FuryBench

FuryBench Testing Framework

home to Integral PlentyChess Obsidian Caissa Quanticade Serendipity Spaghet Schoenemann Clarity Illumina Pawnocchio Vine Anura

Finished
Fury	Integral	aggressive_policy_with_ti	diff	10.0+0.10	LLR -2.28 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -10.12 ± 4.79 (-14.91, -5.33) 95% GAMES 5,222 (24.0% 1,254W 49.1% 2,562D 26.9% 1,406L) PENTA [19⁻² 678⁻¹ 1,368⁰ 528⁺¹ 18⁺²]
Witek	Caissa	fix-bmi2	diff	5.0+0.05	LLR +2.95 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO +3.01 ± 1.84 (+1.17, +4.85) 95% GAMES 37,274 (24.9% 9,297W 51.0% 19,003D 24.1% 8,974L) PENTA [203⁻² 4,093⁻¹ 9,721⁰ 4,418⁺¹ 202⁺²]	Fix pext/pdep and MulHi64 falling back to software on non-Windows builds Also, with pext working, restore -DUSE_BMI2 in the makefile native probe. No functional change.
Fury	Integral	ti-vf-768	diff	N=20000	LLR -2.43 (-2.25, +2.89) bounds [-3.00, 1.00] elo ELO -21.25 ± 9.23 (-30.47, -12.05) 95% GAMES 2,374 (27.4% 650W 39.1% 929D 33.5% 795L) PENTA [67⁻² 338⁻¹ 508⁰ 221⁺¹ 53⁺²]
Fury	Integral	[Obsidian] main	diff	40.0+0.40	ELO -27.00 ± 5.04 (-32.04, -21.98) 95% GAMES 5,002 of 5,000 (21.7% 1,083W 48.9% 2,448D 29.4% 1,471L) PENTA [37⁻² 781⁻¹ 1,223⁰ 453⁺¹ 7⁺²]
DarkNe	Quanticade	net58 *	diff	16.0+0.16	LLR -0.93 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO -1.52 ± 3.62 (-5.14, +2.10) 95% GAMES 8,000 (23.3% 1,864W 53.0% 4,237D 23.7% 1,899L) PENTA [3⁻² 912⁻¹ 2,204⁰ 879⁺¹ 2⁺²]	1024L1 with 50% more SBs on both stages compared to master
Fury	Integral	lmp_hist	diff	10.0+0.10	LLR -2.25 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -1.17 ± 1.98 (-3.14, +0.81) 95% GAMES 30,656 (24.8% 7,612W 50.0% 15,329D 25.2% 7,715L) PENTA [81⁻² 3,679⁻¹ 7,913⁰ 3,572⁺¹ 83⁺²]
Fury	Integral	aggressive_policy_with_ti	diff	10.0+0.10	LLR -2.25 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -9.68 ± 4.79 (-14.47, -4.89) 95% GAMES 5,492 (23.7% 1,304W 49.7% 2,731D 26.5% 1,457L) PENTA [29⁻² 726⁻¹ 1,372⁰ 607⁺¹ 12⁺²]
Fury	Integral	lmp_hist	diff	10.0+0.10	LLR -2.03 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -0.96 ± 1.97 (-2.93, +1.01) 95% GAMES 31,458 (24.9% 7,840W 49.9% 15,691D 25.2% 7,927L) PENTA [96⁻² 3,802⁻¹ 8,015⁰ 3,725⁺¹ 91⁺²]
Fury	Integral	qs_see_catchup	diff	10.0+0.10	LLR -2.30 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -2.49 ± 2.57 (-5.06, +0.07) 95% GAMES 17,982 (24.5% 4,399W 50.4% 9,055D 25.2% 4,528L) PENTA [47⁻² 2,179⁻¹ 4,663⁰ 2,060⁺¹ 42⁺²]
Witek	Caissa	avxvnni_no_bmi2	diff	5.0+0.05	LLR +2.95 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO +3.86 ± 2.15 (+1.71, +6.00) 95% GAMES 28,568 (25.1% 7,166W 50.9% 14,553D 24.0% 6,849L) PENTA [148⁻² 3,252⁻¹ 7,222⁰ 3,459⁺¹ 203⁺²]	AVX-VNNI support
Fury	Integral	qs_see_catchup	diff	10.0+0.10	LLR -2.34 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -21.78 ± 6.74 (-28.52, -15.05) 95% GAMES 2,556 (22.2% 568W 49.3% 1,260D 28.5% 728L) PENTA [8⁻² 379⁻¹ 659⁰ 229⁺¹ 3⁺²]
Fury	Integral	lmr_laterality	diff	10.0+0.10	LLR -1.20 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -3.03 ± 3.90 (-6.92, +0.87) 95% GAMES 8,154 (24.0% 1,956W 51.2% 4,171D 24.9% 2,027L) PENTA [18⁻² 1,051⁻¹ 2,012⁰ 976⁺¹ 20⁺²]
