Deshaun Watson

Cut to QB32 at Matt's direction. Native QB18 on 15.38/g -- ABOVE every season he has played since 2021 and 20% above his own decay-weighted history of 12.85. His actual line since returning: 14.26/g on 6 games, 14.47 on 6, 10.96 on 7, and ZERO games in 2025 after the Achilles. A declining rate on a shrinking sample, and the model reads the thin sample as uncertainty and pulls him UP toward the positional prior, which is how a man who has not played in a year outprojects his own best recent season. Availability is not pricing it either: injury_games_mult is 1.0, so missing an entire season cost him nothing. Both markets have written him off -- no ADP at all and ECR QB32, which is last -- and this lands on ECR. 138.0 over the native 11.1 games is 12.43/g, just under his own weighted history, which is the most that a quarterback coming off two Achilles tears with a rookie behind him should carry. TARGET ADOPTED AT QB35 ON 3 SEP, not re-priced back to QB32. The gap is 8.4 points, which is material by the points test, and it is still not worth serving. He sits at vor -40.2 among draftable quarterbacks running -28 to -52, so QB32 and QB35 are the same thing in a one-QB league: nobody drafts either. And the drift runs the way the argument did -- Sleeper projects 102.06 against this 138.0, so re-pricing UP by 8.4 would move away from the one source that says the cut has not gone far enough. Adopted deliberately rather than silenced: ECR and consensus both say QB32 and are recorded here as disagreeing.

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

No 2025-season weekly record for Deshaun Watson — rookies and players who did not record a stat line have none.