Stockfish Testing Queue

Pending - 4 tests 1.4 hrs

25-05-16 sn doubled_and_passed diff
LLR: -0.66 (-2.94,2.94) [-3.00,1.00]
Total: 14941 W: 2647 L: 2707 D: 9587
sprt @ 10+0.1 th 1 Simplify code for doubled passers : evaluate d6-d7 passers like d5-d7. See Alain's remark on https://github.com/official-stockfish/Stockfish/pull/677 . Tested against passed "weak_pawns4" patch.
24-05-16 Ro SafeBishop diff
LLR: -2.46 (-2.94,2.94) [0.00,5.00]
Total: 11779 W: 2118 L: 2172 D: 7489
sprt @ 10+0.1 th 1 Take 2
22-05-16 Fi lazyDepth diff
LLR: -1.75 (-2.94,2.94) [0.00,5.00]
Total: 23000 W: 3711 L: 3694 D: 15595
sprt @ 5+0.05 th 7 Don't waste helper thread time searching depths lower than main thread.
21-05-16 Ro BishopPawns diff
LLR: -1.77 (-2.94,2.94) [0.00,5.00]
Total: 93166 W: 17112 L: 16788 D: 59266
sprt @ 10+0.1 th 1 When computing the bishop pawn penalty, put more weight on the center pawns

Active - 74 machines 347 cores 1.76M nps (610.36M total nps) 964 games/minute

Machine Cores MNps System Version Running on Last updated
vdv
14
0.00 Linux 3.13.0-74-generic 59 SimpleDouble4 seconds ago
vdv
14
0.00 Linux 3.13.0-74-generic 59 SimpleDouble4 seconds ago
bigpen0r
3
2.21 Linux 3.13.0-74-generic 59 SimpleDouble4 seconds ago
vdv
14
2.27 Linux 3.13.0-74-generic 59 SimpleDouble4 seconds ago
VoyagerOne
5
2.76 Windows 10 59 SimpleDouble4 seconds ago
Thanar
3
1.72 Windows 8 59 SimpleDouble4 seconds ago
Thanar
3
0.93 Windows 7 59 SimpleDouble4 seconds ago
vlad312
1
2.04 Linux 4.2-5.dmz.1-liquorix-amd64 59 SimpleDouble4 seconds ago
JojoM
1
2.18 Linux 3.19.0-59-generic 59 SimpleDouble4 seconds ago
Fisherman
6
1.78 Windows 7 59 SimpleDouble4 seconds ago
psk
3
2.69 Linux 3.13.0-85-generic 59 SimpleDouble4 seconds ago
sunu
2
1.64 Linux 4.5.0-2.slh.2-aptosid-amd64 59 SimpleDouble4 seconds ago
CSU_Dynasty
11
1.57 Windows 7 59 SimpleDouble4 seconds ago
vdv
3
2.24 Linux 3.13.0-74-generic 59 SimpleDouble4 seconds ago
Nesa92
3
1.68 Windows 7 59 SimpleDouble4 seconds ago
Bobo1239
1
0.00 Linux 4.5.3-1-ARCH 59 SimpleDouble4 seconds ago
bigpen0r
14
2.11 Linux 3.13.0-74-generic 59 lmrt2 seconds ago
ctoks
3
2.42 Windows 8.1 59 lmrt2 seconds ago
ako027ako
3
2.12 Windows 10 59 hps seconds ago
vlad312
1
1.93 Linux 4.4-3.dmz.2-liquorix-amd64 59 hps seconds ago
vdv
14
2.15 Linux 3.13.0-74-generic 59 hps seconds ago
fastgm
3
1.47 Windows 7 59 hps seconds ago
xor12
3
1.77 Linux 4.4.0-22-generic 59 hps seconds ago
spams
3
2.21 Windows 7 59 hps seconds ago
vdv
3
2.12 Linux 3.13.0-74-generic 59 hps seconds ago
fastgm
3
2.04 Linux 3.13.0-74-generic 59 hps seconds ago
chris
3
2.38 Windows 10 59 hps seconds ago
vdv
14
2.09 Linux 3.13.0-74-generic 59 hps seconds ago
