Tests

Finished 25 shown

Author	Engine	Identifier	Code	Time control	Results
MasterStruct	Eunoia	feat/minor-corrhist #8937 Finished Error	diff ↗	40.0+0.40	LLR: 1.61 (-2.25, 2.89) [0.00, 5.00] Games: 6926 W: 2028 L: 1918 D: 2980 Ptnml(0-2): 163, 844, 1388, 856, 212
Add Minor corrhist
Ellie	WakWak	repetition-fix #8940 Passed Duck960	diff ↗	8.0+0.08	LLR: 2.90 (-2.25, 2.89) [0.00, 4.20] Games: 1294 W: 584 L: 410 D: 300 Ptnml(0-2): 26, 68, 343, 126, 84
go back to twofold and repetitions seen before root require 3fold
Ellie	WakWak	repetition-fix #8939 Passed Duck960	diff ↗	8.0+0.08	LLR: 2.93 (-2.25, 2.89) [-4.20, 0.00] Games: 1082 W: 507 L: 338 D: 237 Ptnml(0-2): 20, 54, 278, 115, 74
include da duck in repetition. holy fuck we're dumb.
Ellie	sable	bunch-of-shite #8938 Finished FRC	diff ↗	8.0+0.08	LLR: -0.52 (-2.25, 2.89) [0.00, 4.20] Games: 5472 W: 1726 L: 1731 D: 2015 Ptnml(0-2): 40, 488, 1683, 487, 38
history-adjusted futility, correction-aware lmr, improving rfp, fail-low beta narrowing and softened qsearch stand-pat. Implemented in batch inspired by patches from wakwak and JustBot BENCH: 2557503
MasterStruct	Eunoia	feat/minor-corrhist #8936 Passed	diff ↗	8.0+0.08	LLR: 2.90 (-2.25, 2.89) [0.00, 5.00] Games: 12430 W: 3761 L: 3556 D: 5113 Ptnml(0-2): 379, 1411, 2486, 1504, 435
Add Minor corrhist
Ellie	WakWak	main #8926 Failed Duck960	diff ↗	40.0+0.40	Elo: -3.20 +- 3.90 (95%) [N=10000] Games: 10002 W: 3341 L: 3433 D: 3228 Ptnml(0-2): 216, 874, 2865, 878, 168
Challah 4 net (#213)
Ellie	WakWak	mantou #8927 Finished Duck960	diff ↗	8.0+0.08	LLR: 0.51 (-2.25, 2.89) [-4.20, 0.00] Games: 7168 W: 2678 L: 2679 D: 1811 Ptnml(0-2): 279, 608, 1787, 655, 255
simd cleanup
MasterStruct	Eunoia	feat/nonpawn-corrhist #8935 Passed	diff ↗	8.0+0.08	LLR: 2.91 (-2.25, 2.89) [0.00, 5.00] Games: 2300 W: 782 L: 629 D: 889 Ptnml(0-2): 60, 213, 479, 310, 88
Add nonpawn corrhists
nocturn9x	Heimdall	optional-eval-qsearch-evasions #8933 Finished	diff ↗	8.0+0.08	LLR: -0.37 (-2.25, 2.89) [0.00, 3.00] Games: 3028 W: 734 L: 750 D: 1544 Ptnml(0-2): 17, 392, 707, 386, 12
Search all legal evasions in quiescence (bench 5983062) Quiescence previously generated only captures and applied losing-capture SEE pruning in check. Generate every legal evasion and bypass that pruning while checked; return matedIn(ply) when none exist. Keep capture-only generation outside check and retain bound-only TT storage. Add regressions for checkmate without an allocated NNUE state, all four quiet king evasions from a rook check, and a SEE-negative capture evasion. These changes are independent of optional static-eval representation and consumers. Validation: focused AVX2 checked/danger and scalar checked search regressions, including unchanged TT/stack layouts; Python suite (99 run, 5 skipped). Optimized depth-13; depth 14 visited 8366978 nodes. Co-authored-by: GPT 6 <codex@openai.com>
nocturn9x	Heimdall	optional-eval-in-check #8932 Finished	diff ↗	8.0+0.08	LLR: -1.64 (-2.25, 2.89) [0.00, 3.00] Games: 1316 W: 276 L: 374 D: 666 Ptnml(0-2): 14, 193, 331, 117, 3
Guard search heuristics and skip static evaluation in check (bench 5401484) Checked nodes carry noEval through the search stack and TT. Main-search futility pruning and the fail-low extension after a reduced parent search previously consumed static eval in check; require present evaluations before using them. Guard improving, reverse futility, null move and correction-history arithmetic likewise, and compute TT-adjusted eval only where needed. Disable qsearch stand pat when static eval is absent. Preserve capture-only move generation and SEE pruning here; full check evasions and qsearch mate recognition are isolated in a follow-up commit. Until then, checked qsearch