Alternate title: C'mon Bobble, let's have an argument.
http://www.espn.com/nfl/story/_/id/1...drian-peterson
EPA: expected points addedIn 2015, the spread among the 32 teams was 2.3 EPA when passing, measured in terms of standard deviation. When running, the spread among the 32 teams was 0.7 EPA, less than a third of the impact of passing. In terms of WPA, the difference was only slightly less stark. The standard deviation among teams for passing was 1.1 WPA, and for running it was 0.4 WPA. In other words, a team should greatly prefer being a top passing team than a top running team by a factor of 3 to 1. Another way to look at this disparity is that being the most dominant running offense in the league would equate to being a modestly above-average passing offense.
WPA: win probability added
EPA is our best measure of overall scoring productivity, measuring an offense's ability to move the ball, limit turnovers, and even suppress opponents' opportunities to score. WPA does the same thing, but also considers game situation factors like score, time, and timeouts remaining.