jjoshua2
3
2.52 Windows 10 59 hps seconds ago
vdv
3
2.04 Linux 3.13.0-74-generic 59 hps seconds ago
cw
5
0.79 Windows 7 59 hps seconds ago
JojoM
3
1.96 Linux 3.19.0-59-generic 59 hps seconds ago
cw
3
1.02 Windows 10 59 hps seconds ago
spams
3
2.69 Windows 7 59 hps seconds ago
ctoks
2
1.11 Windows 8.1 59 hps 3 minutes ago
mhoram
2
0.77 Linux 3.16.0-4-amd64 59 hps seconds ago
IgorLeMasson
3
1.03 Windows 7 59 hps seconds ago
vdv
3
2.07 Linux 3.13.0-74-generic 59 hps seconds ago
mgrabiak
3
1.10 Windows 8 59 hps seconds ago
Freja
3
1.71 Darwin 15.4.0 59 hps seconds ago
velislav
1
1.77 Linux 3.13.0-83-generic 59 hps seconds ago
velislav
1
1.75 Linux 3.19.0-58-generic 59 hps seconds ago
cw
3
1.33 Windows 7 59 hps seconds ago
velislav
1
1.80 Linux 3.19.0-58-generic 59 hps seconds ago
velislav
1
1.83 Linux 3.19.0-58-generic 59 hps 2 minutes ago
lantonov
1
0.82 Windows 8 59 hps seconds ago
fastgm
14
2.10 Linux 3.13.0-74-generic 59 lmr_depth seconds ago
velislav
2
1.68 Linux 3.10.0-123.8.1.el7.x86_64 59 lmr_depth seconds ago
vdv
3
2.13 Linux 3.13.0-74-generic 59 lmr_depth seconds ago
fastgm
4
1.58 Windows 8 59 lmr_depth seconds ago
vdv
3
2.07 Linux 3.13.0-74-generic 59 lmr_depth seconds ago
leszek
3
1.46 Windows 2003Server 59 lmr_depth seconds ago
Freja
2
2.22 Darwin 15.5.0 59 lmr_depth seconds ago
vdbergh
5
1.58 Linux 2.6.32-504.23.4.el6.x86_64 59 bvqs seconds ago
Hjax
3
2.09 Windows 7 59 bvqs seconds ago
robnjr
5
2.84 Windows 7 59 bvqs 50 minutes ago
joster
3
2.15 Linux 3.19.0-26-generic 59 bvqs seconds ago
vdbergh
5
1.58 Linux 2.6.32-504.23.4.el6.x86_64 59 bvqs 22 minutes ago
Mysseno
4
1.48 Windows 8 59 bvqs seconds ago
cw
3
0.83 Windows 7 59 bvqs seconds ago
tvijlbrief
3
2.17 Linux 3.13.0-30-generic 59 os seconds ago
vdv
3
2.11 Linux 3.13.0-74-generic 59 os seconds ago
sqrt2
3
2.28 Linux 3.16.0-4-amd64 59 os seconds ago
fastgm
3
2.03 Linux 3.13.0-74-generic 59 os seconds ago
velislav
1
1.83 Linux 3.19.0-58-generic 59 os seconds ago
cw
1
1.99 Windows 7 59 os seconds ago
vdv
14
2.14 Linux 3.13.0-74-generic 59 lazyDepth2 seconds ago
fastgm
14
2.04 Linux 3.13.0-74-generic 59 lazyDepth2 seconds ago
crunchy
7
1.44 Windows 7 59 lazyDepth2 seconds ago
leszek
7
1.09 Windows 2003Server 59 lazyDepth2 seconds ago
vdv
14
2.11 Linux 3.13.0-74-generic 59 yellowcombo seconds ago
vdv
3
2.08 Linux 3.13.0-74-generic 59 yellowcombo seconds ago
cw
3
0.72 Windows 7 59 yellowcombo seconds ago
vdv
14
2.20 Linux 3.13.0-74-generic 59 yellowcombo seconds ago
26-05-16 Vo bvqs diff
LLR: -0.46 (-2.94,2.94) [0.00,5.00]
Total: 32827 W: 6023 L: 5902 D: 20902
sprt @ 10+0.1 th 1 Use SEE value to update bestValue in QS futility pruning.
