Dev Builds » 20230703-1624

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 06:53:31 583862 4000 1372 711 1917 +57.95 ± 5.01 3 176 982 835 4 +119.11 ± 10.81
ncm-dbt-02 06:54:11 587705 4000 1321 766 1913 +48.52 ± 5.21 4 227 990 768 11 +97.69 ± 10.8
ncm-dbt-03 06:54:10 586694 4000 1329 738 1933 +51.71 ± 5.15 4 213 976 802 5 +105.63 ± 10.88
ncm-dbt-04 06:54:54 570237 4012 1345 731 1936 +53.59 ± 5.17 5 201 989 797 14 +108.15 ± 10.79
ncm-dbt-05 06:55:27 582624 3988 1343 715 1930 +55.17 ± 5.01 5 173 1011 799 6 +113.08 ± 10.63
20000 6710 3661 9629 +53.38 ± 2.29 21 990 4948 4001 40 +108.68 ± 4.82

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
378389 ncm-dbt-04 570228 12 5 1 6 +120.31 ± 75.55 0 0 2 4 0 +279.41 ± 548.26
378388 ncm-dbt-05 579744 488 160 80 248 +57.48 ± 14.47 1 19 125 97 2 +116.69 ± 30.23
378387 ncm-dbt-03 586435 500 178 106 216 +50.38 ± 13.86 2 18 136 94 0 +106.01 ± 28.73
378386 ncm-dbt-02 586604 500 167 102 231 +45.42 ± 14.55 0 30 126 93 1 +90.97 ± 30.33
378385 ncm-dbt-01 584790 500 185 99 216 +60.36 ± 14.07 0 20 126 102 2 +121.46 ± 30.15
378384 ncm-dbt-04 569829 500 160 87 253 +51.09 ± 14.98 0 32 113 105 0 +104.49 ± 32.14
378383 ncm-dbt-05 582110 500 162 108 230 +37.67 ± 14.61 0 33 132 83 2 +73.34 ± 29.56
378382 ncm-dbt-03 586520 500 173 90 237 +58.21 ± 14.66 1 24 116 109 0 +121.46 ± 31.7
378381 ncm-dbt-02 588472 500 166 87 247 +55.36 ± 14.39 1 23 122 104 0 +115.23 ± 30.81
378380 ncm-dbt-01 584076 500 167 105 228 +43.3 ± 14.29 0 30 128 92 0 +88.0 ± 30.06
378379 ncm-dbt-04 568156 500 169 99 232 +48.96 ± 14.87 1 27 125 95 2 +98.44 ± 30.45
378378 ncm-dbt-03 586774 500 143 99 258 +30.65 ± 13.75 0 32 142 76 0 +61.79 ± 28.21
378377 ncm-dbt-05 583992 500 168 82 250 +60.36 ± 12.9 1 11 139 99 0 +126.18 ± 27.93
378376 ncm-dbt-02 585169 500 171 99 230 +50.38 ± 14.94 0 30 120 98 2 +99.95 ± 31.16
378375 ncm-dbt-01 583824 500 170 88 242 +57.5 ± 13.51 1 16 133 100 0 +119.89 ± 29.04
378374 ncm-dbt-04 570308 500 176 88 236 +61.79 ± 13.99 0 18 129 100 3 +123.02 ± 29.65
378373 ncm-dbt-05 582235 500 177 92 231 +59.64 ± 14.59 1 23 116 110 0 +124.6 ± 31.69
378372 ncm-dbt-03 587325 500 165 91 244 +51.8 ± 13.66 0 22 132 96 0 +106.01 ± 29.35
378371 ncm-dbt-02 589454 500 169 102 229 +46.84 ± 14.22 0 28 127 95 0 +95.44 ± 30.17
378370 ncm-dbt-04 570589 500 168 84 248 +58.93 ± 14.69 0 23 124 99 4 +115.23 ± 30.51
378369 ncm-dbt-01 585000 500 178 86 236 +64.67 ± 14.39 1 20 115 114 0 +135.76 ± 31.8
378368 ncm-dbt-05 584117 500 174 87 239 +61.07 ± 14.38 0 23 118 108 1 +124.6 ± 31.38
378367 ncm-dbt-03 582569 500 171 89 240 +57.5 ± 15.54 0 32 106 110 2 +115.23 ± 33.15
378366 ncm-dbt-02 587367 500 172 103 225 +48.25 ± 15.72 2 31 115 100 2 +98.44 ± 31.86
378365 ncm-dbt-01 583363 500 173 85 242 +61.79 ± 15.34 1 27 106 115 1 +127.76 ± 33.19
378364 ncm-dbt-04 572195 500 158 85 257 +51.09 ± 14.98 1 26 125 95 3 +101.46 ± 30.43
378363 ncm-dbt-05 582360 500 163 88 249 +52.51 ± 14.11 0 25 125 100 0 +107.54 ± 30.4
378362 ncm-dbt-03 587155 500 163 84 253 +55.36 ± 15.06 0 31 109 110 0 +113.68 ± 32.72
378361 ncm-dbt-02 586223 500 151 100 249 +35.56 ± 14.61 1 30 139 77 3 +68.99 ± 28.6
378360 ncm-dbt-04 571310 500 176 98 226 +54.65 ± 15.02 1 28 113 108 0 +113.68 ± 32.15
378359 ncm-dbt-01 583028 500 158 71 271 +61.07 ± 13.38 0 17 129 104 0 +126.18 ± 29.62
378358 ncm-dbt-03 588260 500 178 91 231 +61.08 ± 14.24 1 20 120 109 0 +127.76 ± 31.05
378357 ncm-dbt-02 588984 500 155 93 252 +43.3 ± 14.56 0 32 124 94 0 +88.0 ± 30.62
378356 ncm-dbt-01 584748 500 180 95 225 +59.64 ± 14.46 0 24 118 107 1 +121.46 ± 31.39
378355 ncm-dbt-04 567205 500 171 87 242 +58.93 ± 14.15 1 19 126 103 1 +121.46 ± 30.15
378354 ncm-dbt-05 582027 500 175 86 239 +62.51 ± 14.44 1 20 119 109 1 +129.35 ± 31.2
378353 ncm-dbt-02 589368 500 170 80 250 +63.23 ± 14.75 0 23 117 107 3 +126.18 ± 31.53
378352 ncm-dbt-05 584411 500 164 92 244 +50.38 ± 13.58 1 19 137 93 0 +104.49 ± 28.59
378351 ncm-dbt-03 588515 500 158 88 254 +48.96 ± 15.5 0 34 115 98 3 +95.44 ± 31.86
378350 ncm-dbt-01 582068 500 161 82 257 +55.36 ± 13.84 0 22 127 101 0 +113.68 ± 30.06
378349 ncm-dbt-04 572316 500 162 102 236 +41.89 ± 14.48 1 28 132 88 1 +85.04 ± 29.5

