Dev Builds » 20181227-2038

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 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 10:22:55 1971292 4013 2273 79 1661 +213.22 ± 8.22
ncm-et-4 10:24:00 1939859 3936 2194 102 1640 +205.76 ± 8.3
ncm-et-9 10:23:03 1975284 4019 2312 96 1611 +215.54 ± 8.39
ncm-et-10 10:23:03 1969045 4002 2341 78 1583 +222.65 ± 8.45
ncm-et-13 10:22:58 1974775 4030 2291 91 1648 +212.81 ± 8.28
20000 11411 446 8143 +213.98 ± 3.72

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
64448 ncm-et-10 2018-12-31 11:46 00:00:58 1962473 2 1 0 1 +190.5 ± 634.21
64447 ncm-et-3 2018-12-31 11:44 00:02:39 1965688 13 7 1 5 +173.46 ± 173.59
64446 ncm-et-9 2018-12-31 11:43 00:03:25 1980065 19 12 0 7 +258.46 ± 140.62
64445 ncm-et-13 2018-12-31 11:41 00:04:58 1973123 30 16 0 14 +206.63 ± 90.34
64444 ncm-et-4 2018-12-31 10:37 01:09:57 1928388 436 260 8 168 +229.11 ± 26.04
64443 ncm-et-3 2018-12-31 10:27 01:15:59 1974426 500 289 8 203 +220.88 ± 23.55
64442 ncm-et-9 2018-12-31 10:26 01:16:01 1992279 500 291 13 196 +217.85 ± 24.17
64441 ncm-et-10 2018-12-31 10:26 01:18:48 1964491 500 283 9 208 +213.85 ± 23.25
64440 ncm-et-13 2018-12-31 10:24 01:16:46 1970954 500 271 20 209 +191.78 ± 23.45
64439 ncm-et-4 2018-12-31 09:17 01:18:12 1976986 500 276 11 213 +205.04 ± 22.99
64438 ncm-et-9 2018-12-31 09:08 01:17:10 1966465 500 289 12 199 +216.85 ± 23.94
64437 ncm-et-3 2018-12-31 09:08 01:18:07 1965080 500 282 7 211 +214.85 ± 22.98
64436 ncm-et-10 2018-12-31 09:06 01:18:27 1964315 500 300 10 190 +230.16 ± 24.52
64435 ncm-et-13 2018-12-31 09:06 01:16:42 1974987 500 281 7 212 +213.85 ± 22.92
64434 ncm-et-4 2018-12-31 07:57 01:19:05 1932911 500 286 13 201 +212.86 ± 23.83
64433 ncm-et-3 2018-12-31 07:49 01:17:40 1977608 500 294 15 191 +218.86 ± 24.57
64432 ncm-et-9 2018-12-31 07:48 01:18:25 1972339 500 272 10 218 +202.15 ± 22.64
64431 ncm-et-10 2018-12-31 07:48 01:17:18 1963546 500 295 8 197 +227.04 ± 23.96
64430 ncm-et-13 2018-12-31 07:47 01:18:03 1970324 500 293 8 199 +224.97 ± 23.82
64429 ncm-et-4 2018-12-31 06:39 01:16:47 1964764 500 287 13 200 +213.85 ± 23.9
64428 ncm-et-3 2018-12-31 06:31 01:17:03 1995906 500 289 13 198 +215.85 ± 24.03
64427 ncm-et-10 2018-12-31 06:30 01:16:40 1974298 500 298 12 190 +226.0 ± 24.57
64426 ncm-et-9 2018-12-31 06:28 01:19:34 1975451 500 300 13 187 +227.04 ± 24.82
64425 ncm-et-13 2018-12-31 06:28 01:17:58 1978197 500 284 7 209 +216.85 ± 23.11
64424 ncm-et-4 2018-12-31 05:20 01:17:54 1966152 500 277 15 208 +202.15 ± 23.42
64423 ncm-et-3 2018-12-31 05:12 01:18:00 1966928 500 275 9 216 +206.01 ± 22.73
64422 ncm-et-9 2018-12-31 05:11 01:15:36 1970022 500 272 12 216 +200.24 ± 22.83
64421 ncm-et-10 2018-12-31 05:11 01:18:29 1970300 500 292 11 197 +220.88 ± 24.05
64420 ncm-et-13 2018-12-31 05:09 01:17:15 1974669 500 273 9 218 +204.07 ± 22.61
64419 ncm-et-4 2018-12-31 03:58 01:21:01 1940940 500 260 16 224 +185.33 ± 22.45
64418 ncm-et-3 2018-12-31 03:53 01:17:33 1963686 500 271 10 219 +201.19 ± 22.58
64417 ncm-et-10 2018-12-31 03:53 01:16:46 1971000 500 304 12 184 +232.26 ± 25.02
64416 ncm-et-13 2018-12-31 03:52 01:16:07 1974694 500 285 18 197 +206.98 ± 24.2
64415 ncm-et-9 2018-12-31 03:52 01:17:47 1970184 500 278 15 207 +203.11 ± 23.48
64414 ncm-et-10 2018-12-31 02:35 01:17:11 1983294 500 276 10 214 +206.01 ± 22.89
64413 ncm-et-9 2018-12-31 02:34 01:17:28 1975299 500 312 12 176 +240.82 ± 25.63
64412 ncm-et-3 2018-12-31 02:34 01:18:26 1967077 500 284 7 209 +216.85 ± 23.11
64411 ncm-et-4 2018-12-31 02:34 01:23:34 1846279 500 279 15 206 +204.07 ± 23.55
64410 ncm-et-13 2018-12-31 02:33 01:18:13 1977626 500 308 13 179 +235.45 ± 25.42
64409 ncm-et-3 2018-12-31 01:15 01:17:28 1965231 500 282 9 209 +212.86 ± 23.18
64408 ncm-et-9 2018-12-31 01:15 01:17:37 1975455 500 286 9 205 +216.85 ± 23.45
64407 ncm-et-10 2018-12-31 01:15 01:18:26 1967694 500 292 6 202 +226.0 ± 23.54
64406 ncm-et-4 2018-12-31 01:15 01:17:30 1962457 500 269 11 220 +198.34 ± 22.55
64405 ncm-et-13 2018-12-31 01:15 01:16:56 1978409 500 280 9 211 +210.89 ± 23.05

