NFL wide receiver ratings

Average of four season numbers, each compared with other qualified players that season (0 is average): targets, share of the team's air yards, efficiency with the quarterback's quality taken out, and yards per game.

Each receiver is compared only with receivers this far into the season, so early scores swing a lot.

The four numbers are weighted equally, so a season with more passing overall does not lift everyone's score. Short seasons (seven games, say) are hidden from the default view.

Taking the quarterback out gives a good receiver on a bad passing team the credit a plain efficiency number would miss.

It describes the season; it is not a projection. It predicts next-game yards somewhat better than a simple recent average (about 6.5% smaller errors in testing), but not by enough to beat the sportsbooks' margin on player props, so it is not published as a betting number.

The weekly grid is the season itself; the score above is the season's verdict. It goes back to 2017. A receiver's week is EPA per target, split into what was earned before the catch and what he added after it. Measured across a change of team, this number barely travels with the player. It substantially reflects the offense he plays in.

Projected usage is the preseason prior, updated as games are played. Until mid-season it is the steadier read of who matters; the scores describe only the weeks played so far.

Style. Style is where a wide receiver sits on two axes: how deep his targets are on average (past the line of scrimmage) and how often he gets manufactured touches (screens, throws behind the line and carries). Each axis is cut at a fixed line (12.5 yards and 12%), which gives four boxes. It is a scale, not a type: players near a cut are marked close, and nobody is a pure example of a box. Style is separate from the score and says nothing about how good he is.