Stockfish Testing Queue

Finished - 42781 tests

16-12-03 SC secondBest diff
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 87683 W: 15397 L: 14912 D: 57374
sprt @ 10+0.1 th 1 Use second best move in see pruning before LMR. Take 1.
16-12-05 pb0 verification_search diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 59453 W: 10505 L: 10382 D: 38566
sprt @ 10+0.1 th 1 take 2: trigger verification search earlier
16-12-05 jos easy_pt diff
LLR: -1.28 (-2.94,2.94) [0.00,4.00]
Total: 18792 W: 3285 L: 3290 D: 12217
sprt @ 10+0.1 th 1 Make easy move time dependant on captured piece type. (I hope this still qualifies as tuning patch!?)
16-12-05 sg connected_passed diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 21794 W: 3850 L: 3884 D: 14060
sprt @ 10+0.1 th 1 Give bonus only if pawn not blocked
16-12-02 Fis EasyTweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 140686 W: 24825 L: 24512 D: 91349
sprt @ 10+0.1 th 1 Take 2
16-12-04 SC easyPv2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 74733 W: 13194 L: 13008 D: 48531
sprt @ 10+0.1 th 1 I we just played an easy move, increase multipv by 1 so that in the case that we would be prevented from playing another easy move, at least the main thread spends time searching more thorougly. Take 1. (Please some of the seniors: could you check whether it makes sense before approving? It seemed to be sane locally, but I am always uncertain with this kind of stuff)
16-12-05 Voy pawn-psqt diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 64397 W: 11409 L: 11349 D: 41639
sprt @ 10+0.1 th 1 stc
16-12-05 pb0 verification_search diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13893 W: 2411 L: 2478 D: 9004
sprt @ 10+0.1 th 1 Verification search experiment: disable nmp for the the first half of a further reducted search tree
16-12-05 sni connected_passed2 diff
LLR: -2.99 (-2.94,2.94) [0.00,5.00]
Total: 16792 W: 2938 L: 2994 D: 10860
sprt @ 10+0.1 th 1 Try to apply Stefan's connected passed bonus to any pawn connected to a neighbour friend pawn, even if this neighbour is not passed itself.
16-12-04 hex unsupported_pawns_t3 diff
16338/20000 iterations
32966/40000 games played
40000 @ 20+0.2 th 1 Unsupportable and unsupported tuning
16-12-05 cib cutnode_killer2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 14171 W: 2482 L: 2547 D: 9142
sprt @ 10+0.1 th 1 Decrease cutNode reduction for killer move.
16-12-05 Roc RookOutpost2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 23266 W: 4123 L: 4150 D: 14993
sprt @ 10+0.1 th 1 Rook outpost in relation to passed pawn creation or improvement.
16-12-04 sni flank_attack2 diff
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 13844 W: 1832 L: 1650 D: 10362
sprt @ 60+0.6 th 1 LTC: Take 2
16-12-04 sg connected_passed diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 27682 W: 4944 L: 4952 D: 17786
sprt @ 10+0.1 th 1 bonus +50%. Take 3
16-12-04 Voy pt2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10452 W: 1788 L: 1869 D: 6795
sprt @ 10+0.1 th 1 take 2
16-12-04 sg connected_passed diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15169 W: 2575 L: 2637 D: 9957
sprt @ 10+0.1 th 1 bonus -50%
16-12-04 Voy pt3 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 36090 W: 6327 L: 6302 D: 23461
sprt @ 10+0.1 th 1 take 3
16-12-04 sni flank_attack2 diff
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 34997 W: 6328 L: 6055 D: 22614
sprt @ 10+0.1 th 1 Take 2
16-12-04 Roc SafeMoves diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9478 W: 1645 L: 1730 D: 6103
sprt @ 10+0.1 th 1 Penalty for each minor with no safe square for next move.
16-12-04 sg connected_passed diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 16616 W: 2901 L: 2956 D: 10759
sprt @ 10+0.1 th 1 Swap mid- and endgame bonus. Take 2
16-12-04 sg connected_passed diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 42951 W: 7648 L: 7592 D: 27711
sprt @ 10+0.1 th 1 Bonus for connected passed pawns
16-12-04 sni open_files6 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 3978 W: 651 L: 759 D: 2568
sprt @ 10+0.1 th 1 Initiative midgame malus for the attacking side, for each open file
16-12-04 vdv KingTune diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 25118 W: 4365 L: 4437 D: 16316
sprt @ 10+0.1 th 1 KingTune param tweaks.
16-12-04 IIv fmc diff
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 18125 W: 3103 L: 3289 D: 11733
sprt @ 10+0.1 th 1 Take 3. As the previous test is struggling, I lowered the priority, hoping for a better test.
