Dev Builds » 20220709-0858

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 11:20:10 1959887 3359 2893 3 463 +449.86 +/- 15.83
ncm-et-4 11:20:05 1884745 3237 2819 10 408 +460.01 +/- 16.9
ncm-et-9 11:20:03 1962188 3352 2901 11 440 +452.27 +/- 16.27
ncm-et-10 11:13:33 1957496 3317 2860 5 452 +450.31 +/- 16.03
ncm-et-13 11:20:13 1958405 3350 2889 10 451 +448.56 +/- 16.06
ncm-et-15 11:20:22 1961483 3385 2915 7 463 +448.14 +/- 15.84
20000 17277 46 2677 +451.43 +/- 6.58

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
159411 ncm-et-4 2022-07-10 00:13 00:53:54 1760386 237 210 0 27 +487.56 +/- 68.21
159410 ncm-et-13 2022-07-09 23:55 01:11:18 1947517 350 299 0 51 +441.86 +/- 48.51
159409 ncm-et-9 2022-07-09 23:55 01:11:36 1965081 352 301 1 50 +439.3 +/- 49.13
159408 ncm-et-3 2022-07-09 23:53 01:13:34 1956488 359 311 0 48 +457.93 +/- 50.12
159407 ncm-et-15 2022-07-09 23:48 01:18:54 1956569 385 339 3 43 +467.1 +/- 53.15
159406 ncm-et-10 2022-07-09 23:42 01:24:20 1972069 415 361 0 54 +462.98 +/- 47.15
159405 ncm-et-4 2022-07-09 22:32 01:40:25 1969395 500 429 3 68 +438.95 +/- 41.92
159404 ncm-et-13 2022-07-09 22:14 01:40:19 1963387 500 425 0 75 +436.43 +/- 39.73
159403 ncm-et-9 2022-07-09 22:12 01:41:44 1960776 500 428 1 71 +441.5 +/- 40.95
159402 ncm-et-3 2022-07-09 22:11 01:41:02 1959858 500 427 0 73 +441.5 +/- 40.29
159401 ncm-et-15 2022-07-09 22:04 01:43:05 1951769 500 432 0 68 +454.76 +/- 41.82
159400 ncm-et-10 2022-07-09 21:58 01:40:04 1962498 498 429 1 68 +448.61 +/- 41.89
159399 ncm-et-4 2022-07-09 20:40 01:50:45 1790707 500 439 1 60 +471.92 +/- 44.73
159398 ncm-et-13 2022-07-09 20:34 01:40:06 1963382 500 441 2 57 +474.93 +/- 45.97
159397 ncm-et-10 2022-07-09 20:32 01:22:30 1959085 404 353 0 51 +468.6 +/- 48.59
159396 ncm-et-3 2022-07-09 20:30 01:40:42 1959698 500 425 0 75 +436.43 +/- 39.73
159395 ncm-et-9 2022-07-09 20:28 01:44:08 1955111 500 434 2 64 +454.76 +/- 43.27
159394 ncm-et-15 2022-07-09 20:24 01:39:34 1961075 500 436 1 63 +463.16 +/- 43.6
159393 ncm-et-4 2022-07-09 18:57 01:42:46 1927898 500 423 0 77 +431.48 +/- 39.18
159392 ncm-et-13 2022-07-09 18:51 01:42:04 1963714 500 435 1 64 +460.32 +/- 43.24
159391 ncm-et-10 2022-07-09 18:49 01:41:59 1949787 500 432 1 67 +452.04 +/- 42.22
159390 ncm-et-3 2022-07-09 18:48 01:41:05 1964329 500 434 1 65 +457.52 +/- 42.89
159389 ncm-et-9 2022-07-09 18:48 01:39:26 1961599 500 432 2 66 +449.35 +/- 42.58
159388 ncm-et-15 2022-07-09 18:44 01:38:58 1969870 500 420 2 78 +419.61 +/- 39.03
159387 ncm-et-4 2022-07-09 17:09 01:47:20 1870150 500 442 4 54 +471.92 +/- 47.15
159386 ncm-et-13 2022-07-09 17:08 01:42:15 1957098 500 429 4 67 +436.43 +/- 42.22
159385 ncm-et-9 2022-07-09 17:07 01:40:12 1964555 500 422 3 75 +421.93 +/- 39.85
159384 ncm-et-3 2022-07-09 17:06 01:41:21 1964468 500 440 1 59 +474.93 +/- 45.13
159383 ncm-et-10 2022-07-09 17:06 01:42:53 1944051 500 420 2 78 +419.61 +/- 39.03
159382 ncm-et-15 2022-07-09 17:04 01:39:34 1960154 500 429 0 71 +446.7 +/- 40.89
159381 ncm-et-9 2022-07-09 15:26 01:40:01 1971408 500 436 2 62 +460.32 +/- 43.99
159380 ncm-et-13 2022-07-09 15:26 01:41:50 1956485 500 422 0 78 +429.05 +/- 38.92
159379 ncm-et-4 2022-07-09 15:25 01:43:20 1905288 500 441 1 58 +477.99 +/- 45.54
159378 ncm-et-3 2022-07-09 15:24 01:41:26 1952226 500 426 0 74 +438.95 +/- 40.01
159377 ncm-et-15 2022-07-09 15:23 01:40:02 1963694 500 425 1 74 +433.94 +/- 40.08
159376 ncm-et-10 2022-07-09 15:23 01:41:43 1955743 500 437 0 63 +468.95 +/- 43.54
159375 ncm-et-9 2022-07-09 13:43 01:42:56 1956788 500 448 0 52 +504.32 +/- 48.17
159374 ncm-et-13 2022-07-09 13:43 01:42:21 1957252 500 438 3 59 +463.16 +/- 45.12
159373 ncm-et-3 2022-07-09 13:43 01:41:00 1962148 500 430 1 69 +446.7 +/- 41.57
159372 ncm-et-10 2022-07-09 13:43 01:40:04 1959244 500 428 1 71 +441.5 +/- 40.95
159371 ncm-et-4 2022-07-09 13:43 01:41:35 1969396 500 435 1 64 +460.32 +/- 43.24
159370 ncm-et-15 2022-07-09 13:43 01:40:15 1967253 500 434 0 66 +460.32 +/- 42.48

