Dev Builds » 20180501-2155

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 06:04:01 2001474 2397 1116 72 1209 +162.15 +/- 9.5
ncm-et-4 06:04:22 2002543 2397 1107 70 1220 +160.9 +/- 9.44
ncm-et-5 01:48:03 2022549 725 312 27 386 +144.35 +/- 16.76
ncm-et-6 01:47:45 2022959 711 317 19 375 +155.18 +/- 16.9
ncm-et-7 01:48:01 1985458 715 332 30 353 +156.55 +/- 17.77
ncm-et-8 01:47:50 2023120 713 345 17 351 +172.8 +/- 17.63
ncm-et-9 06:03:20 2001402 2386 1073 75 1238 +154.82 +/- 9.35
ncm-et-10 06:04:23 1970735 2366 1090 62 1214 +161.7 +/- 9.43
ncm-et-11 01:47:44 2008813 710 342 19 349 +170.56 +/- 17.72
ncm-et-12 01:47:54 1969216 692 319 22 351 +159.44 +/- 17.65
ncm-et-13 06:03:20 1966872 2357 1097 74 1186 +161.5 +/- 9.6
ncm-et-14 01:48:04 2021246 721 342 19 360 +167.53 +/- 17.4
ncm-et-15 06:03:58 1981793 2382 1075 76 1231 +155.29 +/- 9.39
ncm-et-16 01:47:59 2020187 728 336 24 368 +159.18 +/- 17.26
20000 9203 606 10191 +159.72 +/- 3.27

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
50988 ncm-et-10 2018-08-24 02:36 00:22:58 1981381 147 70 3 74 +170.93 +/- 38.35
50987 ncm-et-13 2018-08-24 02:35 00:23:59 1917645 149 62 7 80 +134.6 +/- 37.12
50986 ncm-et-15 2018-08-24 02:33 00:25:44 1981851 169 68 6 95 +133.69 +/- 33.48
50985 ncm-et-4 2018-08-24 02:32 00:26:32 1989874 175 75 10 90 +135.53 +/- 35.39
50984 ncm-et-3 2018-08-24 02:31 00:27:19 1990981 179 77 8 94 +141.22 +/- 34.31
50983 ncm-et-9 2018-08-24 02:30 00:28:43 1988296 181 90 5 86 +177.04 +/- 36.04
50982 ncm-et-10 2018-08-24 01:20 01:14:57 1977029 500 224 12 264 +157.24 +/- 20.11
50981 ncm-et-4 2018-08-24 01:15 01:15:59 1992722 500 240 18 242 +165.8 +/- 21.46
50980 ncm-et-13 2018-08-24 01:14 01:19:21 1895956 500 245 16 239 +171.9 +/- 21.58
50979 ncm-et-3 2018-08-24 01:14 01:16:21 1987193 500 234 14 252 +164.07 +/- 20.81
50978 ncm-et-9 2018-08-24 01:13 01:15:46 1989876 500 233 19 248 +158.93 +/- 21.16
50977 ncm-et-15 2018-08-24 01:13 01:19:02 1918124 500 230 20 250 +155.54 +/- 21.07
50976 ncm-et-10 2018-08-24 00:01 01:17:55 1951641 500 241 13 246 +171.02 +/- 21.1
50975 ncm-et-9 2018-08-23 23:57 01:15:17 1989246 500 211 14 275 +144.72 +/- 19.6
50974 ncm-et-4 2018-08-23 23:56 01:18:00 1988296 500 234 14 252 +164.07 +/- 20.81
50973 ncm-et-13 2018-08-23 23:56 01:17:14 1967536 500 224 15 261 +154.7 +/- 20.36
50972 ncm-et-15 2018-08-23 23:55 01:16:07 1972216 500 211 14 275 +144.72 +/- 19.6
50971 ncm-et-3 2018-08-23 23:55 01:17:47 1989401 500 230 16 254 +158.93 +/- 20.76
50970 ncm-et-9 2018-08-23 22:40 01:15:54 1990509 500 217 10 273 +153.02 +/- 19.57
50969 ncm-et-13 2018-08-23 22:40 01:15:05 1969279 500 238 16 246 +165.8 +/- 21.19
50968 ncm-et-15 2018-08-23 22:39 01:15:12 1983891 500 236 17 247 +163.21 +/- 21.16
50967 ncm-et-10 2018-08-23 22:39 01:20:22 1873498 500 233 15 252 +162.35 +/- 20.84
50966 ncm-et-4 2018-08-23 22:39 01:15:40 1992406 500 229 7 264 +165.8 +/- 19.92
50965 ncm-et-3 2018-08-23 22:39 01:14:45 1991930 500 232 16 252 +160.64 +/- 20.86
25806 ncm-et-12 2018-05-03 15:01 00:29:26 2019696 192 96 5 91 +178.99 +/- 34.99
25805 ncm-et-11 2018-05-03 14:58 00:31:56 2016446 210 107 8 95 +177.85 +/- 34.65
25804 ncm-et-9 2018-05-03 14:58 00:31:55 2025735 205 99 5 101 +172.14 +/- 32.98
25803 ncm-et-13 2018-05-03 14:58 00:32:06 2026391 208 98 4 106 +169.24 +/- 31.88
25802 ncm-et-6 2018-05-03 14:58 00:32:17 2025245 211 94 8 109 +150.34 +/- 31.8
25801 ncm-et-15 2018-05-03 14:58 00:32:25 2017421 213 103 7 103 +168.71 +/- 32.93
25800 ncm-et-7 2018-05-03 14:58 00:32:41 1971490 215 103 10 102 +160.88 +/- 33.4
25799 ncm-et-4 2018-05-03 14:57 00:33:16 2026065 222 98 6 118 +153.19 +/- 30.17
25798 ncm-et-10 2018-05-03 14:57 00:33:23 2018395 219 98 4 117 +159.45 +/- 30.04
25797 ncm-et-3 2018-05-03 14:57 00:33:19 2024099 218 95 4 119 +154.46 +/- 29.66
25796 ncm-et-8 2018-05-03 14:57 00:33:16 2023120 213 100 5 108 +166.67 +/- 31.72
25795 ncm-et-14 2018-05-03 14:57 00:33:27 2020351 221 105 6 110 +167.52 +/- 31.61
25794 ncm-et-5 2018-05-03 14:57 00:33:28 2022304 225 111 6 108 +175.73 +/- 32.06
25793 ncm-et-16 2018-05-03 14:57 00:33:47 2021163 228 102 7 119 +154.14 +/- 30.23
25792 ncm-et-3 2018-05-03 13:41 01:14:30 2025244 500 248 14 238 +176.33 +/- 21.58
25791 ncm-et-11 2018-05-03 13:41 01:15:48 2001181 500 235 11 254 +167.53 +/- 20.6
25790 ncm-et-6 2018-05-03 13:41 01:15:28 2020674 500 223 11 266 +157.24 +/- 19.97
25789 ncm-et-13 2018-05-03 13:41 01:15:35 2024426 500 230 16 254 +158.93 +/- 20.76
25788 ncm-et-12 2018-05-03 13:41 01:18:28 1918737 500 223 17 260 +152.18 +/- 20.47
25787 ncm-et-15 2018-05-03 13:41 01:15:28 2017258 500 227 12 261 +159.78 +/- 20.26
25786 ncm-et-8 2018-05-03 13:41 01:14:34 2023120 500 245 12 243 +175.44 +/- 21.24
25785 ncm-et-9 2018-05-03 13:41 01:15:45 2024754 500 223 22 255 +148.02 +/- 20.85
25784 ncm-et-16 2018-05-03 13:41 01:14:12 2019212 500 234 17 249 +161.49 +/- 21.05
25783 ncm-et-7 2018-05-03 13:41 01:15:20 1999427 500 229 20 251 +154.7 +/- 21.02
25782 ncm-et-10 2018-05-03 13:41 01:14:48 2022471 500 224 15 261 +154.7 +/- 20.36
25781 ncm-et-5 2018-05-03 13:41 01:14:35 2022794 500 201 21 278 +130.94 +/- 19.65
25780 ncm-et-14 2018-05-03 13:41 01:14:37 2022141 500 237 13 250 +167.53 +/- 20.88
25779 ncm-et-4 2018-05-03 13:41 01:14:55 2025899 500 231 15 254 +160.64 +/- 20.73

