Stockfish Testing Queue

Finished - 52819 tests

17-05-11 sni overprotection diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 10969 W: 1959 L: 2038 D: 6972
sprt @ 10+0.1 th 1 Small bonus for overprotecting our weak pawns
17-05-11 Voy qs-pt diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 29099 W: 5252 L: 5253 D: 18594
sprt @ 10+0.1 th 1 stc
17-05-11 sg qs_nullmove diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 18415 W: 3325 L: 3372 D: 11718
sprt @ 10+0.1 th 1 Fix bug. Forgot to correct my tempo to get the same eval as static eval if no good opponent move exists. Stop other test
17-05-11 sg qs_nullmove diff
LLR: -1.93 (-2.94,2.94) [0.00,5.00]
Total: 5136 W: 915 L: 975 D: 3246
sprt @ 10+0.1 th 1 Try in qsearch a null move search to get a safer stand pat value (this recognize threats like hanging pieces)
17-05-11 SC quadraticLoss01 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7334 W: 1284 L: 1378 D: 4672
sprt @ 10+0.1 th 1 And take 3.
17-05-11 SC quadraticLoss01 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5664 W: 986 L: 1087 D: 3591
sprt @ 10+0.1 th 1 A locally tuned version. Take 2.
17-05-10 Voy ept diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 58835 W: 10578 L: 10454 D: 37803
sprt @ 10+0.1 th 1 Evasion Prunable Tweak
17-05-10 vdv int16 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 54476 W: 9949 L: 9886 D: 34641
sprt @ 10+0.1 th 1 stc (just test the int16 bit for no regression).
17-05-11 Sap LMRCutNodeCheck diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 40092 W: 7192 L: 7148 D: 25752
sprt @ 10+0.1 th 1 Make a special case for checks in LMR.
17-05-11 vdv int16v2 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 27019 W: 4905 L: 4794 D: 17320
sprt @ 10+0.1 th 1 take 2
17-05-11 mar excluded2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 31933 W: 5758 L: 5748 D: 20427
sprt @ 10+0.1 th 1 Check some tweaks on how excludedMove is handled
17-05-11 pb0 flatten_out diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 11924 W: 2096 L: 2171 D: 7657
sprt @ 10+0.1 th 1 Reduce bonus for higher depths. Local STC test after 1k: W: 199 L: 162 D: 639
17-05-10 Sap ReduceDepthForRetreat diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 33935 W: 6186 L: 6166 D: 21583
sprt @ 10+0.1 th 1 Excluding reductions for checks and evasions should improve tactics and specifically perpetual check.
17-05-10 pb0 quiet_checks diff
LLR: -3.56 (-2.94,2.94) [0.00,5.00]
Total: 28043 W: 5102 L: 5133 D: 17808
sprt @ 10+0.1 th 1 Take 2
17-05-10 Voy ept2 diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 15778 W: 2976 L: 2783 D: 10019
sprt @ 10+0.1 th 1 Ver. 2
17-05-10 sg quiets2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8435 W: 1493 L: 1582 D: 5360
sprt @ 10+0.1 th 1 Fix problem with triggered assert. See my comment here https://github.com/locutus2/Stockfish/compare/0c1f119...078aac2
17-05-10 vdv tosq diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8621 W: 1499 L: 1588 D: 5534
sprt @ 10+0.1 th 1 stc
17-05-10 SC quadraticLoss01 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 4222 W: 749 L: 857 D: 2616
sprt @ 10+0.1 th 1 Many weaknesses is worse than many strengths is good. Sanity check.
17-05-10 Roc LooseP diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 21812 W: 3906 L: 3939 D: 13967
sprt @ 10+0.1 th 1 Take 3: Do not x-ray through opponent blockers of same type.
17-05-10 Sap LMROurNullMove diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16894 W: 3133 L: 3185 D: 10576
sprt @ 10+0.1 th 1 Reduce depth if we made a null move previously.
17-05-10 sg quiets diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15320 W: 2760 L: 2820 D: 9740
sprt @ 10+0.1 th 1 Exclude quiet checks for negative stats update.
17-05-10 vdv fromto diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 2946 W: 470 L: 583 D: 1893
sprt @ 10+0.1 th 1 stc
17-05-10 pb0 quiete_checks diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 6732 W: 1188 L: 1285 D: 4259
sprt @ 10+0.1 th 1 Take 3: to measure if the additional logic in take 2 brings a benefit at all
17-05-10 Fis catchUpDepth diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 31855 W: 5284 L: 5037 D: 21534
sprt @ 5+0.05 th 7 Take 1 is struggling so set it to pri -1. This version is +5.7 elo after 8k local games.
17-05-09 sni qs4 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 87947 W: 15890 L: 15883 D: 56174
sprt @ 10+0.1 th 1 Always use ttDepth = -1 in quiescence search.
