Dev Builds » 20181014-1815

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 08:42:21 1990282 3403 1816 82 1505 +195.3 +/- 8.61
ncm-et-4 08:42:11 1989875 3395 1796 78 1521 +193.66 +/- 8.55
ncm-et-9 08:41:30 1946054 3217 1774 173 1270 +189.77 +/- 9.55
ncm-et-10 08:42:19 1964023 3320 1741 82 1497 +190.71 +/- 8.62
ncm-et-13 08:41:43 1971029 3362 1765 79 1518 +191.54 +/- 8.55
ncm-et-15 08:42:03 1966049 3303 1756 87 1460 +193.31 +/- 8.76
20000 10648 581 8771 +192.4 +/- 3.58

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
57027 ncm-et-9 2018-10-19 01:32 00:34:35 1951462 217 119 13 85 +185.55 +/- 37.16
57026 ncm-et-15 2018-10-19 01:19 00:47:19 1982478 303 156 7 140 +187.05 +/- 28.18
57025 ncm-et-10 2018-10-19 01:16 00:50:40 1982950 320 172 7 141 +198.16 +/- 28.2
57024 ncm-et-13 2018-10-19 01:08 00:57:59 1937001 362 191 14 157 +185.77 +/- 27.0
57023 ncm-et-4 2018-10-19 01:05 01:01:07 1990349 395 204 6 185 +191.44 +/- 24.3
57022 ncm-et-3 2018-10-19 01:04 01:02:49 1993198 403 191 8 204 +170.19 +/- 22.97
57021 ncm-et-9 2018-10-19 00:08 01:22:55 1952520 500 279 19 202 +200.24 +/- 23.89
57020 ncm-et-15 2018-10-18 23:58 01:20:12 1953224 500 258 8 234 +190.85 +/- 21.61
57019 ncm-et-10 2018-10-18 23:58 01:17:18 1964226 500 261 13 226 +189.0 +/- 22.25
57018 ncm-et-13 2018-10-18 23:50 01:16:53 1991297 500 251 14 235 +179.0 +/- 21.75
57017 ncm-et-4 2018-10-18 23:46 01:18:05 1991141 500 249 9 242 +181.7 +/- 21.19
57016 ncm-et-3 2018-10-18 23:46 01:16:36 1990507 500 265 14 221 +191.78 +/- 22.58
57015 ncm-et-9 2018-10-18 22:46 01:21:09 1949028 500 275 26 199 +189.92 +/- 24.16
57014 ncm-et-15 2018-10-18 22:36 01:20:23 1941104 500 270 8 222 +202.15 +/- 22.33
57013 ncm-et-10 2018-10-18 22:36 01:20:55 1911402 500 249 13 238 +178.11 +/- 21.55
57012 ncm-et-13 2018-10-18 22:28 01:20:50 1946456 500 261 6 233 +195.51 +/- 21.59
57011 ncm-et-3 2018-10-18 22:27 01:17:55 1990507 500 262 8 230 +194.57 +/- 21.85
57010 ncm-et-4 2018-10-18 22:27 01:18:06 1988769 500 257 15 228 +183.51 +/- 22.19
57009 ncm-et-9 2018-10-18 21:24 01:20:46 1956789 500 255 37 208 +162.35 +/- 23.66
57008 ncm-et-15 2018-10-18 21:18 01:17:23 1980442 500 278 12 210 +206.01 +/- 23.21
57007 ncm-et-10 2018-10-18 21:18 01:16:57 1981381 500 277 14 209 +203.11 +/- 23.33
57006 ncm-et-13 2018-10-18 21:10 01:17:13 1988298 500 265 10 225 +195.51 +/- 22.21
57005 ncm-et-3 2018-10-18 21:10 01:16:28 1990507 500 298 13 189 +224.97 +/- 24.67
57004 ncm-et-4 2018-10-18 21:09 01:17:11 1990506 500 278 9 213 +208.93 +/- 22.92
57003 ncm-et-9 2018-10-18 20:03 01:19:03 1952520 500 311 24 165 +227.04 +/- 26.61
57002 ncm-et-15 2018-10-18 19:59 01:17:44 1982635 500 264 13 223 +191.78 +/- 22.43
57001 ncm-et-10 2018-10-18 19:59 01:17:29 1981070 500 266 15 219 +191.78 +/- 22.72
57000 ncm-et-3 2018-10-18 19:53 01:16:01 1989085 500 261 14 225 +188.08 +/- 22.34
56999 ncm-et-13 2018-10-18 19:51 01:17:29 1959610 500 287 6 207 +220.88 +/- 23.2
56998 ncm-et-4 2018-10-18 19:51 01:16:27 1988769 500 266 10 224 +196.45 +/- 22.28
56997 ncm-et-9 2018-10-18 18:40 01:21:51 1911628 500 255 40 205 +159.78 +/- 23.84
56996 ncm-et-10 2018-10-18 18:40 01:18:09 1980129 500 249 9 242 +181.7 +/- 21.19
56995 ncm-et-15 2018-10-18 18:37 01:21:14 1941081 500 255 17 228 +179.9 +/- 22.23
56994 ncm-et-3 2018-10-18 18:36 01:15:41 1989401 500 279 8 213 +210.89 +/- 22.89
56993 ncm-et-13 2018-10-18 18:35 01:15:14 1986406 500 255 10 235 +186.25 +/- 21.63
56992 ncm-et-4 2018-10-18 18:34 01:16:03 1989717 500 279 16 205 +203.11 +/- 23.63
56991 ncm-et-9 2018-10-18 17:18 01:21:11 1948432 500 280 14 206 +206.01 +/- 23.52
56990 ncm-et-15 2018-10-18 17:18 01:17:48 1981382 500 275 22 203 +193.64 +/- 23.86
56989 ncm-et-13 2018-10-18 17:18 01:16:05 1988141 500 255 19 226 +178.11 +/- 22.4
56988 ncm-et-4 2018-10-18 17:18 01:15:12 1989875 500 263 13 224 +190.85 +/- 22.37
56987 ncm-et-10 2018-10-18 17:18 01:20:51 1947007 500 267 11 222 +196.45 +/- 22.43
56986 ncm-et-3 2018-10-18 17:18 01:16:51 1988769 500 260 17 223 +184.42 +/- 22.53

