Dev Builds » 20140405-0926

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 05:59:43 1990891 2399 173 970 1256 -119.98 +/- 9.46
ncm-et-4 06:00:08 1972159 2355 150 927 1278 -119.09 +/- 9.3
ncm-et-5 01:47:28 1989261 715 47 280 388 -117.5 +/- 16.9
ncm-et-6 01:47:09 2025163 723 52 268 403 -107.06 +/- 16.55
ncm-et-7 01:47:24 2010135 722 53 284 385 -115.2 +/- 17.06
ncm-et-8 01:47:03 1990412 719 52 284 383 -116.26 +/- 17.1
ncm-et-9 06:00:15 1991540 2384 144 955 1285 -123.1 +/- 9.27
ncm-et-10 06:00:08 1987552 2380 177 958 1245 -118.39 +/- 9.51
ncm-et-11 01:47:05 1980374 692 53 275 364 -115.54 +/- 17.6
ncm-et-12 01:47:06 2021244 714 40 284 390 -123.71 +/- 16.77
ncm-et-13 06:00:17 1989931 2382 134 945 1303 -123.21 +/- 9.17
ncm-et-14 01:46:50 2018720 717 36 289 392 -128.1 +/- 16.67
ncm-et-15 05:59:55 1987012 2396 155 958 1283 -121.12 +/- 9.3
ncm-et-16 01:47:11 2012447 702 41 272 389 -118.74 +/- 16.76
20000 1307 7949 10744 -119.93 +/- 3.21

