List outcome pairings

List the tracked Outcomes pairings, oldest first, one page, each measured now from the stored daily numbers: how strongly the outcome moved after the signal moved. Weeks (Monday to Sunday) are rebuilt from daily counts, each series becomes its week-over-week change, and the outcome's change is set against the signal's change 0 to 8 weeks earlier; an outcome read from AI assistants is compared with its other channel, which carries the season and promotions both share. The lag is chosen on the earlier weeks and checked on the later ones, and the 95% range comes from a seeded moving-block bootstrap. Each pairing answers a label: collecting (fewer than 16 paired weeks), inconclusive (the range spans no link and a meaningful one), no_meaningful_link (the range lies inside ±0.3), suggestive, or consistent (the pinned pairing only: its range excludes zero and held on the later weeks); exploratory pairings are corrected together for false discoveries. It also answers the correlation, the percent change in the outcome per point (or per 1%) of the signal with its range, the weeks, the smallest correlation those weeks can detect, the signal's reliability, the weekly series, and one sentence. Every result is an association between numbers, never proof of cause.