Commit

Commit ID 69dc5568b3bb9a02c71bdd5bb0e87135e892ea19
Author 31m059
Date 2018-12-27 20:38:31 UTC
Always initialize and evaluate king safety Recent tests by @xoto10, @Vizvezdenec, and myself seemed to hint that Elo could be gained by expanding the number of cases where king safety is applied. Several users (@Spliffjiffer, @Vizvezdenec) have anticipated benefits specifically in evaluation of tactics. It appears that we actually do not need to restrict the cases in which we initialize and evaluate king safety at all: initializing and evaluating it in every position appears roughly Elo-neutral at STC and possibly a substantial Elo gain at LTC. Any explanation for this scaling is, at this point, conjecture. Assuming it is not due to chance, my hypothesis is that initialization of king safety in all positions is a mild slowdown, offset by an Elo gain of evaluating king safety in all positions. At STC this produces Elo gains and losses that offset each other, while at longer time control the slowdown is much less important, leaving only the Elo gain. It probably helps SF to explore king attacks much earlier in search with high numbers of enemy pieces concentrating but not essentially attacking king ring. Thanks to @xoto10 and @Vizvezdenec for helping run my LTC! Closes https://github.com/official-stockfish/Stockfish/pull/1906 STC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 35432 W: 7815 L: 7721 D: 19896 http://tests.stockfishchess.org/tests/view/5c24779d0ebc5902ba131b26 LTC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 12887 W: 2217 L: 2084 D: 8586 http://tests.stockfishchess.org/tests/view/5c25049a0ebc5902ba132586 Bench: 3163951 ------------------ How to continue from there? * Next step will be to tune all the king danger terms once more after that :-)
Copyright 2011–2024 Next Chess Move LLC