Stockfish Testing Queue

Pending - 0 tests 0.0 hrs

None

Active - 0 tests

Finished - 921 tests

15-07-17 mc simp_sf diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 26828 W: 4683 L: 4884 D: 17261
sprt @ 10+0.1 th 1 Code coverage with gcov shows that this path is almost never taken, so try to simplify it out.
09-07-17 mc phase256 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 48309 W: 6108 L: 6132 D: 36069
sprt @ 60+0.6 th 1 LTC: Double game phase resolution to 256 steps
09-07-17 mc noqmi diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 49390 W: 6223 L: 6434 D: 36733
sprt @ 60+0.6 th 1 LTC: Try reverting the material imbalance hack.
09-07-17 mc noqmi diff
LLR: 3.25 (-2.94,2.94) [-3.00,1.00]
Total: 82403 W: 14885 L: 14853 D: 52665
sprt @ 10+0.1 th 1 Try reverting the material imbalance hack.
09-07-17 mc phase256 diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 48612 W: 8976 L: 8653 D: 30983
sprt @ 10+0.1 th 1 Double game phase resolution to 256 steps
09-07-17 mc imb32 diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 15242 W: 2769 L: 2873 D: 9600
sprt @ 10+0.1 th 1 Double imbalance resolution
24-06-17 mc ct diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 15942 W: 2306 L: 2179 D: 11457
sprt @ 10+0.1 th 7 Regression test for check time simplification (this code is smp so test with 7 threads)
24-06-17 mc ct diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 11895 W: 1741 L: 1608 D: 8546
sprt @ 10+0.1 th 7 Regression test for check time simplification (this code is smp so test with 7 threads): full patch, no races.
27-05-17 mc master diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 33986 W: 6206 L: 5935 D: 21845
sprt @ 10+0.1 th 1 Check resolution with slowdown below 1%
26-05-17 mc master diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 53930 W: 9778 L: 9751 D: 34401
sprt @ 10+0.1 th 1 Validate slowdown code: test is expected to fail
26-05-17 mc master diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 4668 W: 935 L: 762 D: 2971
sprt @ 10+0.1 th 1 Another attempt at slowdown code: -4%
14-05-17 mc master diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 11312 W: 2202 L: 2005 D: 7105
sprt @ 10+0.1 th 1 Validate std::chrono machinery with 0% slowdown. This test is expected to fail.
14-05-17 mc master diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 13640 W: 2518 L: 2316 D: 8806
sprt @ 10+0.1 th 1 Use std::chrono::high_resolution_clock. Slowdown of 1%
14-05-17 mc master diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 17268 W: 3264 L: 3048 D: 10956
sprt @ 10+0.1 th 1 Slowdown of 1%. Retest with time compensation (https://github.com/mcostalba/Stockfish/commit/2c987c331)
13-05-17 mc master diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 21866 W: 4107 L: 3875 D: 13884
sprt @ 10+0.1 th 1 slowdown 1%. (repeat the test to verify realibility): Run 2
13-05-17 mc master diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 9373 W: 1851 L: 1660 D: 5862
sprt @ 10+0.1 th 1 slowdown 1%. (repeat the test to verify realibility): Run 3
13-05-17 mc master diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 14640 W: 2743 L: 2537 D: 9360
sprt @ 10+0.1 th 1 Slowdown by 2%
13-05-17 mc master diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 4897 W: 983 L: 809 D: 3105
sprt @ 10+0.1 th 1 Slowdown SF of 5%. Test to check if sprt[0, 4] can find it
13-05-17 mc slowdown diff
LLR: -0.04 (-2.94,2.94) [0.00,4.00]
Total: 16 W: 2 L: 4 D: 10
sprt @ 10+0.1 th 1 Slowdown SF of 5%. Test to check if sprt[0, 4] can find it.
31-12-16 mc master diff
ELO: 5.21 +-1.5 (95%) LOS: 100.0%
Total: 40000 W: 4256 L: 3656 D: 32088
40000 @ 60+0.6 th 1 Regression test
27-12-16 mc 3fold^ diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 115663 W: 14904 L: 14906 D: 85853
sprt @ 60+0.6 th 1 LTC: Redo Sergei's test with simplified patch: root excluded from extended draw detection
27-12-16 mc 3fold diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 80523 W: 10452 L: 10409 D: 59662
sprt @ 60+0.6 th 1 LTC: Redo Sergei's test with simplified patch: root included from extended draw detection.
26-12-16 mc 3fold^ diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 51562 W: 9314 L: 9245 D: 33003
sprt @ 10+0.1 th 1 Redo Sergei's test with simplified patch: root excluded from extended draw detection
26-12-16 mc 3fold diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 24519 W: 4439 L: 4324 D: 15756
sprt @ 10+0.1 th 1 Redo Sergei's test with simplified patch: root included from extended draw detection.
29-11-16 mc king_th^ diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 38648 W: 6878 L: 6904 D: 24866
sprt @ 10+0.1 th 1 Tweak king safety threshold: take 1
