rung rung-seed   knob recency_half_life_days   probes 26 (13 current / 13 history)
baseline at the shipped default 0.0: 21 / 26 right

   value    right  fixed  broken      cur     hist   verdict
     0.0   21 / 26      0       0   12 / 13    9 / 13   baseline
   730.0   14 / 26      1       8   13 / 13    1 / 13   BREAKS 8: p03 p04 p07 p18 p20 p22 p23 p24
   365.0   13 / 26      1       9   13 / 13    0 / 13   BREAKS 9: p03 p04 p07 p08 p18 p20 p22 p23 p24
   180.0   13 / 26      1       9   13 / 13    0 / 13   BREAKS 9: p03 p04 p07 p08 p18 p20 p22 p23 p24
    90.0   13 / 26      1       9   13 / 13    0 / 13   BREAKS 9: p03 p04 p07 p08 p18 p20 p22 p23 p24
    30.0   13 / 26      1       9   13 / 13    0 / 13   BREAKS 9: p03 p04 p07 p08 p18 p20 p22 p23 p24

HEADROOM, observed, from the rows (ADR-RS decision 22b):
  improvement — probes not right at the default: 5 / 26
  regression  — probes right at the default:     21 / 26

🔴 NO single global value clears the 0-broken bar. Every value that fixes a probe breaks another, which is the pre-registered answer NO — and it is a result, not a failed run.

per-probe rows (156) -> work/regression/2026-09-12-priors-and-tables/evidence/sweep-rung-seed-recency_half_life_days.jsonl
