Dev Builds » 20220209-1617

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 15. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo
ncm-dbt-01 06:58:37 584324 3990 934 1114 1942 -15.68 ± 5.25 5 546 1076 360 8 -31.96 ± 10.34
ncm-dbt-02 06:58:30 585921 4000 942 1096 1962 -13.38 ± 5.16 5 520 1104 366 5 -26.81 ± 10.19
ncm-dbt-03 06:57:22 587479 4000 919 1146 1935 -19.74 ± 5.18 9 554 1098 333 6 -39.08 ± 10.21
ncm-dbt-04 06:57:54 571138 4012 927 1167 1918 -20.81 ± 5.11 6 563 1105 329 3 -41.24 ± 10.17
ncm-dbt-05 06:58:19 583968 3998 919 1090 1989 -14.87 ± 5.19 6 531 1096 360 6 -29.79 ± 10.23
20000 4641 5613 9746 -16.9 ± 2.32 31 2714 5479 1748 28 -33.77 ± 4.57

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
404710 ncm-dbt-04 570709 12 2 3 7 -29.01 ± 131.8 0 3 1 2 0 -58.45 ± 350.58
404709 ncm-dbt-01 584411 490 114 134 242 -14.19 ± 14.63 1 62 139 42 1 -28.43 ± 28.65
404708 ncm-dbt-02 589155 500 122 143 235 -14.6 ± 14.78 0 70 131 49 0 -29.25 ± 29.78
404707 ncm-dbt-05 586646 498 112 133 253 -14.66 ± 14.71 0 69 132 48 0 -29.37 ± 29.64
404706 ncm-dbt-04 571833 500 120 158 222 -26.45 ± 14.39 2 71 140 37 0 -50.38 ± 28.55
404705 ncm-dbt-03 588600 500 116 134 250 -12.51 ± 14.36 0 64 141 44 1 -26.46 ± 28.48
404704 ncm-dbt-01 583740 500 121 140 239 -13.21 ± 14.15 0 63 144 42 1 -27.85 ± 28.07
404703 ncm-dbt-02 586816 500 106 144 250 -26.46 ± 13.99 0 73 143 33 1 -54.65 ± 28.11
404702 ncm-dbt-05 588217 500 110 136 254 -18.08 ± 13.75 0 65 146 39 0 -36.26 ± 27.78
404701 ncm-dbt-04 571712 500 110 151 239 -28.56 ± 14.41 0 78 136 35 1 -58.93 ± 29.06
404700 ncm-dbt-03 587962 500 120 144 236 -16.69 ± 14.44 1 65 142 41 1 -33.46 ± 28.33
404699 ncm-dbt-01 586858 500 117 141 242 -16.69 ± 14.94 0 73 128 49 0 -33.46 ± 30.16
404698 ncm-dbt-05 586097 500 127 134 239 -4.86 ± 16.4 2 68 117 61 2 -9.73 ± 31.5
404697 ncm-dbt-02 586393 500 120 127 253 -4.86 ± 14.98 1 61 132 56 0 -8.34 ± 29.66
404696 ncm-dbt-04 569470 500 115 141 244 -18.08 ± 14.67 1 68 138 42 1 -36.26 ± 28.86
404695 ncm-dbt-03 586520 500 122 140 238 -12.51 ± 15.24 1 67 133 47 2 -26.46 ± 29.52
404694 ncm-dbt-01 584327 500 109 142 249 -22.97 ± 14.91 0 75 136 36 3 -50.38 ± 29.09
404693 ncm-dbt-05 582444 500 117 140 243 -15.99 ± 14.64 1 67 136 46 0 -30.65 ± 29.13
404692 ncm-dbt-02 585126 500 120 138 242 -12.51 ± 14.09 0 62 145 42 1 -26.46 ± 27.94
404691 ncm-dbt-04 571351 500 107 138 255 -21.57 ± 13.62 0 67 147 36 0 -43.3 ± 27.61
404690 ncm-dbt-03 587622 500 109 149 242 -27.85 ± 13.82 1 70 148 30 1 -56.07 ± 27.39
404689 ncm-dbt-05 580945 500 115 134 251 -13.21 ± 15.17 1 67 134 46 2 -27.85 ± 29.39
404688 ncm-dbt-02 583154 500 115 153 232 -26.45 ± 14.13 2 69 144 35 0 -50.38 ± 27.99
404687 ncm-dbt-01 585632 500 115 148 237 -22.96 ± 14.4 2 67 144 36 1 -44.72 ± 28.02
404686 ncm-dbt-04 571351 500 112 147 241 -24.36 ± 14.76 1 75 132 42 0 -47.55 ± 29.63
404685 ncm-dbt-03 588345 500 104 134 262 -20.87 ± 15.38 4 68 132 46 0 -36.26 ± 29.64
404684 ncm-dbt-01 582945 500 111 129 260 -12.51 ± 14.99 1 67 131 51 0 -23.66 ± 29.78
404683 ncm-dbt-02 584622 500 116 135 249 -13.21 ± 13.75 0 60 150 39 1 -27.85 ± 27.25
404682 ncm-dbt-05 583028 500 111 135 254 -16.69 ± 13.5 0 61 153 35 1 -34.86 ± 26.8
404681 ncm-dbt-04 571270 500 113 154 233 -28.56 ± 13.6 0 72 148 29 1 -58.93 ± 27.37
404680 ncm-dbt-03 587325 500 111 145 244 -23.66 ± 14.45 1 72 137 40 0 -46.13 ± 28.97
404679 ncm-dbt-02 586139 500 131 128 241 +2.08 ± 15.24 0 59 131 58 2 +1.39 ± 29.79
404678 ncm-dbt-01 582485 500 126 147 227 -14.6 ± 15.4 1 71 127 50 1 -29.25 ± 30.28
404677 ncm-dbt-04 569270 500 120 144 236 -16.69 ± 14.94 1 70 131 48 0 -32.05 ± 29.77
404676 ncm-dbt-05 579661 500 110 136 254 -18.08 ± 14.54 0 70 137 42 1 -37.67 ± 28.99
404675 ncm-dbt-03 585970 500 118 151 231 -22.96 ± 15.03 1 75 131 42 1 -46.13 ± 29.76
404674 ncm-dbt-04 573284 500 128 131 241 -2.08 ± 14.99 1 59 132 58 0 -2.78 ± 29.66
404673 ncm-dbt-01 584201 500 121 133 246 -8.34 ± 15.27 0 68 127 54 1 -18.08 ± 30.28
404672 ncm-dbt-02 585970 500 112 128 260 -11.12 ± 15.37 2 66 128 54 0 -19.48 ± 30.16
404671 ncm-dbt-05 584706 500 117 142 241 -17.38 ± 14.49 2 64 141 43 0 -32.05 ± 28.47
404670 ncm-dbt-03 587494 500 119 149 232 -20.87 ± 14.5 0 73 134 43 0 -41.89 ± 29.37

