Stockfish Testing Queue

Finished - 40562 tests

17-02-04 sg nmp diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 11688 W: 2061 L: 2137 D: 7490
sprt @ 10+0.1 th 1 no nullmove pruning if the last two positions (with same side to move) are in check
17-02-03 sg nmp diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9664 W: 1720 L: 1804 D: 6140
sprt @ 10+0.1 th 1 no nullmove pruning if last position was in check
17-02-03 Gua CloseEnemies3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4543 W: 808 L: 915 D: 2820
sprt @ 10+0.1 th 1 if to use not only king flank for close enemies v3
17-02-03 Gua CloseEnemies2 diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40833 W: 7583 L: 7813 D: 25437
sprt @ 10+0.1 th 1 more simple
17-02-03 SC moreR diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 20328 W: 3657 L: 3696 D: 12975
sprt @ 10+0.1 th 1 Completely remove history correction factor.
17-02-03 sni windmill3' diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 19025 W: 3394 L: 3439 D: 12192
sprt @ 10+0.1 th 1 Use a Score for windmill, as discovered checks are dangerous in endgame too
17-02-03 sni entry_points diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16028 W: 2856 L: 2913 D: 10259
sprt @ 10+0.1 th 1 Take 2
17-02-03 Voy statT2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 4304 W: 726 L: 868 D: 2710
sprt @ 10+0.1 th 1 Try a cubic type formula
17-02-03 Voy captureSortT diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 9114 W: 1583 L: 1670 D: 5861
sprt @ 10+0.1 th 1 stc
17-02-03 sni entry_points diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3149 W: 522 L: 634 D: 1993
sprt @ 10+0.1 th 1 Bonus for entry points in the opponent camp
17-02-03 sni tune_king_danger2 diff
5952/10000 iterations
12242/20000 games played
20000 @ 20+0.2 th 1 Tune Windmill as a score
17-02-03 Anb always-calculate-space diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 58914 W: 10684 L: 10628 D: 37602
sprt @ 10+0.1 th 1 Check if always using space calculation is useful.
17-02-03 Gua CloseEnemies diff
LLR: -2.92 (-2.94,2.94) [-3.00,1.00]
Total: 6206 W: 1090 L: 1257 D: 3859
sprt @ 10+0.1 th 1 CloseEnemies without king flank rule
17-02-03 sni capturing_checks' diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 20579 W: 3643 L: 3681 D: 13255
sprt @ 10+0.1 th 1 Take 3, bonus=32
17-02-03 sni static diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5932 W: 1033 L: 1133 D: 3766
sprt @ 10+0.1 th 1 Use qsearch() instead of evaluate() for some pruning decisions during the search.
17-02-03 pb0 capture_escape diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 11854 W: 2083 L: 2158 D: 7613
sprt @ 10+0.1 th 1 Take 3 (Take 2 put to prio -1)
17-02-03 pb0 capture_escape diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17741 W: 3159 L: 3209 D: 11373
sprt @ 10+0.1 th 1 Take 2
17-02-03 Mys ts diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 12756 W: 2230 L: 2302 D: 8224
sprt @ 10+0.1 th 1 see if queen blocks are more a burden for us
17-02-03 fau xo diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 23454 W: 4195 L: 4272 D: 14987
sprt @ 10+0.1 th 1 Just a strange experiment I would like to do since framework is empty (What is good in noe phase cannot be too harmful in the other phase)
17-02-01 Gua tweak3 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 156856 W: 28317 L: 27942 D: 100597
sprt @ 10+0.1 th 1 do easy move somemore faster
17-02-01 Gua tweak2 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 91341 W: 16411 L: 16258 D: 58672
sprt @ 10+0.1 th 1 tweak2
17-02-02 Gua tweak3b diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 79406 W: 14354 L: 14241 D: 50811
sprt @ 10+0.1 th 1 easy move still more faster than in version 3
17-02-01 Gua tweak4 diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 72460 W: 13183 L: 13092 D: 46185
sprt @ 10+0.1 th 1 do easy move somemore slower
17-02-02 SC moreR02 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17368 W: 3063 L: 3115 D: 11190
sprt @ 10+0.1 th 1 Reduce less if I escaped futility pruning by a large margin.
17-02-02 SC moreR diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 37536 W: 6789 L: 6755 D: 23992
sprt @ 10+0.1 th 1 And compensate a bit in history. Take 3 and call it a day.
17-02-02 SC moreR diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 35954 W: 6474 L: 6447 D: 23033
sprt @ 10+0.1 th 1 Reduce checks a bit less.