29-11-16 mc king_th diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 5573 W: 926 L: 1063 D: 3584
sprt @ 10+0.1 th 1 Tweak king safety threshold: take 2
06-11-16 mc no_piecelist diff
LLR: -3.42 (-2.94,2.94) [-3.00,1.00]
Total: 33769 W: 4178 L: 4391 D: 25200
sprt @ 60+0.6 th 1 LTC: Retire piecelist: take 2 (added quick access to king square)
06-11-16 mc no_piecelist diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 12666 W: 2285 L: 2151 D: 8230
sprt @ 10+0.1 th 1 Retire piecelist: take 2 (added quick access to king square)
05-11-16 mc no_piecelist diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 113491 W: 19779 L: 20124 D: 73588
sprt @ 10+0.1 th 1 Respin an old patch from Lucas to retire piece lists
09-10-16 mc master diff
ELO: 76.60 +-1.8 (95%) LOS: 100.0%
Total: 40000 W: 10250 L: 1571 D: 28179
40000 @ 60+0.6 th 1 Regression test
09-10-16 mc pieces_no_type diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5329 W: 893 L: 995 D: 3441
sprt @ 10+0.1 th 1 Remove byTypeBB and byColorBB[]: speed optimization test
02-10-16 mc null diff
ELO: -10.73 +-3.6 (95%) LOS: 0.0%
Total: 9100 W: 1016 L: 1297 D: 6787
20000 @ 60+0.6 th 1 Measure the value of null move search at high depths (tested at LTC because at STC the change is not effective)
25-09-16 mc fixmovecount diff
LLR: -2.94 (-2.94,2.94) [-3.00,1.00]
Total: 5199 W: 839 L: 1003 D: 3357
sprt @ 10+0.1 th 1 Fix moveCount to count only legal moves. See https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/9mcmjnyqbAQ
23-09-16 mc numa diff
ELO: -1.74 +-17.0 (95%) LOS: 42.1%
Total: 200 W: 12 L: 13 D: 175
2000 @ 60+0.6 th 32 Numa re-spin: test another numa version fully equivalent to master, but NUMA allocation. This one comes from Texel.
19-09-16 mc cachedsee2 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 11529 W: 2000 L: 2076 D: 7453
sprt @ 10+0.1 th 1 Cache see results: take 2
18-09-16 mc cachedsee2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 10324 W: 1783 L: 1865 D: 6676
sprt @ 10+0.1 th 1 Cache see results
16-09-16 mc lmrmc diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 950 W: 131 L: 254 D: 565
sprt @ 10+0.1 th 1 Crazy idea: take 2
16-09-16 mc lmrmc^ diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 509 W: 38 L: 158 D: 313
sprt @ 10+0.1 th 1 Crazy idea: take 1
12-09-16 mc numa diff
ELO: 2.02 +-7.4 (95%) LOS: 70.4%
Total: 1379 W: 116 L: 108 D: 1155
2000 @ 60+0.6 th 32 NUMA vs per-thread counterMoveHistory: same conditions as http://tests.stockfishchess.org/tests/view/57c896200ebc59030fbe4ce9
12-09-16 mc correctpin diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 103929 W: 18475 L: 18807 D: 66647
sprt @ 10+0.1 th 1 Fix pin-aware SEE to use correct pinners.
08-09-16 mc master diff
ELO: 66.53 +-1.8 (95%) LOS: 100.0%
Total: 40000 W: 9460 L: 1893 D: 28647
40000 @ 60+0.6 th 1 Regression test
08-09-16 mc prune_depth diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 12537 W: 2251 L: 2365 D: 7921
sprt @ 10+0.1 th 1 Tweak prune depth: take 2
08-09-16 mc prune_depth diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 5939 W: 962 L: 1098 D: 3879
sprt @ 10+0.1 th 1 Tweak prune depth: take 3
08-09-16 mc prune_depth^ diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 9499 W: 1690 L: 1814 D: 5995
sprt @ 10+0.1 th 1 Tweak prune depth: take 1
08-09-16 mc tropism diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3381 W: 671 L: 785 D: 1925
sprt @ 10+0.1 th 1 Disable king tropism when using standard king safety
07-09-16 mc prune_dangerous diff
ELO: 10.32 +-4.1 (95%) LOS: 100.0%
Total: 10000 W: 1970 L: 1673 D: 6357
10000 @ 10+0.1 th 1 see prune even dangerous moves at very low depths: I am curious to see how much it's worth this patch that passed SPRT blazingly fast.
01-09-16 mc thread_cm diff
ELO: 8.34 +-6.9 (95%) LOS: 99.1%
Total: 2000 W: 229 L: 181 D: 1590
2000 @ 60+0.6 th 32 Test per-thread counterMoveHistory: Take 3 with 32 threads
01-09-16 mc thread_cm diff
ELO: -1.95 +-4.1 (95%) LOS: 17.7%
Total: 5000 W: 444 L: 472 D: 4084
5000 @ 180+0.6 th 7 Test per-thread counterMoveHistory: take 4, very long TC
29-08-16 mc thread_cm diff
ELO: -2.46 +-3.2 (95%) LOS: 6.5%
Total: 9896 W: 1037 L: 1107 D: 7752
10000 @ 60+0.6 th 7 Test per-thread counterMoveHistory, people reported that in case of big NUMA machines, like TCEC superfinal one, this could lead to big nps improvement. We need at least 7 threads and at least 60 secs/game to have a sensible measurement of the possible ELO loss. Set with high priority because TCEC superfinal is near and we need many tests to validate this solution.
30-08-16 mc thread_cm diff
ELO: 3.49 +-4.3 (95%) LOS: 94.6%
Total: 4880 W: 490 L: 441 D: 3949
5000 @ 60+0.6 th 22 Test per-thread counterMoveHistory: Take 2 with 22 threads