Slate Stats › NFL › Positions

Quarterbacks: A Change at the Position Moves Receivers by Nothing

The quarterback matters enormously to a team and surprisingly little to the fantasy value of the receivers around him. Measured against a control on 2,489 changed player-weeks, a change at the position is null on receptions, yards and targets alike.

A quarterback change is null for receivers, and the control is the whole result

Receptions +0.013 (t=0.28), receiving yards -0.305 (t=-0.49), targets +0.049 (t=0.82), across 2,489 changed player-weeks against 18,737 control. WITHOUT a control the same data says receivers gain +0.637 receptions (t=+2.84) on a quarterback change — which is almost entirely the control's own drift, since any 40%-snap player with four trailing games gains +0.223. The uncontrolled version is not merely bigger, it is significant.

A rookie quarterback is much worse and it still does not reach receptions

Rookie passers are 0.13 EPA per play worse than third-year ones, which is large. Against each receiver's own baseline, rookie quarterback minus veteran is -0.001 receptions for receivers, -0.017 for tight ends and +0.000 for backs. The mechanism is an offset: catch rate is much lower with a rookie (59.9% against 63.8%) while target share is slightly higher.

Pass volume is in secular decline

League-wide pass attempts per game have fallen from 36.05 in 2018 to 33.54 in 2025, about -0.34 a season. Any model fitted across seasons without a trend term must over-predict volume, and this one did until the term was added.

A backup quarterback favors the checkdown, barely

Net of control, when the starter differs from the season's usual one, backs gain +0.156 receptions (t=1.75), tight ends +0.065 and receivers -0.088. The direction matches the checkdown theory and nothing reaches significance: a backup quarterback is worth about a sixth of a reception.

Every figure above was measured against a control and carries its sample size and t-statistic. Verified 2026-09-29. A lift measured without a control is not merely larger here — it is often significant and wrong, which is why each one states what it was compared against.