Dev Builds » 20230903-0728

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 14. 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 09:57:16 1100565 3358 1459 306 1593 +124.35 +/- 5.22 0 35 498 1104 42 +276.5 +/- 15.26
ncm-dbt-02 09:50:17 1196503 3306 1394 264 1648 +123.73 +/- 5.19 0 34 490 1094 35 +276.95 +/- 15.38
ncm-dbt-03 09:54:28 1229606 3362 1483 274 1605 +130.79 +/- 5.24 0 32 460 1137 52 +293.47 +/- 15.89
ncm-dbt-04 09:56:02 1227490 3348 1443 275 1630 +126.52 +/- 5.12 0 29 487 1119 39 +284.49 +/- 15.42
ncm-dbt-05 09:48:48 1229499 3288 1404 279 1605 +123.87 +/- 5.31 1 37 481 1086 39 +276.19 +/- 15.54
ncm-dbt-06 09:56:31 1223365 3338 1428 287 1623 +123.74 +/- 5.48 2 41 494 1078 54 +270.87 +/- 15.33
20000 8611 1685 9704 +125.51 +/- 2.15 3 208 2910 6618 261 +279.67 +/- 6.31

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
201833 ncm-dbt-05 1231963 288 122 20 146 +128.62 +/- 16.63 0 2 40 100 2 +297.57 +/- 54.78
201832 ncm-dbt-02 1196259 306 130 33 143 +114.06 +/- 17.43 0 5 48 98 2 +252.41 +/- 49.86
201831 ncm-dbt-06 1252837 338 140 32 166 +115.04 +/- 16.96 0 8 46 114 1 +259.41 +/- 50.78
201830 ncm-dbt-04 1212098 348 142 22 184 +124.92 +/- 15.6 0 3 51 117 3 +283.21 +/- 48.28
201829 ncm-dbt-01 1112270 358 151 37 170 +114.62 +/- 15.48 0 4 59 114 2 +255.13 +/- 44.7
201828 ncm-dbt-03 1227558 362 154 26 182 +128.39 +/- 14.83 0 3 49 127 2 +298.71 +/- 49.38
201827 ncm-dbt-05 1235086 500 218 40 242 +129.35 +/- 13.9 0 5 71 165 9 +285.49 +/- 40.81
201826 ncm-dbt-02 1199018 500 208 45 247 +117.55 +/- 13.53 0 4 86 153 7 +254.16 +/- 36.71
201825 ncm-dbt-04 1204260 500 224 43 233 +131.74 +/- 12.74 0 2 71 171 6 +301.33 +/- 40.6
201824 ncm-dbt-06 1238593 500 214 41 245 +125.38 +/- 13.61 0 5 74 164 7 +277.93 +/- 39.92
201823 ncm-dbt-01 1098713 500 218 45 237 +125.38 +/- 13.43 0 3 79 160 8 +275.45 +/- 38.39
201822 ncm-dbt-03 1237671 500 218 33 249 +134.95 +/- 13.22 0 5 61 178 6 +312.48 +/- 44.19
201821 ncm-dbt-05 1237800 500 216 54 230 +116.78 +/- 14.37 1 8 74 162 5 +258.75 +/- 39.96
201820 ncm-dbt-02 1194937 500 210 42 248 +121.45 +/- 12.95 0 5 75 167 3 +275.45 +/- 39.63
201819 ncm-dbt-04 1250881 500 215 36 249 +130.14 +/- 13.7 0 2 78 159 11 +282.94 +/- 38.55
201818 ncm-dbt-06 1223022 500 216 44 240 +124.6 +/- 14.96 1 5 77 155 12 +265.78 +/- 39.12
201817 ncm-dbt-03 1222525 500 228 44 228 +134.15 +/- 14.67 0 7 64 167 12 +293.29 +/- 43.07
201816 ncm-dbt-01 1114665 500 214 39 247 +126.97 +/- 12.66 0 3 73 170 4 +290.66 +/- 40.08
201815 ncm-dbt-05 1235709 500 219 47 234 +124.6 +/- 13.8 0 7 70 167 6 +277.93 +/- 41.14
201814 ncm-dbt-02 1200373 500 212 41 247 +123.81 +/- 13.1 0 3 79 162 6 +275.45 +/- 38.39
201813 ncm-dbt-04 1217649 500 209 40 251 +122.24 +/- 13.3 0 6 73 167 4 +275.45 +/- 40.24
201812 ncm-dbt-06 1204197 500 223 52 225 +123.81 +/- 14.81 0 6 80 151 13 +258.75 +/- 38.32
201811 ncm-dbt-03 1226670 500 223 32 245 +139.81 +/- 13.26 0 2 65 173 10 +318.25 +/- 42.59
201810 ncm-dbt-01 1096750 500 223 34 243 +138.18 +/- 14.06 0 5 62 172 11 +309.64 +/- 43.82
201809 ncm-dbt-05 1213895 500 216 31 253 +134.95 +/- 13.22 0 3 67 172 8 +306.84 +/- 41.99
201808 ncm-dbt-02 1192556 500 208 43 249 +119.11 +/- 14.03 0 9 72 164 5 +263.42 +/- 40.52
201807 ncm-dbt-03 1237669 500 222 48 230 +126.17 +/- 14.29 0 5 77 157 11 +270.57 +/- 39.07
201806 ncm-dbt-04 1236756 500 223 47 230 +127.76 +/- 13.02 0 4 71 170 5 +290.66 +/- 40.76
201805 ncm-dbt-06 1221409 500 207 37 256 +123.02 +/- 13.82 0 5 78 159 8 +268.17 +/- 38.8
201804 ncm-dbt-01 1101799 500 216 45 239 +123.81 +/- 13.81 0 7 71 166 6 +275.45 +/- 40.84
201803 ncm-dbt-05 1220484 500 204 46 250 +113.68 +/- 14.05 0 8 82 154 6 +245.2 +/- 37.88
201802 ncm-dbt-02 1195911 500 216 30 254 +135.76 +/- 13.01 0 4 62 178 6 +315.35 +/- 43.81
201801 ncm-dbt-03 1218631 500 215 40 245 +126.97 +/- 12.85 0 4 71 171 4 +290.66 +/- 40.76
201800 ncm-dbt-04 1220354 500 222 43 235 +130.14 +/- 13.52 0 7 62 176 5 +298.62 +/- 43.77
201799 ncm-dbt-06 1208964 500 209 39 252 +123.02 +/- 14.65 1 9 65 169 6 +275.45 +/- 42.59
201798 ncm-dbt-01 1078652 500 213 55 232 +113.68 +/- 13.72 0 7 83 155 5 +247.41 +/- 37.61
201797 ncm-dbt-05 1231562 500 209 41 250 +121.45 +/- 12.77 0 4 77 166 3 +275.45 +/- 39.01
201796 ncm-dbt-02 1196472 500 210 30 260 +130.94 +/- 13.14 0 4 68 172 6 +298.62 +/- 41.71
201795 ncm-dbt-03 1236520 500 223 51 226 +124.6 +/- 13.8 0 6 73 164 7 +275.45 +/- 40.24
201794 ncm-dbt-04 1250437 500 208 44 248 +118.33 +/- 13.35 0 5 81 159 5 +261.07 +/- 38.02
201793 ncm-dbt-06 1214539 500 219 42 239 +128.55 +/- 13.19 0 3 74 166 7 +288.06 +/- 39.79
201792 ncm-dbt-01 1101106 500 224 51 225 +125.38 +/- 13.61 0 6 71 167 6 +280.42 +/- 40.83