26-05-16 Ro SimpleDouble4 diff
LLR: -0.52 (-2.94,2.94) [-3.00,1.00]
Total: 626 W: 104 L: 133 D: 389
sprt @ 10+0.1 th 1 Completely remove double pawns, with small adjustment in evaluate_passed_pawns
26-05-16 sn lmr_depth diff
LLR: 0.11 (-2.94,2.94) [0.00,4.00]
Total: 24522 W: 4513 L: 4423 D: 15586
sprt @ 10+0.1 th 1 LMR-depth = 2 * ONE_PLY
23-05-16 Vo hps diff
LLR: 1.88 (-2.94,2.94) [-3.00,1.00]
Total: 80235 W: 10634 L: 10646 D: 58955
sprt @ 60+0.6 th 1 LTC: Simplification for history pruning. (Hoping for a bit of elo as well)
25-05-16 My os diff
LLR: 0.43 (-2.94,2.94) [0.00,5.00]
Total: 32008 W: 5934 L: 5778 D: 20296
sprt @ 10+0.1 th 1 Similar to the recent checks patch, penalty for unguarded (fix bench)
22-05-16 El lmrt2 diff
LLR: -0.66 (-2.94,2.94) [0.00,5.00]
Total: 125240 W: 23064 L: 22555 D: 79621
sprt @ 10+0.1 th 1 Take 2: combine with jcalovski's idea of decreasing reduction for killers
25-05-16 Fi lazyDepth2 diff
LLR: 2.16 (-2.94,2.94) [0.00,5.00]
Total: 12743 W: 2154 L: 2014 D: 8575
sprt @ 5+0.05 th 7 Don't let any helper thread start searching a depth lower or equal to what has already been completed. (lazyDepth priority has been set to -1)
24-05-16 th yellowcombo diff
LLR: -0.38 (-2.94,2.94) [0.00,4.00]
Total: 9689 W: 1320 L: 1313 D: 7056
sprt @ 60+0.6 th 1 LTC for yellowcombo by snicolet's proposition (see commit comments). I strongly doubt that this patch will pass. Log throughput and I'll stop if negative after 5k games.

Finished - 13592 tests

25-05-16 sn queen_pins3 diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 32020 W: 5914 L: 5652 D: 20454
sprt @ 10+0.1 th 1 Pins or discovered attacks on the opponent's queen: take 2 (bonus=S(35,0))
23-05-16 lb history diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 66905 W: 12229 L: 12068 D: 42608
sprt @ 10+0.1 th 1 clamp
25-05-16 Ro SimpleDouble2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10566 W: 1915 L: 1995 D: 6656
sprt @ 10+0.1 th 1 Score only the frontmost pawn as "passed pawn" as before.
25-05-16 Vo bmu diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 21933 W: 4031 L: 3912 D: 13990
sprt @ 10+0.1 th 1 Don't update bestValue at futility pruning. (I believe this shouldn't be updated)
21-05-16 go master diff
ELO: 29.44 +-1.6 (95%) LOS: 100.0%
Total: 40040 W: 6104 L: 2719 D: 31217
40000 @ 60+0.6 th 1 Regression test (71bfbb22fce23f56)
24-05-16 sn weak_pawns4 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 26739 W: 3562 L: 3449 D: 19728
sprt @ 60+0.6 th 1 LTC: Simplify doubled pawns
25-05-16 ci lmr_incheck diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 10568 W: 1826 L: 1907 D: 6835
sprt @ 10+0.1 th 1 Skip LMR if in check (at low depth)
24-05-16 Vo sp-killer diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 12889 W: 2322 L: 2392 D: 8175
sprt @ 10+0.1 th 1 Don't SEE prune main killer. (Sometimes killer moves act like a bait)
24-05-16 jo null3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9897 W: 1796 L: 1879 D: 6222
sprt @ 10+0.1 th 1 Calculate null reduction based on depth, iteration and value. How bad is this untuned, first try?
