Dev Builds » 20130613-1746

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 06:01:03 1982322 2469 65 1540 864 -239.42 +/- 11.55
ncm-et-4 06:00:35 1983370 2503 63 1515 925 -230.22 +/- 11.13
ncm-et-5 01:55:09 2023037 802 19 481 302 -228.11 +/- 19.47
ncm-et-6 01:55:29 2025327 802 30 482 290 -221.69 +/- 20.0
ncm-et-7 01:55:15 2024835 792 20 501 271 -244.83 +/- 20.66
ncm-et-9 04:56:37 1977827 2032 42 1275 715 -244.53 +/- 12.66
ncm-et-10 06:00:40 1979872 2450 58 1527 865 -240.6 +/- 11.52
ncm-et-11 01:54:57 2025981 795 25 501 269 -240.14 +/- 20.78
ncm-et-12 01:55:24 2006775 791 21 487 283 -234.98 +/- 20.19
ncm-et-13 06:00:31 1989015 2475 70 1562 843 -242.36 +/- 11.71
ncm-et-14 01:55:24 2019208 805 26 483 296 -223.79 +/- 19.76
ncm-et-15 06:00:46 1985207 2471 60 1590 821 -251.43 +/- 11.86
ncm-et-16 01:55:07 2019136 813 21 515 277 -244.99 +/- 20.44
20000 520 12459 7021 -239.17 +/- 4.05

