Dev Builds » 20200927-0624

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 08:32:50 1962942 3344 2603 27 714 +354.78 ± 12.74
ncm-et-4 08:39:33 1946831 3316 2577 14 725 +357.0 ± 12.61
ncm-et-9 08:36:28 1966800 3353 2625 27 701 +358.66 ± 12.86
ncm-et-10 08:31:40 1957723 3308 2568 23 717 +353.94 ± 12.71
ncm-et-13 08:34:34 1961151 3345 2643 25 677 +365.57 ± 13.09
ncm-et-15 08:36:39 1968280 3334 2617 19 698 +362.53 ± 12.88
20000 15633 135 4232 +358.72 ± 5.23

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
139317 ncm-et-4 2020-09-28 03:24 00:48:13 1966947 316 248 1 67 +364.66 ± 42.0
139316 ncm-et-10 2020-09-28 03:24 00:48:42 1964465 308 241 1 66 +362.51 ± 42.32
139315 ncm-et-15 2020-09-28 03:20 00:52:01 1969409 334 261 0 73 +364.47 ± 40.05
139314 ncm-et-3 2020-09-28 03:20 00:52:08 1960923 344 254 5 85 +318.13 ± 37.27
139313 ncm-et-13 2020-09-28 03:19 00:52:48 1959704 345 270 3 72 +357.86 ± 40.6
139312 ncm-et-9 2020-09-28 03:18 00:54:23 1963378 353 287 3 63 +386.12 ± 43.55
139311 ncm-et-10 2020-09-28 02:07 01:15:32 1960769 500 381 0 119 +347.77 ± 31.15
139310 ncm-et-4 2020-09-28 02:03 01:20:07 1914519 500 386 3 111 +351.11 ± 32.49
139309 ncm-et-15 2020-09-28 02:02 01:17:03 1968019 500 401 4 95 +375.98 ± 35.27
139308 ncm-et-3 2020-09-28 02:02 01:17:26 1963687 500 387 2 111 +354.5 ± 32.44
139307 ncm-et-13 2020-09-28 02:00 01:18:06 1965071 500 402 8 90 +370.41 ± 36.22
139306 ncm-et-9 2020-09-28 02:00 01:16:59 1962776 500 382 5 113 +341.24 ± 32.24
139305 ncm-et-10 2020-09-28 00:49 01:17:04 1935890 500 390 5 105 +354.5 ± 33.5
139304 ncm-et-15 2020-09-28 00:44 01:17:28 1970479 500 394 5 101 +361.43 ± 34.18
139303 ncm-et-3 2020-09-28 00:44 01:17:05 1963226 500 399 7 94 +366.78 ± 35.45
139302 ncm-et-4 2020-09-28 00:43 01:18:48 1929182 500 390 1 109 +361.43 ± 32.7
139301 ncm-et-13 2020-09-28 00:43 01:16:22 1959700 500 392 5 103 +357.94 ± 33.83
139300 ncm-et-9 2020-09-28 00:41 01:18:16 1968317 500 394 4 102 +363.2 ± 33.99
139299 ncm-et-10 2020-09-27 23:31 01:17:21 1962772 500 393 8 99 +354.5 ± 34.52
139298 ncm-et-3 2020-09-27 23:25 01:17:06 1962305 500 387 2 111 +354.5 ± 32.44
139297 ncm-et-4 2020-09-27 23:25 01:17:01 1962918 500 396 3 101 +368.59 ± 34.13
139296 ncm-et-15 2020-09-27 23:25 01:18:00 1969550 500 390 2 108 +359.68 ± 32.92
139295 ncm-et-13 2020-09-27 23:25 01:17:01 1960006 500 393 2 105 +364.98 ± 33.41
139294 ncm-et-9 2020-09-27 23:23 01:16:59 1966770 500 385 4 111 +347.77 ± 32.52
139293 ncm-et-10 2020-09-27 22:12 01:16:59 1962921 500 395 5 100 +363.2 ± 34.35
139292 ncm-et-3 2020-09-27 22:08 01:15:50 1963996 500 391 4 105 +357.94 ± 33.48
139291 ncm-et-13 2020-09-27 22:07 01:16:12 1964796 500 394 3 103 +364.98 ± 33.79
139290 ncm-et-15 2020-09-27 22:07 01:16:45 1959694 500 382 2 116 +346.12 ± 31.7
139289 ncm-et-4 2020-09-27 22:05 01:19:05 1950122 500 384 2 114 +349.43 ± 31.99
139288 ncm-et-9 2020-09-27 22:05 01:17:02 1971567 500 381 3 116 +342.85 ± 31.74
139287 ncm-et-10 2020-09-27 20:53 01:18:24 1957860 500 385 3 112 +349.43 ± 32.33
139286 ncm-et-3 2020-09-27 20:50 01:17:03 1962924 500 393 3 104 +363.2 ± 33.62
139285 ncm-et-13 2020-09-27 20:49 01:17:21 1958472 500 395 2 103 +368.59 ± 33.75
139284 ncm-et-15 2020-09-27 20:48 01:17:16 1969867 500 389 2 109 +357.94 ± 32.76
139283 ncm-et-4 2020-09-27 20:47 01:16:46 1968629 500 392 2 106 +363.2 ± 33.24
139282 ncm-et-9 2020-09-27 20:47 01:16:35 1966469 500 398 3 99 +372.25 ± 34.49
139281 ncm-et-10 2020-09-27 19:34 01:17:38 1959388 500 383 1 116 +349.43 ± 31.64
139280 ncm-et-3 2020-09-27 19:33 01:16:12 1963535 500 392 4 104 +359.68 ± 33.64
139279 ncm-et-13 2020-09-27 19:31 01:16:44 1960308 500 397 2 101 +372.25 ± 34.1
139278 ncm-et-9 2020-09-27 19:30 01:16:14 1968326 500 398 5 97 +368.59 ± 34.9
139277 ncm-et-15 2020-09-27 19:29 01:18:06 1970945 500 400 4 96 +374.11 ± 35.07
139276 ncm-et-4 2020-09-27 19:26 01:19:33 1935505 500 381 2 117 +344.48 ± 31.55

Commit

Commit ID 1dbd2a1ad548b3ca676f7da949e1a998c64b836b
Author Stéphane Nicolet
Date 2020-09-27 06:24:50 UTC
Tweak nnue scaling to keep more material Current master uses a constant scale factor of 5/4 = 1.25 for the output of the NNUE network, for compatibility with search and classical evaluation. We modify this scale factor to make it dependent on the phase of the game, going from about 1.5 in the opening to 1.0 for pure pawn endgames. This helps Stockfish to avoid exchanges of pieces (heavy pieces in particular) when she has the advantage, keeping more material on the board when attacking. Passed STC: LLR: 2.95 (-2.94,2.94) {-0.25,1.25} Total: 14744 W: 1771 L: 1599 D: 11374 Ptnml(0-2): 87, 1184, 4664, 1344, 93 https://tests.stockfishchess.org/tests/view/5f6fb0a63b22d6afa506904f Passed LTC: LLR: 2.95 (-2.94,2.94) {0.25,1.25} Total: 8912 W: 512 L: 393 D: 8007 Ptnml(0-2): 7, 344, 3637, 459, 9 https://tests.stockfishchess.org/tests/view/5f6fcf533b22d6afa5069066 closes https://github.com/official-stockfish/Stockfish/pull/3154 Bench: 3943952
Copyright 2011–2024 Next Chess Move LLC