Dev Builds » 20230203-1918

You are viewing an old NCM Stockfish dev build test. You may find the most recent dev build tests using Stockfish 15 as the baseline here.

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 04:59:54 1207549 1672 671 164 837 +108.77 ± 7.5 0 30 279 517 10 +237.85 ± 20.42
ncm-dbt-02 04:58:45 1242578 1666 690 153 823 +116.13 ± 7.56 0 26 260 531 16 +254.99 ± 21.17
ncm-dbt-03 05:00:47 1236532 1664 680 152 832 +114.18 ± 7.82 0 32 259 522 19 +247.29 ± 21.21
ncm-dbt-04 05:00:20 1225167 1676 675 160 841 +110.32 ± 7.57 0 27 285 510 16 +238.37 ± 20.19
ncm-dbt-05 04:58:41 1244662 1648 670 144 834 +114.91 ± 7.51 0 27 255 531 11 +254.73 ± 21.38
ncm-dbt-06 05:00:54 1227381 1674 685 170 819 +110.46 ± 7.65 0 29 281 510 17 +238.11 ± 20.34
ncm-et-3 06:23:16 1307608 1650 688 151 811 +117.34 ± 7.86 0 31 246 528 20 +255.68 ± 21.77
ncm-et-4 06:24:18 1305272 1670 685 165 820 +111.9 ± 7.56 0 30 267 526 12 +245.41 ± 20.89
ncm-et-9 06:23:32 1302517 1676 688 180 808 +108.72 ± 7.73 0 32 283 506 17 +233.28 ± 20.28
ncm-et-10 06:24:15 1297774 1666 684 155 827 +114.27 ± 7.54 0 21 282 510 20 +246.88 ± 20.26
ncm-et-13 06:23:33 1313905 1674 700 189 785 +109.55 ± 7.68 0 29 286 504 18 +234.92 ± 20.16
ncm-et-15 06:23:44 1300578 1664 663 167 834 +106.8 ± 7.73 0 31 291 493 17 +227.93 ± 19.98
20000 8179 1950 9871 +111.93 ± 2.21 0 345 3274 6188 193 +242.78 ± 5.95

