Stockfish Testing Queue

Finished - 31484 tests

16-04-20 sni weak_pawn diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 24101 W: 4347 L: 4546 D: 15208
sprt @ 10+0.1 th 1 Try doubled pawn simplification
16-04-21 Roc MinorControl2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5200 W: 892 L: 995 D: 3313
sprt @ 10+0.1 th 1 (re) Take 2.
16-04-20 jos probcut diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 14122 W: 2487 L: 2552 D: 9083
sprt @ 10+0.1 th 1 Try to be more careful in ProbCut.
16-04-20 Elb piece_checked diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 27227 W: 3629 L: 3704 D: 19894
sprt @ 60+0.6 th 1 Adjusted values for checks from Minor pieces. As this is kingsafety related and thus time control dependant, try at LTC.
16-04-20 sg lmr diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 28183 W: 4989 L: 5050 D: 18144
sprt @ 10+0.1 th 1 With the fmh and fmh2 patches the move ordering is now better. So try more aggressive lmr (+10%).
16-04-20 sni tweak_connected diff
LLR: -1.47 (-2.94,2.94) [0.00,4.00]
Total: 5065 W: 906 L: 967 D: 3192
sprt @ 10+0.1 th 1 Try simpler connected pawns value for endgame
16-04-19 sni opposed_levers4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10979 W: 1977 L: 2055 D: 6947
sprt @ 10+0.1 th 1 Lever bonus by weakness flag and rank (a lever is weak when our lever pawn is unsupported or opposed by an enemy pawn on the same file).
16-04-19 Elb piece_checked diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 72458 W: 12980 L: 12891 D: 46587
sprt @ 10+0.1 th 1 Adjusted values for checks from Minor pieces
16-04-19 sni captures diff
LLR: -1.69 (-2.94,2.94) [0.00,5.00]
Total: 18833 W: 3434 L: 3425 D: 11974
sprt @ 10+0.1 th 1 Prefer captures near our king during move ordering, take 2
16-04-19 Voy fm2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 15370 W: 2775 L: 2879 D: 9716
sprt @ 10+0.1 th 1 Don't update fm2 for fail-low and prior TT move. Use [0,4] as I don't consider this as a "simplification".
16-04-19 sni captures diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 23404 W: 4230 L: 4255 D: 14919
sprt @ 10+0.1 th 1 Prefer captures near our king during move ordering
16-04-19 sg king_mobility diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18501 W: 3351 L: 3397 D: 11753
sprt @ 10+0.1 th 1 Give penalty if king has no mobility (only for endgame)
16-04-19 sni captures diff
LLR: -1.80 (-2.94,2.94) [0.00,5.00]
Total: 10725 W: 1972 L: 2002 D: 6751
sprt @ 10+0.1 th 1 Prefer captures from external files during move ordering
16-04-19 the loose_bonus diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7097 W: 1267 L: 1362 D: 4468
sprt @ 10+0.1 th 1 Take 2: lower values.
16-04-18 pec tm diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 27516 W: 4917 L: 4980 D: 17619
sprt @ 10+0.1 th 1 See if recent search improvements made more time available
16-04-18 sni captures diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 16360 W: 2926 L: 2982 D: 10452
sprt @ 10+0.1 th 1 Prefer captures on central files during move ordering
16-04-19 the loose_bonus diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 10869 W: 1932 L: 2011 D: 6926
sprt @ 10+0.1 th 1 Take 1: add bonus for extra loose pieces. Corrected bench number.
16-04-18 sni king_walk2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15825 W: 2851 L: 2909 D: 10065
sprt @ 10+0.1 th 1 Take 3: try another formula
16-04-19 Mys SMP1 diff
LLR: -0.35 (-2.94,2.94) [0.00,5.00]
Total: 81 W: 7 L: 21 D: 53
sprt @ 5+0.05 th 7 Second take on min thread depth skip size, this time limited to root which I hope will improve base search quality.
16-04-18 Mys up diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 38981 W: 7082 L: 7041 D: 24858
sprt @ 10+0.1 th 1 Expand on an attempt by VoyagerOne on updating fmh at leaf. Include fmh, fmh2 & cmh.
16-04-18 Elb scale_tweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 24050 W: 4353 L: 4427 D: 15270
sprt @ 10+0.1 th 1 Tweaked evaluate_scale_factor values based on local tuning. (Take 2)
16-04-18 Mys yc diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 17758 W: 3177 L: 3273 D: 11308
sprt @ 10+0.1 th 1 Come on yellow combo! Each commit contains the link to the yellow test.
16-04-18 sni king_walk2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9826 W: 1755 L: 1838 D: 6233
sprt @ 10+0.1 th 1 Take 2: only count the squares which strictly decrease the king distance
16-04-18 sni opposed_levers4 diff
LLR: -2.42 (-2.94,2.94) [0.00,5.00]
Total: 17152 W: 3140 L: 3169 D: 10843
sprt @ 10+0.1 th 1 Take 4
16-04-18 Elb scale_tweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 14139 W: 2487 L: 2595 D: 9057
sprt @ 10+0.1 th 1 Tweaked evaluate_scale_factor values based on local tuning.
16-04-17 Mys SMP diff
LLR: -2.22 (-2.94,2.94) [0.00,4.00]
Total: 12840 W: 2029 L: 2104 D: 8707
sprt @ 5+0.05 th 7 Would like to test effect of a constant larger skip size for thread depth allocation.
