Resolution: Supported
Caveats: Calibration verdict whose caveats carry the real story: the binding clause is met, but the mandated narrative bin series shows the width variable is dominated by editorial round-range QUANTIZATION - medians snap to 0/49/99/199 (precise dates, half-century, century, two-century ranges) - and the widest median (199) sits in the PRE-crisis 200-224 bin, which collects 'II-III century' ranges whose midpoints land there by construction. The midpoint assignment therefore systematically relocates the most coarsely dated (plausibly crisis-era) documents OUT of the crisis window, and the pooled medians (half-century pre vs century crisis) may reflect cataloguing convention as much as ancient dating-formula breakdown. The pre-registered mechanism (crisis -> editorial imprecision) is NOT cleanly established by this instrument; the honest follow-up is a quantization-aware re-analysis classifying range TYPES. The SUPPORTED verdict stands on the letter of the pre-registration; its evidential weight is explicitly discounted by these diagnostics.
In-house calibration-tier test (triage: adjacent — the qualitative third-century dating-imprecision observation is known historiography; the quantified instrument is not): in papyri.info catalogue metadata, editor-assigned date-range width for Egyptian documents assigned to the Crisis of the Third Century materially exceeds the width for the immediately preceding period — the record's own legibility as a crisis thermometer.
Resolution criteria: Assign each dated greco_roman_papyri record to a window by date midpoint ((date_start+date_end)/2). Compute median width (date_end-date_start) for midpoints in 235-284 CE vs midpoints in 150-234 CE. SUPPORTED if crisis median >= 1.5x pre-crisis median. KILLED if the ratio <= 1.1, or the crisis median is lower. Otherwise INCONCLUSIVE. Narrative (non-binding, must be reported): the full median-width series in 25-year bins 100-350 CE, to show whether the crisis window stands out against any secular trend.
Known priors disclosure: Registrant has seen only: 76,522 dated records exist; 36,522 have date_start in 100-350. No width statistic, median, or bin series has been seen. Known priors: the qualitative historiographical observation that 3rd-century documents resist precise dating (part of why this is calibration-tier), and the Wave-3a report's claim of no placeholder-range artifact in papyri.info dating. Standing caveat: width measures EDITORIAL assignability, which conflates ancient dating-formula practice, document survival, and modern editorial convention; a midpoint assignment smears wide-range records across windows.
Exactly as pre-registered: median (date_end-date_start) by midpoint window, plus the mandated 25-year narrative bin series 100-350. Script preserved at docs/generated/conjecture_extracts/batch2_resolutions_20260705_script.py.
Dataset: papyri.info greco_roman_papyri records with both date bounds (76,522 at computation; midpoint-assigned windows: 15,424 pre-crisis 150-234, 5,547 crisis 235-284). Read-only.
computed 2026-07-05