16-12-04 goo beta_alpha diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 16956 W: 2988 L: 3042 D: 10926
sprt @ 10+0.1 th 1 pruning other branch
16-12-04 Roc NoMoveKing diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10956 W: 1854 L: 1933 D: 7169
sprt @ 10+0.1 th 1 Endgame penalty for low mobility king
16-12-04 hex movepick_evasion diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8229 W: 1407 L: 1497 D: 5325
sprt @ 10+0.1 th 1 Movepick evasion, use least valuable defender first
16-12-04 SC improving diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5470 W: 940 L: 1042 D: 3488
sprt @ 10+0.1 th 1 Use improving margin for tweaking reductions. Take 1.
16-12-01 Gua PawnFlankTunes diff
34191/35000 iterations
70000/70000 games played
70000 @ 20+0.2 th 1 PawnFlankTunes
16-12-03 Voy pt diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 51164 W: 9031 L: 8942 D: 33191
sprt @ 10+0.1 th 1 stc
16-12-03 hex pawns_unsupp diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 4075 W: 671 L: 779 D: 2625
sprt @ 10+0.1 th 1 Supported pawn, opposite from previous test
16-12-01 sni flank_attack diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 94803 W: 16907 L: 16634 D: 61262
sprt @ 10+0.1 th 1 Combo of two parameters tweaks by GuardianRM and myself
16-12-03 Fis EasyTweak diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 29952 W: 5253 L: 5309 D: 19390
sprt @ 10+0.1 th 1 Take 3(Take 2 set to pri -1)
16-12-03 sni open_files5 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 28718 W: 4890 L: 4897 D: 18931
sprt @ 10+0.1 th 1 Scale down midgame eval by number of open files
16-11-30 fau tunebp diff
19561/20000 iterations
39574/40000 games played
40000 @ 60+0.6 th 1 Tuning for "BishopPairAttack" together with the Base Bishop values. Tuning at LTC since we need to find the best values to scale good in LTC since passing STC is easy for this patch, what we need is better values for LTC
16-12-03 Fis EasyTweak diff
LLR: -0.31 (-2.94,2.94) [0.00,4.00]
Total: 645 W: 87 L: 101 D: 457
sprt @ 10+0.1 th 1 Take 3(Take2 on pri -1)
16-12-03 fau bishoppair4 diff
LLR: -2.90 (-2.94,2.94) [0.00,5.00]
Total: 22440 W: 3819 L: 3849 D: 14772
sprt @ 10+0.1 th 1 since tuned values turned a green patch into a red patch let me try the exact opposite changes
16-12-03 fau bishoppair3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17370 W: 2960 L: 3013 D: 11397
sprt @ 10+0.1 th 1 Bishop Pair Attack bonus, with decreased values for Base Bishops both in mg and eg. (Tuned values)
16-12-03 sg pruning2 diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 16014 W: 2839 L: 2710 D: 10465
sprt @ 10+0.1 th 1 Another simplification try.
16-12-03 goo PvNode diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 2605 W: 419 L: 533 D: 1653
sprt @ 10+0.1 th 1 Not pruning for PvNode
16-12-01 sni connexion3 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 28238 W: 4997 L: 5058 D: 18183
sprt @ 10+0.1 th 1 Take 3 (priority -1)
16-12-02 Elb escape diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 62413 W: 7948 L: 7881 D: 46584
sprt @ 60+0.6 th 1 Can we get rid of the !pawn condition in decrease lmr reduction? (LTC)
16-12-02 nep max_element_opt diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 86663 W: 15310 L: 15074 D: 56279
sprt @ 10+0.1 th 1 Optimize pick_best() by saving one conditional jump for each invocation. No functional change
16-12-01 sni connexion3 diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 92560 W: 16576 L: 16420 D: 59564
sprt @ 10+0.1 th 1 Take 4 (priority -1)
16-12-02 Voy lmrmc diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 28833 W: 5074 L: 5078 D: 18681
sprt @ 10+0.1 th 1 stc
16-12-02 nep sse_sort_opt2 diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 14299 W: 2545 L: 2609 D: 9145
sprt @ 10+0.1 th 1 Optimize insertion_sort() using SSE2 (2 elements per cycle)
16-12-02 Elb escape diff
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40866 W: 7251 L: 7164 D: 26451
sprt @ 10+0.1 th 1 Can we get rid of the !pawn condition in decrease lmr reduction?
16-12-01 Gua IsolatedFlagTunes diff
14822/15000 iterations
30000/30000 games played
30000 @ 20+0.2 th 1 Tunes
16-12-02 mbo rank_threats_sp diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 38144 W: 6709 L: 6675 D: 24760
sprt @ 10+0.1 th 1 Rank based safe pawn threats.
16-12-02 SC secondBest diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 22569 W: 3963 L: 3993 D: 14613
sprt @ 10+0.1 th 1 This was performing very good locally. Take 5.