Dev Builds » 20210805-0647

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 06:55:22 577710 4000 494 1511 1995 -90.32 ± 5.15 37 1055 798 108 2 -186.7 ± 12.05
ncm-dbt-02 06:55:47 585438 4000 522 1486 1992 -85.41 ± 5.08 29 1023 831 117 0 -176.1 ± 11.79
ncm-dbt-03 06:54:30 584518 4000 514 1531 1955 -90.32 ± 5.08 25 1080 784 109 2 -189.46 ± 12.16
ncm-dbt-04 06:54:53 569309 4000 517 1483 2000 -85.6 ± 5.02 22 1033 837 105 3 -178.78 ± 11.74
ncm-dbt-05 06:54:30 580807 4000 531 1524 1945 -88.09 ± 5.18 34 1046 801 117 2 -181.93 ± 12.03
20000 2578 7535 9887 -87.94 ± 2.28 147 5237 4051 556 9 -182.56 ± 5.34

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
418190 ncm-dbt-01 578753 500 53 196 251 -102.22 ± 13.04 1 148 94 7 0 -223.94 ± 35.14
418189 ncm-dbt-04 569949 500 70 188 242 -83.57 ± 15.02 4 129 98 19 0 -171.02 ± 34.58
418188 ncm-dbt-02 584579 500 64 192 244 -90.96 ± 15.05 7 129 99 15 0 -183.51 ± 34.39
418187 ncm-dbt-03 585801 500 72 195 233 -87.26 ± 15.11 2 139 90 18 1 -185.33 ± 36.06
418186 ncm-dbt-05 578547 500 62 195 243 -94.69 ± 15.4 9 130 96 15 0 -189.0 ± 34.95
418185 ncm-dbt-01 578259 500 55 191 254 -96.94 ± 14.72 7 134 97 12 0 -198.34 ± 34.72
418184 ncm-dbt-04 570589 500 58 182 260 -88.0 ± 13.66 4 125 112 9 0 -181.7 ± 31.89
418183 ncm-dbt-02 585464 500 72 200 228 -90.96 ± 14.32 4 133 100 13 0 -189.0 ± 34.18
418182 ncm-dbt-03 583950 500 59 191 250 -93.95 ± 13.13 1 138 103 8 0 -202.15 ± 33.42
418181 ncm-dbt-05 584201 500 68 182 250 -80.63 ± 14.8 1 132 98 18 1 -171.02 ± 34.58
418180 ncm-dbt-01 577069 500 56 189 255 -94.69 ± 14.24 4 137 97 12 0 -198.34 ± 34.72
418179 ncm-dbt-03 584243 500 62 175 263 -79.9 ± 13.77 1 125 110 14 0 -167.53 ± 32.45
418178 ncm-dbt-05 577274 500 66 183 251 -82.83 ± 14.72 3 128 103 15 1 -172.78 ± 33.69
418177 ncm-dbt-04 568752 500 53 179 268 -89.48 ± 14.0 3 131 106 9 1 -189.0 ± 32.98
418176 ncm-dbt-02 585885 500 60 179 261 -84.3 ± 15.04 3 133 94 20 0 -174.55 ± 35.29
418175 ncm-dbt-03 583698 500 61 191 248 -92.46 ± 14.79 4 137 95 13 1 -194.57 ± 35.13
418174 ncm-dbt-01 576251 500 78 193 229 -81.36 ± 15.77 8 120 101 21 0 -158.93 ± 34.05
418173 ncm-dbt-05 582569 500 60 195 245 -96.19 ± 14.71 3 145 86 16 0 -204.07 ± 36.93
418172 ncm-dbt-02 584117 500 62 181 257 -84.3 ± 13.88 2 128 107 13 0 -176.33 ± 32.93
418171 ncm-dbt-04 569190 500 64 184 252 -85.04 ± 14.05 3 126 110 10 1 -178.11 ± 32.33
418170 ncm-dbt-03 582611 500 64 199 237 -96.19 ± 14.26 7 130 104 9 0 -196.45 ± 33.3
418169 ncm-dbt-01 578136 500 53 181 266 -90.97 ± 14.32 4 132 103 10 1 -190.85 ± 33.57
