Dak Prescott

RE-EXPRESSED 22 Aug, same total, honest split. The 6 Aug row said '13.5 projected games at age 33; the availability spread is wider than the record supports' -- an argument about GAMES -- and then spent itself entirely on RATE, because his injury_games_mult is 1.0 so split_override had no damping to undo. That is the exact defect the `games` column exists to fix, and it inflated his rate to 19.59/g, above anything his record supports. Both factors are now stated. GAMES 14.0: he has played 42 of 51 across 2023-25 (17, 8, 17), which is 82% and 14.0 games, so the model's 13.5 was mildly low exactly as the original row argued. RATE 18.87/g: his own seasons at the model's QB decay of 0.80 (2025 18.93, 2024 15.82, 2023 20.70), rather than the 19.59 the rate-only override manufactured. 264.2 total, within 0.2 of what the row already delivered -- the answer barely moves, the CLAIM becomes checkable. No rank target: this is a claim about his two factors, not about where he should sit, so board_audit should not check him against one.

This is a view, not a correction — the model said something different and was overruled. See how it works.

Pace extrapolates the selected weeks over 17 games. It is what that stretch was worth, not a projection — no availability, matchup or regression adjustment is applied.