Dev Builds » 20210620-2117

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 06:47:02 581883 4000 453 1560 1987 -98.73 ± 5.11 47 1102 763 87 1 -205.28 ± 12.31
ncm-dbt-02 06:48:49 584983 4000 465 1589 1946 -100.33 ± 5.05 40 1131 742 87 0 -210.89 ± 12.49
ncm-dbt-03 06:47:54 585172 4000 426 1589 1985 -104.02 ± 5.05 53 1129 746 72 0 -217.35 ± 12.44
ncm-dbt-04 06:48:41 567350 4000 436 1601 1963 -104.21 ± 5.05 40 1169 708 82 1 -221.39 ± 12.8
ncm-dbt-05 06:49:25 581418 4000 472 1607 1921 -101.37 ± 5.16 47 1135 726 90 2 -212.37 ± 12.64
20000 2252 7946 9802 -101.73 ± 2.27 227 5666 3685 418 4 -213.41 ± 5.6

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
440801 ncm-dbt-05 578177 500 59 202 239 -102.22 ± 14.02 5 142 94 9 0 -215.85 ± 35.23
440800 ncm-dbt-02 583614 500 49 194 257 -103.73 ± 13.55 4 144 95 7 0 -221.9 ± 34.93
440799 ncm-dbt-04 567165 500 48 198 254 -107.54 ± 13.4 4 148 92 6 0 -232.26 ± 35.5
440798 ncm-dbt-03 586139 500 48 180 272 -93.95 ± 13.13 4 129 112 5 0 -196.45 ± 31.58
440797 ncm-dbt-01 580655 500 63 195 242 -93.95 ± 14.07 2 141 94 13 0 -200.24 ± 35.32
440796 ncm-dbt-05 583698 500 54 200 246 -104.49 ± 14.35 5 147 87 11 0 -221.9 ± 36.76
440795 ncm-dbt-04 566888 500 60 198 242 -98.44 ± 14.74 5 142 89 14 0 -206.01 ± 36.33
440794 ncm-dbt-02 582277 500 54 191 255 -97.69 ± 14.88 4 145 85 16 0 -206.01 ± 37.14
440793 ncm-dbt-03 584285 500 46 202 252 -112.14 ± 13.73 6 150 88 6 0 -240.82 ± 36.39
440792 ncm-dbt-01 580075 500 70 201 229 -93.2 ± 14.95 8 128 101 13 0 -187.16 ± 33.99
440791 ncm-dbt-05 582819 500 60 191 249 -93.2 ± 14.06 4 134 101 11 0 -194.57 ± 33.94
440790 ncm-dbt-04 569190 500 65 202 233 -97.69 ± 15.45 7 139 89 14 1 -202.15 ± 36.32
440789 ncm-dbt-02 586689 500 59 193 248 -95.44 ± 13.94 2 142 94 12 0 -204.07 ± 35.31
440788 ncm-dbt-01 582694 500 52 192 256 -99.95 ± 14.46 5 141 94 9 1 -211.87 ± 35.26
440787 ncm-dbt-03 583447 500 51 203 246 -109.07 ± 14.68 7 149 83 11 0 -230.16 ± 37.67
440786 ncm-dbt-02 585464 500 69 199 232 -92.45 ± 14.35 7 126 107 10 0 -187.16 ± 32.81
440785 ncm-dbt-05 579207 500 67 202 231 -96.19 ± 14.41 3 143 90 14 0 -204.07 ± 36.13
440784 ncm-dbt-04 568792 500 59 197 244 -98.44 ± 14.29 4 142 92 12 0 -207.95 ± 35.71
440783 ncm-dbt-01 584453 500 54 205 241 -108.3 ± 15.43 12 138 89 11 0 -217.85 ± 36.33
440782 ncm-dbt-03 585253 500 67 190 243 -87.26 ± 14.83 5 129 100 16 0 -178.11 ± 34.22
440781 ncm-dbt-02 584874 500 50 205 245 -111.37 ± 14.22 6 152 83 9 0 -238.66 ± 37.66
440780 ncm-dbt-05 581818 500 53 195 252 -101.46 ± 14.32 7 137 97 9 0 -209.91 ± 34.63
