Stockfish Testing Queue

Finished - 40849 tests

16-11-16 sni unstoppable diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25778 W: 4481 L: 4680 D: 16617
sprt @ 10+0.1 th 1 Remove unstoppable passed pawns calculation, and compensate in the endgame values of the Passed[] array.
16-11-16 Gua Assorted-ThreatByPawnPu diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 13189 W: 2270 L: 2382 D: 8537
sprt @ 10+0.1 th 1 Assorted-ThreatByPawnPush
16-11-15 Voy hp diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 29430 W: 5174 L: 5176 D: 19080
sprt @ 10+0.1 th 1 history pruning tweak
16-11-15 goo No-Depth diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 71511 W: 12691 L: 12655 D: 46165
sprt @ 10+0.1 th 1 No Depth
16-11-15 Voy rf2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 10982 W: 1884 L: 2003 D: 7095
sprt @ 10+0.1 th 1 Take 2...
16-11-15 Voy rf diff
LLR: -3.07 (-2.94,2.94) [0.00,4.00]
Total: 5560 W: 902 L: 1045 D: 3613
sprt @ 10+0.1 th 1 asymmetric reduction formula... i.e. increase reduction more for move count.
16-11-15 Elb rook_pawn diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 46159 W: 8176 L: 8108 D: 29875
sprt @ 10+0.1 th 1 Take 2
16-11-15 Gua Assorted-(40k) diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 43110 W: 7532 L: 7544 D: 28034
sprt @ 10+0.1 th 1 Assorted-(40k)
16-11-15 Roc Tropism9 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13099 W: 2319 L: 2389 D: 8391
sprt @ 10+0.1 th 1 Last attempt to improve the Tropism squares
16-11-15 Roc KSafetyReloaded diff
LLR: -3.17 (-2.94,2.94) [-3.00,1.00]
Total: 9770 W: 1183 L: 1354 D: 7233
sprt @ 60+0.6 th 1 Retry this simplification idea from July 2016 which was green at STC but never tried at LTC !
16-11-15 hex evasionPrunable diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17168 W: 3094 L: 2967 D: 11107
sprt @ 10+0.1 th 1 Remove local variable evasionPrunable, change condition evaluation order
16-11-15 Elb rook_pawn diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 15945 W: 2763 L: 2865 D: 10317
sprt @ 10+0.1 th 1 Rook on pawn tweak
16-11-15 Mys pwn diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16943 W: 2920 L: 2974 D: 11049
sprt @ 10+0.1 th 1 Small simple score bonus for passed pawns that have assistance
16-11-15 Mys pwnn diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 10071 W: 1763 L: 1846 D: 6462
sprt @ 10+0.1 th 1 Also score the block square for us when it's held by us
16-11-09 Roc PawnShelterTuned1 diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 273312 W: 35991 L: 35131 D: 202190
sprt @ 60+0.6 th 1 A) Testing all 4 SPSA sessions values at once, "as is"
16-11-15 Fis SMPLeadMaster diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 9586 W: 1431 L: 1516 D: 6639
sprt @ 5+0.05 th 7 Force the 2nd thread to always be 1 ahead of master.
16-11-15 Roc KSafetyReloaded diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33684 W: 6022 L: 5923 D: 21739
sprt @ 10+0.1 th 1 Retry this simplification idea from July 2016 which was green at STC but never tried at LTC !
16-11-14 Gua Tune-Assorted-bonuses-2 diff
19684/20000 iterations
40000/40000 games played
40000 @ 20+0.2 th 1 Tune ThreatByHangingPawn and tune Hanging
16-11-14 Elb pawncombo diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 99202 W: 17798 L: 17619 D: 63785
sprt @ 10+0.1 th 1 Combo of PawnShelterTuned1 and Lever-tuned1 parameter tweaks
16-11-14 sni more_space diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 19423 W: 3450 L: 3540 D: 12433
sprt @ 10+0.1 th 1 Take 2
16-11-14 sni pawnless diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 19756 W: 3422 L: 3512 D: 12822
sprt @ 10+0.1 th 1 Increase pawnless flank penalty by 50% : S(30,120)
16-11-14 pec fail_high_a diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 12536 W: 1887 L: 1961 D: 8688
sprt @ 5+0.05 th 7 stc 7 th: Another try
16-11-14 Gua Assorted-(30k)-(without diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 26163 W: 4581 L: 4649 D: 16933
sprt @ 10+0.1 th 1 Assorted-(30k tunes without Pawnless-flank tunes)
16-11-14 sni more_space diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 41431 W: 7340 L: 7567 D: 26524
sprt @ 10+0.1 th 1 Simplify space formula
16-11-14 sni pawnless diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 41993 W: 7331 L: 7347 D: 27315
sprt @ 10+0.1 th 1 Double the pawnless flank penalty: S(40,160)
16-11-14 Voy nullRt2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 8484 W: 1434 L: 1561 D: 5489
