Loading
Loading the next page.
Loading
MBStatsLoading
Loading
A living notebook by Riel St. Amand
The reproducible trail for the earned-dropbacks null. The decision-priced expectation ships in the public qb_altitude_week view, next to what each week actually produced.
Every cell re-runs in your browser against current data. Covers 1999 through 2025. Data refreshed 1 day ago.
We asked: what happens when you grade a quarterback's menu of decisions instead of his cooking?
The decision-priced expectation ships in the public qb_altitude_week view next to the de-noised EPA each week actually produced. The cell below runs their live association across every 2024 quarterback week of 15 or more dropbacks. A strong grade would track production; the article is about how weakly this one does, and who it crowned.
Rigor check
Decision EPA per dropback against Clean EPA per dropback
Pearson correlation, re-run in your browser on current data.
The wreckage and the upgrade are scripted: the laugh-test board that crowned 2024 Anthony Richardson, the sentinel checks at every blend weight, and the shrinkage sweep that carried reliability from .128 to .147 while validity rose from .505 to .518 all live in the article backtest block. The expectation column ships here precisely so anyone can extend the sweep.
Built on data from the nflverse project.