17-05-10 Voy recap diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 87023 W: 15807 L: 15562 D: 55654
sprt @ 10+0.1 th 1 stc
17-05-08 IIv king_safety diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 20992 W: 2593 L: 2687 D: 15712
sprt @ 60+0.6 th 1 I'm sorry for a direct LTC, but the tuning was done at 30+0.3 and the main detected contibuting factor (- 814 * !pos.count<QUEEN>(Them)) went into direction that favors LTC - throughput 200.
17-05-08 fau tuneAT1 diff
48523/50000 iterations
100000/100000 games played
100000 @ 10+0.1 th 1 Tuning King Distance Attacker (Step 2)
17-05-09 pb0 quiete_checks diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 89991 W: 16512 L: 16252 D: 57227
sprt @ 10+0.1 th 1 Quiet checking moves aren't very quiet actually. Do we overemphasize them when updating stats?
17-05-10 mbo bestThreadTT diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5838 W: 954 L: 1054 D: 3830
sprt @ 5+0.05 th 3 Allow best thread to overwrite more TT entries. Second try with three threads, because local results were better with three.
17-05-10 mbo bestThreadTT diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 1620 W: 218 L: 335 D: 1067
sprt @ 5+0.05 th 7 Allow best thread to overwrite more TT entries.
17-05-10 sni qs4 diff
LLR: -1.38 (-2.94,2.94) [-3.00,1.00]
Total: 4659 W: 805 L: 886 D: 2968
sprt @ 10+0.1 th 1 Take 2: Always use ttDepth = -2 in quiescence search
17-05-09 sg passed_pawn_break3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 56109 W: 10123 L: 10010 D: 35976
sprt @ 10+0.1 th 1 Rewrite my last passed pawn break patch (see https://github.com/official-stockfish/Stockfish/pull/1110) and add the case of a pawn break defended by two opponent pawns but also supported by an own pawn.
17-05-10 Roc LooseP diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 34906 W: 6330 L: 6306 D: 22270
sprt @ 10+0.1 th 1 Take 2(fixed), include xrays through first blocker (including pawn) and stopping at 2nd blocker.
17-05-09 Sap LMROpponentNullMove diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 47272 W: 8558 L: 8482 D: 30232
sprt @ 10+0.1 th 1 LMR reduce if opponent played a null move
17-05-09 sg passed_pawn_break3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 41821 W: 7596 L: 7543 D: 26682
sprt @ 10+0.1 th 1 An optimized version of my last try. Speedup 0.4%. Set other test to prio -1.
17-05-08 sg master diff
ELO: 20.12 +-1.9 (95%) LOS: 100.0%
Total: 40000 W: 7460 L: 5146 D: 27394
40000 @ 10+0.1 th 1 Regression test until "Linear Protector bonus by distance" @ STC for comparison (see https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/sn6zUU8uebs)
17-05-09 Voy onlyOneRecap diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 29964 W: 5409 L: 5407 D: 19148
sprt @ 10+0.1 th 1 stc
17-05-09 Sap LMRBadCaptures diff
LLR: -3.06 (-2.94,2.94) [0.00,5.00]
Total: 33531 W: 6060 L: 6047 D: 21424
sprt @ 10+0.1 th 1 Do not extend LMR searches of bad captures in endgames.
17-05-09 Voy recapPromotion diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 12420 W: 2232 L: 2304 D: 7884
sprt @ 10+0.1 th 1 Include promotions with recaputres
17-05-09 fau KATT diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 24701 W: 4427 L: 4447 D: 15827
sprt @ 10+0.1 th 1 First test with King Distance Attacker, I expect it to fail, still need some work but lets see how it fares.
17-05-09 Voy removeSearchTemplates diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 16723 W: 2949 L: 3049 D: 10725
sprt @ 10+0.1 th 1 stc
17-05-09 Voy rcMoveGen diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 19349 W: 2467 L: 2522 D: 14360
sprt @ 60+0.6 th 1 LTC: Move Generation for recaptures
17-05-09 Elb npmscale diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 36019 W: 6428 L: 6401 D: 23190
sprt @ 10+0.1 th 1 Last try in this series
17-05-09 Sap TemplateCutNode diff
LLR: -1.99 (-2.94,2.94) [-3.00,1.00]
Total: 60711 W: 10963 L: 11173 D: 38575
sprt @ 10+0.1 th 1 I measured a 1.8% speedup on bench. Testing for no regression.
17-05-08 Voy rcMoveGen diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 53757 W: 9790 L: 9438 D: 34529
sprt @ 10+0.1 th 1 Move Generation for recaptures
17-05-09 Elb npmscale diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 4317 W: 754 L: 861 D: 2702
sprt @ 10+0.1 th 1 Try out tuned values
17-05-08 Elb nmpscale_tune diff
19383/20000 iterations
39478/40000 games played
40000 @ 20+0.2 th 1 Tune a variant of the nmpscale idea
17-05-09 Roc LooseP diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 20591 W: 3726 L: 3763 D: 13102
sprt @ 10+0.1 th 1 More xrays
17-05-08 sni qs4 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 23381 W: 4184 L: 4261 D: 14936
sprt @ 10+0.1 th 1 Change TT depth used in quiescence search