| 19-08-18 |
MJZ |
Alpha-ReductionB2 |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 25869 W: 5664 L: 5701 D: 14504
|
sprt
@ 10+0.1 th 1 |
Increase reduction at PvNode if we already had a good move. bestValue = alpha > 320, d > 6 |
| 19-08-18 |
MJZ |
Alpha-ReductionB2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 31576 W: 7064 L: 7071 D: 17441
|
sprt
@ 10+0.1 th 1 |
Increase reduction at PvNode if we already had a good move. bestValue = alpha > 280, d > 6 |
| 19-08-18 |
MJZ |
Alpha-ReductionB |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 53434 W: 11916 L: 11815 D: 29703
|
sprt
@ 10+0.1 th 1 |
Increase reduction at PvNode if we already had a good move. bestValue = alpha > 300, d > 6 |
| 19-08-18 |
MJZ |
Alpha-ReductionB2 |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 29716 W: 6621 L: 6638 D: 16457
|
sprt
@ 10+0.1 th 1 |
Increase reduction at PvNode if we already had a good move. bestValue = alpha > 300, d > 4 |
| 19-08-18 |
MJZ |
Knight_EG2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 50993 W: 11366 L: 11277 D: 28350
|
sprt
@ 10+0.1 th 1 |
EG penalty for lonely Knight with pawns in both flanks. (0, 20) - only vs Rook, 12 > nb.pawns > 3 |
| 19-08-18 |
MJZ |
Alpha-ReductionB2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 41757 W: 9411 L: 9367 D: 22979
|
sprt
@ 10+0.1 th 1 |
Increase reduction at PvNode if we already had a good move. bestValue = alpha > 300, d > 7 |
| 19-08-18 |
MJZ |
Alpha-Ext-tune |
diff
|
15517/20000 iterations
32391/40000 games played
|
40000
@ 10+0.1 th 1 |
Alpha-Extension tuning parameters - second try with different initial values |
| 19-08-18 |
MJZ |
Alpha-Ext-tune |
diff
|
2440/20000 iterations
5117/40000 games played
|
40000
@ 10+0.1 th 1 |
Alpha-Extension tuning parameters |
| 19-08-17 |
MJZ |
Knight_EG2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 12050 W: 2668 L: 2773 D: 6609
|
sprt
@ 10+0.1 th 1 |
EG penalty for lonely Knight with pawns in both flanks. (0, 20) - only vs Rook, nb.pawns > 5 |
| 19-08-17 |
MJZ |
Knight_EG |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 41482 W: 9228 L: 9187 D: 23067
|
sprt
@ 10+0.1 th 1 |
EG penalty for lonely Knight with pawns in both flanks. (0, 20) - only vs Rook |
| 19-08-17 |
MJZ |
Knight_EG2 |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 19934 W: 4404 L: 4470 D: 11060
|
sprt
@ 10+0.1 th 1 |
EG penalty for lonely Knight with pawns in both flanks. (0, 20) - only vs Rook, nb.pawns < 10 |
| 19-08-17 |
MJZ |
Knight_EG |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 36473 W: 8054 L: 8038 D: 20381
|
sprt
@ 10+0.1 th 1 |
EG penalty for lonely Knight with pawns in both flanks. (0, 10) - only vs Rook |
| 19-08-17 |
MJZ |
Knight_EG |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 16496 W: 3581 L: 3664 D: 9251
|
sprt
@ 10+0.1 th 1 |
EG penalty for lonely Knight with pawns in both flanks. (0, 10) |
| 19-08-16 |
MJZ |
Reduc-depth_MN |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 63395 W: 14053 L: 13904 D: 35438
|
sprt
@ 10+0.1 th 1 |
Try to separate depth and move number coefficients in reduction formula. 23.2 / 23.6 |
| 19-08-16 |
MJZ |
LMR-TTcapture2 |
diff
|
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 13150 W: 2845 L: 2944 D: 7361
|
sprt
@ 10+0.1 th 1 |
Skip quiet moves if TTmove is a capture and staticEval is below alpha-400 - shallow depths (< 4) |
| 19-08-16 |
MJZ |
LMR-TTcapture2 |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 9524 W: 2073 L: 2191 D: 5260
|
sprt
@ 10+0.1 th 1 |
Skip quiet moves if TTmove is a capture and staticEval is below alpha-200 - shallow depths (< 4) |
| 19-08-16 |
MJZ |
LMR-TTcapture |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 8643 W: 1881 L: 2003 D: 4759
