A few years back, the fine folks at Football Outsiders introduced us to SackSEER, a regression-based formula developed to predict the NFL success of edge rushers selected in the NFL Draft. Last year, with a bit of experience under their belt, they upgraded their methodology and introduced SackSEER 2.0
The new formula is based on four metrics: an explosion index combining forty-yard dash, vertical jump and broad jump results; SRAM or adjusted sacks per game in college (with some playing time adjustments); passes defensed per college game played and how many eligible games worth of NCAA football the player missed.
Yesterday, Football Outsiders published their SackSEER numbers for college edge rushers in the 2013 NFL Draft class. SackSEER is not the be-all and end-all of statistical analysis, and FO themselves have argued that it is more accurate at identifying busts than it is at singling out potential stars, but it is definitely worth a detailed look. Which is exactly what we'll do below.
If you're not familiar with SackSEER, here's a brief outline of how it works: Using the four metrics outlined above, the SackSEER formula projects each prospect's total sacks through five NFL seasons. Although there are always outliers in the individual projections, when cumulating all the individual numbers, FO have found that the formula projects sack production about three times more accurately than simply going by a player's draft position within the first two rounds.
The model is not without its detractors, and the initial model famously missed on the Giants' Jason Pierre-Paul, who met his five-year sack projection in his first year in the league. It also overrated the Colts' Jerry Hughes, who was touted as the best pass rushing prospect of the 2010 class but has only five sacks in the three years since. Last year, the model completely missed out on Bruce Irvin, who had eight sacks for the Seahawks. FO have since improved their model further, and have added the three-cone drill to the metric along with some other tweaks.
Harping on a few high-profile misses is always easier than looking at the overall accuracy of the model. Applying the model to edge rushers drafted into the NFL since 1999 yields more accurate predictions than misses. So don't discard the model just because of some high profile misses. For the most part, the model is fairly accurate.
Note that the model applies only to 4-3 defensive ends and 3-4 outside linebackers. Here's how the top eight edge rushers of this year's draft class stacked up:
|Damontre Moore||Texas A&M||1-2||23.5|
The folks at FO have calculated the SackSEER values for all combine participants, including those of Cowboys mock draft favorites David Bass, Devin Taylor and Tank Carradine, all of whom fail to make the top eight list above.
The model suggests that players like Dion Jordan and Jarvis Jones who are regularly mocked at the very top of the draft, may end up being overdrafted. At the same time, players like Barkevious Mingo and Bjoern Werner in particular have frequently fallen to the Cowboys in mock drafts, and the model suggests either would be pretty good value at #18. If you're in the Cowboys front office and looking to improve the Cowboys' pass rush, what do you do?
Do you go after an edge rusher if one of the premier guys falls to you, or do you go after a an interior pass rusher? After all, interior pass rushers are becoming more and more important, as teams are increasingly adapting to the more traditional outside edge rushers with short drops and quick throws. If all a QB has to do to avoid the pass rush is step up in the pocket, you've got a pass rushing problem. One that could leave you with another 8-8 record.
Dane Brugler of NFLDraftScout and CBSSports arrived at a similar, if likely unpopular, assessment:
But the ideal scenario for the Cowboys in the first round, besides a trade down to pick up more draft picks, might be if one of the top-three defensive tackles falls to them. Florida's Sharrif Floyd, Utah's Star Lotulelei and Missouri's Sheldon Richardson are all good fits in the Cowboys new four-man front under defensive coordinator Monte Kiffin. While all three could go top-10, there is a chance one of them, likely Richardson, could fall to the 18th pick. And if so, that makes the pick a fairly easy decision.
Assuming Cowboys fan favorites Jonathan Cooper or Chance Warmack are gone, would you be more likely to draft one of the premier edge rushers listed in the table, or would you target one of the top defensive tackles?