Commit

Commit ID fa143922aecaab6f22fe818a5ef23b6ac42fe307
Author peregrineshahin
Date 2023-07-03 16:24:41 UTC
Fix pruning to (in TB loss) in Null move pruning. Current logic can apply Null move pruning on a dead-lost position returning an unproven loss (i.e. in TB loss score or mated in losing score) on nonPv nodes. on a default bench, this can be observed by adding this debugging line: ``` if (nullValue >= beta) { // Do not return unproven mate or TB scores nullValue = std::min(nullValue, VALUE_TB_WIN_IN_MAX_PLY-1); dbg_hit_on(nullValue <= VALUE_TB_LOSS_IN_MAX_PLY); // Hit #0: Total 73983 Hits 1 Hit Rate (%) 0.00135166 if (thisThread->nmpMinPly || depth < 14) return nullValue; ``` This fixes this very rare issue (happens at ~0.00135166% of the time) by eliminating the need to try Null Move Pruning with dead-lost positions and leaving it to be determined by a normal searching flow. The previous try to fix was not as safe enough because it was capping the returned value to (out of TB range) thus reviving the dead-lost position based on an artificial clamp (i.e. the in TB score/mate score can be lost on that nonPv node): https://tests.stockfishchess.org/tests/view/649756d5dc7002ce609cd794 Final fix: Passed STC: https://tests.stockfishchess.org/tests/view/649a5446dc7002ce609d1049 LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 577280 W: 153613 L: 153965 D: 269702 Ptnml(0-2): 1320, 60594, 165190, 60190, 1346 Passed LTC: https://tests.stockfishchess.org/tests/view/649cd048dc7002ce609d4801 LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 246432 W: 66769 L: 66778 D: 112885 Ptnml(0-2): 83, 22105, 78847, 22100, 81 closes https://github.com/official-stockfish/Stockfish/pull/4649 Bench: 2425978
Copyright 2011–2024 Next Chess Move LLC