Dev Builds » 20211222-0718

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 06:53:10 582936 4010 898 1180 1932 -24.47 ± 5.16 3 605 1072 321 4 -49.37 ± 10.36
ncm-dbt-02 06:54:21 585982 3998 876 1190 1932 -27.34 ± 5.14 5 613 1075 303 3 -54.67 ± 10.33
ncm-dbt-03 06:52:56 584615 4000 840 1176 1984 -29.25 ± 5.19 7 632 1052 308 1 -57.86 ± 10.47
ncm-dbt-04 06:53:57 567937 4014 860 1177 1977 -27.5 ± 5.19 8 619 1065 312 3 -54.45 ± 10.4
ncm-dbt-05 06:52:45 581143 3978 888 1197 1893 -27.04 ± 5.39 10 636 999 341 3 -53.16 ± 10.77
20000 4362 5920 9718 -27.12 ± 2.33 33 3105 5263 1585 14 -53.9 ± 4.68

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
434365 ncm-dbt-01 582152 10 2 4 4 -70.41 ± 134.07 0 3 1 1 0 -147.19 ± 653.91 ↓
434364 ncm-dbt-04 568951 14 4 3 7 +24.82 ± 45.76 0 0 6 1 0 +49.92 ± 94.39 ↓
434363 ncm-dbt-05 582277 478 110 137 231 -19.65 ± 15.66 1 71 122 44 1 -39.42 ± 30.89 ↓
434362 ncm-dbt-02 584748 498 111 151 236 -27.97 ± 15.43 0 86 117 46 0 -56.3 ± 31.55 ↓
434361 ncm-dbt-03 586435 500 107 160 233 -36.97 ± 14.7 1 86 128 35 0 -73.34 ± 30.1 ↓
434360 ncm-dbt-01 582277 500 110 162 228 -36.27 ± 14.4 0 85 133 31 1 -74.79 ± 29.41 ↓
434359 ncm-dbt-04 567958 500 106 150 244 -30.65 ± 14.55 1 79 133 37 0 -60.36 ± 29.46 ↓
434358 ncm-dbt-05 581735 500 117 153 230 -25.06 ± 15.79 3 78 121 48 0 -46.13 ± 31.04 ↓
434357 ncm-dbt-02 584411 500 104 143 253 -27.15 ± 14.95 0 82 125 43 0 -54.65 ± 30.53 ↓
434356 ncm-dbt-03 586139 500 99 132 269 -22.96 ± 14.53 0 75 133 42 0 -46.13 ± 29.5 ↓
434355 ncm-dbt-01 583866 500 108 158 234 -34.86 ± 14.31 0 84 132 34 0 -70.44 ± 29.57 ↓
434354 ncm-dbt-04 566691 500 119 144 237 -17.39 ± 14.36 1 64 146 37 2 -36.26 ± 27.78 ↓
434353 ncm-dbt-05 581028 500 118 149 233 -21.57 ± 14.68 0 75 131 44 0 -43.3 ± 29.76 ↓
434352 ncm-dbt-02 585674 500 108 158 234 -34.86 ± 14.04 1 79 139 31 0 -68.99 ± 28.6 ↓
434351 ncm-dbt-03 584076 500 100 146 254 -32.05 ± 14.52 2 77 136 35 0 -61.79 ± 29.05 ↓
434350 ncm-dbt-01 583321 500 121 136 243 -10.43 ± 14.31 0 63 139 48 0 -20.87 ± 28.75 ↓
434349 ncm-dbt-04 566533 500 108 140 252 -22.26 ± 14.21 1 69 141 39 0 -43.3 ± 28.43 ↓
434348 ncm-dbt-05 581277 500 99 161 240 -43.3 ± 15.07 1 94 122 32 1 -88.0 ± 30.9 ↓
434347 ncm-dbt-02 587792 500 108 143 249 -24.36 ± 14.37 1 72 138 39 0 -47.55 ± 28.83 ↓
434346 ncm-dbt-03 584076 500 100 141 259 -28.55 ± 15.3 3 76 131 39 1 -54.65 ± 29.74 ↓
434345 ncm-dbt-04 568752 500 111 151 238 -27.86 ± 15.12 0 83 125 41 1 -57.5 ± 30.53 ↓
434344 ncm-dbt-01 582569 500 108 130 262 -15.3 ± 14.45 0 67 139 43 1 -32.05 ± 28.73 ↓
434343 ncm-dbt-05 577110 500 107 150 243 -29.95 ± 16.34 2 90 107 51 0 -57.5 ± 32.78 ↓