440779 ncm-dbt-04 567047 500 53 196 251 -102.22 ± 13.86 4 144 93 9 0 -217.85 ± 35.44
440778 ncm-dbt-01 580862 500 66 198 236 -93.95 ± 14.67 5 136 95 14 0 -194.57 ± 35.13
440777 ncm-dbt-03 585000 500 52 190 258 -98.44 ± 14.29 6 136 98 10 0 -204.07 ± 34.47
440776 ncm-dbt-02 584117 500 58 203 239 -103.73 ± 14.19 5 145 90 10 0 -219.87 ± 36.1
440775 ncm-dbt-05 583740 500 49 202 249 -109.83 ± 14.37 6 151 83 10 0 -234.38 ± 37.67
440774 ncm-dbt-04 565823 500 55 205 240 -107.54 ± 14.52 6 149 84 11 0 -228.08 ± 37.44
440773 ncm-dbt-01 581902 500 50 190 260 -99.95 ± 13.52 4 139 100 7 0 -211.87 ± 33.93
440772 ncm-dbt-03 587070 500 44 195 261 -108.3 ± 14.05 5 150 86 9 0 -232.26 ± 36.96
440771 ncm-dbt-02 585000 500 52 203 245 -108.3 ± 14.52 6 150 83 11 0 -230.16 ± 37.67
440770 ncm-dbt-05 577684 500 70 210 220 -99.95 ± 16.18 11 134 91 12 2 -202.15 ± 35.92
440769 ncm-dbt-01 582444 500 50 191 259 -100.7 ± 13.85 5 139 98 8 0 -211.87 ± 34.38
440768 ncm-dbt-04 567007 500 49 208 243 -114.45 ± 14.38 8 151 83 8 0 -243.0 ± 37.64
440767 ncm-dbt-03 583154 500 50 211 239 -116.0 ± 15.16 13 143 86 8 0 -236.51 ± 36.93
440766 ncm-dbt-04 566888 500 47 197 256 -107.54 ± 13.4 2 154 86 8 0 -236.51 ± 36.93
440765 ncm-dbt-02 587834 500 74 201 225 -90.22 ± 14.46 6 127 105 12 0 -183.51 ± 33.24
440764 ncm-dbt-05 584201 500 60 205 235 -103.73 ± 14.95 6 147 83 14 0 -217.85 ± 37.63
440763 ncm-dbt-01 581985 500 48 188 264 -99.95 ± 14.61 6 140 92 12 0 -207.95 ± 35.71
440762 ncm-dbt-03 587028 500 68 218 214 -107.54 ± 14.05 7 143 93 7 0 -226.0 ± 35.35

Commit

Commit ID 70ac5ecbb62b68382073b2367644945c1327ddfc
Author Stéphane Nicolet
Date 2021-06-20 21:17:07 UTC
Keep more pawns and pieces when attacking This patch increase the weight of pawns and pieces from 28 to 32 in the scaling formula we apply to the output of the NNUE pure eval. Increasing this gradient for pawns and pieces means that Stockfish will try a little harder to keep material when she has the advantage, and try a little bit harder to escape into an endgame when she is under pressure. STC: LLR: 2.93 (-2.94,2.94) <-0.50,2.50> Total: 53168 W: 4371 L: 4177 D: 44620 Ptnml(0-2): 160, 3389, 19283, 3601, 151 https://tests.stockfishchess.org/tests/view/60cefd1d457376eb8bcab7ab LTC: LLR: 2.94 (-2.94,2.94) <0.50,3.50> Total: 10888 W: 386 L: 288 D: 10214 Ptnml(0-2): 3, 260, 4821, 356, 4 https://tests.stockfishchess.org/tests/view/60cf709d2114332881e7352b closes https://github.com/official-stockfish/Stockfish/pull/3571 Bench: 4965430
Copyright 2011–2025 Next Chess Move LLC