Stockfish Testing Queue

Finished - 3598 tests

18-10-21 Viz SFmaxDist1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9113 W: 1451 L: 1538 D: 6124
sprt @ 60+0.6 th 1 LTC
18-10-20 pro ps_mobility_all1 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 28279 W: 4731 L: 4622 D: 18926
sprt @ 60+0.6 th 1 LTC: mobility equations for all pieces. . fixed rook exception. I'd like to pass this and do a PR so we can put it on the table and have a better discussion about it. This could likely be further simplified. Low tp.
18-10-20 vdv deltaDepth2^ diff
LLR: -3.05 (-2.94,2.94) [0.00,5.00]
Total: 23719 W: 3811 L: 3845 D: 16063
sprt @ 60+0.6 th 1 LTC: Take 4.
18-10-20 Viz MobDng3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 16266 W: 2622 L: 2681 D: 10963
sprt @ 60+0.6 th 1 Seems like this version of idea is at least capable of producing pretty good yellows (like 70 and 90k for 2 attempts). So let me try a speculative LTC to see if it worth trying to find better constants for it or if it scales like crap and is useless. 100 throughput.
18-10-20 vdv deltaDepth diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 25577 W: 4072 L: 4095 D: 17410
sprt @ 60+0.6 th 1 LTC: Make delta depending on expected depth to completion.
18-10-20 31m EscapeBishop^ diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17194 W: 2701 L: 2757 D: 11736
sprt @ 60+0.6 th 1 LTC for S(20, 0).
18-10-18 vdv abDraw^ diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 47182 W: 7617 L: 7556 D: 32009
sprt @ 60+0.6 th 1 spec. LTC. low tp. Adjust ab on draw. Take 2.
18-10-18 Kur rookpsqt5 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 38278 W: 6094 L: 6128 D: 26056
sprt @ 60+0.6 th 1 Speculative LTC for somewhat promising rookpsqt5, low tp (100).
18-10-19 MJZ PvChange4 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13217 W: 2134 L: 2205 D: 8878
sprt @ 60+0.6 th 1 LTC - Don't prune PvNode until BestValue >= alpha && extend PvNode if it change. Test with low throughput = 160
18-10-19 pro ps_backward20 diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25052 W: 3943 L: 4135 D: 16974
sprt @ 60+0.6 th 1 LTC: simplified backward with a bit MORE penalty.
18-10-19 Viz MobDngInit2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 18012 W: 2856 L: 2909 D: 12247
sprt @ 60+0.6 th 1 LTC
18-10-19 vdv PVprune diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 37638 W: 6003 L: 5980 D: 25655
sprt @ 60+0.6 th 1 LTC: Take 2.
18-10-19 SFi qv_wk_combo diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7458 W: 1148 L: 1241 D: 5069
sprt @ 60+0.6 th 1 Q_value4 is likely yellow so combo it with my other yellow LTC weakQueen5, test directly at LTC
18-10-19 SFi qv_wk_combo diff
Pending...
sprt @ 60+0.6 th 1 Q_value is likely yellow so combo it my other yellow LTC weakQueen5, test directly at LTC
18-10-18 vdv PVprune diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 49282 W: 7909 L: 7841 D: 33532
sprt @ 60+0.6 th 1 LTC: Variant MJZ idea.
18-10-18 pro ps_backward20 diff
LLR: -0.06 (-2.94,2.94) [-3.00,1.00]
Total: 49159 W: 7943 L: 8023 D: 33193
sprt @ 60+0.6 th 1 LTC: simplify backward pawns.
18-10-18 pb0 adjustedFH diff
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 8123 W: 1363 L: 1210 D: 5550
sprt @ 60+0.6 th 1 Restest #1768 with restriction to singlethread, LTC
18-10-17 Viz DIrGB2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15469 W: 2421 L: 2484 D: 10564
sprt @ 60+0.6 th 1 Speculative LTC just to have an idea if it worth to try some more variants of this. Extremely low TP.
18-10-17 xot delta3b diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10108 W: 1558 L: 1641 D: 6909
sprt @ 60+0.6 th 1 LTC. Perhaps only modifying alpha we will scale better than modifying delta. subtract 2*delta on fail low.
18-10-17 xot delta3 diff
LLR: -3.37 (-2.94,2.94) [0.00,5.00]
Total: 51336 W: 8282 L: 8222 D: 34832
sprt @ 60+0.6 th 1 LTC. Increase delta to 27 if previousScore < -24.
18-10-16 pro ps_passed71 diff
LLR: -1.57 (-2.94,2.94) [-3.00,1.00]
Total: 61003 W: 9827 L: 10004 D: 41172
sprt @ 60+0.6 th 1 Ltc: simplified passed pawn rules. v2 (don't consider levers).
18-10-17 pb0 fhBestMoveChange diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 16000 W: 2604 L: 2664 D: 10732
sprt @ 60+0.6 th 1 Take 2, LTC
18-10-16 pro ps_passed70 diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35121 W: 5623 L: 5831 D: 23667
sprt @ 60+0.6 th 1 LTC: If stoppers are levers, a simple pawn push will pass. . . can we ignore levers? just do leverPush.
18-10-16 Voy lmrT diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9516 W: 1508 L: 1593 D: 6415
sprt @ 60+0.6 th 1 would like to get an idea how this scales... low throughput
18-10-16 SFi Q_value diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 25873 W: 4169 L: 4241 D: 17463
sprt @ 60+0.6 th 1 LTC +MG+EG
18-10-15 pro ps_mobility20 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 41964 W: 6781 L: 6803 D: 28380
sprt @ 60+0.6 th 1 Speculative LTC: convert mobility equations to values (except rook). Try 0,4. Low tp.
