Stockfish Testing Queue

Finished - 17428 tests

27-04-17 Sa RecursiveLMR diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5852 W: 968 L: 1068 D: 3816
sprt @ 10+0.1 th 1 LMR cutoff if any of iterated search depths fails low.
27-04-17 Vo kingDanger2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5715 W: 966 L: 1067 D: 3682
sprt @ 10+0.1 th 1 stc
27-04-17 Vo qsPruning diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 2310 W: 370 L: 520 D: 1420
sprt @ 10+0.1 th 1 Don't do SEE pruning at qs root. Test against recent passed patch.
27-04-17 sn singular9 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7527 W: 1309 L: 1402 D: 4816
sprt @ 10+0.1 th 1 Take 2
27-04-17 sn singular9 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 11214 W: 1984 L: 2062 D: 7168
sprt @ 10+0.1 th 1 Singular search tweak
27-04-17 pb root_move_lmr diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 18516 W: 3448 L: 3244 D: 11824
sprt @ 10+0.1 th 1 As queue rather empty, testing a weird root-move-lmr idea.
17-04-17 vd rsf diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 136547 W: 17839 L: 18167 D: 100541
sprt @ 60+0.6 th 1 ltc
25-04-17 Vo lmrT diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17748 W: 2247 L: 2308 D: 13193
sprt @ 60+0.6 th 1 Would like to see how this fare at LTC... Low Throughput.
26-04-17 vd killerIsBonus diff
LLR: -2.98 (-2.94,2.94) [-3.00,1.00]
Total: 37391 W: 6607 L: 6828 D: 23956
sprt @ 10+0.1 th 1 stc, alternative
25-04-17 Ro MinAttackers diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35382 W: 4650 L: 4549 D: 26183
sprt @ 60+0.6 th 1 Remove another cap
25-04-17 II king_safety diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 119975 W: 21533 L: 21284 D: 77158
sprt @ 10+0.1 th 1 King safety - take 3 (tuned values).
26-04-17 sn passed_pawns2' diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 45061 W: 5879 L: 5850 D: 33332
sprt @ 60+0.6 th 1 LTC: last try, exclude candidate passers. Tested on top of master with patch "copy killers".
24-04-17 pb smp_skip_move diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 19375 W: 3069 L: 3116 D: 13190
sprt @ 5+0.05 th 7 On helpers skip root move, if it has poor chances to raise alpha. Turns out when another thread with higher completed depth reports given move as best, but it's score is already overtrumped.
27-04-17 Vo kingDanger diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 42260 W: 7584 L: 7530 D: 27146
sprt @ 10+0.1 th 1 stc
26-04-17 Vo pruning2 diff
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 24474 W: 3274 L: 3076 D: 18124
sprt @ 60+0.6 th 1 LTC: Don't do InCheck evasion pruning at the root of QS tree
27-04-17 Fi adaptiveSort diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7857 W: 1391 L: 1483 D: 4983
sprt @ 10+0.1 th 1 Only sort quiets within a margin of best quiet.
26-04-17 fa MixTune diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 37728 W: 6833 L: 6861 D: 24034
sprt @ 10+0.1 th 1 A mix of some tuning of mine
26-04-17 vd ExclStat diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 36133 W: 6538 L: 6509 D: 23086
sprt @ 10+0.1 th 1 stc, against PR1085.
26-04-17 vd moveInit diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 28060 W: 5020 L: 5081 D: 17959
sprt @ 10+0.1 th 1 stc, as discussed in #1085 delay copy
26-04-17 vd threading2 diff
ELO: -4.17 +-5.9 (95%) LOS: 8.3%
Total: 4000 W: 578 L: 626 D: 2796
4000 @ 5+0.05 th 11 elo estimate v4
26-04-17 Vo pruning2 diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 34603 W: 6441 L: 6167 D: 21995
sprt @ 10+0.1 th 1 fixed patch...
26-04-17 sn passed_pawns2 diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 36029 W: 6667 L: 6388 D: 22974
sprt @ 10+0.1 th 1 Last try: exclude candidate passers
25-04-17 II tune_stats diff
27681/110000 iterations
56207/220000 games played
220000 @ 10+0.1 th 1 Continue tuning history stats with better ck values.
26-04-17 fa BPtry2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15876 W: 2842 L: 2900 D: 10134
sprt @ 10+0.1 th 1 Second try for a very simple BP(Them)
26-04-17 Vo qsNoPvPrune diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 20601 W: 3675 L: 3713 D: 13213
sprt @ 10+0.1 th 1 No PV pruning in QS
26-04-17 fa BPP diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23585 W: 4127 L: 4153 D: 15305
sprt @ 10+0.1 th 1 Adding a condition to BishopPawns penalty:We need to have at least 4 pawns OTB
26-04-17 jo outpost_depth1 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 17355 W: 3060 L: 3158 D: 11137
sprt @ 10+0.1 th 1 Now with some more tuning on pawns, each of them with not less than 1 million games.
