Stockfish Testing Queue

Finished - 27341 tests

30-03-16 Vo dhp diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 7934 W: 1401 L: 1493 D: 5040
sprt @ 10+0.1 th 1 dynamic history pruning
30-03-16 lb bitboard_clean diff
LLR: -2.96 (-2.94,2.94) [-4.00,0.00]
Total: 24443 W: 4867 L: 5120 D: 14456
sprt @ 5+0.05 th 1 No functional change. But let's do a sanity check to be sure. Let's see what 32 bit workers think of this (or rather 64 bit machines running 32 bit OS).
30-03-16 sn tune-levers diff
9922/10000 iterations
20000/20000 games played
20000 @ 20+0.2 th 1 Tune levers
28-03-16 El fmh_variant diff
LLR: 0.88 (-2.94,2.94) [0.00,5.00]
Total: 62342 W: 8494 L: 8250 D: 45598
sprt @ 60+0.6 th 1 LTC: last try (against latest master)
29-03-16 Vo StatRebound2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 40797 W: 7464 L: 7415 D: 25918
sprt @ 10+0.1 th 1 Take 2 on this idea.
29-03-16 sn supported_levers diff
LLR: -2.32 (-2.94,2.94) [0.00,5.00]
Total: 36000 W: 6598 L: 6542 D: 22860
sprt @ 10+0.1 th 1 Supported/unsupported levers
30-03-16 fb EvaluateSpace036 diff
LLR: -1.61 (-2.94,2.94) [0.00,5.00]
Total: 2854 W: 480 L: 536 D: 1838
sprt @ 10+0.1 th 1 End game value. Take 2
27-03-16 n_ Combo3 diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 66294 W: 9065 L: 8739 D: 48490
sprt @ 60+0.6 th 1 LTC: Combo patch of two tuning patches which failed yellow at LTC.
29-03-16 El loose_tweak diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 21607 W: 3894 L: 3977 D: 13736
sprt @ 10+0.1 th 1 Take 2
29-03-16 My pv diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 22329 W: 4115 L: 4195 D: 14019
sprt @ 10+0.1 th 1 Curious to also see if reduction (~5%) of MG values will compliment the EG bump.
29-03-16 Vo lmrIdea diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7423 W: 1306 L: 1400 D: 4717
sprt @ 10+0.1 th 1 Fixed Bench
28-03-16 sn space3 diff
LLR: -2.35 (-2.94,2.94) [-3.00,1.00]
Total: 25370 W: 4595 L: 4764 D: 16011
sprt @ 10+0.1 th 1 Simplify space formula (count only our minors). Idea by Alain Savard.
28-03-16 mb tune_pp diff
17544/25000 iterations
35495/50000 games played
50000 @ 20+0.2 th 1 Tune PP and EG material at 20+0.2 because PP values are time dependent. (fixed PSQT initialization)
28-03-16 El loose_tweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 23477 W: 4266 L: 4342 D: 14869
sprt @ 10+0.1 th 1 Try a small bonus for LooseEnemies in MG
29-03-16 My ks2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 2463 W: 434 L: 551 D: 1478
sprt @ 10+0.1 th 1 Include kingring idea in undefended
28-03-16 sn king_ring4 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9044 W: 1606 L: 1693 D: 5745
sprt @ 10+0.1 th 1 Count pawns attacks in the whole kingRing. For instance, this triggers the black king safety calculation if there is a white pawn in g5 against a black king in g8.
27-03-16 lb onepawn diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 68806 W: 9884 L: 9831 D: 49091
sprt @ 10+0.1 th 1 Retire SCALE_FACTOR_ONEPAWN
26-03-16 Vo lmrSimple diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 71350 W: 13016 L: 13297 D: 45037
sprt @ 10+0.1 th 1 stc
28-03-16 fb EvaluateSpace008 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23050 W: 4218 L: 4244 D: 14588
sprt @ 10+0.1 th 1 End game value
28-03-16 sn king_ring2 diff
LLR: -1.87 (-2.94,2.94) [0.00,5.00]
Total: 15869 W: 2996 L: 3006 D: 9867
sprt @ 10+0.1 th 1 More king safety evaluations. Take 2: extended king ring, material >= QueenValueMg (with bug fix, thanks Alain!)
28-03-16 sn loose5 diff
LLR: -1.54 (-2.94,2.94) [0.00,4.00]
Total: 9614 W: 1683 L: 1732 D: 6199
sprt @ 10+0.1 th 1 Try LooseEnemies bonus = S(0,30)
28-03-16 mb tune_pp diff
22/25000 iterations
42/50000 games played
50000 @ 20+0.2 th 1 Tune PP and EG material at 20+0.2 because PP values are time dependent.
27-03-16 sn king_ring diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 18318 W: 3421 L: 3467 D: 11430
sprt @ 10+0.1 th 1 More king safety evaluations. Take 1: extended kingRing, more than two bishops.
