a full queen but ~1.6 p less from the tuner. So the trade with white's queen is good and from there on the EGT integration might have saved the draw. It got very lucky, because I know it misevaluates Q vs RR.

This is a new Rookie version for the first time in a very long time (v3.9). That version is now also running on FICS. It was a full version behind there (running v3.7, no Syzygy) because that computer didn't have an SSD yet for EGTs and now it does. It is not too much stronger, but it supports heuristic draws and has better Syzygy EGT integration and transposition table usage. Maybe 15 elo better overall in HGM's tourney, and a bit more on FICS.

[flok]

@aart: note that this is bikjump on an rpi. actually on the previous model. so it looks like my cluster of 8 chess programs running on the 200mhz faster new rpi is +/- 23 elo points stronger.

[hgm]

No, Aart was playing his own version. This is why I haven't joined yours (rpiBikJump). My policy is to allow each engine to participate only once. And the author has priority over others that might want to run it.

[abik]

No, Aart was playing his own version. This is why I haven't joined yours (rpiBikJump). My policy is to allow each engine to participate only once. And the author has priority over others that might want to run it.

Thanks for clarying H.G.

Folkert, sorry for stopping your BikJump experiment. With less time to spent on Chess for Android, I am back to my original plan of improving BikJump; I had stopped development for a very long time, always postponing it for something else. The version that played Saturday (2.5 dev version) has a completely rewritten evaluation function, using the old search code. My own tests so far showed a minor improvement over older versions, but nothing major yet.

[flok]

No, Aart was playing his own version. This is why I haven't joined yours (rpiBikJump). My policy is to allow each engine to participate only once. And the author has priority over others that might want to run it.

Thanks for clarying H.G.

Folkert, sorry for stopping your BikJump experiment. With less time to spent on Chess for Android, I am back to my original plan of improving BikJump; I had stopped development for a very long time, always postponing it for something else. The version that played Saturday (2.5 dev version) has a completely rewritten evaluation function, using the old search code. My own tests so far showed a minor improvement over older versions, but nothing major yet.

Not to worry! Had not expected a new version already. Very cool that you're working on it again!

Did you see this page: http://csvn.nl/index.php/nieuws/51-toer ... eiten-2015

[abik]

Did you see this page: http://csvn.nl/index.php/nieuws/51-toer ... eiten-2015

I saw it, and I am very interested! But Leiden is a lot farther nowadays then it was for me back in 1996