Stockfish Testing Queue

Finished - 3733 tests

19-05-13 sg singular_margin diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 6294 W: 1323 L: 1456 D: 3515
sprt @ 10+0.1 th 1 1/16 more singular margin.
19-05-13 sg root_order diff
LLR: -1.57 (-2.94,2.94) [0.50,4.50]
Total: 102500 W: 23105 L: 22681 D: 56714
sprt @ 10+0.1 th 1 Collect the two last best quiet moves at root (like killers) and use them at root as killers.
19-05-13 sg root_order diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 16679 W: 3692 L: 3774 D: 9213
sprt @ 10+0.1 th 1 Collect the three last best quiet moves at root (like killers) and use the last two of them as root killers (the first move is already the ttMove).
19-05-13 sg root_orderx diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 20071 W: 4461 L: 4526 D: 11084
sprt @ 10+0.1 th 1 Fix: avoid adding duplicate moves.
19-05-13 sg root_order2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 14679 W: 3237 L: 3329 D: 8113
sprt @ 10+0.1 th 1 Collect the two last best quiet moves at root (like killers) and use them for initializaion of root killers.
19-05-12 sg lmr_moves diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 14690 W: 3314 L: 3406 D: 7970
sprt @ 10+0.1 th 1 Less reduction for each 1024th move.
19-05-12 sg lmr_singular diff
LLR: -2.94 (-2.94,2.94) [0.00,3.50]
Total: 103929 W: 17874 L: 17753 D: 68302
sprt @ 60+0.6 th 1 LTC: The opposite seems bad so try less reductions if singular extension was done at this node.
19-05-12 sg lmr_singular2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 41758 W: 9398 L: 9354 D: 23006
sprt @ 10+0.1 th 1 Only at cut nodes.
19-05-12 sg passed4 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 37566 W: 8528 L: 8504 D: 20534
sprt @ 10+0.1 th 1 Add number of passed pawns to weight to add some synergetic effect.
19-05-12 sg lmr_singular2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 15840 W: 3444 L: 3530 D: 8866
sprt @ 10+0.1 th 1 Only at non-cut nodes.
19-05-12 sg passed6 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 12811 W: 2859 L: 2960 D: 6992
sprt @ 10+0.1 th 1 Adjust weight according to the difference of number of passed pawns of both side.
19-05-12 sg lmr_singular2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 11048 W: 2464 L: 2574 D: 6010
sprt @ 10+0.1 th 1 Only at PV nodes.
19-05-12 sg lmr_ply3 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 17210 W: 3776 L: 3855 D: 9579
sprt @ 10+0.1 th 1 Less reduction each eight ply.
19-05-12 sg passed4 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 26193 W: 5826 L: 5860 D: 14507
sprt @ 10+0.1 th 1 Limit additional weight to 1.
19-05-12 sg lmr_singular diff
LLR: 2.95 (-2.94,2.94) [0.50,4.50]
Total: 15224 W: 3524 L: 3283 D: 8417
sprt @ 10+0.1 th 1 The opposite seems bad so try less reductions if singular extension was done at this node.
19-05-12 sg lmr_singular diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 13286 W: 2945 L: 3044 D: 7297
sprt @ 10+0.1 th 1 More reductions if singular extension was done at this node.
19-05-12 sg passed4a diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 23964 W: 5398 L: 5443 D: 13123
sprt @ 10+0.1 th 1 Decrease weight additionally so a single passed pawn get smaller weight.
19-05-12 sg passed5 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 9115 W: 2031 L: 2151 D: 4933
sprt @ 10+0.1 th 1 Add additionally 1 to weight (So the weight changes also for a single passed pawn).
19-05-12 sg passed5 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 2196 W: 436 L: 592 D: 1168
sprt @ 10+0.1 th 1 Decrease weight additionally so a single passed pawn get smaller weight.
19-05-12 sg passed5^ diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 1777 W: 351 L: 510 D: 916
sprt @ 10+0.1 th 1 Use number of passed pawns directly with rank (Weight only changed for multiple passed pawns).
19-05-12 sg passed4 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 18904 W: 4155 L: 4245 D: 10504
sprt @ 10+0.1 th 1 weight+2.
19-05-12 sg passed4^ diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 50247 W: 11412 L: 11375 D: 27460
sprt @ 10+0.1 th 1 weight+1.
19-05-12 sg passed4 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 14382 W: 3173 L: 3281 D: 7928
sprt @ 10+0.1 th 1 weight-1.
19-05-12 sg passed3 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 44341 W: 10041 L: 9984 D: 24316
sprt @ 10+0.1 th 1 Use king proximity also for rank <= 3. Adjust weight to maintain monotonicity by rank.
19-05-12 sg passed4 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 12265 W: 2731 L: 2835 D: 6699
sprt @ 10+0.1 th 1 9/8 quadratic factor in weight formula. maintain max value of 18 for rank 7. rank 4 value decrease from 3 to 2.