24-05-16 My mo diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15830 W: 2911 L: 2968 D: 9951
sprt @ 10+0.1 th 1 Try the reverse rank for evasions
24-05-16 sn queen_pins3 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 13301 W: 2388 L: 2457 D: 8456
sprt @ 10+0.1 th 1 Bonus for pins or discovered attacks on the opponent's queen.
24-05-16 sn weak_pawns4 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7678 W: 1469 L: 1325 D: 4884
sprt @ 10+0.1 th 1 Simplify doubled pawns
23-05-16 Vo mcp-stats diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 46574 W: 8464 L: 8391 D: 29719
sprt @ 10+0.1 th 1 Use history stats to assist with move count pruning.
24-05-16 Fi uniformTimeStop diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8836 W: 1557 L: 1645 D: 5634
sprt @ 10+0.1 th 1 Take 2
23-05-16 sn checks diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 3567 W: 647 L: 759 D: 2161
sprt @ 10+0.1 th 1 Checks in quiescence search: take 2, sorting quiet checks
23-05-16 sn checks diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 3802 W: 638 L: 782 D: 2382
sprt @ 10+0.1 th 1 Examine more checks in quiescence search.
20-05-16 Vo fmpt' diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 84794 W: 11214 L: 11138 D: 62442
sprt @ 60+0.6 th 1 LTC: Futility Margin Tweak
21-05-16 Fi uniformTimeStop diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 75461 W: 13741 L: 13545 D: 48175
sprt @ 10+0.1 th 1 Uniform time stop for normal and easy moves.
23-05-16 pe tuned_tm diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 28604 W: 5158 L: 5217 D: 18229
sprt @ 10+0.1 th 1 Check spsa tuning though iteration counter was broken
23-05-16 lb historyMax diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 2445 W: 395 L: 545 D: 1505
sprt @ 10+0.1 th 1 double the ceiling
22-05-16 Ro SafeBishop diff
LLR: -1.24 (-2.94,2.94) [0.00,5.00]
Total: 2055 W: 367 L: 411 D: 1277
sprt @ 10+0.1 th 1 Bonus for bishop with more than one "safe square"
21-05-16 Vo hps diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 59551 W: 11015 L: 10961 D: 37575
sprt @ 10+0.1 th 1 Simplification for history pruning. (Hoping for a bit of elo as well)
21-05-16 sn unsupported diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 57302 W: 10436 L: 10397 D: 36469
sprt @ 10+0.1 th 1 Doubling the endgame penalty for unsupported pawns.
21-05-16 sn queen_pins3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 55829 W: 10275 L: 10160 D: 35394
sprt @ 10+0.1 th 1 Last try, bonus = S(45, 0).
21-05-16 th yellowcombo diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 25069 W: 4740 L: 4496 D: 15833
sprt @ 10+0.1 th 1 Combo of two yellow patches (see commit notes).
19-05-16 sn initiative8 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 47548 W: 8707 L: 8700 D: 30141
sprt @ 10+0.1 th 1 Raise pawn weight in initiative
20-05-16 El lmrt2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 50621 W: 9368 L: 9275 D: 31978
sprt @ 10+0.1 th 1 Don't decrease reduction for cut nodes
21-05-16 pe tune_tm diff
60837/20000 iterations
40000/40000 games played
40000 @ 10+0.1 th 1 Tune splitting on failedLow. Previous tuning did not produce much change, so I substantially change limits, to narrow region to where it must be more sensitive
21-05-16 ci missing_cms2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17859 W: 3247 L: 3296 D: 11316
sprt @ 10+0.1 th 1 Use cmh2 if cmh is null
19-05-16 El combo3b diff
LLR: -1.41 (-2.94,2.94) [0.00,4.00]
Total: 65789 W: 11999 L: 11849 D: 41941
sprt @ 10+0.1 th 1 Combo patch (see commit notes). Running against passed hist_pruning patch.
