Stockfish Testing Queue

Finished - 43823 tests

17-09-25 SC moveCount2 diff
LLR: -3.13 (-2.94,2.94) [0.00,4.00]
Total: 10495 W: 1807 L: 1937 D: 6751
sprt @ 10+0.1 th 1 Let us try a logarithmic attempt... testing with [0, 4] as a compromise between [-3, 1] and [0, 5]
17-09-25 Voy lmrT diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 28791 W: 5124 L: 5183 D: 18484
sprt @ 10+0.1 th 1 stc
17-09-25 Gua statScore6 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5634 W: 918 L: 1019 D: 3697
sprt @ 10+0.1 th 1 v6
17-09-25 sni protection diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7586 W: 1295 L: 1388 D: 4903
sprt @ 10+0.1 th 1 Take 3, malus=S(0,2)
17-09-25 Gua statScore5 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3415 W: 577 L: 688 D: 2150
sprt @ 10+0.1 th 1 v5
17-09-25 Gua BlockedPawnsTunes2 diff
21614/22500 iterations
43677/45000 games played
45000 @ 20+0.2 th 1 Continuation of settings with updated values + bonus shift by Rocky640
17-09-25 pb0 extraPenalty diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 35390 W: 6339 L: 6315 D: 22736
sprt @ 10+0.1 th 1 Correct logic for refusal
17-09-25 sni protection diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 11480 W: 1967 L: 2044 D: 7469
sprt @ 10+0.1 th 1 Take 2, malus=S(0,4)
17-09-22 And statscoring diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 95885 W: 17069 L: 16792 D: 62024
sprt @ 10+0.1 th 1 Keep extending stat score for reductions
17-09-25 Voy captR diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 14229 W: 2479 L: 2544 D: 9206
sprt @ 10+0.1 th 1 stc
17-09-25 Voy resetStat diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13198 W: 2240 L: 2310 D: 8648
sprt @ 10+0.1 th 1 stc
17-09-24 mco NTB4 diff
LLR: -0.42 (-2.94,2.94) [-3.00,1.00]
Total: 90373 W: 15827 L: 16000 D: 58546
sprt @ 10+0.1 th 1 Natural TB: regression test vs master when no TB is used
17-09-25 Gua BlockedPawns3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8053 W: 1380 L: 1471 D: 5202
sprt @ 10+0.1 th 1 v3
17-09-24 And goodstats2Fixed diff
LLR: -0.15 (-2.94,2.94) [0.00,5.00]
Total: 30594 W: 3804 L: 3715 D: 23075
sprt @ 60+0.6 th 1 Proper version, updated from http://tests.stockfishchess.org/tests/view/59c4758b0ebc5916ff64b5ff. Scored .002 on voyager's tool, NOT .0002
17-09-25 Elb kiap diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 39520 W: 7059 L: 7017 D: 25444
sprt @ 10+0.1 th 1 My take on ianfab's king attack weight for pawns idea: use slightly bigger weights.
17-09-25 fau FauziPsqt2 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 19803 W: 3430 L: 3520 D: 12853
sprt @ 10+0.1 th 1 Fauzi Psqt Tune#2
17-09-23 fau FauziPsqt diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 123259 W: 22022 L: 21763 D: 79474
sprt @ 10+0.1 th 1 Fauzi Psqt Tune#1
17-09-24 sni x-ray2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7204 W: 1258 L: 1353 D: 4593
sprt @ 10+0.1 th 1 Take 3
17-09-21 sg initiative diff
LLR: -2.40 (-2.94,2.94) [0.00,4.00]
Total: 75161 W: 9486 L: 9415 D: 56260
sprt @ 60+0.6 th 1 I used VoyagerOne proposal on the forum and calculated a slope of 0.0014964 for my 'Initiave + 50%' patch so perhaps it scales good. So try LTC with low throughput. Based on current based. http://tests.stockfishchess.org/tests/view/59bb4bb80ebc5916ff64b313
17-09-24 vdv furtherExcept diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 86940 W: 15112 L: 15412 D: 56416
sprt @ 10+0.1 th 1 stc
17-09-24 fau KingMGINC diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 79491 W: 14096 L: 13986 D: 51409
sprt @ 10+0.1 th 1 King Psqt +5% for MG
17-09-23 And statscoring2 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 76150 W: 13397 L: 13205 D: 49548
sprt @ 10+0.1 th 1 Another variant. Has better search path on the bench at high depths, which may be an indicator.