434342 ncm-dbt-02 586097 500 115 154 231 -27.16 ± 14.44 0 77 136 36 1 -56.07 ± 29.07 ↓
434341 ncm-dbt-04 567958 500 102 147 251 -31.35 ± 13.94 0 78 139 33 0 -63.23 ± 28.63 ↓
434340 ncm-dbt-01 582360 500 117 155 228 -26.46 ± 14.39 0 77 134 39 0 -53.22 ± 29.35 ↓
434339 ncm-dbt-03 586435 500 102 155 243 -36.97 ± 14.7 1 86 128 35 0 -73.34 ± 30.1 ↓
434338 ncm-dbt-05 580323 500 115 154 231 -27.15 ± 13.5 0 71 147 32 0 -54.65 ± 27.55 ↓
434337 ncm-dbt-04 567442 500 103 150 247 -32.75 ± 14.95 1 84 126 39 0 -64.66 ± 30.39 ↓
434336 ncm-dbt-01 582819 500 103 139 258 -25.06 ± 14.81 0 78 131 40 1 -51.8 ± 29.75 ↓
434335 ncm-dbt-03 583614 500 102 151 247 -34.16 ± 14.52 0 85 129 36 0 -68.99 ± 29.98 ↓
434334 ncm-dbt-02 585506 500 115 141 244 -18.08 ± 13.62 1 60 154 34 1 -36.26 ± 26.65 ↓
434333 ncm-dbt-05 581028 500 102 153 245 -35.56 ± 14.22 1 81 136 32 0 -70.44 ± 29.01 ↓
434332 ncm-dbt-01 584034 500 119 146 235 -18.78 ± 14.98 2 69 133 46 0 -34.86 ± 29.51 ↓
434331 ncm-dbt-04 567561 500 108 145 247 -25.76 ± 14.98 3 72 134 41 0 -47.55 ± 29.36 ↓
434330 ncm-dbt-03 584622 500 119 148 233 -20.17 ± 14.32 0 71 137 42 0 -40.48 ± 28.98 ↓
434329 ncm-dbt-02 587749 500 101 157 242 -39.08 ± 14.44 2 83 134 31 0 -76.25 ± 29.26 ↓
434328 ncm-dbt-05 584369 500 120 140 240 -13.9 ± 16.4 2 76 113 58 1 -26.46 ± 31.99 ↓
434327 ncm-dbt-01 583028 500 110 150 240 -27.85 ± 15.0 1 79 130 39 1 -56.07 ± 29.87 ↓
434326 ncm-dbt-03 581527 500 111 143 246 -22.27 ± 14.73 0 76 130 44 0 -44.72 ± 29.89 ↓
434325 ncm-dbt-04 569589 500 99 147 254 -33.46 ± 15.61 1 90 115 44 0 -66.1 ± 31.81 ↓
434324 ncm-dbt-02 585885 500 114 143 243 -20.18 ± 14.83 0 74 132 43 1 -41.89 ± 29.63 ↓

Commit

Commit ID 0a6168089de8df15339e2c8f6604158d7434b678
Author Michael Chaly
Date 2021-12-22 07:18:35 UTC
Fall back to NNUE if classical evaluation is much lower than threshold The idea is that if classical eval returns a value much lower than the threshold of its usage it most likely means that position isn't that simple so we need the more precise NNUE evaluation. passed STC: https://tests.stockfishchess.org/tests/view/61bf3e7557a0d0f327c3c47a LLR: 2.95 (-2.94,2.94) <0.00,2.50> Total: 108072 W: 28007 L: 27604 D: 52461 Ptnml(0-2): 352, 12147, 28650, 12520, 367 passed LTC: https://tests.stockfishchess.org/tests/view/61c0581657a0d0f327c3fa0c LLR: 2.95 (-2.94,2.94) <0.50,3.00> Total: 155096 W: 40392 L: 39841 D: 74863 Ptnml(0-2): 88, 15983, 44843, 16558, 76 closes https://github.com/official-stockfish/Stockfish/pull/3869 bench 4310422
Copyright 2011–2025 Next Chess Move LLC
Feedback & Support Terms of Service Privacy Policy Desktop Version