|
sprt
@ 10+0.1 th 1 |
Skip quiet moves if TTmove is a capture at shallow depths (< 4) |
| 19-08-16 |
MJZ |
Reduc-depth |
diff
|
ELO: 0.02 +-3.8 (95%) LOS: 50.4%
Total: 19161 W: 5720 L: 5719 D: 7722
|
20000
@ 2+0.02 th 1 |
Test : evaluating the impact of less reduction at VSTC. (21.4) |
| 19-08-16 |
MJZ |
Reduc-depth |
diff
|
ELO: 1.73 +-3.7 (95%) LOS: 81.7%
Total: 19705 W: 5904 L: 5806 D: 7995
|
20000
@ 2+0.02 th 1 |
Test : evaluating the impact of less reduction at VSTC. (22.4) |
| 19-08-16 |
MJZ |
Reduc-depth_MN2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 23286 W: 5100 L: 5149 D: 13037
|
sprt
@ 10+0.1 th 1 |
Try to separate depth and move number coefficients in reduction formula. 23.0 / 23.8 |
| 19-08-16 |
MJZ |
Reduc-depth_MN2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 13806 W: 3017 L: 3113 D: 7676
|
sprt
@ 10+0.1 th 1 |
Try to separate depth and move number coefficients in reduction formula. 23.6 / 23.2 |
| 19-08-15 |
MJZ |
TT-Save2 |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 10712 W: 2329 L: 2441 D: 5942
|
sprt
@ 10+0.1 th 1 |
Set ttPv=true if we have a close upper and lower bounds. depth > 6, abs(ttValue-beta) < 50 verification |
| 19-08-15 |
MJZ |
TT-Save2 |
diff
|
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 15167 W: 3315 L: 3404 D: 8448
|
sprt
@ 10+0.1 th 1 |
Set ttPv=true if we have a close upper and lower bounds. depth > 6 |
| 19-08-15 |
MJZ |
TT-Save2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 6393 W: 1369 L: 1502 D: 3522
|
sprt
@ 10+0.1 th 1 |
Set ttPv=true if we have a close upper and lower bounds. |
| 19-08-15 |
MJZ |
TT-Save |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 2147 W: 394 L: 547 D: 1206
|
sprt
@ 10+0.1 th 1 |
Don't save into TT if TTdepth > depth. |
| 19-08-13 |
MJZ |
Futility_test |
diff
|
LLR: -2.03 (-2.94,2.94) [0.50,4.50]
Total: 107250 W: 32508 L: 31985 D: 42757
|
sprt
@ 2+0.02 th 1 |
Test @VSTC for information : futility max depth = 5 |
| 19-08-13 |
MJZ |
WeakPieces |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 18161 W: 4008 L: 4083 D: 10070
|
sprt
@ 10+0.1 th 1 |
Bonus for enemy weak pieces that can be easily attacked by ou queen. (0, 4) |
| 19-08-12 |
MJZ |
Check_Extension |
diff
|
LLR: -0.21 (-2.94,2.94) [0.00,3.50]
Total: 25992 W: 4412 L: 4351 D: 17229
|
sprt
@ 60+0.6 th 1 |
Limit check extension to lower depths. d < 12. Trying Spec LTC (I will stop it if no clear improvement after 10k games). |
| 19-08-13 |
MJZ |
WeakPieces-tune |
diff
|
10843/20000 iterations
22765/40000 games played
|
40000
@ 10+0.1 th 1 |
Bonus for enemy weak pieces - tuning with different parameters |
| 19-08-13 |
MJZ |
WeakPieces-tune |
diff
|
5439/20000 iterations
11391/40000 games played
|
40000
@ 10+0.1 th 1 |
Bonus for enemy weak pieces - tuning |
| 19-08-13 |
MJZ |
Futility_test |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 21669 W: 4757 L: 4814 D: 12098
|
sprt
@ 10+0.1 th 1 |
Test : futility max depth = (6 + ply > 5) |
| 19-08-13 |
MJZ |
Futility_test |
diff
|
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 13151 W: 2858 L: 2957 D: 7336
|
sprt
@ 10+0.1 th 1 |
Test : increase futility max depth if abs(eval - staticEval) > 200 |
| 19-08-13 |
MJZ |
Futility_test2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 12633 W: 2777 L: 2879 D: 6977
|
sprt
@ 10+0.1 th 1 |
Test : increase futility max depth if abs(eval - staticEval) < 200 |
| 19-08-12 |
MJZ |
Check_Extension |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 34466 W: 7728 L: 7721 D: 19017
|
sprt
@ 10+0.1 th 1 |
Limit check extension to lower depths. d < 12 |
| 19-08-12 |
MJZ |
Check_Extension |
diff
|
