Composite skill θ for every GM in Buffalo Bills history, alongside the team's regular-season W-L and playoff result by year.
Years covered
1980–2025
first → last draft
GMs
9
tenures, newest first
Best θ
+0.70
Bill Polian
Worst θ
-0.92
Stew Barber
θ trajectory by GM
Each line is one GM's support estimate at successive season-end snapshots, color-coded by tenure. Lines carry forward past a GM's tenure: drafted players' careers keep playing out, so the model keeps revising its read of an old GM long after they're gone (a 7th-round flier turning into a Pro Bowler in year 6 still credits the GM who drafted him). Background bands mark each GM's actual tenure span. These paths show evidence accumulating over time; they are not the maturity-adjusted headline ranking and deliberately do not display a public interval.
Support snapshot per year is the point estimate from a full NUTS MCMC refit on picks-through-that-year, computed by R/06_incremental_theta.R. It shows evidence accumulation and is not a separate published ranking. The current headline uses the maturity-adjusted deep v30 composite through 2025.
Tenure attribution follows data/gm_tenures.csv. De-facto draft heads (HCs with personnel power, owner-GMs) are credited too. See each card's note line for the role.
Pre-1999 W-L records come from Wikipedia {year}_NFL_season standings; 1999+ from nflverse schedules. Pre-1999 playoff exit isn't yet detected per-team and shows "—".
Pre-1994 model coverage is included through the 1980 expansion. Older GMs still carry wider uncertainty because their evidence is farther from the modern snap-count era.
Era effects: HOF voting standards, Pro Bowl culture, and AV calibration shift across decades. Cross-era θ comparisons should be read with a grain of salt; same-team θ deltas are more reliable than league-wide ones.
