by Nick Jaroszewicz
With running backs like Marshawn Lynch, Frank Gore, BenJarvis Green-Ellis showing that feature backs are back in style, and Chris Johnson finally coming around, many are saying that the pendulum is swinging back to the running game. But are teams running the ball more often now than before?
Turns out, looking at the distribution of pass to run percentages of the last 5 years in the NFL, there is not even the slightest sign of a pendulum shift back to running the football. In fact, we could even make the argument that this season is amongst the most pass happy the league has seen. In 58.3% of offensive plays this season, a quarterback has dropped back to pass the ball. That is the single largest amount of passes, percentage-wise, of the last 5 years, up from the 2011 previous high of 57.1%.
Furthermore, when we break it down on a week-to-week basis, we see a general trend upward, at the slow but sure rate of .03 percentage points a week.
While the NFL may be hungry for a story line of the return of feature backs, it seems that they will continue to air out the football for a long time to come. There has been nothing in the last 6 weeks of the season to say otherwise.
All statistics from pro-football-reference.com
In each season there seems to be a general trend of passing less and less as the season goes on. So does the in season data support this? Essentially, if you drew 4 separate lines of best fit for each season, would they each slope downward? If there is anything to this I think it can be explained by colder weather, but not sure what else.
Would be interesting to see this team by team, perhaps there are a handful of teams that take a heavy passing approach (i.e. 75 pass 25 run) but on the opposite site side most teams are evenly split or lean running. If the passing teams keep increasing there passing rate year by year, the trend would shift upwards without any change in overall league mentality. Also, I’d imagine that every year there could be more teams of the first variety because of the lack of feature backs and therefore skewing the trend upwards.
Also would be interesting to see the trend remains if we remove screen or backwards passes. A lot of this may have to do with the emergence of the receiving RBs like Sproles/Bush or backfield WR like Harvin
Out of curiosity, as it doesn’t really affect the premise of your piece, is that week-to-week slope statistically significant?
Good idea.
Including a fixed effect for week of the season (or some within season time effect, like month), along with increasing the number of years in the study, might make for a more interesting model.