17-02-01 Gua tweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 27547 W: 4903 L: 4966 D: 17678
sprt @ 10+0.1 th 1 tweak in search bench same but must have a diffrence in game
17-02-02 SC moreR diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 17199 W: 3032 L: 3085 D: 11082
sprt @ 10+0.1 th 1 Only if we dont sacrifice to much
17-02-02 sni pinners2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7402 W: 1282 L: 1376 D: 4744
sprt @ 10+0.1 th 1 Double the WeakQueen penalty if there is a pin which is very likely to loose material
17-02-02 SC moreR03 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 703 W: 87 L: 212 D: 404
sprt @ 10+0.1 th 1 Was sane locally, let us try it on the framework.
17-02-02 Voy queenMobS diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25333 W: 3370 L: 3256 D: 18707
sprt @ 60+0.6 th 1 LTC: Queen mob simplification...it might be good that the queen is defending those attack squares...
17-02-02 sni capturing_checks diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 11335 W: 2002 L: 2079 D: 7254
sprt @ 10+0.1 th 1 Take 2, bonus = 128
17-02-02 sni capturing_checks diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 57025 W: 10355 L: 10238 D: 36432
sprt @ 10+0.1 th 1 Bonus for safe capturing checks
17-02-02 sg rookMob diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3316 W: 548 L: 659 D: 2109
sprt @ 10+0.1 th 1 Retry first version but with MobilityBonus compensation (based on measurements)
17-02-02 Voy queenMobS diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 85997 W: 15550 L: 15540 D: 54907
sprt @ 10+0.1 th 1 Queen mob simplification...it might be good that the queen is covering those attacked squares...
17-02-02 sni windmill3' diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13841 W: 2486 L: 2552 D: 8803
sprt @ 10+0.1 th 1 Tuned value for windmill
17-02-02 Gua t2+t3 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 20160 W: 3598 L: 3686 D: 12876
sprt @ 10+0.1 th 1 easy move some more faster and more narrow range
17-02-02 jos udm1 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9071 W: 1539 L: 1626 D: 5906
sprt @ 10+0.1 th 1 My try on piece development. Take 1.
17-02-01 Voy statT diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 51702 W: 6616 L: 6630 D: 38456
sprt @ 60+0.6 th 1 ltc
17-02-01 sni tune_king_danger2 diff
4910/5000 iterations
9980/10000 games played
10000 @ 20+0.2 th 1 The Windmill value was not moving in the previous tuning run. Now use a factor 10 for Windmill
: will this help SPSA ?
17-02-02 Fis lazyTweak diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13165 W: 2323 L: 2392 D: 8450
sprt @ 10+0.1 th 1 Lazy eval tweak.
17-02-01 SC simpScaleFac diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 7448 W: 1281 L: 1375 D: 4792
sprt @ 10+0.1 th 1 Evaluate scale_factor has 2 problems: 1) it does not evaluate if loop if phase == 128, so one condition is not needed (simplification) 2) If a specialized scale function is equal to one of the specified scale factors, it will be overwritten by the ones specified in the function body, which is unintended. Does fixing these give an Elo gain?
17-02-01 fau PiecesDevelopmentA diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8210 W: 1426 L: 1516 D: 5268
sprt @ 10+0.1 th 1 Same as Take 1, but this time compensating the rook psqt in rank 1 so we will prefer to move out the minor pieces rather than the rooks.
17-02-01 fau PiecesDevelopment diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 18702 W: 3374 L: 3420 D: 11908
sprt @ 10+0.1 th 1 Encouraging minor pieces development by giving penalty if our rook attacks our minor in the first rank, which means that we have at least 1 minor and 1 rook in the first rank in the early phase of the game (Thanks to Rocky for the help)
17-02-01 fau PiecesDevelopment1 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 17948 W: 3250 L: 3299 D: 11399
sprt @ 10+0.1 th 1 Version 2, without the rook, just checking for the minors.
17-02-01 fau PiecesDevelopment2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 4777 W: 821 L: 926 D: 3030
sprt @ 10+0.1 th 1 Version 3 (Penalty only if more than 1 minor pieces is still in rank 1) and penalty value is 30
17-01-29 lbr master diff
ELO: 8.82 +-1.5 (95%) LOS: 100.0%
Total: 40000 W: 4595 L: 3580 D: 31825
40000 @ 60+0.6 th 1 Regression test until "Simplify TT penalty stat"
17-02-01 sni tune_king_danger2 diff
3202/5000 iterations
6589/10000 games played
10000 @ 20+0.2 th 1 Tune windmill
17-02-01 Voy statT diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 109084 W: 19934 L: 19406 D: 69744
sprt @ 10+0.1 th 1 stc
17-02-01 sg null_history diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 22382 W: 3989 L: 4019 D: 14374
sprt @ 10+0.1 th 1 Try the opposite and relax the condition eval >= beta if eval is only a little bit smaller than beta and history value is under the 1% highest value to allow more pruning