18-10-16 pro ps_mobility_all1 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 13020 W: 2096 L: 2208 D: 8716
sprt @ 60+0.6 th 1 LTC: mobility equations for all pieces. . fixed rook exception. Try 0,4
18-10-15 vdv fHb^ diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 50207 W: 8102 L: 8030 D: 34075
sprt @ 60+0.6 th 1 LTC Tweak alpha on fH. Take 2.
18-10-15 MJZ Instability_pruning3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 21546 W: 3423 L: 3462 D: 14661
sprt @ 60+0.6 th 1 LTC rerun - correlate pruning and position "instability" - 200 + instab / 4 + keep track after correction STC : http://tests.stockfishchess.org/tests/view/5bc4fefb0ebc592439f81359
18-10-15 pro ps_mobility_all1 diff
LLR: 1.22 (-2.94,2.94) [-3.00,1.00]
Total: 11165 W: 1830 L: 1784 D: 7551
sprt @ 60+0.6 th 1 LTC: mobility equations for all pieces. . fixed rook exception.
18-10-14 SFi twk_QueenValue diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 58880 W: 9474 L: 9444 D: 39962
sprt @ 60+0.6 th 1 changes to queen value do better in LTC, speculative LTC, prio -1, low tp 166
18-10-15 31m NvsR1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 40538 W: 6540 L: 6505 D: 27493
sprt @ 60+0.6 th 1 LTC for @Vizvezdenec. Take 1.
18-10-14 sni master diff
ELO: 49.01 +-1.9 (95%) LOS: 100.0%
Total: 40000 W: 9388 L: 3783 D: 26829
40000 @ 60+0.6 th 1 Regression/progression test against SF9 after "Simplify check extensions" of October, 14th
18-10-14 vdv pvDist4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 20659 W: 3293 L: 3335 D: 14031
sprt @ 60+0.6 th 1 LTC: pvDist in probcut. Take 2.
18-10-14 SFi QueenWithLessPawns5 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 77969 W: 12604 L: 12424 D: 52941
sprt @ 60+0.6 th 1 speculative LTC, low tp 166
18-10-13 vdv IIDdepth2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 33493 W: 5350 L: 5343 D: 22800
sprt @ 60+0.6 th 1 Spec. LTC. Param 12. low tp
18-10-13 31m combo_ver_wq diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 46525 W: 7449 L: 7458 D: 31618
sprt @ 60+0.6 th 1 The STC has already outperformed the one that led to my recent successful speculative LTC. Therefore, before taking any further steps, run a LTC. (I am submitting this now because I probably will not be available when the STC finishes.) Please raise to normal throughput (1000) from the reduced, speculative LTC throughput (166) if the STC passes.
18-10-13 SFi QueenWithLessPawns diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10599 W: 1684 L: 1765 D: 7150
sprt @ 60+0.6 th 1 take 3, S(60,30)
18-10-13 pb0 iid2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 14371 W: 2247 L: 2314 D: 9810
sprt @ 60+0.6 th 1 No iid for expected all-nodes at higher depths. Speculative LTC empty Framework, Prio -1 The idea behind: if a that node all moves always failed low
18-10-13 31m qsFut^^ diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 42126 W: 6880 L: 6837 D: 28409
sprt @ 60+0.6 th 1 LTC for @vondele. Take 5, param 9500.
18-10-12 pro ps_mobility13_bishops diff
LLR: 2.62 (-2.94,2.94) [-3.00,1.00]
Total: 22334 W: 3644 L: 3543 D: 15147
sprt @ 60+0.6 th 1 Ltc: Bishops mobility equation (last piece). v1
18-10-12 pro ps_depthwindow diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 45874 W: 7324 L: 7270 D: 31280
sprt @ 60+0.6 th 1 increase search window for deeper depths. 16+rootDepth/4. LTC because it will have little effect at low depths.
18-10-12 pro ps_mobility12_rooks diff
LLR: 0.25 (-2.94,2.94) [-3.00,1.00]
Total: 16073 W: 2593 L: 2605 D: 10875
sprt @ 60+0.6 th 1 LTC: rooks mobility equation: 6.5x-31, 1 exception.
18-10-12 pb0 adjustedFh2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9371 W: 1465 L: 1551 D: 6355
sprt @ 60+0.6 th 1 Skip resolving fail high at low depths, take 3. Test on top of PR #1768
18-10-12 Viz weakQueen5 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 66695 W: 10694 L: 10640 D: 45361
sprt @ 60+0.6 th 1 LTC for SFisaGod
18-10-11 SFi twk_pawnpush2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 18913 W: 2970 L: 3064 D: 12879
sprt @ 60+0.6 th 1 empty framework, speculative LTC , prio -1, tp 166
18-10-12 pro ps_mobility12_rooks diff
LLR: -1.32 (-2.94,2.94) [-3.00,1.00]
Total: 59829 W: 9505 L: 9666 D: 40658
sprt @ 60+0.6 th 1 LTC: rooks mobility equation, linear, y = 6.5x - 31.
18-10-12 Roc UnsafePin diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 10231 W: 1615 L: 1698 D: 6918
sprt @ 60+0.6 th 1 LTC: Bonus for retaliating against a pinner
18-10-11 31m BsafeCh1 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 17584 W: 2799 L: 2853 D: 11932
sprt @ 60+0.6 th 1 I'm curious to see how this very promising STC (108K yellow) from @Vizvezdenec scales, and the framework is completely empty. Speculative LTC, low throughput (166).
18-10-11 SFi weakQueen2 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 24382 W: 3924 L: 4001 D: 16457
sprt @ 60+0.6 th 1 speculative LTC for this 146k yellow, prio -1