Dev Builds » 20210702-1253

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 15. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo
ncm-dbt-01 11:39:34 584706 6680 851 2554 3275 -90.57 ± 3.91 42 1802 1315 179 2 -189.88 ± 9.38
ncm-dbt-02 11:39:12 586676 6670 839 2591 3240 -93.45 ± 3.9 51 1815 1305 163 1 -195.67 ± 9.41
ncm-dbt-03 11:39:09 587058 6650 814 2592 3244 -95.21 ± 3.97 59 1835 1257 173 1 -198.95 ± 9.6
20000 2504 7737 9759 -93.07 ± 2.27 152 5452 3877 515 4 -194.81 ± 5.46

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
370589 ncm-dbt-03 587197 150 16 55 79 -92.45 ± 26.15 2 38 32 3 0 -187.77 ± 60.75
370588 ncm-dbt-02 589155 170 25 66 79 -85.47 ± 25.3 1 45 33 6 0 -177.48 ± 60.26
370587 ncm-dbt-01 584579 180 23 64 93 -80.55 ± 24.09 2 42 41 5 0 -161.21 ± 53.43
370586 ncm-dbt-03 588899 500 56 194 250 -98.44 ± 13.98 3 143 93 11 0 -209.91 ± 35.49
370585 ncm-dbt-01 585295 500 63 188 249 -88.74 ± 13.68 2 132 105 11 0 -187.16 ± 33.21
370584 ncm-dbt-02 587325 500 66 200 234 -95.44 ± 14.1 4 137 98 11 0 -200.24 ± 34.5
370583 ncm-dbt-03 587070 500 63 189 248 -89.48 ± 15.02 1 145 83 21 0 -190.85 ± 37.38
370582 ncm-dbt-01 581194 500 71 198 231 -90.22 ± 14.46 4 133 99 14 0 -187.16 ± 34.38
370581 ncm-dbt-02 583698 500 58 188 254 -92.45 ± 14.35 5 132 101 12 0 -190.85 ± 33.97
370580 ncm-dbt-03 584076 500 64 186 250 -86.52 ± 14.95 3 135 93 19 0 -179.9 ± 35.49
370579 ncm-dbt-01 586774 500 62 203 235 -100.7 ± 14.47 2 152 81 15 0 -217.85 ± 38.05
370578 ncm-dbt-02 583237 500 61 199 240 -98.44 ± 14.44 4 142 93 10 1 -209.91 ± 35.49
370577 ncm-dbt-03 585421 500 59 199 242 -99.95 ± 14.91 5 145 85 15 0 -209.91 ± 37.16
370576 ncm-dbt-01 584327 500 64 192 244 -90.96 ± 14.17 3 135 99 13 0 -190.85 ± 34.36
370575 ncm-dbt-02 582235 500 60 199 241 -99.2 ± 13.35 4 137 103 6 0 -209.91 ± 33.29
370574 ncm-dbt-03 585548 500 66 188 246 -86.52 ± 13.78 1 133 103 13 0 -183.51 ± 33.63
370573 ncm-dbt-01 584327 500 75 191 234 -82.1 ± 14.27 1 131 101 17 0 -172.78 ± 34.05
370572 ncm-dbt-02 588600 500 60 193 247 -94.69 ± 13.78 5 131 106 8 0 -196.45 ± 32.87
370571 ncm-dbt-03 588472 500 68 196 236 -90.96 ± 14.02 5 128 107 10 0 -187.16 ± 32.81
370570 ncm-dbt-01 584117 500 67 196 237 -91.71 ± 13.42 1 137 102 10 0 -196.45 ± 33.71
370569 ncm-dbt-02 590096 500 78 183 239 -74.06 ± 15.01 5 115 110 20 0 -147.19 ± 32.57
370568 ncm-dbt-03 588047 500 45 193 262 -106.01 ± 14.04 7 141 95 7 0 -221.9 ± 34.93
370567 ncm-dbt-01 585632 500 56 192 252 -96.94 ± 15.16 8 133 97 11 1 -198.34 ± 34.72
370566 ncm-dbt-02 588515 500 56 191 253 -96.19 ± 13.95 3 140 96 11 0 -204.07 ± 34.89
370565 ncm-dbt-03 586985 500 69 194 237 -88.74 ± 13.98 5 125 110 10 0 -181.7 ± 32.28
370564 ncm-dbt-01 585632 500 64 184 252 -85.04 ± 14.2 0 136 99 14 1 -183.51 ± 34.39
370563 ncm-dbt-02 588686 500 62 201 237 -99.2 ± 14.75 4 146 85 15 0 -209.91 ± 37.16
370562 ncm-dbt-03 588302 500 51 201 248 -107.54 ± 15.13 9 144 85 12 0 -221.9 ± 37.21
370561 ncm-dbt-02 587028 500 62 190 248 -90.96 ± 13.72 3 132 105 10 0 -190.85 ± 33.16
370560 ncm-dbt-01 585169 500 56 183 261 -90.22 ± 15.18 6 132 95 17 0 -183.51 ± 35.13
370559 ncm-dbt-03 585548 500 68 197 235 -91.71 ± 14.78 4 136 96 13 1 -192.71 ± 34.94
370558 ncm-dbt-02 585970 500 65 192 243 -90.22 ± 14.75 3 138 92 17 0 -189.0 ± 35.7
370557 ncm-dbt-01 583279 500 68 200 232 -93.95 ± 13.61 1 141 97 11 0 -202.15 ± 34.7
370556 ncm-dbt-03 587664 500 57 209 234 -109.07 ± 14.84 7 150 81 12 0 -230.16 ± 38.13
370555 ncm-dbt-02 586646 500 70 210 220 -99.95 ± 14.61 6 140 92 12 0 -207.95 ± 35.71
370554 ncm-dbt-01 583028 500 69 180 251 -78.44 ± 14.46 0 131 99 20 0 -165.8 ± 34.39
370553 ncm-dbt-02 586816 500 61 193 246 -93.95 ± 13.77 3 136 101 10 0 -198.34 ± 33.9
370552 ncm-dbt-03 588089 500 68 191 241 -87.26 ± 14.69 5 128 102 15 0 -178.11 ± 33.85
370551 ncm-dbt-01 585084 500 53 202 245 -106.77 ± 14.21 7 143 92 8 0 -223.94 ± 35.61
370550 ncm-dbt-01 587452 500 60 181 259 -85.78 ± 14.36 5 124 108 13 0 -174.55 ± 32.75
370549 ncm-dbt-02 585464 500 55 186 259 -93.2 ± 14.21 1 144 90 15 0 -200.24 ± 36.12
370548 ncm-dbt-03 587494 500 64 200 236 -96.94 ± 13.96 2 144 92 12 0 -207.95 ± 35.71

