Stockfish Testing Queue

Pending - 0 tests 0.0 hrs

None

Active - 0 tests

Finished - 533 tests

18-07-07 jdo dbldPawns1 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7807 W: 1700 L: 1793 D: 4314
sprt @ 10+0.1 th 1 Tweak doubled definition to exclude phalanx pawns
18-05-27 jdo lmrX2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8578 W: 1237 L: 1326 D: 6015
sprt @ 60+0.6 th 1 More captures for LMR
18-05-26 jdo lmrX2 diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 197047 W: 40009 L: 38972 D: 118066
sprt @ 10+0.1 th 1 More captures for LMR
18-05-26 jdo lmrX diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 66445 W: 13390 L: 13305 D: 39750
sprt @ 10+0.1 th 1 Always apply extra reduction LMR capture experiment.
18-05-26 jdo lmrX2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 6860 W: 1323 L: 1420 D: 4117
sprt @ 10+0.1 th 1 Less captures for LMR
18-05-26 jdo lmrX1 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 25119 W: 5047 L: 5059 D: 15013
sprt @ 10+0.1 th 1 Second ply of reduction at high depths only
18-04-27 jdo tuneConnected diff
99136/100000 iterations
200000/200000 games played
200000 @ 40+0.4 th 1 Fix by Alain: Tuning the connected bonuses since framework getting empty. Tuning at a longer TC as I think this might be TC-dependent. Throughput reduced to compensate, may stop early if converging.
18-04-28 jdo connectedTune diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23401 W: 4799 L: 4818 D: 13784
sprt @ 10+0.1 th 1 Manual tuning try #1
18-04-28 jdo connectedTune diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 47420 W: 9641 L: 9549 D: 28230
sprt @ 10+0.1 th 1 Endgame values only Endgame values only
18-04-28 jdo connectedTune diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 23481 W: 4755 L: 4775 D: 13951
sprt @ 10+0.1 th 1 Midgame values only
18-04-28 jdo connectedTune diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 12260 W: 2475 L: 2547 D: 7238
sprt @ 10+0.1 th 1 Tuned values Tuned values
18-04-18 jdo cmb2 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 32815 W: 6621 L: 6660 D: 19534
sprt @ 10+0.1 th 1 Combo of parameter tweaks, part 2
18-04-18 jdo cmb1 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 63217 W: 12908 L: 12833 D: 37476
sprt @ 10+0.1 th 1 Combo of two yellow LTC parameter tweaks
18-04-17 jdo captLMR diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 151269 W: 22942 L: 22647 D: 105680
sprt @ 60+0.6 th 1 LTC: Tweak LMR of captures, inspired by V1 tests.
18-04-17 jdo captLMR diff
LLR: 2.94 (-2.94,2.94) [0.00,4.00]
Total: 32287 W: 6810 L: 6527 D: 18950
sprt @ 10+0.1 th 1 Tweak LMR of captures, inspired by V1 tests. Long yellow, so try with correct [0, 4] bounds.
18-04-17 jdo captLMR diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 17257 W: 3467 L: 3564 D: 10226
sprt @ 10+0.1 th 1 Take 4. Using [0, 4] bounds as suggested by Stefan.
18-04-17 jdo captLMR diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 39966 W: 8036 L: 8049 D: 23881
sprt @ 10+0.1 th 1 Take 3.
18-04-17 jdo captLMR diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 19962 W: 4046 L: 4132 D: 11784
sprt @ 10+0.1 th 1 Tweak #2 to LMR of captures. A bit simpler than master so let's go for [0, 4].
18-04-17 jdo captLMR diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 76169 W: 15493 L: 15268 D: 45408
sprt @ 10+0.1 th 1 Tweak LMR of captures, inspired by V1 tests.
18-04-16 jdo futMarg diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 42691 W: 6440 L: 6404 D: 29847
sprt @ 60+0.6 th 1 LTC: LTC positive but struggling, try -32 and depth < 6 3rd LTC: as it was inadvertently run twice for one quick pass and one quick fail.
18-04-15 jdo futMarg diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 62727 W: 12987 L: 12565 D: 37175
sprt @ 10+0.1 th 1 Try -40, depth < 6
18-04-15 jdo futMarg diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 11501 W: 1698 L: 1777 D: 8026
sprt @ 60+0.6 th 1 LTC: try -32 and depth < 6
18-04-15 jdo futMarg diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 27452 W: 5678 L: 5422 D: 16352
sprt @ 10+0.1 th 1 LTC positive but struggling, try -32 and depth < 6
18-04-15 jdo futMarg diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 45079 W: 6825 L: 6780 D: 31474
sprt @ 60+0.6 th 1 LTC: Take 2. Reduce depth as well
18-04-14 jdo futMarg diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 22718 W: 4701 L: 4467 D: 13550
