Of the served 2026 board's availability miss, how much sat on men whose Out/Doubtful designation for the TARGET week was in inj_snapshot before the served generation was built, after it but before kickoff, only after kickoff, or never -- and would an on-time capture have reached the board?

Owner lane: serving

Question

Of the served 2026 board's availability miss, how much sat on men whose Out/Doubtful designation for the TARGET week was in inj_snapshot before the served generation was built, after it but before kickoff, only after kickoff, or never -- and would an on-time capture have reached the board?

Prediction

Frozen before any outcome was read (run/h245_protocol.json, sha256 2a5fd777...): week 1 (b) < 1% of total miss; week 2 (a) = 0 by construction and (b)+(b_post) >= 5% of total miss; S4 -- with the shipped call league_designations(con, 2026, 1) zero week-2/3 inj_snapshot designations reach the overlay through nfl_report, with week=W they do; (b)/(b_post) rows mostly carry proj >= 0.5; week 3 reported, not judged.

Finding and verdict

TESTED. The capture-timing measurement holds as registered (wk2 designated miss 114.5 = 5.8% of 1,987.2, all captured after the build and 108.1 of it after kickoff; wk1 0), but the row's causal counterfactual is refuted: the overlay reads week 1 by a literal, so an on-time capture reaches 1 of 50 week-2 and 2 of 70 week-3 designated men. Week-1 control is structurally weak (no designated non-player on the recoverable board).

Reasoning

Separates the three serving defects that H-245, H-227 and the week-3 non-rebuild each describe, on the object that actually served each kickoff.