Pass Rusher Performance
A few weeks back, while working through some numbers for the Dolphins Cameron Wake, I tried to come up with a scoring metric to value pass rush productivity. I wanted to expand on that a little further, explain the metric and apply it to the league. Since this is primarily a Jets site I’ll try to use some examples from the team in explaining how this works. And of course, once again, thanks to Pro Football Focus and all the amazing work they do charting plays. This is my interpretation of their numbers and I’m sure there are plenty other interpretations one can make if they access their premium content. Highly recommended.
No Pressure vs Under Pressure Performance
When most people look at a pass rush they simply look at a sack as a measure of dominance. This is where the Wake article came from, because his sack numbers were down and people claimed he was losing his effectiveness. But beyond a sack a pass rusher often is put into a situation where he hurries the QB by getting a hand in his face or knocking him down to the ground as he throws the football. There is a clear effect that occurs by pressuring a QB, even if you do not get credited for a sack on the play. Going through 4 years worth of data for all QB’s who played in at least 25% of a teams snaps, and adjusting for spike plays which are give ups, we can determine the change in performance in a number of key categories caused by a pressure.
So there clearly there is a major benefit to a pressure. If we define success on a play as a completion, the odds of success drop by 30.9%. The threat of an interception increases by 60% and the amount of times a QB has to take off running nearly doubles. While in some cases, such as Mike Vick and Cam Newton a run may be a failure for the defense, in most cases the QB running is a clear success for the defense. Some QB’s handle pressure better than others, but the only player last season to see almost no decline was Jay Cutler who completed 58.9% of his passes under normal conditions and 56.2% under pressure. Jason Campbell saw the biggest drop in the NFL in 2011 when he fell from 74.6% to 33.3%.
Valuing the Pressure
With all the data in place we can now put some numbers into play. Based on 100 dropbacks here is what we would expect the average performance to be for an NFL QB under no pressure and under pressure.
Keeping those figures in mind we can now develop a scoring metric for a player based on the increased failures he creates based on rushing the passer. To illustrate we will use Aaron Maybin, who was the Jets most productive pass rusher in 2011.
Maybin saw action on 195 pass plays in 2011. He was credited with 18 pressures (16 hurries and 2 hits) and 6 sacks. That means Maybin was in on 171 regular pass plays where he created no pressure. Using the numbers above we can come up with the average expected performance for a QB, assuming Maybin didn’t pressure the QB at all, and compare that with what should occur when he pressured the QB.
This allows us to guesstimate just how much effect Maybin now has on the outcome of a football play. Without pressure the QB would be expected to fail 70.77 times in 195 dropbacks. Maybin caused the QB to fail 78.3 times. We say that Maybin has increased failures by 10.64% due to his rushing skills. He also increases the chance of an interception by 2.16% due to pressuring the QB. That was 14th in the NFL last season.
Grading all of the players in 2011 who participated in at least 25% pass rush snaps (there are 331 who qualify) we can see the average positional performance for pass rush plays.
No real big surprises here as the positions most often rushing the passer are the players most often causing the QB to play below average. But what if we just take pass rushes into account rather than passing plays? The results become very different.
What causes the large difference is the fact that the pressure caused in scenario one is, for the most part, talent related. In scenario two we move more into scheme related pressures. Secondary blitzes, for the most part, are highly effective because you are overloading the line with rushers and coming from angles where blockers may not see it. Teams aren’t greatly prepared for the LB blitz in a 43 defense. If Eric Smith rushed every play for the Jets teams would be prepared and he would no longer be more successful on a per play basis as a Tamba Hali. But that is the chess game that goes on every week and the job of the GM is to get the best pass rushers at that position to increase the odds of success.
So how did the Jets do?
The overall performance was 16th in the NFL with an average failure of 4.17% per qualified player. The poor categories were from the line where the Jets were below average across the board. The secondary was above average and the ILB rushes were exceptional, 2nd in the NFL to the Colts, with a 3.57% average grade. That is probably high praise for Ryan’s defense. The Jets rankings were as follows based on total pass snaps:
If you adjust to see just how effective players were only when they actually rushed the passer the big changes were that Strickland and Wilson were two of the worst corner rushers in the league (number 11 and 12 respectively), while Scott was 11th and Harris 23rd at ILB. Everyone else was about the same. A big reason is that the Jets sometimes lose the element of surprise since they rush certain players more often (Smith for example rushed more than anyone at his position other than Roman Harper) from the non-rush spots so teams are prepared for it.
What is interesting is that two of the Jets better situational rushers are now gone (Westerman and Pitoitua) and Scott is likely not going to be rushing as much. They are going to need rookie Quinton Coples and 2nd year pros Muhammad Wilkerson and Kenrick Ellis to step up this season, especially if they transition to more 43 sets. Laron Landry was the best ranked rusher at safety last season with a 3.2% so he may provide a boost there, though he had a limited set of snaps. Yeremiah Bell was more effective on a per rush basis than Smith, and both did have a good number of chances so he could be a small upgrade in that category as well.
So who are the top guys in the NFL?
Here are the top 30 players in the NFL for 2011 based on increased failures. I also included their interception increases for reference. You'll notice two Eagles right near the top and they were far and away the best pressure team in the NFL. If you have any questions or comments on it feel free to post or email me.