sprt @ 10+0.1 th 1 Now try to reduce margins by 5, depth < 5
18-04-14 jdo futMarg diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16760 W: 3490 L: 3284 D: 9986
sprt @ 10+0.1 th 1 Now with depth < 6.
18-04-14 jdo futMarg diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 14657 W: 3049 L: 2853 D: 8755
sprt @ 10+0.1 th 1 Last test: -50 only
18-04-14 jdo futMarg diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 13323 W: 2817 L: 2626 D: 7880
sprt @ 10+0.1 th 1 Reduce depth even further
18-04-14 jdo futMarg diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 47177 W: 7121 L: 7068 D: 32988
sprt @ 60+0.6 th 1 Try again on the futility margins. -25
18-04-14 jdo futMarg diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 21438 W: 4455 L: 4227 D: 12756
sprt @ 10+0.1 th 1 Take 2. Reduce depth as well
18-04-14 jdo futMarg diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 17611 W: 3664 L: 3454 D: 10493
sprt @ 10+0.1 th 1 Try again on the futility margins. -25
18-04-14 jdo searchParams2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16176 W: 3273 L: 3326 D: 9577
sprt @ 10+0.1 th 1 Another tweak
18-04-14 jdo searchParams diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 32282 W: 6511 L: 6552 D: 19219
sprt @ 10+0.1 th 1 Another tweak of search parameters.
18-04-13 jdo eelco diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 98105 W: 14704 L: 14697 D: 68704
sprt @ 60+0.6 th 1 For information, see how this does at LTC non-regression. See discussion at GitHub page. Eelco de Groot's suggested patch: https://github.com/official-stockfish/Stockfish/issues/1544
18-04-13 jdo seePrune diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 40908 W: 8259 L: 8268 D: 24381
sprt @ 10+0.1 th 1 Margin -10%
18-04-13 jdo seePrune diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16642 W: 3343 L: 3394 D: 9905
sprt @ 10+0.1 th 1 Another formula
18-04-13 jdo seePrune diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 11314 W: 2269 L: 2388 D: 6657
sprt @ 10+0.1 th 1 Margin +10%
18-04-13 jdo nullT diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 39891 W: 8104 L: 8116 D: 23671
sprt @ 10+0.1 th 1 10% less depth reduction
18-04-13 jdo nullT diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 36134 W: 7275 L: 7302 D: 21557
sprt @ 10+0.1 th 1 10% more depth reduction
18-04-13 jdo nullT diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 24199 W: 4850 L: 4921 D: 14428
sprt @ 10+0.1 th 1 Experiment with null move reduction. +5% static reduction
18-04-13 jdo nullT diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 24642 W: 5023 L: 5092 D: 14527
sprt @ 10+0.1 th 1 Tweak null reduction: -5% static reduction
18-04-13 jdo eelco diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35749 W: 7288 L: 7194 D: 21267
sprt @ 10+0.1 th 1 For information, see how this does at LTC non-regression. See discussion at GitHub page. Eelco de Groot's suggested patch: https://github.com/official-stockfish/Stockfish/issues/1544
18-04-13 jdo pruneSSMC diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8996 W: 1267 L: 1355 D: 6374
sprt @ 60+0.6 th 1 LTC: Another tweak.
18-04-12 jdo eelco diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 43251 W: 8820 L: 8747 D: 25684
sprt @ 10+0.1 th 1 Eelco de Groot's suggested patch: https://github.com/official-stockfish/Stockfish/issues/1544
18-04-12 jdo pruneTTCapt diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 38311 W: 7770 L: 7721 D: 22820
sprt @ 10+0.1 th 1 Tweak to low depth pruning.
18-04-12 jdo pruneCN diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 16847 W: 3377 L: 3428 D: 10042
sprt @ 10+0.1 th 1 One more idea.
18-04-12 jdo pruneSSMC diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 6602 W: 1403 L: 1244 D: 3955
sprt @ 10+0.1 th 1 Another tweak.
18-04-12 jdo futM diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 11268 W: 2212 L: 2288 D: 6768
sprt @ 10+0.1 th 1 Another try with the futility margins
18-04-11 jdo extCh diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15177 W: 3067 L: 3125 D: 8985
sprt @ 10+0.1 th 1 Check extension tweak.
18-04-11 jdo resonance diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 41333 W: 8476 L: 8482 D: 24375
sprt @ 10+0.1 th 1 Another try on Stephane's resonance idea: use more multiples of 4 for evaluation parameters.