Dev Builds » 20200824-1003

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 06:29:21 583169 3994 163 1988 1843 -171.44 ± 4.33 110 1618 256 13 0 -447.74 ± 21.36
ncm-dbt-02 06:27:57 585782 3992 152 1936 1904 -167.05 ± 4.31 96 1606 280 14 0 -431.11 ± 20.42
ncm-dbt-03 06:27:10 584327 4000 185 1978 1837 -167.64 ± 4.63 129 1549 308 14 0 -415.05 ± 19.46
ncm-dbt-04 06:28:01 566133 4000 172 1972 1856 -168.4 ± 4.5 118 1578 290 14 0 -425.46 ± 20.06
ncm-dbt-05 06:28:56 582484 4014 139 2005 1870 -174.94 ± 4.33 128 1617 255 7 0 -457.48 ± 21.41
20000 811 9879 9310 -169.89 ± 1.98 581 7968 1389 62 0 -434.81 ± 9.14

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
460708 ncm-dbt-05 580240 14 0 7 7 -190.68 ± 7.94 0 7 0 0 0 -1199.83 ± 201.74
460707 ncm-dbt-02 585548 492 13 246 233 -178.81 ± 10.76 11 212 22 1 0 -516.01 ± 76.56
460706 ncm-dbt-01 584285 494 21 252 221 -176.15 ± 14.01 22 189 34 2 0 -431.67 ± 60.2
460705 ncm-dbt-04 567640 500 21 253 226 -174.55 ± 12.37 17 198 35 0 0 -449.34 ± 59.17
460704 ncm-dbt-03 582277 500 16 250 234 -176.32 ± 12.46 16 204 28 2 0 -466.04 ± 66.79
460703 ncm-dbt-05 584327 500 17 242 241 -168.4 ± 11.9 12 202 35 1 0 -438.95 ± 59.31
460702 ncm-dbt-02 586731 500 13 249 238 -178.11 ± 11.53 14 209 26 1 0 -490.71 ± 69.77
460701 ncm-dbt-03 586731 500 22 249 229 -170.15 ± 13.18 18 192 39 1 0 -419.61 ± 55.94
460700 ncm-dbt-04 566099 500 20 243 237 -166.66 ± 12.51 13 199 36 2 0 -424.28 ± 58.4
460699 ncm-dbt-01 585464 500 27 251 222 -167.53 ± 10.69 8 208 34 0 0 -454.75 ± 60.11
460698 ncm-dbt-05 580447 500 17 257 226 -181.7 ± 12.22 19 202 29 0 0 -484.24 ± 65.6
460697 ncm-dbt-02 585000 500 18 243 239 -168.4 ± 11.9 12 202 35 1 0 -438.95 ± 59.31
460696 ncm-dbt-01 583112 500 12 242 246 -172.78 ± 12.98 16 201 30 3 0 -444.09 ± 64.08
460695 ncm-dbt-03 586477 500 26 243 231 -161.49 ± 12.41 13 191 46 0 0 -397.71 ± 51.02
460694 ncm-dbt-04 560093 500 17 245 238 -171.02 ± 12.89 16 198 34 2 0 -433.94 ± 60.2
460693 ncm-dbt-05 581485 500 15 253 232 -179.9 ± 11.61 15 209 25 1 0 -497.4 ± 71.3
460692 ncm-dbt-02 585759 500 27 245 228 -162.35 ± 13.87 17 187 43 3 0 -385.59 ± 53.09
460691 ncm-dbt-01 579578 500 21 249 230 -171.02 ± 12.42 14 202 32 2 0 -444.09 ± 62.18
460690 ncm-dbt-04 566336 500 22 238 240 -160.64 ± 12.91 12 195 40 3 0 -397.72 ± 55.14
460689 ncm-dbt-03 583196 500 20 252 228 -174.55 ± 13.53 21 191 37 1 0 -429.05 ± 57.55
460688 ncm-dbt-05 580489 500 17 248 235 -173.66 ± 11.44 13 205 32 0 0 -466.03 ± 62.14
460687 ncm-dbt-02 585042 500 18 237 245 -163.21 ± 12.97 12 199 35 4 0 -410.58 ± 58.88
460686 ncm-dbt-03 585885 500 24 240 236 -160.64 ± 13.13 10 202 32 6 0 -406.2 ± 60.76
460685 ncm-dbt-04 566336 500 25 247 228 -165.79 ± 12.11 11 202 35 2 0 -429.05 ± 59.28
460684 ncm-dbt-05 579951 500 16 260 224 -185.33 ± 11.29 17 210 23 0 0 -526.7 ± 74.68
460683 ncm-dbt-01 584285 500 20 244 236 -167.53 ± 10.96 7 212 29 2 0 -460.32 ± 65.55
460682 ncm-dbt-02 586435 500 20 235 245 -159.78 ± 11.6 8 200 41 1 0 -410.58 ± 54.45
460681 ncm-dbt-03 583740 500 33 249 218 -160.64 ± 14.55 19 182 45 4 0 -370.41 ± 51.77
460680 ncm-dbt-04 567244 500 17 241 242 -167.53 ± 11.73 11 203 35 1 0 -438.95 ± 59.31
460679 ncm-dbt-05 579001 500 21 246 233 -168.4 ± 13.07 17 192 40 1 0 -415.05 ± 55.18
460678 ncm-dbt-01 582444 500 20 254 226 -176.32 ± 12.21 16 203 30 1 0 -466.04 ± 64.5
460677 ncm-dbt-02 588174 500 23 241 236 -162.35 ± 11.4 9 200 41 0 0 -419.6 ± 54.3
460676 ncm-dbt-03 584327 500 22 248 230 -169.27 ± 12.55 16 194 40 0 0 -424.27 ± 55.03
460675 ncm-dbt-05 591168 500 14 247 239 -175.43 ± 10.71 11 211 28 0 0 -490.7 ± 66.89
460674 ncm-dbt-04 566849 500 23 253 224 -172.78 ± 13.44 19 194 35 2 0 -429.05 ± 59.28
460673 ncm-dbt-01 583908 500 23 250 227 -170.15 ± 12.72 14 202 31 3 0 -438.95 ± 62.99
460672 ncm-dbt-03 581985 500 22 247 231 -168.4 ± 12.62 16 193 41 0 0 -419.6 ± 54.3
460671 ncm-dbt-02 583572 500 20 240 240 -164.07 ± 12.92 13 197 37 3 0 -410.58 ± 57.43
460670 ncm-dbt-04 568474 500 27 252 221 -168.4 ± 13.73 19 189 40 2 0 -406.2 ± 55.21
460669 ncm-dbt-05 585253 500 22 245 233 -166.66 ± 15.23 24 179 43 4 0 -377.87 ± 52.99
460668 ncm-dbt-01 582277 500 19 246 235 -170.14 ± 11.76 13 201 36 0 0 -444.08 ± 58.27