26-04-17 Vo pruning2 diff
LLR: -0.29 (-2.94,2.94) [0.00,5.00]
Total: 336 W: 54 L: 65 D: 217
sprt @ 10+0.1 th 1 stc
25-04-17 fa TuneBPE diff
19669/20000 iterations
40000/40000 games played
40000 @ 10+0.1 th 1 I also noticed that we need to have at least 3 pawns eligible for the BP penalty in order for this penalty to be accurate, a very fast tune to check if the values of the penalty itself needs to be changed (Very different than BPP tune, so we can normally run them in parallel)
26-04-17 fa BPE diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 5584 W: 958 L: 1060 D: 3566
sprt @ 10+0.1 th 1 Adding a condition to BishopPawns penalty:We need to have at least 3 pawns eligible for the penalty
26-04-17 sn passed_pawns2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 24515 W: 4389 L: 4410 D: 15716
sprt @ 10+0.1 th 1 relative rank >= RANK_3 (submitted with priority -1)
25-04-17 fa TuneBPP diff
19740/20000 iterations
40000/40000 games played
40000 @ 10+0.1 th 1 I realized that we need to have at least 4 pawns in order for this penalty to be accurate, but a very fast tune to check if the values of the penalty itself needs to be changed
26-04-17 sn passed_pawns2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15469 W: 2768 L: 2827 D: 9874
sprt @ 10+0.1 th 1 factor 1/3
26-04-17 Vo pruning diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 19038 W: 3346 L: 3438 D: 12254
sprt @ 10+0.1 th 1 stc
26-04-17 sn passed_pawns2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5582 W: 920 L: 1021 D: 3641
sprt @ 10+0.1 th 1 factor 1/5
25-04-17 Vo movePlayed diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 57916 W: 10420 L: 10300 D: 37196
sprt @ 10+0.1 th 1 stc
25-04-17 sn passed_pawns2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 45850 W: 8244 L: 8175 D: 29431
sprt @ 10+0.1 th 1 Take 3
25-04-17 vd killerCopy diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 28368 W: 3730 L: 3619 D: 21019
sprt @ 60+0.6 th 1 ltc, test as bugfix
22-04-17 II tune_king diff
19341/20000 iterations
39114/40000 games played
40000 @ 30+0.3 th 1 Further tune king safety. Throughput 300 because of the used TC.
25-04-17 st null_worked diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 976 W: 130 L: 252 D: 594
sprt @ 10+0.1 th 1 Record in the TTEntry if a null move search resulted in a cutoff. If not, record it with a flag so that next time we meet this position we don't try null move.
24-04-17 fa BPtu diff
37971/40000 iterations
80000/80000 games played
80000 @ 10+0.1 th 1 Tuning BP starting from 0 based on the number of pawns eligible for the penalty so we can try to find a more accurate formula to use.
24-04-17 Ro MinAttackers diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 52048 W: 9397 L: 9329 D: 33322
sprt @ 10+0.1 th 1 Remove another cap
25-04-17 vd killerCopy diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 70845 W: 12752 L: 12716 D: 45377
sprt @ 10+0.1 th 1 stc, test as bugfix
22-04-17 II tune_stats diff
38402/150000 iterations
78987/300000 games played
300000 @ 10+0.1 th 1 Trying to reveal true structure of history stats. Throughput 300 because of the number of games used. I cut updates at depth 12 because it is appropriate so for STC, and plan to extend to higher depths by analogy.
24-04-17 pb smp_skip_move diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 32240 W: 5231 L: 5228 D: 21781
sprt @ 5+0.05 th 5 Take 2
25-04-17 sn passed_pawns2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 26504 W: 4720 L: 4733 D: 17051
sprt @ 10+0.1 th 1 Take 2
23-04-17 jo lazy_skip2 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 20272 W: 3171 L: 3215 D: 13886
sprt @ 5+0.05 th 7 Also skip lower depths than the main thread already finished searching. Seems to be a non-issue at shallow depths.
23-04-17 sg sort_tweak diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 21408 W: 3833 L: 3917 D: 13658
sprt @ 10+0.1 th 1 My LTC test seems not good so far so perhaps more sorting is needed at higher depths. Test against passed sortLinearRebase patch.
25-04-17 Vo cmhStatsTT diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 16883 W: 2963 L: 3017 D: 10903
sprt @ 10+0.1 th 1 see how this goes...
25-04-17 Fi d2a90288b0fe7d44f77a387 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 12318 W: 2117 L: 2232 D: 7969
sprt @ 10+0.1 th 1 Same as http://tests.stockfishchess.org/tests/view/58fbc2400ebc59035df33e4b except 0 value quiets will not be sorted at depth 1.