Dev Builds » 20210415-1045

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 7. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games Wins Losses Draws Elo
ncm-et-3 08:21:09 1964894 3351 2703 10 638 +385.24 +/- 13.46
ncm-et-4 08:20:47 1961951 3351 2700 16 635 +382.62 +/- 13.51
ncm-et-9 08:20:52 1959822 3325 2642 19 664 +371.21 +/- 13.21
ncm-et-10 08:20:46 1956716 3313 2669 9 635 +384.51 +/- 13.49
ncm-et-13 08:21:03 1962947 3349 2667 14 668 +374.27 +/- 13.16
ncm-et-15 08:20:51 1954001 3311 2682 15 614 +387.07 +/- 13.74
20000 16063 83 3854 +380.73 +/- 5.47

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
145866 ncm-et-15 2021-04-16 18:26 00:46:58 1942356 311 245 2 64 +364.4 +/- 43.11
145865 ncm-et-10 2021-04-16 18:26 00:47:00 1957705 313 252 0 61 +386.68 +/- 44.02
145864 ncm-et-9 2021-04-16 18:25 00:48:25 1967791 325 261 2 62 +378.75 +/- 43.86
145863 ncm-et-4 2021-04-16 18:22 00:51:42 1967545 351 283 1 67 +385.02 +/- 42.07
145862 ncm-et-13 2021-04-16 18:20 00:53:15 1960775 349 273 1 75 +362.64 +/- 39.62
145861 ncm-et-3 2021-04-16 18:20 00:53:19 1961232 351 271 0 80 +356.28 +/- 38.17
145860 ncm-et-15 2021-04-16 17:11 01:14:21 1970481 500 402 3 95 +379.78 +/- 35.24
145859 ncm-et-9 2021-04-16 17:09 01:14:53 1968800 500 399 4 97 +372.25 +/- 34.89
145858 ncm-et-10 2021-04-16 17:09 01:17:05 1956944 500 404 2 94 +385.59 +/- 35.4
145857 ncm-et-4 2021-04-16 17:06 01:14:39 1956187 500 407 3 90 +389.56 +/- 36.25
145856 ncm-et-3 2021-04-16 17:05 01:14:20 1964618 500 425 2 73 +431.48 +/- 40.4
145855 ncm-et-13 2021-04-16 17:05 01:14:30 1965076 500 396 4 100 +366.78 +/- 34.34
145854 ncm-et-15 2021-04-16 15:54 01:16:40 1951764 500 403 1 96 +385.59 +/- 34.96
145853 ncm-et-10 2021-04-16 15:54 01:14:19 1956214 500 398 2 100 +374.11 +/- 34.27
145852 ncm-et-9 2021-04-16 15:53 01:15:23 1957253 500 397 3 100 +370.41 +/- 34.31
145851 ncm-et-4 2021-04-16 15:52 01:14:09 1960310 500 397 2 101 +372.25 +/- 34.1
145850 ncm-et-13 2021-04-16 15:51 01:13:58 1965998 500 402 0 98 +385.59 +/- 34.52
145849 ncm-et-3 2021-04-16 15:50 01:14:36 1968622 500 404 2 94 +385.59 +/- 35.4
145848 ncm-et-9 2021-04-16 14:38 01:14:27 1959542 500 392 5 103 +357.94 +/- 33.83
145847 ncm-et-15 2021-04-16 14:37 01:16:11 1947109 500 412 2 86 +401.92 +/- 37.09
145846 ncm-et-10 2021-04-16 14:37 01:16:30 1948575 500 396 1 103 +372.25 +/- 33.69
145845 ncm-et-4 2021-04-16 14:36 01:14:52 1969395 500 402 2 96 +381.7 +/- 35.02
145844 ncm-et-3 2021-04-16 14:35 01:14:32 1958945 500 399 0 101 +379.77 +/- 33.98
145843 ncm-et-13 2021-04-16 14:34 01:15:35 1965227 500 395 4 101 +364.98 +/- 34.16
145842 ncm-et-15 2021-04-16 13:21 01:15:13 1955603 500 414 2 84 +406.2 +/- 37.55
145841 ncm-et-9 2021-04-16 13:21 01:16:46 1951167 500 400 1 99 +379.78 +/- 34.4
145840 ncm-et-4 2021-04-16 13:20 01:15:08 1963698 500 400 1 99 +379.78 +/- 34.4
145839 ncm-et-10 2021-04-16 13:20 01:16:10 1957707 500 408 0 92 +397.72 +/- 35.69
145838 ncm-et-13 2021-04-16 13:20 01:14:20 1961861 500 400 2 98 +377.87 +/- 34.64
145837 ncm-et-3 2021-04-16 13:19 01:15:19 1964038 500 401 3 96 +377.87 +/- 35.05
145836 ncm-et-10 2021-04-16 12:06 01:13:30 1962921 500 408 1 91 +395.65 +/- 35.96
145835 ncm-et-9 2021-04-16 12:05 01:15:20 1959700 500 390 2 108 +359.68 +/- 32.92
145834 ncm-et-3 2021-04-16 12:05 01:13:47 1968797 500 404 1 95 +387.57 +/- 35.16
145833 ncm-et-15 2021-04-16 12:04 01:16:43 1953134 500 408 3 89 +391.57 +/- 36.46
145832 ncm-et-13 2021-04-16 12:04 01:15:05 1959236 500 408 1 91 +395.65 +/- 35.96
145831 ncm-et-4 2021-04-16 12:04 01:16:04 1956357 500 417 1 82 +415.05 +/- 37.98
145830 ncm-et-13 2021-04-16 10:49 01:14:20 1962462 500 393 2 105 +364.98 +/- 33.41
145829 ncm-et-4 2021-04-16 10:49 01:14:13 1960165 500 394 6 100 +359.68 +/- 34.36
145828 ncm-et-10 2021-04-16 10:49 01:16:12 1956949 500 403 3 94 +381.7 +/- 35.44
145827 ncm-et-9 2021-04-16 10:49 01:15:38 1954503 500 403 2 95 +383.64 +/- 35.21
145826 ncm-et-15 2021-04-16 10:49 01:14:45 1957562 500 398 2 100 +374.11 +/- 34.27
145825 ncm-et-3 2021-04-16 10:49 01:15:16 1968008 500 399 2 99 +375.98 +/- 34.46

