Stockfish Testing Queue

Finished - 4556 tests

17-06-30 sg tune_candidate_passers diff
24139/25000 iterations
48942/50000 games played
50000 @ 10+0.1 th 1 Tune scaling factors for candidate passers (ck=5)
17-06-29 sg candidate_passers3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13125 W: 2301 L: 2371 D: 8453
sprt @ 10+0.1 th 1 Give only quarter bonus if sacrificing square is blocked or well attacked
17-06-29 sg candidate_passers4 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 25347 W: 4499 L: 4517 D: 16331
sprt @ 10+0.1 th 1 Half bonus for passed pawn sacrifice only if sacrificing square is blocked or well attacked by the opponent
17-06-29 sg candidate_passers5 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 9117 W: 1608 L: 1695 D: 5814
sprt @ 10+0.1 th 1 Half bonus for passed pawn sacrifice only if candidate passer is not well defended. (Park with prio -1)
17-06-29 sg candidate_passers2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 41431 W: 7465 L: 7415 D: 26551
sprt @ 10+0.1 th 1 Consider the sacrificing pawn as candidate passer (so we can additionally remove a condition)
17-06-28 sg candidate_passers diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 53027 W: 9672 L: 9571 D: 33784
sprt @ 10+0.1 th 1 Generalize candidate passer breaks to cases where the sacrificing pawn is more than one push away.
17-06-28 sg candidate_passers diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 34234 W: 6189 L: 6169 D: 21876
sprt @ 10+0.1 th 1 Even more general
17-06-22 sg ttCapture diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 14969 W: 2634 L: 2696 D: 9639
sprt @ 10+0.1 th 1 Allow ttCapture only for depth >= 8
17-06-23 sg ttCapture2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 31305 W: 5602 L: 5595 D: 20108
sprt @ 10+0.1 th 1 Allow ttCapture only if not in check
17-06-23 sg ttCapture4 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 13330 W: 2341 L: 2410 D: 8579
sprt @ 10+0.1 th 1 ttDepth >= depth-3
17-06-23 sg ttCapture2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 26229 W: 4769 L: 4782 D: 16678
sprt @ 10+0.1 th 1 Do ttCapture stuff at pv nodes only if move is good (value > alpha).
17-06-22 sg ttCapture diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 50545 W: 9065 L: 8977 D: 32503
sprt @ 10+0.1 th 1 Allow ttCapture only for depth >= 4
17-06-23 sg ttCapture3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 33903 W: 6182 L: 6162 D: 21559
sprt @ 10+0.1 th 1 Reduce only if statScore <= 0. (Park with prio -1)
17-06-22 sg ttCapture2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10578 W: 1898 L: 1978 D: 6702
sprt @ 10+0.1 th 1 Allow ttCapture only for depth >= 5
17-06-22 sg ttCapture diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9049 W: 1573 L: 1660 D: 5816
sprt @ 10+0.1 th 1 Allow ttCapture only for depth <= 8
17-06-21 sg ttCapture2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7624 W: 1362 L: 1455 D: 4807
sprt @ 10+0.1 th 1 Trigger the ttCapture stuff only if not at an expected ALL node. Test against VoyagerOne passed patch.
17-06-21 sg ttCapture2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15703 W: 2711 L: 2770 D: 10222
sprt @ 10+0.1 th 1 Trigger the ttCapture stuff only if not at an expected cut node. Test against VoyagerOne passed patch.
17-06-21 sg ttCapture2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 5041 W: 847 L: 951 D: 3243
sprt @ 10+0.1 th 1 Trigger the ttCapture stuff only if not at an expected PV node. Test against VoyagerOne passed patch.
17-06-20 sg outposts2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 30242 W: 5440 L: 5437 D: 19365
sprt @ 10+0.1 th 1 Count this time only reachable knight outposts. (park with prio -1)
17-06-20 sg outposts diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 40804 W: 7444 L: 7395 D: 25965
sprt @ 10+0.1 th 1 Count as reachable outpost if square only blocked by own bishop or knight
17-06-20 sg outposts2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7921 W: 1407 L: 1499 D: 5015
sprt @ 10+0.1 th 1 Surprisingly only using knight outposts goes really bad, so try the opposite and use only bishop outposts.
17-06-20 sg pinners2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 12018 W: 2104 L: 2178 D: 7736
sprt @ 10+0.1 th 1 Do only increase kingAttackersWeight but only with quarter bonus.
17-06-19 sg pinners2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8759 W: 1546 L: 1634 D: 5579
sprt @ 10+0.1 th 1 Do only increase kingAttackersWeight. (Park with prio -1)
17-06-19 sg pinners2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 8934 W: 1597 L: 1684 D: 5653
sprt @ 10+0.1 th 1 Treat pinners as king ring attackers if pinned piece is not defended by a pawn. Do only increase kingAttackersCount.