Fury	Integral	lmr_laterality	diff	10.0+0.10	LLR -0.86 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO -1.19 ± 3.19 (-4.38, +2.00) 95% GAMES 11,424 (24.7% 2,817W 50.3% 5,751D 25.0% 2,856L) PENTA [32⁻² 1,330⁻¹ 3,019⁰ 1,307⁺¹ 24⁺²]
Mergen	Illumina	pre-v4-tune	diff	40.0+0.40	LLR +2.89 (-2.25, +2.89) bounds [0.00, 5.00] elo ELO +3.15 ± 2.52 (+0.64, +5.67) 95% GAMES 18,394 (21.0% 3,865W 58.9% 10,831D 20.1% 3,698L) PENTA [36⁻² 2,074⁻¹ 4,817⁰ 2,227⁺¹ 43⁺²]	Tuning
DarkNe	Quanticade	net58	diff	N=5000	LLR +1.78 (-2.25, +2.89) bounds [0.00, 3.00] elo ELO +5.60 ± 4.67 (+0.93, +10.28) 95% GAMES 12,032 (33.9% 4,080W 33.8% 4,066D 32.3% 3,886L) PENTA [504⁻² 1,253⁻¹ 2,349⁰ 1,365⁺¹ 545⁺²]	net58
Vast	Clarity	tb-filtering	diff	8.0+0.08	LLR -2.27 (-2.25, +2.89) bounds [-3.00, 1.00] elo ELO -2.26 ± 2.51 (-4.76, +0.25) 95% GAMES 21,546 (22.6% 4,874W 54.1% 11,658D 23.3% 5,014L) PENTA [201⁻² 2,459⁻¹ 5,541⁰ 2,423⁺¹ 149⁺²]	Trying it again with TBs, hopefully I can get some workers Issues to be fixed with clarity & fastchess (no SyzygyProbeLimit and bestmove doesnt match last pv)
Fury	Integral	ti_tweaks	diff	20.0+0.20	LLR +2.91 (-2.25, +2.89) bounds [0.00, 4.00] elo ELO +12.31 ± 5.43 (+6.88, +17.74) 95% GAMES 3,868 (26.6% 1,029W 50.3% 1,947D 23.1% 892L) PENTA [4⁻² 395⁻¹ 998⁰ 534⁺¹ 3⁺²]
Witek	Caissa	eval-84-144B	diff	10.0+0.10	LLR +2.97 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO +1.35 ± 1.06 (+0.30, +2.41) 95% GAMES 110,222 (23.7% 26,151W 52.9% 58,349D 23.3% 25,722L) PENTA [337⁻² 13,024⁻¹ 28,013⁰ 13,347⁺¹ 390⁺²]	Optimize StaticExchangeEvaluation Measured bench speedup: +0.233% No functional change.
Witek	Caissa	eval-84-144B	diff	10.0+0.10	LLR +2.95 (-2.94, +2.94) bounds [0.00, 2.00] elo ELO +8.26 ± 3.10 (+5.16, +11.35) 95% GAMES 10,860 (27.1% 2,946W 48.1% 5,226D 24.8% 2,688L) PENTA [1⁻² 1,087⁻¹ 2,998⁰ 1,341⁺¹ 3⁺²]	Optimize StaticExchangeEvaluation Measured bench speedup: +0.233% No functional change.
Mergen	Illumina	pre-v4-tune	diff	40.0+0.40	TUNE 18 parameters ITER 3,342 / 25,000 GAMES 53,480 / 400,000	Pre-release tune
Fury	Integral	ti_tweaks	diff	40.0+0.40	LLR +2.90 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO +5.94 ± 2.98 (+2.97, +8.92) 95% GAMES 13,274 (25.7% 3,415W 50.3% 6,671D 24.0% 3,188L) PENTA [16⁻² 1,507⁻¹ 3,363⁰ 1,736⁺¹ 15⁺²]
Witek	Caissa	see_speedup	diff	4.0+0.04	LLR +2.96 (-2.94, +2.94) bounds [-2.00, 0.00] elo ELO +0.37 ± 1.12 (-0.75, +1.49) 95% GAMES 105,178 (24.6% 25,923W 50.8% 53,445D 24.5% 25,810L) PENTA [735⁻² 11,966⁻¹ 27,065⁰ 12,097⁺¹ 726⁺²]	Optimize StaticExchangeEvaluation Measured bench speedup: +0.233% No functional change.
Fury	Integral	main	diff	10.0+0.10 *	LLR +2.93 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO +56.35 ± 8.81 (+47.62, +65.16) 95% GAMES 1,474 (32.7% 482W 50.7% 747D 16.6% 245L) PENTA [1⁻² 75⁻¹ 359⁰ 290⁺¹ 12⁺²]	Save draw score to correction history if there's an upcoming repetition (#241)
Fury	Integral	ti_tweaks	diff	N=25000	ELO +39.42 ± 6.22 (+33.23, +45.64) 95% GAMES 5,116 of 5,000 (36.0% 1,840W 39.4% 2,014D 24.7% 1,262L) PENTA [71⁻² 439⁻¹ 1,072⁰ 793⁺¹ 183⁺²]
Fury	Integral	ti_tweaks	diff	10.0+0.10	LLR +0.10 (-2.25, +2.89) bounds [0.00, 2.50] elo ELO +1.57 ± 6.74 (-5.17, +8.32) 95% GAMES 2,428 (25.5% 620W 49.4% 1,199D 25.1% 609L) PENTA [5⁻² 260⁻¹ 676⁰ 265⁺¹ 8⁺²]

1 2 3 4 5 220 221 222