LLR: -2.94 (-2.94,2.94) [0.50,4.50]
Total: 9764 W: 2138 L: 2254 D: 5372
|
sprt
@ 10+0.1 th 1 |
Limit check extension to lower depths. d < 10 |
| 19-08-08 |
MJZ |
NMP-lastMove2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.00,3.50]
Total: 137288 W: 23018 L: 22809 D: 91461
|
sprt
@ 60+0.6 th 1 |
NMP if staticEval decrease in last move. Threshold 140 - LTC |
| 19-08-09 |
MJZ |
Passed_backpieces |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 22191 W: 4893 L: 4948 D: 12350
|
sprt
@ 10+0.1 th 1 |
Remove back protection or attack for passed pawns if there is an obstruction. New test after parameter tuning. |
| 19-08-09 |
MJZ |
Passed_back-tune |
diff
|
14621/20000 iterations
30497/40000 games played
|
40000
@ 10+0.1 th 1 |
Remove back protection or attack for passed pawns if there is an obstruction. Tuning parameters? |
| 19-08-09 |
MJZ |
Passed_backpieces |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 22290 W: 4877 L: 4931 D: 12482
|
sprt
@ 10+0.1 th 1 |
Remove back protection or attack for passed pawns if there is an obstruction. |
| 19-08-09 |
MJZ |
staticEval-Update |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 3002 W: 587 L: 737 D: 1678
|
sprt
@ 10+0.1 th 1 |
In case of beta cutoff, update staticEval in Qsearch with the next position |
| 19-08-09 |
MJZ |
lastMove-staticEval3 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 10149 W: 2200 L: 2314 D: 5635
|
sprt
@ 10+0.1 th 1 |
Decrease reduction if opponent last move is quiet and improves staticEval. th = 120 |
| 19-08-09 |
MJZ |
lastMove-staticEval3 |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 10852 W: 2314 L: 2425 D: 6113
|
sprt
@ 10+0.1 th 1 |
Increase reduction if opponent last move is quiet and decrease staticEval. th = 160 |
| 19-08-08 |
MJZ |
lastMove-staticEval |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 34561 W: 7730 L: 7723 D: 19108
|
sprt
@ 10+0.1 th 1 |
Decrease reduction if opponent last move is quiet and improves staticEval. th = 160 - new run after correction |
| 19-08-08 |
MJZ |
Attacked_Queen |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 11436 W: 2516 L: 2624 D: 6296
|
sprt
@ 10+0.1 th 1 |
Bonus if opponent queen is attacked (50, 50) |
| 19-08-08 |
MJZ |
lastMove-staticEval2 |
diff
|
LLR: -0.63 (-2.94,2.94) [0.50,4.50]
Total: 1275 W: 282 L: 311 D: 682
|
sprt
@ 10+0.1 th 1 |
Decrease reduction if last move is quiet and improves staticEval. th = 160, small compensation to reduction - new run after correction |
| 19-08-08 |
MJZ |
lastMove-staticEval |
diff
|
LLR: -0.41 (-2.94,2.94) [0.50,4.50]
Total: 1029 W: 220 L: 238 D: 571
|
sprt
@ 10+0.1 th 1 |
Decrease reduction if last move is quiet and improves staticEval. th = 160 - new run after correction |
| 19-08-06 |
MJZ |
NMP-lastMove2 |
diff
|
LLR: 2.96 (-2.94,2.94) [0.50,4.50]
Total: 150356 W: 33983 L: 33068 D: 83305
|
sprt
@ 10+0.1 th 1 |
Back to original idea : NMP if staticEval decrease in last move. Threshold 140 |
| 19-08-08 |
MJZ |
lastMove-staticEval |
diff
|
LLR: -2.96 (-2.94,2.94) [0.50,4.50]
Total: 3708 W: 779 L: 927 D: 2002
|
sprt
@ 10+0.1 th 1 |
Decrease reduction if last move is quiet and improves staticEval. th = 160 |
| 19-08-08 |
MJZ |
lastMove-staticEval2 |
diff
|
LLR: -2.95 (-2.94,2.94) [0.50,4.50]
Total: 3175 W: 656 L: 806 D: 1713
|
sprt
@ 10+0.1 th 1 |
Decrease reduction if last move is quiet and improves staticEval. th = 160, small compensation to reduction |
| 19-08-06 |
MJZ |
NMP-rootDepth2 |
diff
|
LLR: -0.56 (-2.94,2.94) [0.00,3.50]
Total: 6840 W: 1162 L: 1176 D: 4502
|
sprt
@ 60+0.6 th 1 |
NMP formula with relative depth - 300 / 180 - limited test @LTC (I will stop it if no improvement after 10k games) |