Stockfish Testing Queue

Pending - 0 tests 0.0 hrs

None

Active - 0 tests

Finished - 74 tests

08-12-17 xo pawnEndgames2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9002 W: 1564 L: 1651 D: 5787
sprt @ 10+0.1 th 1 skip early pruning if < 3 pieces left on board
07-12-17 xo qless5d diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 49205 W: 8915 L: 8904 D: 31386
sprt @ 10+0.1 th 1 reduce queen value by (20,80)
07-12-17 xo qless5c diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 33463 W: 6057 L: 6099 D: 21307
sprt @ 10+0.1 th 1 reduce queen value by (40,80)
07-12-17 xo qless5b diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 15252 W: 2724 L: 2829 D: 9699
sprt @ 10+0.1 th 1 reduce queen value by (80,80)
06-12-17 xo qless7 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 18432 W: 3336 L: 3430 D: 11666
sprt @ 10+0.1 th 1 reduce queen eg value by 100
05-12-17 xo qless3 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 26493 W: 4811 L: 4877 D: 16805
sprt @ 10+0.1 th 1 Smaller reduction for queen eg value.
05-12-17 xo qless5 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 60150 W: 11054 L: 11004 D: 38092
sprt @ 10+0.1 th 1 looks like first, promising, test was just luck. last try, reduce queen eg value by 80. (correct base branch this time)
06-12-17 xo qless8 diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 26730 W: 4795 L: 4860 D: 17075
sprt @ 10+0.1 th 1 last go: reduce queen eg value by 60
05-12-17 xo qless4 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 9060 W: 1628 L: 1754 D: 5678
sprt @ 10+0.1 th 1 reduce eg queen value by 240
05-12-17 xo qless1 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 5296 W: 891 L: 1030 D: 3375
sprt @ 10+0.1 th 1 Try lowering eg value of queen like Komodo seems to do (corrected bounds)
05-12-17 xo qless1 diff
LLR: 0.52 (-2.94,2.94) [0.00,5.00]
Total: 8721 W: 1636 L: 1576 D: 5509
sprt @ 10+0.1 th 1 Try lowering eg value of queen like Komodo seems to do
03-12-17 xo phased_contempt3_no_dra diff
ELO: -0.41 +-1.9 (95%) LOS: 33.3%
Total: 40000 W: 5903 L: 5950 D: 28147
40000 @ 10+0.1 th 1 Contempt=5 with no draw adjudication at STC. (Deleted earlier in error, this is STC to compare with LTC)
03-12-17 xo phased_contempt3_no_dra diff
ELO: 0.43 +-1.8 (95%) LOS: 67.8%
Total: 40000 W: 5651 L: 5602 D: 28747
40000 @ 10+0.1 th 1 Contempt=0, no draw adjudication. Test no contempt against master with new code and no draw adjudication. Contempt 5 result looks a bit odd so this will add another data point to the graph of results.
02-12-17 xo 2026d5ded75e323845ac036 diff
ELO: -0.59 +-2.1 (95%) LOS: 29.0%
Total: 20000 W: 1872 L: 1906 D: 16222
20000 @ 60+0.6 th 1 Contempt=5, no draw adj. Elo change appears to suffer at LTC compared to STC, maybe we need to use 0 or 5 to reduce this problem.
02-12-17 xo phased_contempt3_no_dra diff
ELO: -0.90 +-1.9 (95%) LOS: 18.0%
Total: 40000 W: 6383 L: 6487 D: 27130
40000 @ 10+0.1 th 1 Contempt=15, no draw adjudication. Test Contempt/4 for endgame to see how this is with no draw adjudication.
02-12-17 xo 2026d5ded75e323845ac036 diff
ELO: 0.89 +-1.9 (95%) LOS: 81.6%
Total: 40000 W: 6621 L: 6518 D: 26861
40000 @ 10+0.1 th 1 Test Contempt 15 with no draw adjudication at STC to get an idea of scaling with time.
02-12-17 xo 2026d5ded75e323845ac036 diff
ELO: 0.97 +-1.9 (95%) LOS: 84.0%
Total: 40000 W: 6381 L: 6269 D: 27350
40000 @ 10+0.1 th 1 Test Contempt 10 with no draw adjudication at STC to get an idea of scaling with time.
29-11-17 xo attackRook1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5792 W: 991 L: 1092 D: 3709
sprt @ 10+0.1 th 1 bonus for possible bishop onto rook attacks
24-11-17 xo conmin diff
ELO: 110.23 +-3.0 (95%) LOS: 100.0%
Total: 19512 W: 7056 L: 1065 D: 11391
40000 @ 10+0.1 th 1 Test minimal change to master against SF7, contempt=20.
22-11-17 xo master diff
ELO: 121.15 +-2.2 (95%) LOS: 100.0%
Total: 40000 W: 15682 L: 2273 D: 22045
40000 @ 10+0.1 th 1 master with contempt 5 vs SF7 (to compare with vdv's reference test)
21-11-17 xo con20 diff
ELO: -4.38 +-2.9 (95%) LOS: 0.2%
Total: 20000 W: 3636 L: 3888 D: 12476
20000 @ 10+0.1 th 1 hardcode uci contempt to 20
21-11-17 xo con5 diff
ELO: -1.22 +-2.9 (95%) LOS: 20.6%
Total: 20000 W: 3620 L: 3690 D: 12690
20000 @ 10+0.1 th 1 test draw ratio with uci contempt = 5
21-11-17 xo con1 diff
ELO: -1.42 +-2.9 (95%) LOS: 16.6%