Commit

Commit ID a7ab92ec25c91e8413630c52cfc2db6b4ecacf0c
Author dsmsgms
Date 2021-04-15 10:45:39 UTC
Use classical eval for Bishop vs Pawns NNUE evaluation is incapable of recognizing trivially drawn bishop endgames (the wrong-colored rook pawn), which are in fact ubiquitous and stock standard in chess analysis. Switching off NNUE evaluation in KBPs vs KPs endgames is a measure that stops Stockfish from trading down to a drawn version of these endings when we presumably have advantage. The patch is able to edge over master in endgame positions. Patch tested for Elo gain with the "endgame.epd" book, and verified for non-regression with our usual book (see the pull request for details). STC: LLR: 2.93 (-2.94,2.94) {-0.20,1.10} Total: 33232 W: 6655 L: 6497 D: 20080 Ptnml(0-2): 4, 2342, 11769, 2494, 7 https://tests.stockfishchess.org/tests/view/6074a52981417533789605b8 LTC: LLR: 2.93 (-2.94,2.94) {0.20,0.90} Total: 159056 W: 29799 L: 29378 D: 99879 Ptnml(0-2): 7, 9004, 61085, 9425, 7 https://tests.stockfishchess.org/tests/view/6074c39a81417533789605ca Closes https://github.com/official-stockfish/Stockfish/pull/3427 Bench: 4503918 blah
Copyright 2011–2024 Next Chess Move LLC