24-03-16 Ro SemiBackward4 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 71308 W: 13075 L: 12895 D: 45338
sprt @ 10+0.1 th 1 Take 2
27-03-16 El fmh_variant diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 30988 W: 5741 L: 5483 D: 19764
sprt @ 10+0.1 th 1 Final try
27-03-16 Vo cmo diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 27909 W: 5068 L: 5074 D: 17767
sprt @ 10+0.1 th 1 Capture Move Order
27-03-16 Vo ap5 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 19212 W: 3499 L: 3542 D: 12171
sprt @ 10+0.1 th 1 Last try...don't razor advance pawn.
27-03-16 lb onepawn^ diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 35366 W: 5159 L: 5207 D: 25000
sprt @ 10+0.1 th 1 SCALE_FACTOR_ONEPAWN=56
27-03-16 n_ Combo3 diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 14223 W: 2700 L: 2494 D: 9029
sprt @ 10+0.1 th 1 A combo patch of two tuning patches which finished LTC yellow. Added "loose3" and "endgame_pp".
27-03-16 Vo ap4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10746 W: 1928 L: 2007 D: 6811
sprt @ 10+0.1 th 1 Take 4...
27-03-16 lb onepawn diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 27008 W: 3838 L: 3911 D: 19259
sprt @ 10+0.1 th 1 SCALE_FACTOR_ONEPAWN=40
26-03-16 Vo ap3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 33317 W: 6100 L: 6082 D: 21135
sprt @ 10+0.1 th 1 Take 3
26-03-16 Fi ConnectedRooks diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 13336 W: 2424 L: 2492 D: 8420
sprt @ 10+0.1 th 1 Connected rooks.
26-03-16 pb see_killer2 diff
ELO: -0.66 +-2.9 (95%) LOS: 32.9%
Total: 20000 W: 3680 L: 3718 D: 12602
20000 @ 10+0.1 th 1 Move Order: more promising killer first, evaluation based on SEE-differences
26-03-16 Ro NoPopcount diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7709 W: 1401 L: 1574 D: 4734
sprt @ 10+0.1 th 1 A few simplifications in evaluate.cpp
26-03-16 sn loose3 diff
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 11936 W: 1651 L: 1493 D: 8792
sprt @ 60+0.6 th 1 LTC: Bonus for loose opponent pieces or pawns
16-03-16 jo values_spsa2 diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 197829 W: 26389 L: 26013 D: 145427
sprt @ 60+0.6 th 1 LTC: Another SPSA2 session on outpost values. Also on RookOnFile bonus.
25-03-16 mb endgame_pp diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 54637 W: 7533 L: 7238 D: 39866
sprt @ 60+0.6 th 1 LTC: Raise endgame PP and material values.
26-03-16 lb checked diff
LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 38867 W: 6908 L: 7131 D: 24828
sprt @ 10+0.1 th 1 take 2: don't touch the popcount
26-03-16 sn loose3 diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 30504 W: 5743 L: 5485 D: 19276
sprt @ 10+0.1 th 1 Bonus for loose opponent pieces or pawns
26-03-16 jo queen_outOfPlay diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23633 W: 4198 L: 4223 D: 15212
sprt @ 10+0.1 th 1 Inspired by Rocky's and Mysseno's tests, retry this old idea of a penalty for a queen out of play. Take 1.
26-03-16 Ro UndefendedKR diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 11925 W: 2123 L: 2198 D: 7604
sprt @ 10+0.1 th 1 Take 2, inspired by jcalovsky recent patch
25-03-16 Ro QSupportedCheck diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 40139 W: 7278 L: 7232 D: 25629
sprt @ 10+0.1 th 1 For queen contact checks, use similar bonus pattern as other safe checks
26-03-16 lb checked diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 23823 W: 4288 L: 4486 D: 15049
sprt @ 10+0.1 th 1 let's try to simplify and harmonize code a bit: remove score -= Checked, and the last popcount (in QueenContactCheck)
25-03-16 Vo ap2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 26720 W: 4878 L: 4889 D: 16953
sprt @ 10+0.1 th 1 Meant to use 7th Rank...(Thanks Rocky)
26-03-16 lb onepawn diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 20347 W: 3660 L: 3747 D: 12940
sprt @ 10+0.1 th 1 SCALE_FACTOR_ONEPAWN=32
25-03-16 Ro UK_Pawns diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10101 W: 1800 L: 1882 D: 6419
sprt @ 10+0.1 th 1 Do not count opponent pawns which stands in our undefended area
25-03-16 lb replace diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 2837 W: 459 L: 572 D: 1806
sprt @ 5+0.05 th 3 STC: re-probe. take 2 (reschedule test deleted by troll)
25-03-16 Vo ap diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 40640 W: 7420 L: 7371 D: 25849
sprt @ 10+0.1 th 1 Skip pruning if prior move is an advance pawn
25-03-16 My ks diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 32026 W: 5778 L: 5767 D: 20481
sprt @ 10+0.1 th 1 Try a variant of Rocky's recent Kingring patch