Total: 20000 W: 3540 L: 3622 D: 12838
20000 @ 10+0.1 th 1 test draw ratio with uci contempt = 1
21-11-17 xo mat3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 6397 W: 1082 L: 1180 D: 4135
sprt @ 10+0.1 th 1 +ve material bonus in eval (bugfixed)
20-11-17 xo mat2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9919 W: 1783 L: 1866 D: 6270
sprt @ 10+0.1 th 1 smaller bonus and remove extra make_score()
20-11-17 xo mat2 diff
LLR: -0.04 (-2.94,2.94) [0.00,5.00]
Total: 115 W: 21 L: 22 D: 72
sprt @ 10+0.1 th 1 smaller bonus
20-11-17 xo opb8 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 27260 W: 4936 L: 4945 D: 17379
sprt @ 10+0.1 th 1 combine mg and eg ideas
20-11-17 xo mat1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 11584 W: 2078 L: 2154 D: 7352
sprt @ 10+0.1 th 1 hold material when even/winning
19-11-17 xo opb7 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23520 W: 4241 L: 4266 D: 15013
sprt @ 10+0.1 th 1 adjust both mg and eg values
19-11-17 xo opb5b diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 24277 W: 4339 L: 4361 D: 15577
sprt @ 10+0.1 th 1 need abs() on value, sf minus five. bench 5252122
18-11-17 xo opb6 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 36288 W: 6643 L: 6613 D: 23032
sprt @ 10+0.1 th 1 lower scale factor for opposite bishops
18-11-17 xo opb5 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 11193 W: 1989 L: 2067 D: 7137
sprt @ 10+0.1 th 1 more cautious in strong winning/maybe won positions
18-11-17 xo opb4 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 13579 W: 2405 L: 2473 D: 8701
sprt @ 10+0.1 th 1 Test one part of Stephane's best opposite bishops patch
16-11-17 xo opb3 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 45275 W: 8210 L: 8143 D: 28922
sprt @ 10+0.1 th 1 bonus/malus (8) no side test
15-11-17 xo opb3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 65263 W: 11751 L: 11600 D: 41912
sprt @ 10+0.1 th 1 bonus/malus (16) remove sidetomove test
16-11-17 xo opb3b diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23317 W: 4178 L: 4204 D: 14935
sprt @ 10+0.1 th 1 bonus/malus 14
15-11-17 xo opb3 diff
LLR: -3.14 (-2.94,2.94) [0.00,5.00]
Total: 47203 W: 8560 L: 8492 D: 30151
sprt @ 10+0.1 th 1 bonus/malus for opp bishops (16)
16-11-17 xo opb3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11750 W: 2104 L: 2179 D: 7467
sprt @ 10+0.1 th 1 bonus/malus (24) no side test
15-11-17 xo opb2 diff
LLR: -0.73 (-2.94,2.94) [0.00,5.00]
Total: 8279 W: 1486 L: 1482 D: 5311
sprt @ 10+0.1 th 1 mg penalty (16) for opp bishops
15-11-17 xo opb1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4799 W: 796 L: 901 D: 3102
sprt @ 10+0.1 th 1 take 2: smaller penalty range
14-11-17 xo opb1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 2550 W: 455 L: 572 D: 1523
sprt @ 10+0.1 th 1 oppbishops1: scale penalty by npm
02-11-17 xo rof diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 43959 W: 5473 L: 5453 D: 33033
sprt @ 60+0.6 th 1 LTC for El
03-11-17 xo initMat5b diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11285 W: 2036 L: 2113 D: 7136
sprt @ 10+0.1 th 1 last go: delay exchanges if score > 80 and initiative > 20
02-11-17 xo initMat4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 24157 W: 4372 L: 4394 D: 15391
sprt @ 10+0.1 th 1 even smaller bonus for holding material
02-11-17 xo initMat3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 6628 W: 1130 L: 1227 D: 4271
sprt @ 10+0.1 th 1 take 3: smaller material bonus
02-11-17 xo initMat2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 4519 W: 767 L: 873 D: 2879
sprt @ 10+0.1 th 1 delay material exchanges if we have the initiative and a positive score
01-11-17 xo initMat diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 10903 W: 1942 L: 2021 D: 6940
sprt @ 10+0.1 th 1 If score and initiative good, increase own piece values to discourage exchanges
31-10-17 xo bw_contempt2b diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 31149 W: 5646 L: 5638 D: 19865
sprt @ 10+0.1 th 1 increased caution for black
31-10-17 xo bw_contempt2c diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 9581 W: 1691 L: 1776 D: 6114
sprt @ 10+0.1 th 1 Take3: increase caution for black to 60
31-10-17 xo bw_contempt3b diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 17623 W: 3136 L: 3186 D: 11301
sprt @ 10+0.1 th 1 make opponent cautious as well as black (idea from sn)