Stockfish Testing Queue

Finished - 33101 tests

18-08-14 sni bad_capture4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11070 W: 2349 L: 2426 D: 6295
sprt @ 10+0.1 th 1 bad captures: SEE < max(0, value)
18-08-14 sni bad_capture4~1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8229 W: 1738 L: 1829 D: 4662
sprt @ 10+0.1 th 1 bad captures : SEE < min(0, value)
18-08-14 Gua DefaultCheck3 diff
ELO: -14.95 +-6.4 (95%) LOS: 0.0%
Total: 5000 W: 1000 L: 1215 D: 2785
5000 @ 10+0.1 th 1 Default check for reducing the value of pawns (125-175)
18-08-14 SC uselessPawns diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 18773 W: 4026 L: 4116 D: 10631
sprt @ 10+0.1 th 1 PMG 120, 210.
18-08-14 SC uselessPawns diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 19221 W: 4093 L: 4182 D: 10946
sprt @ 10+0.1 th 1 PMG 120, 213.
18-08-14 Gua PawnTweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 8792 W: 1806 L: 1935 D: 5051
sprt @ 10+0.1 th 1 PawnTweak (types + psqt)
18-08-14 Gua MidgameLimit2 diff
LLR: 0.13 (-2.94,2.94) [0.00,4.00]
Total: 37008 W: 6481 L: 6352 D: 24175
sprt @ 60+0.6 th 1 Speculative LTC MidgameLimit 15308
18-08-14 Gua DefaultCheck2 diff
ELO: -11.47 +-6.5 (95%) LOS: 0.0%
Total: 5000 W: 1064 L: 1229 D: 2707
5000 @ 10+0.1 th 1 Default check for reducing the value of pawns (only mg)
18-08-14 sni master diff
ELO: 2.22 +-9.0 (95%) LOS: 68.6%
Total: 2500 W: 553 L: 537 D: 1410
1000 @ 10+0.1 th 1 Experiment with fishtest handling of tests with a fixed number of games: what happens if I require 2500 games at the beginning (so that 10 workers of 250 games each are allocated) and *then* lower the limit of the test to 1000 games? Will the server stop the test after 1000 games or 2500 games? The idea would be to find a trick around the current behaviour of current 40000 games regression tests at LTC which taking 4 days to run because of the speed difference of the 250 blocks.
18-08-14 SC pawnTweaks2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 8368 W: 1749 L: 1880 D: 4739
sprt @ 10+0.1 th 1 Lower pawn value even more.
18-08-14 xor fbTweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 23832 W: 5243 L: 5313 D: 13276
sprt @ 10+0.1 th 1 Add 16 from this constant.
18-08-14 sg prune_draw_new diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 15310 W: 3434 L: 3228 D: 8648
sprt @ 10+0.1 th 1 Extend effect only to possible PV nodes.
18-08-14 Gua DefaultCheck diff
ELO: -57.33 +-6.9 (95%) LOS: 0.0%
Total: 5027 W: 889 L: 1711 D: 2427
20000 @ 10+0.1 th 1 Default check for reducing the value of pawns
18-08-14 SC pawnTweaks diff
LLR: 0.45 (-2.94,2.94) [0.00,4.00]
Total: 1000 W: 185 L: 158 D: 657
sprt @ 60+0.6 th 1 -5, 105/100. Speculative LTC
18-08-11 fau FauTune diff
192906/200000 iterations
399888/400000 games played
400000 @ 30+0.3 th 1 Long Tune for mobility
18-08-13 sni capture_history2~1 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 228842 W: 50894 L: 50152 D: 127796
sprt @ 10+0.1 th 1 Double weight of capture history
18-08-13 sni capture_history2 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 31693 W: 7039 L: 7078 D: 17576
sprt @ 10+0.1 th 1 Quadruple weight of capture history
18-08-13 sni pawn_chains10 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 23240 W: 5088 L: 5161 D: 12991
sprt @ 10+0.1 th 1 Pawn chains in the center, again: (A,B) = (4,4)
18-08-13 sni pawn_chains10 diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 69321 W: 15412 L: 15301 D: 38608
sprt @ 10+0.1 th 1 Pawn chains in the center, again: (A,B) = (3,3)
18-08-13 sni mob_initiative5 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 67670 W: 14984 L: 14781 D: 37905
sprt @ 10+0.1 th 1 Pawn mobility and strong pawn center (A,B)=(4,4)
18-08-13 sni demolition7 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 33294 W: 7346 L: 7313 D: 18635
sprt @ 10+0.1 th 1 Demolition in tropism count
18-08-13 sni demolition7 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 23526 W: 5236 L: 5251 D: 13039
sprt @ 10+0.1 th 1 Double tropism coeff in king danger
18-08-13 IIv tune_psqt diff
134927/140000 iterations
280000/280000 games played
280000 @ 10+0.1 th 1 One more try to tune - KING PSQT.