16-04-18 Voy fmh-lmr diff
LLR: 2.99 (-2.94,2.94) [0.00,5.00]
Total: 52232 W: 9654 L: 9304 D: 33274
sprt @ 10+0.1 th 1 Try to use fmhs for lmr.
16-04-17 cib fmh2init2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 22311 W: 3941 L: 3972 D: 14398
sprt @ 10+0.1 th 1 Using counter and follow up move history for previous move at root position.
16-04-18 cib cmh_fix2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13150 W: 2331 L: 2400 D: 8419
sprt @ 10+0.1 th 1 Resubmitting with correct master. Take 2.
16-04-18 Roc EndGameWeakness2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 8871 W: 1619 L: 1707 D: 5545
sprt @ 10+0.1 th 1 Take 2, smaller bonus
16-04-18 Roc EndGameWeakness2 diff
LLR: -0.86 (-2.94,2.94) [0.00,5.00]
Total: 900 W: 164 L: 197 D: 539
sprt @ 10+0.1 th 1 Closing in against pawn weakness (fixed)
16-04-17 cib cmh_fix2 diff
LLR: -1.39 (-2.94,2.94) [0.00,5.00]
Total: 32770 W: 5917 L: 5837 D: 21016
sprt @ 10+0.1 th 1 Take 2.
16-04-17 sg 7eaea3848c9e8a388c0b79c diff
ELO: -2.29 +-4.0 (95%) LOS: 13.3%
Total: 11395 W: 2234 L: 2309 D: 6852
20000 @ 5+0.1 th 1 This is a short test of the state Info patch against itself to check for crashes (see forum)
16-04-17 sg ee7a68ea5fc07e331df462b diff
ELO: 2.86 +-4.2 (95%) LOS: 90.7%
Total: 10433 W: 2161 L: 2075 D: 6197
20000 @ 5+0.1 th 1 As benchmark check the patch before StateInfo for crashes.
16-04-17 sg fmh2_no_update diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 12833 W: 2205 L: 2383 D: 8245
sprt @ 10+0.1 th 1 Simplification: Removing fmh2 completely was a regression. fmh and fmh2 use the same history table, so try using fmh2 only for move ordering and remove the additional fmh2 updates.
16-04-17 Roc KingManeuver diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9637 W: 1720 L: 1804 D: 6113
sprt @ 10+0.1 th 1 Smaller penalty
16-04-17 sni opposed_levers2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 44144 W: 8125 L: 8061 D: 27958
sprt @ 10+0.1 th 1 Take 4. Penalties for opposed levers on rank 2 and 3, because they are often a sign that the opponent is trying to create a passed pawn (for instance White b2, Black c3 and b4).
16-04-17 sni king_walk diff
LLR: -2.43 (-2.94,2.94) [0.00,5.00]
Total: 5686 W: 1062 L: 1141 D: 3483
sprt @ 10+0.1 th 1 Test a variant of Alain's king walk idea
16-04-17 sni knight3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5142 W: 915 L: 1019 D: 3208
sprt @ 10+0.1 th 1 Avoid pawn exchanges when we have knights. Tested on top of the passed fmh2 patch.
16-04-17 Mys SMP diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13416 W: 2161 L: 2231 D: 9024
sprt @ 5+0.05 th 7 Min skip depth for thread allocations
16-04-17 cib cmh_fix diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 72849 W: 13271 L: 13086 D: 46492
sprt @ 10+0.1 th 1 Using same method to get cmh for movepick and Pruning/LMR.
16-04-17 sg fmh2_nullmove diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9739 W: 1707 L: 1791 D: 6241
sprt @ 10+0.1 th 1 Use CounterMoveStats also for null moves
16-04-15 cib fmh2init diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 89871 W: 16182 L: 15928 D: 57761
sprt @ 10+0.1 th 1 Checking if we can use the fmh and fmh2 concept for the first two moves. Testing on top of the passed fmh2 patch
16-04-17 Roc KingManeuver diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13286 W: 2421 L: 2489 D: 8376
sprt @ 10+0.1 th 1 Take 2
16-04-15 Elb attack_unit2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 84017 W: 11245 L: 11169 D: 61603
sprt @ 60+0.6 th 1 Since http://tests.stockfishchess.org/tests/view/570f879b0ebc59301a354554 failed yellow, try to see if it scales better at LTC. Will stop the test if not positive after 5000 games.
16-04-16 Roc KingManeuver diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 3993 W: 677 L: 786 D: 2530
sprt @ 10+0.1 th 1 First attempt to improve the King play in end game
16-04-16 Voy statInterations diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 6023 W: 1032 L: 1132 D: 3859
sprt @ 10+0.1 th 1 Prevent stats from interacting by using two CounterMoveHistory tables. (Maybe this is why CMH2 failed) First test with just fmh, before trying cmh.
16-04-15 sni opposed_levers2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 48341 W: 6438 L: 6396 D: 35507
sprt @ 60+0.6 th 1 LTC: Use opposed flag to score levers, take 3 (only change values for midgame). Tested on top of the passed fmh2 patch.
16-04-16 sg fmh3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5683 W: 995 L: 1096 D: 3592
sprt @ 10+0.1 th 1 Now that fmh2 is established as useful try as next logical step the extension of this idea one more move. Test against passed fmh2 patch.
16-04-15 fbv UndefAlgPwns053 diff
LLR: -0.78 (-2.94,2.94) [0.00,5.00]
Total: 6003 W: 1119 L: 1126 D: 3758
sprt @ 10+0.1 th 1 Bonus if aligned pawns are undefended