Stockfish Testing Queue

Finished - 4029 tests

19-02-21 Elb cmp_tw diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 28084 W: 4592 L: 4690 D: 18802
sprt @ 60+0.6 th 1 Another idea (spec LTC, lower TP, corrected bounds)
19-02-22 pro ps_openfiles3 diff
41651/50000 iterations
86942/100000 games played
100000 @ 40+0.4 th 1 tune QueenOnOpenFile (and behind a rook).
19-02-21 pb0 master diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 15984 W: 2350 L: 2223 D: 11411
sprt @ 60+0.6 th 4 As there is still some resistance against PR #1782 (suspect for multi-thread negative scaling) do another non-regression test at higher tc. (master vs. master without the patch). Lowered throughput.
19-02-21 SFi center_pawns diff
LLR: -1.78 (-2.94,2.94) [0.00,3.50]
Total: 12638 W: 2044 L: 2115 D: 8479
sprt @ 60+0.6 th 1 spec LTC (VLTP)
19-02-21 Elb cmp_tw diff
LLR: -0.01 (-2.94,2.94) [0.50,4.50]
Total: 148 W: 20 L: 20 D: 108
sprt @ 60+0.6 th 1 Another idea (spec LTC, lower TP)
19-02-21 Viz LDB1 diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 27864 W: 4605 L: 4703 D: 18556
sprt @ 60+0.6 th 1 LTC
19-02-21 Voy pruningT2 diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 26720 W: 4362 L: 4463 D: 17895
sprt @ 60+0.6 th 1 Speculative LTC on this patch. As it seems to scale well based off the STC and MTC. Low TP.
19-02-20 MJZ CutNodes2 diff
LLR: 2.96 (-2.94,2.94) [0.00,3.50]
Total: 27549 W: 4614 L: 4363 D: 18572
sprt @ 60+0.6 th 1 Probcutcount limit = 2 + 2 * cutnode - Speculatif LTC with low throughput.
19-02-19 pb0 skip_small diff
LLR: -2.26 (-2.94,2.94) [0.00,3.50]
Total: 79247 W: 11455 L: 11389 D: 56403
sprt @ 60+0.6 th 4 LTC: Encounter skip-scheme scaling problem by skipping plies only as long processed nodes per iteration are below 70000. Test against PR #1972 (correct base branch now)
19-02-17 vdv threadPick diff
LLR: -0.85 (-2.94,2.94) [0.00,3.50]
Total: 9310 W: 1243 L: 1270 D: 6797
sprt @ 60+0.6 th 8 spec LTC of http://tests.stockfishchess.org/tests/view/5c4efb030ebc593af5d4c910 given that the voting is TC sensitive
19-02-18 pro ps_reductions101 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 83388 W: 13920 L: 13808 D: 55660
sprt @ 60+0.6 th 1 Ltc: try this at 0,4 (just to verify)
19-02-18 pro ps_seed100 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 10179 W: 1654 L: 1775 D: 6750
sprt @ 60+0.6 th 1 Ltc: try an equation for pawn Seed.
19-02-15 Cof noskip2 diff
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 30243 W: 4052 L: 3815 D: 22376
sprt @ 120+1.2 th 4 Show Elo gain of noskip (derived from latest master) on VLTC 4 threads
19-02-17 xor nmpT diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 37962 W: 6243 L: 6313 D: 25406
sprt @ 60+0.6 th 1 LTC: tweak.
19-02-17 31m KingMoves diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 19744 W: 3169 L: 3290 D: 13285
sprt @ 60+0.6 th 1 There's no reason to leave the framework empty for hours with a promising test waiting...speculative LTC for @MJZ1977's 170K STC yellow.
19-01-29 Cof novoting diff
LLR: 0.33 (-2.94,2.94) [-3.00,1.00]
Total: 157875 W: 20295 L: 20483 D: 117097
sprt @ 60+0.6 th 8 quick scaling check for no thread voting on 60+0.6 8 threads. Should fail quickly. No functional change on 1 thread.
19-02-14 pro ps_skipquiets1 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 77107 W: 12792 L: 12761 D: 51554
sprt @ 60+0.6 th 1 LTC: remove moveCountPruning variable (just use skipQuiets).