Commit

Commit ID 12ef8f71a2822f91e844e26d352d5fc27ac9559f
Author protonspring
Date 2018-05-01 21:55:30 UTC
Use special rule for BlockedByKing Simplification: remove BlockedByKing from storm array and use a special rule. The BlockedByKing section in the storm array is substantially similar to the Unopposed section except for two extreme values V(-290), V(-274). Turns out removing BlockedByKing and using a special rule for these two values shows no Elo loss. All the other values in the BlockedByKing section are apparently irrelevant. BlockedByKing now falls under unopposed which (to me) is a bit more logical since there is no defending pawn on this file. Also, retuning the Unopposed section may be another improvement. GOOD) This is a simplification because the entire BlockedByKing section of the storm array goes away reducing a few lines of code (and less values to tune). This also brings clarity because the special rule is self documenting. BAD) It takes execution time to apply the special rule. This should be negli- gible because it is based on a template parameter and is boiled down to two bitwise AND's. STC: LLR: 2.96 (-2.94,2.94) [-3.00,1.00] Total: 33470 W: 6820 L: 6721 D: 19929 http://tests.stockfishchess.org/tests/view/5ae7b6e60ebc5926dba90e13 LTC: LLR: 2.96 (-2.94,2.94) [-3.00,1.00] Total: 47627 W: 7045 L: 6963 D: 33619 http://tests.stockfishchess.org/tests/view/5ae859ff0ebc5926dba90e85 Closes https://github.com/official-stockfish/Stockfish/pull/1574 Bench: 5512000 ----------- How to continue after this patch? This patch may open the possibility to move the special rule to evaluate.cpp in the evaluate::king() function, where we could refine the rule using king danger information. For instance, with a king in H2 blocking an opponent pawn in H3, it may be critical to know that the opponent has no safe check in G2 before giving the bonus :-)
Copyright 2011–2024 Next Chess Move LLC