sprt @ 10+0.1 th 1 take 2
16-11-14 Gua Tune-Assorted-bonuses-1 diff
19522/20000 iterations
40000/40000 games played
40000 @ 20+0.2 th 1 Tune ThreatByPawnPush & Pawnless flank
16-11-14 Voy nullRt diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 23096 W: 4004 L: 4083 D: 15009
sprt @ 10+0.1 th 1 stc
16-11-14 Mys pwn diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 27451 W: 4816 L: 4826 D: 17809
sprt @ 10+0.1 th 1 Open calc into pawn array see how this goes
16-11-14 Gua Assorted-(20k) diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 22745 W: 4001 L: 4080 D: 14664
sprt @ 10+0.1 th 1 Assorted tunes (20k)
16-11-14 Elb bishop_flank diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 15007 W: 1964 L: 2033 D: 11010
sprt @ 10+0.1 th 1 Final try
16-11-14 pec fail_high_a diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 10245 W: 1920 L: 1752 D: 6573
sprt @ 10+0.1 th 1 Another try
16-11-14 Mys check diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 754 W: 109 L: 279 D: 366
sprt @ 10+0.1 th 1 Small attempt to simplify down checks
16-11-11 Gua Lever-tuned1 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 208342 W: 37052 L: 36510 D: 134780
sprt @ 10+0.1 th 1 Changing only the basic parameters (after 70K)
16-11-13 Roc Tropism8 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 70693 W: 12593 L: 12421 D: 45679
sprt @ 10+0.1 th 1 Never score squares occupied by pawns which are blocked by pawn.
16-11-12 atu is_draw diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 158015 W: 27941 L: 27570 D: 102504
sprt @ 10+0.1 th 1 Position::is_draw(): start searching from the 4th ply behind
16-11-13 sg pruning3 diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 49533 W: 6215 L: 6426 D: 36892
sprt @ 60+0.6 th 1 LTC: try to replace my new pruning formula through a simple constant (measured on big bench to keep pruning rate), to check if the formula is really useful and not only the less pruning. Test against my passed patch.
16-11-14 cib delta_stat diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15072 W: 2598 L: 2660 D: 9814
sprt @ 10+0.1 th 1 Using opponent delta stats score in LMR
16-11-13 Roc Tropism8 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 24582 W: 4297 L: 4319 D: 15966
sprt @ 10+0.1 th 1 Fixed test signature
16-11-13 Voy flBonus diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 20929 W: 3673 L: 3710 D: 13546
sprt @ 10+0.1 th 1 fail low stat bonus
16-11-13 Roc Tropism8 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 26622 W: 4609 L: 4623 D: 17390
sprt @ 10+0.1 th 1 Do not score twice the pawn-blocked pawns (similar to what we do for the other checks)
16-11-12 Gua tune-ds diff
19703/20000 iterations
40000/40000 games played
40000 @ 20+0.2 th 1 ds tunes (initial start parameters)
16-11-13 vdv nodecount2 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 26643 W: 4766 L: 4655 D: 17222
sprt @ 10+0.1 th 1 Test alternate implementation of PR #864 for no regression.
16-11-13 sg pruning3 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25303 W: 4546 L: 4433 D: 16324
sprt @ 10+0.1 th 1 try to replace my new pruning formula through a simple constant (measured on big bench to keep pruning rate), to check if the formula is really useful and not only the less pruning. Test against my passed patch.
16-11-13 Gua ds-+-lever diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 37585 W: 6587 L: 6555 D: 24443
sprt @ 10+0.1 th 1 ds + lever tuned1
16-11-13 sg pruning_quiet diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 6121 W: 1044 L: 1143 D: 3934
sprt @ 10+0.1 th 1 Try the same trick now with quiet pruning and compensate in the other direction to get less pruning. Test against my passed pruning patch.
16-11-13 Gua Tunes-+-ds diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13088 W: 2312 L: 2382 D: 8394
sprt @ 10+0.1 th 1 less aggressive lever tunes + ds + ds tunes
16-11-12 Mys ds diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 44234 W: 5580 L: 5557 D: 33097
sprt @ 60+0.6 th 1 This is an attempt of a faster implementation of this yellow run: http://tests.stockfishchess.org/tests/view/582199f90ebc5910626b9c45
16-11-11 sg pruning diff
LLR: 3.44 (-2.94,2.94) [0.00,5.00]
Total: 74863 W: 13381 L: 12924 D: 48558
sprt @ 10+0.1 th 1 Try to compensate maked the patch worse, so do the opposite
16-11-12 Fis specialQuiet diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15734 W: 1940 L: 2008 D: 11786
sprt @ 60+0.6 th 1 LTC