Commit

Commit ID b25d68f6ee2d016cc0c14b076e79e6c44fdaea2a
Author Stéphane Nicolet
Date 2023-09-03 07:28:16 UTC
Introduce simple_eval() for lazy evaluations This patch implements the pure materialistic evaluation called simple_eval() to gain a speed-up during Stockfish search. We use the so-called lazy evaluation trick: replace the accurate but slow NNUE network evaluation by the super-fast simple_eval() if the position seems to be already won (high material advantage). To guard against some of the most obvious blunders introduced by this idea, this patch uses the following features which will raise the lazy evaluation threshold in some situations: - avoid lazy evals on shuffling branches in the search tree - avoid lazy evals if the position at root already has a material imbalance - avoid lazy evals if the search value at root is already winning/losing. Moreover, we add a small random noise to the simple_eval() term. This idea (stochastic mobility in the minimax tree) was worth about 200 Elo in the pure simple_eval() player on Lichess. Overall, the current implementation in this patch evaluates about 2% of the leaves in the search tree lazily. -------------------------------------------- STC: LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 60352 W: 15585 L: 15234 D: 29533 Ptnml(0-2): 216, 6906, 15578, 7263, 213 https://tests.stockfishchess.org/tests/view/64f1d9bcbd9967ffae366209 LTC: LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 35106 W: 8990 L: 8678 D: 17438 Ptnml(0-2): 14, 3668, 9887, 3960, 24 https://tests.stockfishchess.org/tests/view/64f25204f5b0c54e3f04c0e7 verification run at VLTC: LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 74362 W: 19088 L: 18716 D: 36558 Ptnml(0-2): 6, 7226, 22348, 7592, 9 https://tests.stockfishchess.org/tests/view/64f2ecdbf5b0c54e3f04d3ae All three tests above were run with adjudication off, we also verified that there was no regression on matetracker (thanks Disservin!). ---------------------------------------------- closes https://github.com/official-stockfish/Stockfish/pull/4771 Bench: 1393714
Copyright 2011–2024 Next Chess Move LLC