Run T-0116 rejected

2026-09-22 · Evidence: SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.

Hypothesis

When improving, ProbCut verifies a capture at one depth shallower, using depth minus 4 minus 1 instead of depth minus 4. This makes the verification search cheaper in positions that are already improving.

Wording edited for publication from the German entry; the register carries the author’s own text.

Stages

Point estimate with the reported 95-percent half-width, per stage. The dashed line is zero. Estimates from a stopped sequential test are biased by the stopping rule; the interval is nominal.

Field	short stage
Verdict	rejected
Elo Strength difference to the opponent, estimated from the games, with a 95-percent half-width where the oracle reported one. Never an absolute rating.	−3.66 ± 7.57
nElo Normalised Elo: the Elo difference divided by the spread of the game results, so that the SPRT bounds mean the same at different time controls.	−5.64 ± 11.65
Games	3,416
Wins / draws / losses	891 / 1,598 / 927
pentanomial Games are played in pairs with swapped colours; the five counts are the pairs scoring 0, ½, 1, 1½ and 2 points.	`80 / 419 / 746 / 383 / 80`
LLR Log-likelihood ratio, the running evidence of an SPRT. It starts at 0; a stage passes at the upper bound and fails at the lower one.	−3.01
Bounds	−2.94 … +2.94
SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.	[0.0, 10.0]
Error rates α / β	0.05 / 0.05
Model	normalized
time control Base time plus increment per move in seconds, e.g. 8+0.08: eight seconds per game plus 0.08 seconds per move.	8+0.08
Book	A
Opponent	vs predecessor
Duration	25 min

Provenance

Date: 2026-09-22
Started: 2026-09-22 15:59:21 UTC
Duration: 26 min
Author: Agent X
Epoch: 5
Gatekeeper: passed
Bench nodes: 3,627