19-05-12 sg passed4 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 11540 W: 2532 L: 2640 D: 6368
sprt @ 10+0.1 th 1 3/4 quadratic factor in weight formula. maintain max value of 18 for rank 7. rank 4 value increase from 3 to 6.
19-05-12 sg passed4 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 5192 W: 1096 L: 1235 D: 2861
sprt @ 10+0.1 th 1 7/8 quadratic factor in weight formula. maintain max value of 18 for rank 7. rank 4 value increase from 3 to 5.
19-05-12 sg passed4 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 3825 W: 807 L: 954 D: 2064
sprt @ 10+0.1 th 1 Half quadratic factor in weight formula. maintain max value of 18 for rank 7. rank 4 value increase from 3 to 7.
19-05-12 sg passed3 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 23220 W: 5241 L: 5313 D: 12666
sprt @ 10+0.1 th 1 Do advanced passed pawn evaluation also for rank 3.
19-05-12 sg passed2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 8643 W: 1893 L: 2015 D: 4735
sprt @ 10+0.1 th 1 Scale king proximity by rank / 6.
19-05-11 sg passed2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 20124 W: 4454 L: 4519 D: 11151
sprt @ 10+0.1 th 1 Decrease king proximity Them bonus by 1/8 and increase king proximity Us by 1/8.
19-05-11 sg passed2^ diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 16837 W: 3750 L: 3831 D: 9256
sprt @ 10+0.1 th 1 Increase king proximity bonus by 1/8.
19-05-11 sg passed2^ diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 10344 W: 2265 L: 2379 D: 5700
sprt @ 10+0.1 th 1 Decrease king proximity Us bonus by 1/8 and increase king proximity Them by 1/8.
19-05-11 sg passed2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 9636 W: 2085 L: 2202 D: 5349
sprt @ 10+0.1 th 1 Decrease king proximity bonus by 1/8.
19-05-11 sg passed2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 10078 W: 2221 L: 2346 D: 5511
sprt @ 10+0.1 th 1 Roughly -50% king proximity bonus.
19-05-11 sg passed2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 12837 W: 2878 L: 2992 D: 6967
sprt @ 10+0.1 th 1 Roughly +50% king proximity bonus.
19-05-11 sg passed2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 3825 W: 817 L: 968 D: 2040
sprt @ 10+0.1 th 1 Double up king proximity bonus.
19-05-11 sg passed2 diff
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 47618 W: 10841 L: 10766 D: 26011
sprt @ 10+0.1 th 1 Double king proximity if pawn endgame.
19-05-11 sg combo_stat_bonus_passed diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 137613 W: 31276 L: 30888 D: 75449
sprt @ 10+0.1 th 1 Try combo of my yellow stats bonus and passed tweak (both over 80k games).
19-05-11 sg lmr_ply diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 25132 W: 5647 L: 5686 D: 13799
sprt @ 10+0.1 th 1 Do no LMR for first 5 moves at ply=1 (4 moves failed already yelloe).
19-05-10 sg passed diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 82838 W: 18620 L: 18454 D: 45764
sprt @ 10+0.1 th 1 Use whole file for defended and safe squares calculation.
19-05-10 sg lmr_ply diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 46767 W: 10582 L: 10513 D: 25672
sprt @ 10+0.1 th 1 Do no LMR for first three moves at ply=1.
19-05-11 sg combo_stat_bonus_knight diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 28850 W: 6500 L: 6549 D: 15801
sprt @ 10+0.1 th 1 Try combo of my yellow stats bonus and knight material tweak (both over 80k games).
19-05-10 sg lmr_ply diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 25421 W: 5722 L: 5760 D: 13939
sprt @ 10+0.1 th 1 Do no LMR for first two moves at ply=1.
19-05-06 sg stats_bonus diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 88432 W: 19587 L: 19402 D: 49443
sprt @ 10+0.1 th 1 Tweak stats bonus: decrease quadratic factor and compensate with linear factor to maintain the same stat bonus for depth 1. Max stat bonus slightly decreased.
19-05-09 sg lmr_ply diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 14373 W: 3202 L: 3295 D: 7876
sprt @ 10+0.1 th 1 Now try the first three plies.
19-05-09 sg lmr_ply diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 33398 W: 7491 L: 7489 D: 18418
sprt @ 10+0.1 th 1 Exclude moves from LMR now also the first two plies (not only at the root).
19-05-09 sg lmr_root diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 28083 W: 6299 L: 6324 D: 15460
sprt @ 10+0.1 th 1 Do no LMR at root if goodMove counter > 2.
19-05-09 sg lmr_ply2 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 12455 W: 2759 L: 2862 D: 6834
sprt @ 10+0.1 th 1 Now test at root and ply=1 only at non-PV nodes (because at PV nodes it seems bad in the test).
19-05-09 sg lmr_ply2 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 9873 W: 2105 L: 2221 D: 5547
sprt @ 10+0.1 th 1 Now test first two plies only at PV nodes.