Commit

Commit ID f7b3f0e8426bbf7414d139ed9d1cfa7a98d7314d
Author Sami Kiminki
Date 2020-08-24 10:03:28 UTC
Allow TT entries with key16==0 to be fetched Fix the issue where a TT entry with key16==0 would always be reported as a miss. Instead, we'll use depth8 to detect whether the TT entry is occupied. In order to do that, we'll change DEPTH_OFFSET to -7 (depth8==0) to distinguish between an unoccupied entry and the otherwise lowest possible depth, i.e., DEPTH_NONE (depth8==1). To prevent a performance regression, we'll reorder the TT entry fields by the access order of TranspositionTable::probe(). Memory in general works fastest when accessed in sequential order. We'll also match the store order in TTEntry::save() with the entry field order, and re-order the 'if-or' expressions in TTEntry::save() from the cheapest to the most expensive. Finally, as we now have a proper TT entry occupancy test, we'll fix a minor corner case with hashfull reporting. To reproduce: - Use a big hash - Either: a. Start 31 very quick searches (this wraparounds generation to 0); or b. Force generation of the first search to 0. - go depth infinite Before the fix, hashfull would incorrectly report nearly full hash immediately after the search start, since TranspositionTable::hashfull() used to consider only the entry generation and not whether the entry was actually occupied. STC: LLR: 2.95 (-2.94,2.94) {-0.25,1.25} Total: 36848 W: 4091 L: 3898 D: 28859 Ptnml(0-2): 158, 2996, 11972, 3091, 207 https://tests.stockfishchess.org/tests/view/5f3f98d5dc02a01a0c2881f7 LTC: LLR: 2.95 (-2.94,2.94) {0.25,1.25} Total: 32280 W: 1828 L: 1653 D: 28799 Ptnml(0-2): 34, 1428, 13051, 1583, 44 https://tests.stockfishchess.org/tests/view/5f3fe77a87a5c3c63d8f5332 closes https://github.com/official-stockfish/Stockfish/pull/3048 Bench: 3760677
Copyright 2011–2026 Next Chess Move LLC