17-06-19 sg pinners2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 15986 W: 2887 L: 2944 D: 10155
sprt @ 10+0.1 th 1 Treat pinners as king ring attackers if pinned piece is not defended by a pawn
17-06-19 sg pinners diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8588 W: 1495 L: 1584 D: 5509
sprt @ 10+0.1 th 1 Treat pinners as king ring attackers. (Inspired by pb's pinners test)
17-06-17 sg lmr3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 7640 W: 1322 L: 1415 D: 4903
sprt @ 10+0.1 th 1 Less LMR for captures at pv nodes (Fixed because of negative reduction)
17-06-17 sg lmr diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 25951 W: 4666 L: 4681 D: 16604
sprt @ 10+0.1 th 1 Less LMR for captures at root.
17-06-17 sg lmr3 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18055 W: 3246 L: 3294 D: 11515
sprt @ 10+0.1 th 1 Less LMR for pawn moves at root.
17-06-17 sg lmr2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 10800 W: 1914 L: 1993 D: 6893
sprt @ 10+0.1 th 1 Less LMR for checks at root
17-06-16 sg lmr diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9715 W: 1790 L: 1874 D: 6051
sprt @ 10+0.1 th 1 Less LMR at root
17-06-14 sg king_safety_tuned diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 51770 W: 6806 L: 6755 D: 38209
sprt @ 60+0.6 th 1 LTC: Final tuned values after nearly 120K games
17-06-14 sg king_shelter_defense diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 4842 W: 853 L: 958 D: 3031
sprt @ 10+0.1 th 1 Try it with compensation in shelter weakness array
17-06-14 sg king_shelter_defense diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 8823 W: 1554 L: 1642 D: 5627
sprt @ 10+0.1 th 1 Bonus for king shelter pawns which defended by a pawn
17-06-14 sg king_safety_tuned diff
LLR: 2.98 (-2.94,2.94) [0.00,5.00]
Total: 12506 W: 2303 L: 2125 D: 8078
sprt @ 10+0.1 th 1 Final tuned values after nearly 120K games
17-06-02 sg tune_king_safety diff
56817/60000 iterations
117695/120000 games played
120000 @ 30+0.3 th 1 After my failed series about weighting with king file occupation i try a more general approach by extending shelterWeakness and stormDanger by an extra dimension notKingFile vs kingFile. Because of tc dependency of king safety i tune at intermediate tc 30+0.3 (176 parameters, ck=10, Low throughput).
17-06-12 sg fix_pet21 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 28850 W: 5258 L: 5260 D: 18332
sprt @ 10+0.1 th 1 Increase moveCount in QS only for non-captures. Take 4
17-06-12 sg fix_pet21 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 23841 W: 4350 L: 4373 D: 15118
sprt @ 10+0.1 th 1 Fix the issue with PET#21 (see forum) by allowing evasion pruning only for material more than a bishop
17-06-12 sg fix_pet21 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5449 W: 930 L: 1032 D: 3487
sprt @ 10+0.1 th 1 Now check material condition only for moveCount = 2. Take 3. Set other test to prio -1.
17-06-12 sg fix_pet21 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 15707 W: 2821 L: 2879 D: 10007
sprt @ 10+0.1 th 1 Now check material condition only for moveCount > 2. Take 2
17-06-11 sg passed_pawn_state diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 19128 W: 3531 L: 3574 D: 12023
sprt @ 10+0.1 th 1 I think more tuning is needed but first test the tuned parameters
17-06-09 sg tune_passed_pawn_state diff
29241/30000 iterations
60000/68000 games played
68000 @ 10+0.1 th 1 Generalize the different bonus by defended and unsafe squares to an bonus array and tune it. Additionally we can use now different values for middle and endgame. The starting values are calculated to natch functional the master.
17-06-09 sg passed_defended diff
LLR: -2.16 (-2.94,2.94) [0.00,5.00]
Total: 23850 W: 4377 L: 4366 D: 15107
sprt @ 10+0.1 th 1 Exclude the case of pawn on 7th row because the condition would be always true.
17-06-09 sg passed_defended diff
LLR: -1.45 (-2.94,2.94) [0.00,5.00]
Total: 21919 W: 3926 L: 3895 D: 14098
sprt @ 10+0.1 th 1 The base idea was significant yellow. Excluding the case of a pawn at 7th rank but a not defended promotion square seems currently going bad. So try opposite and exclude the other part.
17-06-07 sg passed_defended diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 63936 W: 11728 L: 11579 D: 40629
sprt @ 10+0.1 th 1 bonus if all squares except blocking square are defended
17-06-07 sg passed_defended3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 31479 W: 5710 L: 5701 D: 20068
sprt @ 10+0.1 th 1 more bonus if blocking square is defended by pawn
17-06-07 sg passed_defended2 diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 39083 W: 7103 L: 7062 D: 24918
sprt @ 10+0.1 th 1 Bonus if next square after block square is defended
17-06-08 sg passed_defended2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 12938 W: 2332 L: 2402 D: 8204
sprt @ 10+0.1 th 1 Bonus if next square after block square is defended. Try lower bonus.
17-06-08 sg passed_defended2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 19111 W: 3479 L: 3522 D: 12110
sprt @ 10+0.1 th 1 Bonus if next square after block square is defended. Try higher bonus.
17-06-07 sg passed_defended diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 13806 W: 2465 L: 2532 D: 8809
sprt @ 10+0.1 th 1 Bonus if their exists defended unsafe squares