Commit

Commit ID 0370077c375e313cec055c8e52c34f05ae06fdc4
Author Miguel Lahoz
Date 2018-10-14 18:15:16 UTC
Simplify evaluation of blockers_for_king Currently, we have two evaluation terms which account for pinned pieces. One is for all pinned pieces in kingDanger computation and another for just pinned pawns in ThreatByRank. We can increase the relevant bonus for kingDanger calculation and do away with the ThreatByRank, which seems to just add more complexity. STC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 113353 W: 24299 L: 24356 D: 64698 http://tests.stockfishchess.org/tests/view/5ba348c20ebc592cf2766e61 LTC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 96458 W: 15514 L: 15511 D: 65433 http://tests.stockfishchess.org/tests/view/5ba398830ebc592cf2767563 At 100k games, I thought it struggles a bit, but some related [0,4] tests attempting individual tweaks seem to fail: I tried directly tweaking ThreatByRank: http://tests.stockfishchess.org/tests/view/5ba3c6300ebc592cf276791c http://tests.stockfishchess.org/tests/view/5ba3c6190ebc592cf2767917 @Vizveznedec was also recently trying to tweak the same coeffecients for kingDanger calculation: http://tests.stockfishchess.org/tests/view/5ba2c7320ebc592cf27664b2 http://tests.stockfishchess.org/tests/view/5ba2c8220ebc592cf27664b8 http://tests.stockfishchess.org/tests/view/5ba2c7880ebc592cf27664b4 http://tests.stockfishchess.org/tests/view/5ba2c7ce0ebc592cf27664b6 Bench: 4648095
Copyright 2011–2024 Next Chess Move LLC