Project the player.
Expose the risk.

NFSMELL tested 70 qualifying quarterbacks from the 2008-2022 draft classes with chronological holdouts. Adding hand size slightly improved EPA/play error and fumble-rate ranking, but worsened CPOE and turnover-rate error.

The bottom-15% tag remains visible scouting context and adds zero points to QB OVR, team strength, or a betting pick. The fumble interaction is worth monitoring in poor weather; the broad claim that small hands create winning quarterbacks is not supported.

Role-specific college production, age-adjusted breakout, competition, validated scheme fit, preseason usage, snap share, and verified health must be joined with strict timestamps. Current staff identity is available, but a coach's historical tendency cannot become a numeric rookie modifier until the feature wins a cutoff-safe test. Film-only traits are not claimed unless a licensed or independently charted input actually exists.