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
93050 ncm-et-4 2019-08-17 04:18 00:22:01 1937485 138 11 55 72 -114.78 +/- 39.83
93049 ncm-et-10 2019-08-17 04:15 00:25:09 1963843 166 12 60 94 -103.41 +/- 34.27
93048 ncm-et-13 2019-08-17 04:14 00:26:04 1975915 167 8 76 83 -150.17 +/- 36.89
93047 ncm-et-15 2019-08-17 04:14 00:26:12 1974047 177 13 67 97 -109.48 +/- 33.95
93046 ncm-et-3 2019-08-17 04:13 00:26:25 1974373 176 11 72 93 -125.62 +/- 34.74
93045 ncm-et-9 2019-08-17 04:12 00:27:30 1970850 176 12 68 96 -114.52 +/- 34.09
93044 ncm-et-4 2019-08-17 03:01 01:15:43 1966311 500 30 194 276 -118.33 +/- 19.94
93043 ncm-et-10 2019-08-17 02:59 01:14:48 1990509 500 43 197 260 -110.6 +/- 20.91
93042 ncm-et-13 2019-08-17 02:58 01:14:59 1965072 500 28 205 267 -128.55 +/- 20.35
93041 ncm-et-15 2019-08-17 02:57 01:15:20 1970349 500 36 199 265 -117.55 +/- 20.58
93040 ncm-et-3 2019-08-17 02:57 01:15:30 1978882 500 38 209 253 -123.81 +/- 21.21
93039 ncm-et-9 2019-08-17 02:55 01:16:41 1970177 500 39 200 261 -116.0 +/- 20.82
93038 ncm-et-10 2019-08-17 01:41 01:16:43 1966460 500 40 193 267 -109.83 +/- 20.53
93037 ncm-et-13 2019-08-17 01:41 01:15:56 1974830 500 34 185 281 -108.3 +/- 19.76
93036 ncm-et-4 2019-08-17 01:41 01:19:11 1918331 500 36 200 264 -118.33 +/- 20.63
93035 ncm-et-3 2019-08-17 01:40 01:15:17 1975933 500 36 195 269 -114.45 +/- 20.38
93034 ncm-et-15 2019-08-17 01:40 01:16:26 1968937 500 25 210 265 -134.95 +/- 20.4
93033 ncm-et-9 2019-08-17 01:40 01:14:02 1995583 500 23 212 265 -138.18 +/- 20.35
93032 ncm-et-3 2019-08-17 00:24 01:15:28 1966327 500 33 203 264 -123.02 +/- 20.59
93031 ncm-et-4 2019-08-17 00:24 01:15:45 1965379 500 23 197 280 -126.18 +/- 19.6
93030 ncm-et-9 2019-08-17 00:24 01:14:40 1962304 500 34 201 265 -120.67 +/- 20.55
93029 ncm-et-15 2019-08-17 00:24 01:14:49 1973892 500 34 196 270 -116.78 +/- 20.3
93028 ncm-et-13 2019-08-17 00:24 01:15:54 1970661 500 30 201 269 -123.81 +/- 20.29
93027 ncm-et-10 2019-08-17 00:24 01:16:15 1967387 500 38 210 252 -124.6 +/- 21.26
25316 ncm-et-11 2018-05-02 06:53 00:28:38 2017420 192 16 75 101 -110.33 +/- 33.57
25315 ncm-et-9 2018-05-02 06:51 00:30:40 2025573 208 11 82 115 -123.55 +/- 30.84
25314 ncm-et-16 2018-05-02 06:50 00:30:49 2011938 202 13 70 119 -100.77 +/- 30.05
25313 ncm-et-5 2018-05-02 06:50 00:31:37 2021977 215 17 87 111 -117.39 +/- 32.05
25312 ncm-et-4 2018-05-02 06:49 00:32:09 2024590 217 9 90 118 -136.27 +/- 30.34
25311 ncm-et-14 2018-05-02 06:49 00:32:04 2018395 217 16 89 112 -121.61 +/- 31.86
25310 ncm-et-10 2018-05-02 06:49 00:32:05 2017258 214 11 92 111 -138.39 +/- 31.72
25309 ncm-et-13 2018-05-02 06:49 00:32:16 2026718 215 12 74 129 -103.12 +/- 28.59
25308 ncm-et-15 2018-05-02 06:49 00:32:27 2019859 219 15 89 115 -122.2 +/- 31.31
25307 ncm-et-7 2018-05-02 06:48 00:33:14 1996823 222 18 77 127 -94.61 +/- 29.48
25306 ncm-et-3 2018-05-02 06:48 00:33:11 2026226 223 19 89 115 -112.87 +/- 31.53
25305 ncm-et-8 2018-05-02 06:48 00:33:28 1957705 219 18 81 120 -102.85 +/- 30.58
25304 ncm-et-12 2018-05-02 06:48 00:33:23 2021163 214 10 84 120 -125.31 +/- 29.99
25303 ncm-et-6 2018-05-02 06:48 00:33:39 2024100 223 17 86 120 -111.14 +/- 30.62
25302 ncm-et-14 2018-05-02 05:33 01:14:46 2019046 500 20 200 280 -130.94 +/- 19.52
25301 ncm-et-11 2018-05-02 05:33 01:18:27 1943329 500 37 200 263 -117.55 +/- 20.7
25300 ncm-et-15 2018-05-02 05:33 01:14:41 2014992 500 32 197 271 -119.11 +/- 20.22
25299 ncm-et-3 2018-05-02 05:33 01:13:52 2023609 500 36 202 262 -119.89 +/- 20.73
25298 ncm-et-8 2018-05-02 05:33 01:13:35 2023119 500 34 203 263 -122.24 +/- 20.65
25297 ncm-et-6 2018-05-02 05:33 01:13:30 2026227 500 35 182 283 -105.25 +/- 19.68
25296 ncm-et-10 2018-05-02 05:33 01:15:08 2019860 500 33 206 261 -125.39 +/- 20.74
25295 ncm-et-13 2018-05-02 05:33 01:15:08 2026390 500 22 204 274 -132.54 +/- 19.87
25294 ncm-et-7 2018-05-02 05:33 01:14:10 2023447 500 35 207 258 -124.6 +/- 20.92
25293 ncm-et-9 2018-05-02 05:33 01:16:42 2024754 500 25 192 283 -120.67 +/- 19.49
25292 ncm-et-4 2018-05-02 05:33 01:15:19 2020863 500 41 191 268 -107.54 +/- 20.5
25291 ncm-et-5 2018-05-02 05:33 01:15:51 1956545 500 30 193 277 -117.55 +/- 19.89
25290 ncm-et-12 2018-05-02 05:33 01:13:43 2021325 500 30 200 270 -123.02 +/- 20.24
25289 ncm-et-16 2018-05-02 05:33 01:16:22 2012956 500 28 202 270 -126.18 +/- 20.2

Commit

Commit ID be641e881fdfdf3354453381f832fe7822e7c731
Author Lucas Braesch
Date 2014-04-05 09:26:44 UTC
Remove QueenOn7th and QueenOnPawn Small simplification. Passed SPRT(-3,1) both at STC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 17051 W: 3132 L: 3005 D: 10914 and LTC: LLR: 4.55 (-2.94,2.94) [-3.00,1.00] Total: 24890 W: 3842 L: 3646 D: 17402 The rationale behind this is that I've never managed to add a Queen on 7th rank bonus in DiscoCheck, because it never showed to be positive (evne slightly) in testing. The only thing that worked is Rook on 7th rank. In terms of SF code, it seemed natural to group it with QueenOnPawn as well as those are done together. I know you're against groupping in general, but when it comes to non regression test, you are being more conservative by groupping. If the group passes SPRT(-3,1) it's safer to commit, than test every component in SPRT(-3,1) and end up with the risk of commiting several -1 elo regression instead of just one -1 elo regression. In chess terms, perhaps it's just easier to manouver a Queen (which can more also diagonaly) than a Rook. Therefore you can let the search do its job without needing eval ad-hoc terms to guide it. For the Rook which takes more moves to manouver such eval terms can be (marginally) useful. bench: 7473314
Copyright 2011–2024 Next Chess Move LLC