Dev Builds » 20220419-1759

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 06:53:55 584394 4000 1091 996 1913 +8.25 ± 5.2 2 414 1071 513 0 +16.86 ± 10.38
ncm-dbt-02 06:55:53 586231 3996 1017 981 1998 +3.13 ± 5.03 1 415 1129 453 0 +6.43 ± 10.05
ncm-dbt-03 06:53:40 586142 4000 1035 1013 1952 +1.91 ± 5.19 0 452 1076 470 2 +3.47 ± 10.35
ncm-dbt-04 06:54:23 570835 4000 1018 1024 1958 -0.52 ± 5.24 4 461 1075 457 3 -0.87 ± 10.36
ncm-dbt-05 06:55:07 583217 4004 1040 1065 1899 -2.17 ± 5.06 4 441 1135 420 2 -3.99 ± 10.02
20000 5201 5079 9720 +2.12 ± 2.3 11 2183 5486 2313 7 +4.38 ± 4.58

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
404580 ncm-dbt-05 581194 4 1 1 2 -0.0 ± 296.81 0 1 0 1 0 -0.0 ± 1199.83 ↓
404579 ncm-dbt-02 583321 496 129 115 252 +9.81 ± 14.37 0 48 138 62 0 +19.63 ± 28.85 ↓
404578 ncm-dbt-01 585253 500 142 119 239 +15.99 ± 14.64 0 47 133 70 0 +32.05 ± 29.52 ↓
404577 ncm-dbt-04 571913 500 132 126 242 +4.17 ± 14.8 0 56 132 62 0 +8.34 ± 29.66 ↓
404576 ncm-dbt-03 585970 500 130 131 239 -0.69 ± 14.99 0 61 129 60 0 -1.39 ± 30.04 ↓
404575 ncm-dbt-05 585506 500 128 128 244 -0.0 ± 13.49 0 49 152 49 0 -0.0 ± 27.02 ↓
404574 ncm-dbt-02 586646 500 126 129 245 -2.08 ± 14.23 0 56 141 53 0 -4.17 ± 28.5 ↓
404573 ncm-dbt-01 584495 500 131 129 240 +1.39 ± 14.42 0 55 138 57 0 +2.78 ± 28.89 ↓
404572 ncm-dbt-04 571471 500 119 117 264 +1.39 ± 13.21 0 46 156 48 0 +2.78 ± 26.46 ↓
404571 ncm-dbt-03 585928 500 126 122 252 +2.78 ± 14.16 0 52 142 56 0 +5.56 ± 28.37 ↓
404570 ncm-dbt-05 583866 500 147 135 218 +8.34 ± 14.78 0 52 135 62 1 +15.3 ± 29.27 ↓
404569 ncm-dbt-02 587113 500 129 128 243 +0.69 ± 14.74 0 58 133 59 0 +1.39 ± 29.53 ↓
404568 ncm-dbt-01 587070 500 135 120 245 +10.43 ± 14.95 1 50 132 67 0 +22.27 ± 29.65 ↓
404567 ncm-dbt-03 587537 500 131 130 239 +0.69 ± 14.36 0 55 139 56 0 +1.39 ± 28.76 ↓
404566 ncm-dbt-04 570829 500 122 124 254 -1.39 ± 15.17 0 62 129 58 1 -4.17 ± 30.04 ↓
404565 ncm-dbt-05 581028 500 140 144 216 -2.78 ± 14.8 2 55 138 55 0 -2.78 ± 28.89 ↓
404564 ncm-dbt-02 586139 500 137 125 238 +8.34 ± 14.26 1 46 143 60 0 +18.08 ± 28.23 ↓
404563 ncm-dbt-01 582527 500 141 131 228 +6.95 ± 15.03 1 53 131 65 0 +15.3 ± 29.78 ↓
404562 ncm-dbt-04 569470 500 134 131 235 +2.08 ± 14.86 1 54 137 57 1 +4.17 ± 29.02 ↓
404561 ncm-dbt-03 587537 500 127 120 253 +4.86 ± 14.73 0 54 136 59 1 +8.34 ± 29.15 ↓
404560 ncm-dbt-05 580987 500 125 118 257 +4.86 ± 14.6 0 53 138 58 1 +8.34 ± 28.89 ↓
404559 ncm-dbt-02 586985 500 120 123 257 -2.08 ± 13.83 0 53 147 50 0 -4.17 ± 27.7 ↓
404558 ncm-dbt-01 583992 500 144 125 231 +13.21 ± 14.8 0 50 131 69 0 +26.46 ± 29.78 ↓
404557 ncm-dbt-04 570509 500 131 141 228 -6.95 ± 15.4 2 63 128 57 0 -11.12 ± 30.16 ↓
404556 ncm-dbt-03 583028 500 129 135 236 -4.17 ± 14.54 0 60 136 54 0 -8.34 ± 29.15 ↓
404555 ncm-dbt-05 585759 500 130 134 236 -2.78 ± 13.89 0 54 146 50 0 -5.56 ± 27.84 ↓
404554 ncm-dbt-01 583028 500 134 122 244 +8.34 ± 14.39 0 50 138 62 0 +16.69 ± 28.89 ↓
404553 ncm-dbt-02 585169 500 131 120 249 +7.64 ± 13.67 0 45 149 56 0 +15.3 ± 27.42 ↓
404552 ncm-dbt-03 585885 500 134 121 245 +9.03 ± 14.71 0 51 136 62 1 +16.69 ± 29.14 ↓
404551 ncm-dbt-04 570188 500 128 137 235 -6.25 ± 14.98 0 65 129 56 0 -12.51 ± 30.04 ↓
404550 ncm-dbt-05 582485 500 119 138 243 -13.21 ± 13.48 0 59 151 40 0 -26.46 ± 27.12 ↓
404549 ncm-dbt-01 586097 500 135 127 238 +5.56 ± 14.28 0 51 140 59 0 +11.12 ± 28.63 ↓
404548 ncm-dbt-02 587664 500 126 118 256 +5.56 ± 14.79 0 55 132 63 0 +11.12 ± 29.66 ↓
404547 ncm-dbt-04 571150 500 129 136 235 -4.86 ± 14.86 1 60 134 55 0 -8.34 ± 29.41 ↓
404546 ncm-dbt-05 581652 500 116 133 251 -11.82 ± 13.9 1 58 148 43 0 -22.27 ± 27.54 ↓
404545 ncm-dbt-03 587409 500 119 122 259 -2.08 ± 15.72 0 68 117 65 0 -4.17 ± 31.5 ↓
404544 ncm-dbt-01 582694 500 129 123 248 +4.17 ± 15.05 0 58 128 64 0 +8.34 ± 30.16 ↓
404543 ncm-dbt-04 571150 500 123 112 265 +7.64 ± 15.09 0 55 130 64 1 +13.9 ± 29.91 ↓
404542 ncm-dbt-05 586477 500 134 134 232 0.0 ± 15.3 1 60 127 62 0 +1.39 ± 30.29 ↓
404541 ncm-dbt-03 585843 500 139 132 229 +4.86 ± 14.22 0 51 141 58 0 +9.73 ± 28.5 ↓
404540 ncm-dbt-02 586816 500 119 123 258 -2.78 ± 13.89 0 54 146 50 0 -5.56 ± 27.84 ↓