20-05-16 jo tune_check_attack diff
12611/20000 iterations
4769/40000 games played
40000 @ 60+0.6 th 1 Try to further tune check and attack units, now at LTC. Start with preliminary tuned values which seem to be much better at STC. Lower throughput.
21-05-16 sn queen_pins3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17899 W: 3344 L: 3392 D: 11163
sprt @ 10+0.1 th 1 Try the opposite: bonus = S(60, 0)
21-05-16 ci missing_cms diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18099 W: 3334 L: 3381 D: 11384
sprt @ 10+0.1 th 1 CounterMovesStats for Null Move.
19-05-16 jo king_safety_clop diff
LLR: -2.32 (-2.94,2.94) [0.00,4.00]
Total: 50002 W: 6830 L: 6811 D: 36361
sprt @ 60+0.6 th 1 LTC: Tuned check and attack unit values. Most values seem to be stable, so let's try a sprt-test.
21-05-16 sn queen_pins3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5520 W: 954 L: 1056 D: 3510
sprt @ 10+0.1 th 1 Same logic than the patch which is struggling at LTC, but half bonus. Also rebased on current master.
20-05-16 pe tune_tm diff
62498/20000 iterations
39339/40000 games played
40000 @ 10+0.1 th 1 Tune splitting on failedLow
15-05-16 sn queen_pins diff
LLR: -1.35 (-2.94,2.94) [0.00,5.00]
Total: 103259 W: 13956 L: 13666 D: 75637
sprt @ 60+0.6 th 1 LTC: Bonus for pins on the opponent queen
20-05-16 ci update_stats diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 35024 W: 6299 L: 6276 D: 22449
sprt @ 10+0.1 th 1 Before updating other quiets moves, check entry with current move.
19-05-16 II impFactor diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 43677 W: 7919 L: 7926 D: 27832
sprt @ 10+0.1 th 1 Removing min/max in improvingFactor. This is "almost" a simplification, but I test with [0,4]. Two new patches included in master.
20-05-16 El lmrt diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 26855 W: 4903 L: 4967 D: 16985
sprt @ 10+0.1 th 1 LMR parameter tweak. Test against passed hist_pruning patch.
20-05-16 Vo fmpt diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 29296 W: 5469 L: 5212 D: 18615
sprt @ 10+0.1 th 1 Futile margin parameter tweak...
20-05-16 sn opposite_bishops2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10355 W: 1834 L: 1915 D: 6606
sprt @ 10+0.1 th 1 Endgames with opposite-colored bishops and heavy pieces are maybe not so drawish
19-05-16 sg outposts3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 27054 W: 4994 L: 5003 D: 17057
sprt @ 10+0.1 th 1 Calculate more precise pawn attack span by consider only squares up to next opposing pawn. This leads to more possible outposts.
19-05-16 pe tune_tm diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 17644 W: 3247 L: 3343 D: 11054
sprt @ 10+0.1 th 1 After tuning
19-05-16 jo king_safety_clop diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 11276 W: 2223 L: 2025 D: 7028
sprt @ 10+0.1 th 1 Tuned check and attack unit values. Most values seem to be stable, so let's try a sprt-test.
18-05-16 pe tune_tm diff
50521/20000 iterations
40000/40000 games played
40000 @ 10+0.1 th 1 Tune large set of parameters with new model
19-05-16 My sc diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 15476 W: 2751 L: 2811 D: 9914
sprt @ 10+0.1 th 1 Modification to scale factor
11-05-16 sn unsafe diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 51656 W: 9546 L: 9524 D: 32586
sprt @ 10+0.1 th 1 Tweak blocked pawn condition when considering "other" checks.
13-05-16 El king_safety diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 171847 W: 31827 L: 31390 D: 108630
sprt @ 10+0.1 th 1 Update king safety values after local tuning (10K games, TC 3+0.03)
19-05-16 pb bishop_knight_exchange diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 18521 W: 3366 L: 3412 D: 11743
sprt @ 10+0.1 th 1 Score bishop-knight exchange as good capture when bishop pair is already gone