Receivers
Wide receivers and tight ends: volume, depth and value per target.
Volume vs. value
Each dot is a qualified receiver. Right = a bigger share of the team’s targets; up = more EPA per target. Dashed line: league average on all targets.
Leaderboard
Click a column to sort. Ranks (EPA per target) are within WR or TE among qualified players.
How these are computed
- Targets are pass attempts with an intended receiver, including plays wiped out by a penalty (for EPA); counting stats use plays that stand.
- Qualified: at least 3 targets per game the team played.
- Target and air-yards shares are of the team’s totals in the games the player played. WOPR = 1.5 × target share + 0.7 × air-yards share.
- Catch % over expected uses nflfastR’s completion probability; YAC over expected uses its expected YAC (both 2006+).
- Stabilised (the Numbers switch): each rate pulled toward what similar players usually do, by an amount set by how much one season of plays can tell apart (fitted on 1999–2017 by splitting seasons in half; on 2018–2025 it predicted the other half of a season 25–56% better than the raw rate). The range after each number is a 90% interval; after the rank, the range of ranks it could plausibly hold. The prior rises with playing time: starters are better than backups.
- All numbers are descriptive: they don’t adjust for teammates, opponents or play-calling (except the teams page’s adjusted columns).
- EPA is nflfastR’s expected points added. WPA is win probability added from our own model (the registered WP model behind the 4th-down tool): per play in WP points, totals in wins. A touchdown earns the score plus the expected try; tries, kickoffs and returns earn their own WPA. For 1999–2025 these are the production model’s in-sample values.
- ± is a game-clustered standard error.
- * Some of this season’s games have play-by-play errors in the source (for example, every 2001 Jaguars home game is scored 21–3), so our WPA totals skip those games; rates use the games that remain.