Commit

Commit ID 6e0680efa0d0653686b8eb8753dc49718b95cb2b
Author Joost VandeVondele
Date 2022-04-19 17:59:04 UTC
Update default net to nn-d0b74ce1e5eb.nnue train a net using training data with a heavier weight on positions having 16 pieces on the board. More specifically, with a relative weight of `i * (32-i)/(16 * 16)+1` (where i is the number of pieces on the board). This is done with the trainer branch https://github.com/glinscott/nnue-pytorch/pull/173 The command used is: ``` python train.py $datafile $datafile $restarttype $restartfile --gpus 1 --threads 4 --num-workers 12 --random-fen-skipping=3 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --features=HalfKAv2_hm^ --lambda=1.00 --max_epochs=$epochs --seed $RANDOM --default_root_dir exp/run_$i ``` The datafile is T60T70wIsRightFarseerT60T74T75T76.binpack, the restart is from the master net. passed STC: LLR: 2.94 (-2.94,2.94) <0.00,2.50> Total: 22728 W: 6197 L: 5945 D: 10586 Ptnml(0-2): 105, 2453, 6001, 2695, 110 https://tests.stockfishchess.org/tests/view/625cf944ff677a888877cd90 passed LTC: LLR: 2.94 (-2.94,2.94) <0.50,3.00> Total: 35664 W: 9535 L: 9264 D: 16865 Ptnml(0-2): 30, 3524, 10455, 3791, 32 https://tests.stockfishchess.org/tests/view/625d3c32ff677a888877d7ca closes https://github.com/official-stockfish/Stockfish/pull/3989 Bench: 7269563
Copyright 2011–2025 Next Chess Move LLC
Feedback & Support Terms of Service Privacy Policy Desktop Version