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
194758 ncm-dbt-05 1240878 148 65 9 74 +138.33 ± 22.59 0 2 14 58 0 +343.47 ± 96.76
194757 ncm-dbt-02 1237419 166 66 16 84 +107.99 ± 23.87 0 2 31 48 2 +229.28 ± 61.99
194756 ncm-dbt-01 1198398 172 73 26 73 +97.41 ± 23.99 0 4 32 49 1 +207.41 ± 61.3
194755 ncm-dbt-03 1236119 164 66 10 88 +123.59 ± 23.91 0 2 24 54 2 +274.55 ± 71.74
194754 ncm-dbt-06 1228972 174 72 15 87 +118.16 ± 25.5 0 4 25 55 3 +252.28 ± 69.94
194753 ncm-dbt-04 1218989 176 70 12 94 +118.93 ± 22.08 0 2 27 58 1 +268.0 ± 67.27
194752 ncm-dbt-05 1240755 500 197 48 255 +106.77 ± 13.72 0 9 86 152 3 +232.26 ± 36.96
194751 ncm-dbt-02 1246754 500 212 45 243 +120.67 ± 14.51 0 10 70 163 7 +263.42 ± 41.06
194750 ncm-dbt-01 1212330 500 205 52 243 +109.83 ± 14.37 0 9 86 148 7 +232.26 ± 36.96
194749 ncm-dbt-03 1225122 500 207 49 244 +113.68 ± 13.55 0 6 85 154 5 +247.41 ± 37.09
194748 ncm-dbt-06 1228577 500 206 46 248 +115.22 ± 13.55 0 7 80 159 4 +254.16 ± 38.36
194747 ncm-dbt-04 1224648 500 196 55 249 +100.7 ± 14.16 0 11 91 144 4 +213.85 ± 35.9
194746 ncm-dbt-01 1197939 500 203 40 257 +117.55 ± 12.46 0 4 80 165 1 +268.17 ± 38.21
194745 ncm-dbt-05 1240161 500 204 42 254 +116.77 ± 14.04 0 8 78 158 6 +254.16 ± 38.89
194744 ncm-dbt-02 1247705 500 211 49 240 +116.77 ± 13.36 0 6 80 160 4 +258.75 ± 38.32
194743 ncm-dbt-03 1247854 500 206 43 251 +117.55 ± 15.0 0 12 71 159 8 +251.89 ± 40.68
194742 ncm-dbt-06 1242675 500 201 55 244 +104.49 ± 14.35 0 9 93 141 7 +217.85 ± 35.44
194741 ncm-dbt-04 1225342 500 202 49 249 +109.83 ± 13.89 0 8 86 151 5 +236.51 ± 36.93
194740 ncm-dbt-01 1221531 500 190 46 264 +102.97 ± 14.02 0 13 81 155 1 +226.0 ± 38.11
194739 ncm-dbt-02 1238437 500 201 43 256 +113.68 ± 13.55 0 8 79 160 3 +251.89 ± 38.63
194738 ncm-dbt-05 1256854 500 204 45 251 +114.45 ± 13.38 0 8 77 163 2 +256.44 ± 39.15
194737 ncm-dbt-04 1231690 500 207 44 249 +117.55 ± 13.7 0 6 81 157 6 +256.44 ± 38.07
194736 ncm-dbt-06 1209303 500 206 54 240 +109.07 ± 13.73 0 9 83 155 3 +238.66 ± 37.66
194735 ncm-dbt-03 1237036 500 201 50 249 +108.3 ± 14.37 0 12 79 155 4 +234.38 ± 38.61
165796 ncm-et-3 1313651 150 67 16 67 +123.02 ± 27.4 0 4 18 51 2 +271.38 ± 82.61
165795 ncm-et-15 1306416 164 74 13 77 +135.73 ± 26.3 0 4 16 59 3 +306.37 ± 87.46
165794 ncm-et-4 1307249 170 73 15 82 +123.48 ± 22.45 0 2 24 58 1 +282.05 ± 71.81
165793 ncm-et-10 1297518 166 70 18 78 +112.62 ± 21.09 0 1 29 53 0 +255.59 ± 64.06
165792 ncm-et-13 1303097 174 77 15 82 +129.47 ± 22.82 0 2 23 60 2 +294.38 ± 73.58
165791 ncm-et-9 1298351 176 67 23 86 +88.74 ± 24.11 0 6 32 50 0 +190.85 ± 61.26
165790 ncm-et-3 1310345 500 206 46 248 +115.22 ± 14.69 0 11 75 157 7 +247.41 ± 39.65
165789 ncm-et-4 1298100 500 212 55 233 +112.91 ± 14.05 0 9 80 156 5 +245.2 ± 38.39
165788 ncm-et-15 1293077 500 205 51 244 +110.6 ± 14.22 0 9 84 151 6 +236.51 ± 37.42
165787 ncm-et-13 1320613 500 209 43 248 +119.89 ± 14.35 0 8 76 158 8 +258.75 ± 39.42
165786 ncm-et-9 1299449 500 211 57 232 +110.6 ± 13.73 0 7 87 151 5 +238.66 ± 36.67
165785 ncm-et-10 1302286 500 204 42 254 +116.77 ± 14.37 0 7 83 151 9 +247.41 ± 37.61
165784 ncm-et-3 1304375 500 200 36 264 +118.33 ± 14.03 0 11 67 169 3 +265.78 ± 41.9
165783 ncm-et-4 1304780 500 199 53 248 +104.49 ± 14.04 0 11 85 151 3 +226.0 ± 37.21
165782 ncm-et-15 1307977 500 195 47 258 +106.01 ± 13.39 0 5 97 143 5 +226.0 ± 34.38
165781 ncm-et-13 1308736 500 203 69 228 +95.44 ± 13.94 0 9 103 133 5 +198.34 ± 33.48
165780 ncm-et-9 1299405 500 198 57 245 +100.7 ± 14.16 0 11 91 144 4 +213.85 ± 35.9
165779 ncm-et-10 1304733 500 208 45 247 +117.55 ± 14.04 0 7 80 156 7 +254.16 ± 38.36
165778 ncm-et-3 1302064 500 215 53 232 +116.77 ± 13.88 0 5 86 151 8 +249.64 ± 36.79
165777 ncm-et-13 1323177 500 211 62 227 +106.77 ± 13.89 0 10 84 153 3 +232.26 ± 37.43
165776 ncm-et-4 1310961 500 201 42 257 +114.45 ± 13.55 0 8 78 161 3 +254.16 ± 38.89
165775 ncm-et-15 1294842 500 189 56 255 +94.69 ± 14.24 0 13 94 140 3 +200.24 ± 35.32
165774 ncm-et-9 1312866 500 212 43 245 +122.24 ± 14.33 0 8 73 161 8 +265.78 ± 40.25
165773 ncm-et-10 1286560 500 202 50 248 +109.07 ± 13.4 0 6 90 150 4 +236.51 ± 35.94

Commit

Commit ID 8d3457a9966f8c744ab7f8536be408196ccd8af9
Author pb00067
Date 2023-02-03 19:18:50 UTC
Improve excluded move logic PR consists of 2 improvements on nodes with excludeMove: 1. Remove xoring the posKey with make_key(excludedMove) Since we never call tte->save anymore with excludedMove, the unique left purpose of the xoring was to avoid a TT hit. Nevertheless on a normal bench run this produced ~25 false positives (key collisions) To avoid that we now forbid early TT cutoff's with excludeMove Maybe these accesses to TT with xored key caused useless misses in the CPU caches (L1, L2 ...) Now doing the probe with the same key as the enclosing search does, should hit the CPU cache. 2. Don't probe Tablebases with excludedMove. This can't be tested on fishtest, but it's obvious that tablebases don't deliver any information about suboptimal moves. Side note: Very surprisingly it looks like we cannot use static eval's from TT since they slightly differ over time due to changing optimism. Attempts to use static eval's from TT did loose about 13 ELO. This is something about to investigate. LTC: https://tests.stockfishchess.org/tests/view/63dc0f8de9d4cdfbe672d0c6 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 44736 W: 12046 L: 11733 D: 20957 Ptnml(0-2): 12, 4212, 13617, 4505, 22 An analogue of this passed STC & LTC see PR #4374 (thanks Dubslow for reviewing!) closes https://github.com/official-stockfish/Stockfish/pull/4380 Bench: 4758694
Copyright 2011–2024 Next Chess Move LLC