19-02-14 xor prune diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 122696 W: 20385 L: 20219 D: 82092
sprt @ 60+0.6 th 1 Spec LTC on empty framework to see if this is TC dependent
19-02-13 MJZ Passed_Protection diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 13388 W: 2116 L: 2254 D: 9018
sprt @ 60+0.6 th 1 LTC - MG Bonus for well protected passed pawn. +1/16. STC : http://tests.stockfishchess.org/tests/view/5c64a3880ebc5925cffbf8f4
19-02-12 pro ps_reductions2 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 70765 W: 11617 L: 11575 D: 47573
sprt @ 60+0.6 th 1 LTC: remove a dimension from reductions.
19-02-13 Viz NoMobBlk1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 30264 W: 4935 L: 4969 D: 20360
sprt @ 60+0.6 th 1 Spec. LTC for this - this mobility stuff is more of a strategical thing so can vary with different depths... Using STC bounds for faster converging.
19-02-13 pro ps_reductions2 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 69384 W: 11513 L: 11470 D: 46401
sprt @ 60+0.6 th 1 Ltc v4
19-02-12 MJZ DeepPlies diff
LLR: -0.86 (-2.94,2.94) [0.00,3.50]
Total: 7909 W: 1328 L: 1357 D: 5224
sprt @ 60+0.6 th 1 The idea is to reserve tactics to shallow plies and more positional evaluation to deep plies. This first experiment just prune tactical moves for deep plies. SEE > -300 / LTC limited test
19-02-12 MJZ CST-AddLines3 diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 23144 W: 3815 L: 3926 D: 15403
sprt @ 60+0.6 th 1 Critical Search Tree (CST) : try adding promising lines. parent CST && no counter move (bestvalue <= alpha && d > 7) - speculative LTC to see the behavior at higher depths
19-02-10 31m lmr_pawnPushVsK~9 diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 30554 W: 4985 L: 5076 D: 20493
sprt @ 60+0.6 th 1 Speculative LTC for 165K yellow.
19-02-09 Viz SliderNoPawn4 diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 19743 W: 3229 L: 3350 D: 13164
sprt @ 60+0.6 th 1 LTC
19-02-09 Ala AssymKingPSQT diff
LLR: -1.50 (-2.94,2.94) [0.00,3.50]
Total: 95656 W: 16166 L: 15986 D: 63504
sprt @ 60+0.6 th 1 Spec LTC, as it may be better on longer TC. Low TP.
19-02-09 MJZ AdaptTimeControl diff
ELO: -2.41 +-8.7 (95%) LOS: 29.4%
Total: 1874 W: 281 L: 294 D: 1299
2000 @ 120+1.2 th 1 I just observed that SF plays too fast in large time controls. This limited test is to check if adding some time helps in large time controls.
19-02-09 xor killer1 diff
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 26978 W: 4398 L: 4445 D: 18135
sprt @ 60+0.6 th 1 Spec LTC as suggested
19-02-06 sni anti_shuffing12~1 diff
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 98724 W: 16425 L: 16187 D: 66112
sprt @ 60+0.6 th 1 Sp├ęculative LTC : initiative -= 2 * rule50_count()
19-02-06 Ala 8x8KingPSQTune2 diff
28739/30000 iterations
59714/60000 games played
60000 @ 60+0.6 th 1 STC with raw unsmoothed values did +1.5 elo. Finish tune on rebased master with lower variance to help convergence and some pawn value co-tuning.
19-02-08 pro ps_fallingEval2 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 31382 W: 5291 L: 5187 D: 20904
sprt @ 60+0.6 th 1 LTC: [fixed edges] remove failedLow and simplify fallingEval (mathematically substantially similar to master.)
19-02-07 MJZ Simplify_reduction3 diff
ELO: -1.44 +-9.4 (95%) LOS: 38.3%
Total: 1448 W: 199 L: 205 D: 1044
10000 @ 180+1.8 th 1 Last VLTC verification for passed STC and LTC [-3..1] Simplify reduction formula.r+=22% if not improving.
19-02-07 MJZ Simplify_reduction3 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 52989 W: 8735 L: 8665 D: 35589
sprt @ 60+0.6 th 1 LTC run of http://tests.stockfishchess.org/tests/view/5c5b4c720ebc5925cffb3ee4 Simplify reduction formula.r+=22% if not improving.
