Explaining PFF's grading system
On every play, a PFF analyst will grade each player on a scale of -2 to +2 according to what he did on the play.
At one end of the scale you have a catastrophic game-ending interception or pick-six from a quarterback, and at the other a perfect deep bomb into a tight window in a critical game situation, with the middle of that scale being 0-graded, or ‘expected’ plays that are neither positive nor negative.
Each game is also graded by a second PFF analyst independent of the first, and those grades are compared by a third, Senior Analyst, who rules on any differences between the two. These grades are verified by the Pro Coach Network, a group of former and current NFL coaches with over 700 combined years of NFL coaching experience, to get them as accurate as they can be.
From there, the grades are normalized to better account for game situation; this ranges from where a player lined up to the dropback depth of the quarterback or the length of time he had the ball in his hand and everything in between. They are finally converted to a 0-100 scale and appear in our Player Grades Tool.
Season-level grades aren’t simply an average of every game-grade a player compiles over a season, but rather factor in the duration at which a player performed at that level. Achieving a grade of 90.0 in a game once is impressive, doing it (12) times in a row is more impressive.
It is entirely possible that a player will have a season grade higher than any individual single-game grade he achieved, because playing well for an extended period of time is harder to do than for a short period.Similarly, playing badly for a long time is a greater problem than playing badly once, so the grade can also be compounded negatively.
Each week, grades are subject to change while we run through our extensive review process including All-22 tape runs and coaching audit, so you may notice discrepancies among grades published in earlier articles compared with those in the Player Grades tool until grade lock each week.