418168 ncm-dbt-05 579537 500 73 184 243 -78.43 ± 14.02 3 119 114 14 0 -160.64 ± 31.79
418167 ncm-dbt-04 565626 500 63 181 256 -83.57 ± 14.6 1 134 98 16 1 -178.11 ± 34.58
418166 ncm-dbt-02 586435 500 66 170 264 -73.34 ± 13.57 0 119 116 15 0 -153.86 ± 31.51
418165 ncm-dbt-03 584579 500 60 204 236 -102.97 ± 14.49 6 143 90 11 0 -215.85 ± 36.11
418164 ncm-dbt-01 577807 500 64 181 255 -82.83 ± 14.44 5 121 110 14 0 -167.53 ± 32.45
418163 ncm-dbt-02 584327 500 65 190 245 -88.74 ± 15.28 7 128 98 17 0 -178.11 ± 34.58
418162 ncm-dbt-04 569669 500 67 196 237 -91.71 ± 14.49 3 138 94 15 0 -192.71 ± 35.33
418161 ncm-dbt-05 580738 500 62 191 247 -91.71 ± 13.57 4 129 109 8 0 -190.85 ± 32.34
418160 ncm-dbt-01 577233 500 78 191 231 -79.9 ± 14.64 3 124 107 15 1 -165.8 ± 33.0
418159 ncm-dbt-02 585674 500 58 188 254 -92.45 ± 14.05 5 130 105 10 0 -190.85 ± 33.16
418158 ncm-dbt-05 581361 500 64 200 236 -96.94 ± 15.16 7 137 91 15 0 -198.34 ± 35.92
418157 ncm-dbt-03 584874 500 62 191 247 -91.71 ± 14.78 2 143 87 18 0 -194.57 ± 36.67
418156 ncm-dbt-04 569829 500 67 183 250 -82.1 ± 14.13 4 121 112 13 0 -167.53 ± 32.08
418155 ncm-dbt-02 587028 500 75 186 239 -78.43 ± 13.73 1 123 112 14 0 -164.07 ± 32.12
418154 ncm-dbt-03 586393 500 74 185 241 -78.43 ± 14.46 2 125 105 18 0 -162.35 ± 33.37
418153 ncm-dbt-04 570869 500 75 190 235 -81.37 ± 13.66 0 129 107 14 0 -172.78 ± 32.96
418152 ncm-dbt-01 578177 500 57 189 254 -93.95 ± 15.11 5 139 89 17 0 -194.57 ± 36.29
418151 ncm-dbt-05 582235 500 76 194 230 -83.57 ± 14.6 4 126 104 16 0 -171.02 ± 33.51

Commit

Commit ID 5cd42f6b0b30b7b3332367230aeef2efc93e5aad
Author Stefan Geschwentner
Date 2021-08-05 06:47:33 UTC
Simplify new cmh pruning thresholds by using directly a quadratic formula. This decouples also the stat bonus updates from the threshold which creates less dependencies for tuning of stat bonus parameters. Perhaps a further fine tuning of the now separated coefficients for constHist[0] and constHist[1] could give further gains. STC: LLR: 2.93 (-2.94,2.94) <-2.50,0.50> Total: 78384 W: 6134 L: 6090 D: 66160 Ptnml(0-2): 207, 5013, 28705, 5063, 204 https://tests.stockfishchess.org/tests/view/6106d235afad2da4f4ae3d4b LTC: LLR: 2.93 (-2.94,2.94) <-2.50,0.50> Total: 38176 W: 1149 L: 1095 D: 35932 Ptnml(0-2): 6, 1000, 17030, 1038, 14 https://tests.stockfishchess.org/tests/view/6107a080afad2da4f4ae3def closes https://github.com/official-stockfish/Stockfish/pull/3639 Bench: 5098146
Copyright 2011–2025 Next Chess Move LLC