19-02-06 MJZ Alpha-reduction4 diff
LLR: -0.44 (-2.94,2.94) [0.00,3.50]
Total: 5380 W: 912 L: 923 D: 3545
sprt @ 60+0.6 th 1 Speculative LTC : don't decrease reduction if alpha > 340 && pvHit STC : http://tests.stockfishchess.org/tests/view/5c5aa69e0ebc5925cffb2a79 (Please adjust throughput if necessary)
19-02-05 Ala KingSafetyParams diff
LLR: -2.96 (-2.94,2.94) [0.00,3.50]
Total: 88935 W: 14711 L: 14640 D: 59584
sprt @ 60+0.6 th 1 LTC for king safety params tune (yellow STC was with elo gainer bounds, would have passed with param tweak bounds)
19-02-05 Mat Alpha-reduction2 diff
LLR: 0.01 (-2.94,2.94) [0.50,4.50]
Total: 118 W: 20 L: 19 D: 79
sprt @ 180+0.6 th 1 Very long time control, rescheduled
19-02-05 pro ps_fmc2 diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 61356 W: 9995 L: 10246 D: 41115
sprt @ 60+0.6 th 1 LTC: simplify FutilityMoveCount equations.
19-02-04 Ala KingSafetyTune2 diff
24023/25000 iterations
49975/50000 games played
50000 @ 60+0.6 th 1 Finish king safety tuning with tweaked variances.
19-02-05 pro ps_timereduction1 diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 22858 W: 3832 L: 3715 D: 15311
sprt @ 60+0.6 th 1 LTC: simplify time reduction. 1 + 0.05 * depth.
19-02-04 pro ps_searchinit3 diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 8547 W: 1340 L: 1508 D: 5699
sprt @ 60+0.6 th 1 combine two passing STC simplifications in search init.
19-01-28 Ala 8x8KingPSQTune diff
51225/75000 iterations
106500/150000 games played
150000 @ 60+0.6 th 1 Retry the 8x8 king PSQT idea. Tune starting from Kurt values.
19-02-03 Ala master diff
ELO: 17.71 +-1.8 (95%) LOS: 100.0%
Total: 40000 W: 6661 L: 4624 D: 28715
40000 @ 60+0.6 th 1 Regression/progression test against SF10 after "Less king danger if we have a knight near by to defend it." of February, 3rd.
19-02-02 MJZ Blocked_posNM2 diff
LLR: -0.21 (-2.94,2.94) [0.00,3.50]
Total: 7053 W: 909 L: 905 D: 5239
sprt @ 60+0.6 th 8 Blocked positions detection : we adapt NMP to potentially blocked positions. If a move don't lead a progress, we return a reduced value. Skip MCP for other moves. Needs a high depth search. Please adjust throughput if needed but the idea is promising in limited number of games. http://tests.stockfishchess.org/tests/view/5c54b18b0ebc592fc7bb077a
19-02-02 Ala KingSafetyTune diff
36266/60000 iterations
76026/120000 games played
120000 @ 60+0.6 th 1 Tune king safety, including knight in king ring bonus, with more variance for newly added/changed values.
19-02-02 Fis MinorInKingRing diff
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 96669 W: 16318 L: 15872 D: 64479
sprt @ 60+0.6 th 1 LTC Try lower weight
19-02-01 MJZ Blocked_posNM2 diff
ELO: 4.85 +-11.7 (95%) LOS: 79.1%
Total: 932 W: 135 L: 122 D: 675
2000 @ 60+0.6 th 8 Blocked positions : skip move count pruning if we escape NMP. Limited test at high depth with low throughput.
19-02-01 sni MinorInKingRing diff
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 88518 W: 14690 L: 14619 D: 59209
sprt @ 60+0.6 th 1 LTC for mstembera: Use king attacks instead of kingring. Rebase on current master for LTC.
19-02-01 xot time7 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 58515 W: 9670 L: 9609 D: 39236
sprt @ 60+0.6 th 1 LTC: simplify: remove failedLow. no functional change
19-01-29 xor discoExt diff
LLR: 2.95 (-2.94,2.94) [0.00,3.50]
Total: 227341 W: 37972 L: 37165 D: 152204
sprt @ 60+0.6 th 1 LTC: test rebased to current master (as requested).