Commit

Commit ID b0b31558a236301aa8578030a8a3109856634fa9
Author Michael Chaly
Date 2022-02-09 16:17:00 UTC
Big search tuning Most credits for this patch should go to @candirufish. Based on his big search tuning (1M games at 20+0.1s) https://tests.stockfishchess.org/tests/view/61fc7a6ed508ec6a1c9f4b7d with some hand polishing on top of it, which includes : a) correcting trend sigmoid - for some reason original tuning resulted in it being negative. This heuristic was proven to be worth some elo for years so reversing it sign is probably some random artefact; b) remove changes to continuation history based pruning - this heuristic historically was really good at providing green STCs and then failing at LTC miserably if we tried to make it more strict, original tuning was done at short time control and thus it became more strict - which doesn't scale to longer time controls; c) remove changes to improvement - not really indended :). passed STC https://tests.stockfishchess.org/tests/view/6203526e88ae2c84271c2ee2 LLR: 2.94 (-2.94,2.94) <0.00,2.50> Total: 16840 W: 4604 L: 4363 D: 7873 Ptnml(0-2): 82, 1780, 4449, 2033, 76 passed LTC https://tests.stockfishchess.org/tests/view/620376e888ae2c84271c35d4 LLR: 2.96 (-2.94,2.94) <0.50,3.00> Total: 17232 W: 4771 L: 4542 D: 7919 Ptnml(0-2): 14, 1655, 5048, 1886, 13 closes https://github.com/official-stockfish/Stockfish/pull/3926 bench 5030992
Copyright 2011–2025 Next Chess Move LLC