17-09-24 fau KingEGINC diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 38223 W: 6745 L: 6773 D: 24705
sprt @ 10+0.1 th 1 King Psqt +5% for EG
17-09-24 Elb avoidrep2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 25190 W: 4511 L: 4529 D: 16150
sprt @ 10+0.1 th 1 Have another try at this: strong decrease in reduction for moves that escape a repetition
17-09-23 Gua BlockedPawnsTunes diff
17183/17500 iterations
35000/35000 games played
35000 @ 20+0.2 th 1 space bonuses by blocked pawns rank (tunes)
17-09-24 vdv ScoreTime diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16638 W: 2888 L: 2943 D: 10807
sprt @ 10+0.1 th 1 stc
17-09-24 fau KingMGDEC diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 24828 W: 4361 L: 4434 D: 16033
sprt @ 10+0.1 th 1 King Psqt -5% for MG
17-09-24 fau KingEGDEC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 23299 W: 4032 L: 4110 D: 15157
sprt @ 10+0.1 th 1 King Psqt -5% for EG
17-09-24 ian pawn_attack2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 12841 W: 2248 L: 2319 D: 8274
sprt @ 10+0.1 th 1 tweak definition of king attackers
17-09-24 sni x-ray2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3690 W: 710 L: 822 D: 2158
sprt @ 10+0.1 th 1 Take 2
17-09-24 And consecutiveLosses diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8856 W: 1521 L: 1609 D: 5726
sprt @ 10+0.1 th 1 Increase reduction if we are following a heavily losing line, and the move we are about to try has a poor history.
17-09-24 Gua BlockedPawns2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 22889 W: 4078 L: 4106 D: 14705
sprt @ 10+0.1 th 1 take 2 (close to average between v1 and after tunes)
17-09-24 sni superweak' diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11700 W: 2088 L: 2163 D: 7449
sprt @ 10+0.1 th 1 Also consider superweak empty squares
17-09-22 fau KnightEGDEC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 107902 W: 18967 L: 18764 D: 70171
sprt @ 10+0.1 th 1 Knight Psqt -10% for EG
17-09-24 And consecutiveGains diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 16029 W: 2914 L: 2971 D: 10144
sprt @ 10+0.1 th 1 Lower LMR Reduction for very good moves that follow a successful line of gaining moves.
17-09-22 fau QueenEGINC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 69415 W: 12395 L: 12317 D: 44703
sprt @ 10+0.1 th 1 Queen Psqt +10% for EG
17-09-22 fau BishopEGINC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 58953 W: 10356 L: 10315 D: 38282
sprt @ 10+0.1 th 1 Bishop Psqt -10% for MG
17-09-23 sni statscore_compare diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 27753 W: 4803 L: 4813 D: 18137
sprt @ 10+0.1 th 1 Another try (idea by Michael Byrne)
17-09-22 Voy seReduction diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 55576 W: 9899 L: 9791 D: 35886
sprt @ 10+0.1 th 1 stc
17-09-23 sni x-ray2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8696 W: 1478 L: 1566 D: 5652
sprt @ 10+0.1 th 1 x-ray thru enemy pieces attacked by pawn
17-09-23 And goodroots2Updated diff
LLR: -1.84 (-2.94,2.94) [0.00,5.00]
Total: 2773 W: 451 L: 517 D: 1805
sprt @ 10+0.1 th 1 Yellow STC. Slope of ~.0002 in Voyager's tool. This change looks depth (TC) dependent.
17-09-23 Gua Reductions8 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13408 W: 2277 L: 2346 D: 8785
sprt @ 10+0.1 th 1 Reductions for case if opponent move count >30
17-09-22 fau BishopMGDEC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 48417 W: 8508 L: 8502 D: 31407
sprt @ 10+0.1 th 1 Bishop Psqt -10% for MG
17-09-21 And goodroots2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 76252 W: 13818 L: 13619 D: 48815
sprt @ 10+0.1 th 1 Take 2. For actual root, and opponents 'roots'
17-09-22 fau KnightMGDEC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 47548 W: 8324 L: 8321 D: 30903
sprt @ 10+0.1 th 1 Knight Psqt -10% for MG
17-09-22 fau QueenEGDEC diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 30594 W: 5345 L: 5399 D: 19850
sprt @ 10+0.1 th 1 Queen Psqt -10% for EG
17-09-22 sni factor2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 45343 W: 8289 L: 8220 D: 28834
sprt @ 10+0.1 th 1 Anticipating endgame values more
17-09-23 sni passed_pawns_rank4 diff
LLR: -2.32 (-2.94,2.94) [0.00,4.00]
Total: 23760 W: 4154 L: 4197 D: 15409
sprt @ 10+0.1 th 1 relative rank >= 4
17-09-22 fau QueenMGDEC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 33325 W: 5879 L: 5923 D: 21523
sprt @ 10+0.1 th 1 Queen Psqt -10% for MG
17-09-22 fau BishopMGINC diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 28179 W: 5006 L: 5067 D: 18106
sprt @ 10+0.1 th 1 Bishop Psqt +10% for MG