nodes with no searched capture retain -SCORE_INF. At the hard ply boundary, return a neutral score in check instead of evaluating. Enforce the rawEval precondition with doAssert, including in danger builds. Subprocess tests verify this invariant and absent-eval unwrapping survive there. Validation: focused AVX2 checked/danger and scalar checked tests; exhaustive optional-eval/TT round trips and unchanged layouts; Python suite (99 run, 5 skipped). Optimized depth-13; depth 14 visited 8078278 nodes. Twelve alternating depth-14 pairs on CPU 2 versus the representation-only commit show -1.25% throughput change, 95% bootstrap interval [-7.32%, +3.49%]. Search trees differ, so this measures throughput rather than identical work or playing strength. Co-authored-by: GPT 6 <codex@openai.com>
MasterStruct	Eunoia	refactor/corrhist #8928 Finished	diff ↗	8.0+0.08	LLR: 0.44 (-2.25, 2.89) [-7.00, 0.00] Games: 3068 W: 901 L: 907 D: 1260 Ptnml(0-2): 84, 318, 722, 340, 70
Refactor correction histories
paul	Byte-Knight	fix/tt-lookup #8915 Passed	diff ↗	8.0+0.08	LLR: 2.96 (-2.94, 2.94) [-5.00, 0.00] Games: 25866 W: 7126 L: 7114 D: 11626 Ptnml(0-2): 482, 2890, 6225, 2806, 530
fix: check if entry is empty before ttable get_entry
Ellie	WakWak	multipv #8921 Passed Duck960	diff ↗	8.0+0.08	LLR: 2.95 (-2.25, 2.89) [-4.20, 0.00] Games: 15044 W: 5302 L: 5206 D: 4536 Ptnml(0-2): 378, 1351, 4011, 1361, 421
aha! stale scores were survivng asp retries
Ellie	WakWak	multipv #8920 Finished Duck960	diff ↗	8.0+0.08	LLR: -1.72 (-2.25, 2.89) [-4.20, 0.00] Games: 9060 W: 3107 L: 3233 D: 2720 Ptnml(0-2): 272, 854, 2376, 784, 244
maybe some fixes
MasterStruct	Eunoia	feat/iir #8919 Finished	diff ↗	8.0+0.08	LLR: -0.06 (-2.25, 2.89) [0.00, 5.00] Games: 3372 W: 1014 L: 1000 D: 1358 Ptnml(0-2): 110, 386, 682, 396, 112
Add Internal Iterative Reductions
Ellie	WakWak	challah_4 #8918 Passed Duck960	diff ↗	8.0+0.08	LLR: 3.01 (-2.25, 2.89) [0.00, 4.20] Games: 7826 W: 3021 L: 2815 D: 1990 Ptnml(0-2): 263, 658, 1945, 704, 343
Ellie	WakWak	challah_4 #8917 Finished Error Duck960	diff ↗	8.0+0.08	LLR: 0.00 (-2.25, 2.89) [0.00, 4.20] Games: 0 W: 0 L: 0 D: 0 Ptnml(0-2): 0, 0, 0, 0, 0
Ellie	WakWak	challah_4 #8916 Finished Error Duck960	diff ↗	8.0+0.08	LLR: 0.00 (-2.25, 2.89) [0.00, 4.20] Games: 0 W: 0 L: 0 D: 0 Ptnml(0-2): 0, 0, 0, 0, 0
paul	Byte-Knight	fix/see-pruning #8914 Finished Error	diff ↗	8.0+0.08	LLR: 0.00 (-2.94, 2.94) [-5.00, 0.00] Games: 0 W: 0 L: 0 D: 0 Ptnml(0-2): 0, 0, 0, 0, 0
fix: gate see pruning with moves_seen > 0
DarkNeutrino	WakWak	nmp-has-non-pawns #8909 Passed Duck960	diff ↗	8.0+0.08	LLR: 3.74 (-2.25, 2.89) [-4.20, 0.00] Games: 37496 W: 13106 L: 13055 D: 11335 Ptnml(0-2): 956, 3330, 10105, 3421, 936
Only do NMP if we have minor or major pieces
MasterStruct	Eunoia	feat/aw-centering #8913 Finished	diff ↗	8.0+0.08	LLR: 1.15 (-2.25, 2.89) [0.00, 5.00] Games: 5544 W: 1595 L: 1520 D: 2429 Ptnml(0-2): 118, 569, 1335, 620, 130
Widen aspiration windows based on score
nocturn9x	Heimdall	master #8814 Finished Error	diff ↗	N=12000	Generated 4509392/5000000 Games Elo: -0.08 +- 0.26 (95%) [N=5000000] Games: 4509392 W: 1498806 L: 1499796 D: 1510790
Ellie	WakWak	main #8852 Finished Error Duck960	diff ↗	N=5000	Generated 4151650/10000000 Games Elo: 0.09 +- 0.29 (95%) [N=10000000] Games: 4151650 W: 1546247 L: 1545208 D: 1060195
ZTS439	WakWak	duck_count_history #8910 Passed Duck960	diff ↗	8.0+0.08	LLR: 2.89 (-2.25, 2.89) [0.00, 4.20] Games: 10366 W: 3702 L: 3508 D: 3156 Ptnml(0-2): 261, 944, 2632, 1032, 314
use in lmr instead of dcp
ZTS439	WakWak	duck_count_history #8904 Finished Duck960	diff ↗	8.0+0.08	LLR: -1.84 (-2.25, 2.89) [0.00, 4.20] Games: 3072 W: 1028 L: 1113 D: 931 Ptnml(0-2): 103, 293, 789, 288, 63