Commit

Commit ID 773dff020968f7a6f590cfd53e8fd89f12e15e36
Author Joost VandeVondele
Date 2021-07-02 12:53:30 UTC
Stockfish 14 Official release version of Stockfish 14 Bench: 4770936 --- Today, we have the pleasure to announce Stockfish 14. As usual, downloads will be freely available at https://stockfishchess.org The engine is now significantly stronger than just a few months ago, and wins four times more game pairs than it loses against the previous release version [0]. Stockfish 14 is now at least 400 Elo ahead of Stockfish 7, a top engine in 2016 [1]. During the last five years, Stockfish has thus gained about 80 Elo per year. Stockfish 14 evaluates positions more accurately than Stockfish 13 as a result of two major steps forward in defining and training the efficiently updatable neural network (NNUE) that provides the evaluation for positions. First, the collaboration with the Leela Chess Zero team - announced previously [2] - has come to fruition. The LCZero team has provided a collection of billions of positions evaluated by Leela that we have combined with billions of positions evaluated by Stockfish to train the NNUE net that powers Stockfish 14. The fact that we could use and combine these datasets freely was essential for the progress made and demonstrates the power of open source and open data [3]. Second, the architecture of the NNUE network was significantly updated: the new network is not only larger, but more importantly, it deals better with large material imbalances and can specialize for multiple phases of the game [4]. A new project, kick-started by Gary Linscott and Tomasz Sobczyk, led to a GPU accelerated net trainer written in pytorch.[5] This tool allows for training high-quality nets in a couple of hours. Finally, this release features some search refinements, minor bug fixes and additional improvements. For example, Stockfish is now about 90 Elo stronger for chess960 (Fischer random chess) at short time control. The Stockfish project builds on a thriving community of enthusiasts (thanks everybody!) that contribute their expertise, time, and resources to build a free and open-source chess engine that is robust, widely available, and very strong. We invite our chess fans to join the fishtest testing framework and programmers to contribute to the project on github [6]. Stay safe and enjoy chess! The Stockfish team [0] https://tests.stockfishchess.org/tests/view/60dae5363beab81350aca077 [1] https://nextchessmove.com/dev-builds [2] https://stockfishchess.org/blog/2021/stockfish-13/ [3] https://lczero.org/blog/2021/06/the-importance-of-open-data/ [4] https://github.com/official-stockfish/Stockfish/commit/e8d64af1 [5] https://github.com/glinscott/nnue-pytorch/ [6] https://stockfishchess.org/get-involved/
Copyright 2011–2024 Next Chess Move LLC