Commit

Commit ID 2e02dd79366e6f17df5b0599048937289fd5819e
Author Joost VandeVondele
Date 2022-07-09 08:58:04 UTC
Limit the researching at same depth. If the elapsed time is close to the available time, the time management thread can signal that the next iterations should be searched at the same depth (Threads.increaseDepth = false). While the rootDepth increases, the adjustedDepth is kept constant with the searchAgainCounter. In exceptional cases, when threading is used and the master thread, which controls the time management, signals to not increaseDepth, but by itself takes a long time to finish the iteration, the helper threads can search repeatedly at the same depth. This search finishes more and more quickly, leading to helper threads that report a rootDepth of MAX_DEPTH (245). The latter is not optimal as it is confusing for the user, stops search on these threads, and leads to an incorrect bias in the thread voting scheme. Probably with only a small impact on strength. This behavior was observed almost two years ago, see https://github.com/official-stockfish/Stockfish/issues/2717 This patch fixes #2717 by ensuring the effective depth increases at once every four iterations, even in increaseDepth is false. Depth 245 searches (for non-trivial positions) were indeed absent with this patch, but frequent with master in the tests below: https://discord.com/channels/435943710472011776/813919248455827515/994872720800088095 Total pgns: 2173 Base: 2867 Patch: 0 it passed non-regression testing in various setups: SMP STC: https://tests.stockfishchess.org/tests/view/62bfecc96178ffe6394ba036 LLR: 2.94 (-2.94,2.94) <-2.25,0.25> Total: 37288 W: 10171 L: 10029 D: 17088 Ptnml(0-2): 75, 3777, 10793, 3929, 70 SMP LTC: https://tests.stockfishchess.org/tests/view/62c08f6f49b62510394be066 LLR: 2.94 (-2.94,2.94) <-2.25,0.25> Total: 190568 W: 52125 L: 52186 D: 86257 Ptnml(0-2): 70, 17854, 59504, 17779, 77 LTC: https://tests.stockfishchess.org/tests/view/62c08b6049b62510394bdfb6 LLR: 2.96 (-2.94,2.94) <-2.25,0.25> Total: 48120 W: 13204 L: 13083 D: 21833 Ptnml(0-2): 54, 4458, 14919, 4571, 58 Special thanks to miguel-I, Disservin, ruicoelhopedro and others for analysing the problem, the data, and coming up with the key insight, needed to fix this longstanding issue. closes https://github.com/official-stockfish/Stockfish/pull/4104 Bench: 5182295
Copyright 2011–2024 Next Chess Move LLC