Dev Builds » 20230203-1918

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 11:07:33 1954591 3295 2853 6 436 +454.81 ± 16.33
ncm-et-4 11:10:27 1955418 3332 2853 2 477 +443.62 ± 15.59
ncm-et-9 11:13:52 1955393 3343 2901 2 440 +459.18 ± 16.24
ncm-et-10 11:09:52 1955056 3322 2872 6 444 +453.03 ± 16.18
ncm-et-13 11:14:57 1962715 3386 2921 7 458 +450.16 ± 15.93
ncm-et-15 11:11:47 1946729 3322 2836 4 482 +439.58 ± 15.51
20000 17236 27 2737 +449.96 ± 6.5

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
164778 ncm-et-3 2023-02-04 09:24 00:59:42 1952531 295 251 0 44 +437.49 ± 52.39
164777 ncm-et-15 2023-02-04 09:18 01:05:32 1949940 322 278 0 44 +453.87 ± 52.45
164776 ncm-et-10 2023-02-04 09:18 01:05:44 1952825 322 281 0 41 +467.0 ± 54.48
164775 ncm-et-4 2023-02-04 09:16 01:08:17 1948758 332 291 0 41 +472.67 ± 54.49
164774 ncm-et-9 2023-02-04 09:15 01:08:07 1960617 343 307 0 36 +502.63 ± 58.47
164773 ncm-et-13 2023-02-04 09:08 01:15:58 1967774 386 327 1 58 +429.73 ± 45.45
164772 ncm-et-3 2023-02-04 07:43 01:40:26 1967076 500 434 2 64 +454.76 ± 43.27
164771 ncm-et-10 2023-02-04 07:37 01:39:40 1967105 500 423 0 77 +431.48 ± 39.18
164770 ncm-et-15 2023-02-04 07:35 01:42:45 1921251 500 422 1 77 +426.65 ± 39.25
164769 ncm-et-9 2023-02-04 07:34 01:40:33 1956972 500 434 1 65 +457.52 ± 42.89
164768 ncm-et-4 2023-02-04 07:33 01:41:35 1955329 500 419 0 81 +421.93 ± 38.15
164767 ncm-et-13 2023-02-04 07:28 01:38:55 1963134 500 433 1 66 +454.76 ± 42.55
164766 ncm-et-3 2023-02-04 06:02 01:40:05 1958186 500 427 1 72 +438.95 ± 40.66
164765 ncm-et-10 2023-02-04 05:56 01:40:51 1953303 500 430 1 69 +446.7 ± 41.57
164764 ncm-et-15 2023-02-04 05:52 01:42:13 1947223 500 429 1 70 +444.09 ± 41.26
164763 ncm-et-9 2023-02-04 05:51 01:42:36 1940195 500 433 0 67 +457.52 ± 42.15
164762 ncm-et-4 2023-02-04 05:49 01:43:21 1953630 500 432 0 68 +454.76 ± 41.82
164761 ncm-et-13 2023-02-04 05:48 01:39:31 1964341 500 433 2 65 +452.04 ± 42.92
164760 ncm-et-3 2023-02-04 04:20 01:41:42 1956030 500 432 1 67 +452.04 ± 42.22
164759 ncm-et-10 2023-02-04 04:13 01:42:14 1949225 500 436 0 64 +466.03 ± 43.18
164758 ncm-et-15 2023-02-04 04:13 01:38:50 1954224 500 422 1 77 +426.65 ± 39.25
164757 ncm-et-4 2023-02-04 04:09 01:39:28 1965133 500 421 0 79 +426.65 ± 38.66
164756 ncm-et-9 2023-02-04 04:08 01:41:46 1962022 500 433 1 66 +454.76 ± 42.55
164755 ncm-et-13 2023-02-04 04:07 01:40:05 1975770 500 429 0 71 +446.7 ± 40.89
164754 ncm-et-3 2023-02-04 02:37 01:42:09 1944610 500 432 0 68 +454.76 ± 41.82
164753 ncm-et-10 2023-02-04 02:32 01:39:54 1958942 500 430 1 69 +446.7 ± 41.57
164752 ncm-et-15 2023-02-04 02:32 01:39:54 1958626 500 427 1 72 +438.95 ± 40.66
164751 ncm-et-4 2023-02-04 02:29 01:39:08 1961933 500 437 1 62 +466.04 ± 43.97
164750 ncm-et-9 2023-02-04 02:27 01:40:56 1952011 500 433 0 67 +457.52 ± 42.15
164749 ncm-et-13 2023-02-04 02:26 01:40:55 1958550 500 431 1 68 +449.35 ± 41.89
164748 ncm-et-3 2023-02-04 00:54 01:42:49 1950272 500 434 0 66 +460.32 ± 42.48
164747 ncm-et-10 2023-02-04 00:51 01:41:02 1947968 500 434 1 65 +457.52 ± 42.89
164746 ncm-et-15 2023-02-04 00:50 01:41:00 1943455 500 428 0 72 +444.08 ± 40.59
164745 ncm-et-4 2023-02-04 00:49 01:39:22 1940225 500 437 1 62 +466.04 ± 43.97
164744 ncm-et-13 2023-02-04 00:46 01:39:08 1959789 500 429 1 70 +444.09 ± 41.26
164743 ncm-et-9 2023-02-04 00:45 01:40:44 1947713 500 432 0 68 +454.76 ± 41.82
164742 ncm-et-3 2023-02-03 23:12 01:40:40 1953437 500 443 2 55 +481.1 ± 46.83
164741 ncm-et-10 2023-02-03 23:10 01:40:27 1956025 500 438 3 59 +463.16 ± 45.12
164740 ncm-et-4 2023-02-03 23:09 01:39:16 1962918 500 416 0 84 +415.04 ± 37.43
164739 ncm-et-15 2023-02-03 23:08 01:41:33 1952390 500 430 0 70 +449.35 ± 41.19
164738 ncm-et-9 2023-02-03 23:06 01:39:10 1968224 500 429 0 71 +446.7 ± 40.89
164737 ncm-et-13 2023-02-03 23:05 01:40:25 1949652 500 439 1 60 +471.92 ± 44.73

Commit

Commit ID 8d3457a9966f8c744ab7f8536be408196ccd8af9
Author pb00067
Date 2023-02-03 19:18:50 UTC
Improve excluded move logic PR consists of 2 improvements on nodes with excludeMove: 1. Remove xoring the posKey with make_key(excludedMove) Since we never call tte->save anymore with excludedMove, the unique left purpose of the xoring was to avoid a TT hit. Nevertheless on a normal bench run this produced ~25 false positives (key collisions) To avoid that we now forbid early TT cutoff's with excludeMove Maybe these accesses to TT with xored key caused useless misses in the CPU caches (L1, L2 ...) Now doing the probe with the same key as the enclosing search does, should hit the CPU cache. 2. Don't probe Tablebases with excludedMove. This can't be tested on fishtest, but it's obvious that tablebases don't deliver any information about suboptimal moves. Side note: Very surprisingly it looks like we cannot use static eval's from TT since they slightly differ over time due to changing optimism. Attempts to use static eval's from TT did loose about 13 ELO. This is something about to investigate. LTC: https://tests.stockfishchess.org/tests/view/63dc0f8de9d4cdfbe672d0c6 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 44736 W: 12046 L: 11733 D: 20957 Ptnml(0-2): 12, 4212, 13617, 4505, 22 An analogue of this passed STC & LTC see PR #4374 (thanks Dubslow for reviewing!) closes https://github.com/official-stockfish/Stockfish/pull/4380 Bench: 4758694
Copyright 2011–2024 Next Chess Move LLC