NFL

Can We Predict the Success of NFL Pass-Rushers From College Data?

Cleveland Browns' rookie defensive end Emmanuel Ogbah should continue his sack artistry in the NFL. At least, that's what math says.

We’re all looking for an edge on our opponents in fantasy football leagues. Every incremental advantage we gain stacks up into an even greater total leg up, so it’s hard to discount even the smallest gain that is made; it might end up being the difference between success and mediocrity, or mediocrity and utter and abject failure.

This is the conundrum the Cleveland Browns faced coming into the 2016 NFL Draft: "How do we gain an edge?"

Many have heard that the Browns took a new-age approach this year, hiring former baseball executive Paul DePodesta to head their scouting department. DePodesta, of course, was one of the instrumental characters in instituting the “Moneyball” philosophy into Major League Baseball, which essentially theorized that if professional sports were all just a market, and if every market had inefficiencies -- undervalued assets -- they could find the inefficiencies and exploit them in their sport.

We try to do the same thing here at numberFire. We find you the “misfit toy” assets and find out what makes them tick so that you can capitalize on your market inefficiencies. Today, we’re going to literally look at how to get an edge -- an edge rusher, that is -- and figure out if we can predict NFL pass-rushing success from their college data.

Shaving Points

The question is simple: How do we project edge rushers and the sacks they contribute in the NFL? Is there a way for us to see them coming before they hit the pros?

In a previous article on defensive linemen and whether their athletic measurements can predict success, we found that sacks are the most-predicted measure of success by various athletic measurables. I began to wonder if we could predict sacks even better.

I collected data for defensive linemen drafted from the 2009 to 2013 NFL Draft classes -- giving us a large enough sample size of at least three professional seasons -- and cross-checked their college sack production against their professional careers’. Could we see the best pass-rushers in the league coming before they blossomed?

To prove a possible relationship between college sacks and professional sacks, I ran a statistical correlation. The r-value correlation is measured on a scale of -1 to 1. The closer the value is to 1 or -1, the more direct of a relationship there is between two variables; the closer it is to 0, the more random the association is. The table below shows the correlations between College Sacks and NFL Sacks, and between College Sacks per Year and NFL Sacks per Year. Is there a relationship?

Correlation 2009-13
College vs. NFL Sacks 0.27
College vs. NFL Sacks/Year 0.42


There's at least an indication of a weak relationship between both sets of variables, but it's striking that the College Sacks per Year and NFL Sacks per Year have a relatively decent r-value. Anything from 0.40 to 0.60 can be considered a moderate correlation, which means there's a chance that the two variables are linked.

I wanted to drill a little deeper into this data. What if we look only at the players who have played the most games in the NFL? Things such as injury, bad landing spots, and a number of other issues can result in a lack of simply enough playing time in the pros to rack up good sack numbers. What if we controlled for that?

If we look solely at the players with an average of at least 14 games played per season in the NFL, the correlation between College Sacks per Year and NFL Sacks per Year leaps up to 0.53; even better than the dataset as a whole. The average College Sacks per Year of these players was 3.68; more than this is a good indicator for a player.

This, to me, confirms that there is some correlation between the ability to get sacks in college and the ability to get sacks in the NFL. It may not be an ironclad causation; there are always the players like Tyson Alualu, Jared Crick, and Aaron Maybin, who have great college careers and then do much less in the NFL, but for each of those, there's a Robert Quinn, a Carlos Dunlap, and an Everson Griffen that show us that we should consider this a real skill set.

Pass-Rush Prognostication

So, what do we do with this information? If College Sacks per Year are a good indicator of future pass-rush success, and our previous study on defensive linemen showed that those selected higher in the NFL Draft are most likely to have NFL success, we can apply this information to the 2016 rookie draft class.

The table below shows the 2016 rookies most likely to be impressive sack artists in their careers, with College Sacks Per Year marks higher than the top NFL average (3.68) and in the top two rounds of draft selections.

Name Team College Sack/Yr Drafted
Emmanuel Ogbah CLE 8.83 R2, P1
Joey Bosa SD 8.67 R1, P3
Shaq Lawson BUF 6.67 R1, P19
Noah Spence TB 6.67 R2, P9
Kamalei Correa BAL 6.33 R2, P11
DeForest Buckner SF 4.50 R1, P7
Sheldon Rankins NO 4.50 R1, P12
Kevin Dodd TEN 4.17 R2, P2
Adam Gotsis DEN 3.83 R2, P32


It’s interesting to note that the Browns’ 2016 second-round selection, Emmanuel Ogbah, had more average sacks than the first defensive player selected this year, Joey Bosa. If I had extended this list to include the third-round selections, the Browns would have been just one of two teams with players to make this list. Their fourth-round pick, Wisconsin outside linebacker Joe Schobert, also would have made the list.

Perhaps this is what Moneyball football has taught the Browns’ front office: production begets production. And that’s simple enough math that you don’t even need a calculator to do it.