aggregation-calibration-shared-permutation
IN premise — summaries/2026/08/24/aristotelian-2026-sR-references-chunk-2.md
Created 2026-08-24T17:10:51+00:00
In aggregation-aware null calibration, a single permutation π must be applied across all layer pairs simultaneously; using independent permutations per layer destroys the dependence structure and invalidates the EVT correction
Summary
When calibrating the null distribution across multiple layers, the same random reshuffling must be applied to every layer pair at once. Treating each pair independently scrambles the relationships between layers, which silently breaks the extreme-value correction and makes the resulting calibration unreliable.