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
107543 ncm-et-10 2019-12-13 22:26 00:22:08 1959390 148 5 87 56 -216.87 +/- 45.81
107542 ncm-et-3 2019-12-13 22:25 00:23:14 1960314 158 4 88 66 -205.83 +/- 41.71
107541 ncm-et-13 2019-12-13 22:24 00:24:06 1969550 171 3 119 49 -287.01 +/- 49.33
107540 ncm-et-15 2019-12-13 22:23 00:24:39 1974203 166 3 110 53 -266.12 +/- 47.21
107539 ncm-et-9 2019-12-13 22:22 00:26:03 1964005 179 1 112 66 -251.96 +/- 41.5
107538 ncm-et-4 2019-12-13 22:22 00:26:42 1967874 178 2 105 71 -229.46 +/- 39.95
107537 ncm-et-10 2019-12-13 21:10 01:14:34 1965234 500 11 312 177 -241.91 +/- 25.53
107536 ncm-et-3 2019-12-13 21:10 01:14:08 1965227 500 10 316 174 -247.41 +/- 25.75
107535 ncm-et-13 2019-12-13 21:09 01:13:42 1969413 500 14 303 183 -229.12 +/- 25.13
107534 ncm-et-15 2019-12-13 21:09 01:13:15 1969096 500 8 318 174 -251.89 +/- 25.69
107533 ncm-et-9 2019-12-13 21:08 01:12:42 1971132 500 14 303 183 -229.12 +/- 25.13
107532 ncm-et-4 2019-12-13 21:08 01:12:38 1968314 500 19 291 190 -211.87 +/- 24.7
107531 ncm-et-10 2019-12-13 19:55 01:14:25 1959390 500 11 302 187 -231.21 +/- 24.77
107530 ncm-et-15 2019-12-13 19:54 01:13:36 1959710 500 13 323 164 -251.89 +/- 26.64
107529 ncm-et-9 2019-12-13 19:54 01:13:38 1966006 500 14 316 170 -243.0 +/- 26.15
107528 ncm-et-13 2019-12-13 19:53 01:14:26 1972191 500 14 320 166 -247.41 +/- 26.48
107527 ncm-et-4 2019-12-13 19:53 01:13:31 1965567 500 8 293 199 -224.97 +/- 23.82
107526 ncm-et-3 2019-12-13 19:53 01:15:29 1958777 500 11 303 186 -232.26 +/- 24.84
107525 ncm-et-4 2019-12-13 18:40 01:12:34 1961095 500 10 302 188 -232.26 +/- 24.67
107524 ncm-et-9 2019-12-13 18:39 01:12:59 1963239 500 11 311 178 -240.82 +/- 25.45
107523 ncm-et-10 2019-12-13 18:39 01:14:30 1957292 500 13 315 172 -243.0 +/- 25.97
107522 ncm-et-3 2019-12-13 18:39 01:12:49 1960923 500 15 308 177 -233.32 +/- 25.61
107521 ncm-et-13 2019-12-13 18:39 01:13:10 1973919 500 21 311 168 -230.16 +/- 26.37
107520 ncm-et-15 2019-12-13 18:39 01:13:49 1970957 500 11 306 183 -235.45 +/- 25.07
35524 ncm-et-12 2018-05-27 03:42 00:42:10 2021163 291 6 179 106 -237.85 +/- 33.01
35523 ncm-et-11 2018-05-27 03:42 00:42:17 2024917 295 12 186 97 -235.35 +/- 34.81
35522 ncm-et-7 2018-05-27 03:41 00:43:03 2024263 292 8 171 113 -218.97 +/- 31.95
35521 ncm-et-15 2018-05-27 03:41 00:43:32 2018557 305 8 196 101 -249.86 +/- 34.03
35520 ncm-et-10 2018-05-27 03:41 00:43:40 2017582 302 9 179 114 -221.35 +/- 31.89
35519 ncm-et-3 2018-05-27 03:41 00:43:52 2022304 311 11 191 109 -229.52 +/- 32.76
35518 ncm-et-6 2018-05-27 03:40 00:43:58 2024756 302 7 185 110 -235.13 +/- 32.43
35517 ncm-et-5 2018-05-27 03:40 00:43:57 2022956 302 9 183 110 -228.16 +/- 32.52
35516 ncm-et-14 2018-05-27 03:40 00:43:58 2018232 305 7 170 128 -207.18 +/- 29.77
35515 ncm-et-13 2018-05-27 03:40 00:44:07 2024427 304 8 195 101 -249.16 +/- 34.03
35514 ncm-et-16 2018-05-27 03:40 00:44:43 2022303 313 7 191 115 -234.31 +/- 31.69
35513 ncm-et-4 2018-05-27 03:39 00:45:21 2015396 325 11 212 102 -251.03 +/- 33.94
35512 ncm-et-9 2018-05-27 03:33 00:51:15 2024754 353 2 233 118 -272.02 +/- 31.08
35511 ncm-et-16 2018-05-27 02:28 01:10:24 2015970 500 14 324 162 -251.89 +/- 26.83
35510 ncm-et-5 2018-05-27 02:28 01:11:12 2023119 500 10 298 192 -228.08 +/- 24.38
35509 ncm-et-10 2018-05-27 02:28 01:11:23 2020348 500 9 332 159 -266.97 +/- 27.01
35508 ncm-et-4 2018-05-27 02:28 01:09:49 2021977 500 13 312 175 -239.74 +/- 25.73
35507 ncm-et-13 2018-05-27 02:28 01:11:00 2024590 500 10 314 176 -245.2 +/- 25.59
35506 ncm-et-7 2018-05-27 02:28 01:12:12 2025408 500 12 330 158 -261.07 +/- 27.16
35505 ncm-et-3 2018-05-27 02:28 01:11:31 2026391 500 14 334 152 -263.42 +/- 27.74
35504 ncm-et-12 2018-05-27 02:28 01:13:14 1992388 500 15 308 177 -233.32 +/- 25.61
35503 ncm-et-15 2018-05-27 02:28 01:11:55 2018720 500 17 337 146 -263.42 +/- 28.32
35502 ncm-et-6 2018-05-27 02:28 01:11:31 2025899 500 23 297 180 -213.85 +/- 25.45
35501 ncm-et-14 2018-05-27 02:28 01:11:26 2020185 500 19 313 168 -234.38 +/- 26.36
35500 ncm-et-11 2018-05-27 02:28 01:12:40 2027046 500 13 315 172 -243.0 +/- 25.97

Commit

Commit ID 3b8f66f8accefe86db9296fa276e4b33cdc450e2
Author Marco Costalba
Date 2013-06-13 17:46:49 UTC
Introduce Cut/All node definitions Follow Don Dailey definition of cut/all node: "If the previous node was a cut node, we consider this an ALL node. The only exception is for PV nodes which are a special case of ALL nodes. In the PVS framework, the first zero width window searched from a PV node is by our definition a CUT node and if you have to do a re-search then it is suddenly promoted to a PV nodes (as per PVS search) and only then can the cut and all nodes swap positions. In other words, these internal search failures can force the status of every node in the subtree to swap if it propagates back to the last PV nodes." http://talkchess.com/forum/viewtopic.php?topic_view=threads&p=519741&t=47577 With this definition we have an hit rate higher than 90% on: if (!PvNode && depth > 4 * ONE_PLY) dbg_hit_on_c(cutNode, (bestValue >= beta)); And an hit rate of just 28% on: if (!PvNode && depth > 4 * ONE_PLY) dbg_hit_on_c(!cutNode, (bestValue >= beta)); No functional change.
Copyright 2011–2024 Next Chess Move LLC