18-08-14 Gua MidgameLimit2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 100367 W: 22379 L: 22145 D: 55843
sprt @ 10+0.1 th 1 MidgameLimit 15308
18-08-13 Viz capture_history2~1 diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 43251 W: 7425 L: 7131 D: 28695
sprt @ 60+0.6 th 1 LTC for 180k games yellowish run with LLR close to 0. Double weight of capture history
18-08-14 Gua MidgameLimit diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 45543 W: 10059 L: 10043 D: 25441
sprt @ 10+0.1 th 1 MidgameLimit 15208
18-08-13 sg prune_draw_new2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 23987 W: 5292 L: 5305 D: 13390
sprt @ 10+0.1 th 1 Do no pruning at all.
18-08-13 sg prune_draw_new diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 9953 W: 2199 L: 2282 D: 5472
sprt @ 10+0.1 th 1 Extend effect also to ALL nodes.
18-08-13 SC pawnTweaks diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 64394 W: 13894 L: 13807 D: 36693
sprt @ 10+0.1 th 1 -5, 105/100.
18-08-13 sg lmr_draw_new2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 46072 W: 10257 L: 10160 D: 25655
sprt @ 10+0.1 th 1 Try the opposite and do only for checks less reductions.
18-08-13 sg prune_draw_new diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 31578 W: 5350 L: 5346 D: 20882
sprt @ 60+0.6 th 1 LTC: No move count pruning if best move is draw.
18-08-13 SC pawnTweaks diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 38355 W: 8179 L: 8194 D: 21982
sprt @ 10+0.1 th 1 -10, 11/10.
18-08-13 SC pawnTweaks diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 26020 W: 5570 L: 5632 D: 14818
sprt @ 10+0.1 th 1 -30, 13/10.
18-08-13 SC pawnTweaks diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 12703 W: 2741 L: 2855 D: 7107
sprt @ 10+0.1 th 1 -20, 12/10.
18-08-13 SC pawnTweaks diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 2945 W: 597 L: 751 D: 1597
sprt @ 10+0.1 th 1 -40, 14/10.
18-08-13 sg prune_draw_new diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 51909 W: 11623 L: 11235 D: 29051
sprt @ 10+0.1 th 1 No move count pruning if best move is draw.
18-08-13 SC pawnTweaks diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 33983 W: 7302 L: 7333 D: 19348
sprt @ 10+0.1 th 1 -5, 11/10.
18-08-13 sg lmr_draw_new diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 37376 W: 8480 L: 8424 D: 20472
sprt @ 10+0.1 th 1 Reduce less only for quiet moves.
18-08-13 sg lmr_draw_new2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23492 W: 5190 L: 5205 D: 13097
sprt @ 10+0.1 th 1 No less reductions for checks.
18-08-13 SC pawnTweaks diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 18508 W: 3922 L: 4013 D: 10573
sprt @ 10+0.1 th 1 -5, 4/3.
18-08-13 sg prune_draw_new2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7748 W: 1657 L: 1750 D: 4341
sprt @ 10+0.1 th 1 Try to skip the static eval draw condition.
18-08-13 SC pawnTweaks diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 7077 W: 1450 L: 1586 D: 4041
sprt @ 10+0.1 th 1 -5, 5/4.
18-08-13 sg lmr_draw_new2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5868 W: 1253 L: 1356 D: 3259
sprt @ 10+0.1 th 1 Reduce less only for quiet moves but without the static eval < draw condition.
18-08-13 SC pawnTweaks diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 6335 W: 1345 L: 1485 D: 3505
sprt @ 10+0.1 th 1 Midgame pawns are not that useless.
18-08-13 sg lmr_draw_new diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 19462 W: 4239 L: 4275 D: 10948
sprt @ 10+0.1 th 1 Reduce less only if also static eval < draw.
18-08-13 SC pawnTweaks diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 884 W: 169 L: 343 D: 372
sprt @ 10+0.1 th 1 Midgame pawns are more useless.
18-08-13 SC pawnTweaks diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 3162 W: 672 L: 827 D: 1663
sprt @ 10+0.1 th 1 Midgame pawns are useless.
18-08-13 SC noCaptures diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 6976 W: 1477 L: 1574 D: 3925
sprt @ 10+0.1 th 1 Captures will not be best moves that easily, if we are leading.
18-08-13 SC iidPruning diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5700 W: 1214 L: 1318 D: 3168
sprt @ 10+0.1 th 1 Prune on high margin iid fail highs.
18-08-13 SC noCaptures diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4851 W: 1012 L: 1120 D: 2719
sprt @ 10+0.1 th 1 Captures will not be best moves that easily.