J.J. Watt
Getty Images
By The Numbers

NFL Draft Analytics for DEF

Updated On: February 19, 2019, 1:48 am ET

In this column, we are going to find which college and NFL Combine statistics have been correlated to early NFL success, which I’m measuring using Pro Football Reference’s Approximate Value (AV) statistic. However, I’m only looking at the per game AV (AV/G) in NFL seasons two through four because rookie players often don’t perform and teams see the most return on their picks in the latter portion of their rookie contracts.

 

If you aren’t familiar with correlations here is a brief overview: The chart below will look complicated, but it’s really not. The left side of the chart is the college stats, and the top of the chart is the NFL stats. The darker the color (either green or red) the more correlated the two stats are. If there isn’t a lot of color in the cell, then the two stats have little to zero relationship. The numbers shown in the chart are correlation coefficients. If you want R-squared, just square the number. I chose correlation coefficients to show positive and negative correlations with color. Here’s a short video if you’re confused.

 

EDGE (Under 270 pounds)

 

 

These are your speed rushers, and it shows in the NFL Combine stats. The most correlated stats with AV/G are the cone drill, speed score, 40-yard dash, and agility score. And all of them are measurements of speed. If your trying to predict Pro Bowls, then speed score is the top metric.

 

Of the on-field college metrics, there isn’t a single statistic that has been correlated to overall early NFL success. Instead, we need to go NFL statistic by NFL statistic. For example, let’s look at NFL sacks per game, which is the primary statistic NFL teams are looking for when they draft an EDGE rusher. There are two college stats that have historically been correlated with NFL sacks per game -- tackles for loss and tackles for loss per game.

 

 

That means that by simply weighting college tackles for loss over college sacks, you’ll have an advantage over other draft analysts. In fact, 15 of the 17 EDGE rushers who averaged at least 0.50 sacks per game in NFL seasons two through four averaged at least 1.0 tackles for loss in his final collegiate season.

 

EDGE (Over 270 pounds)

 

 

A lot of the same things that we just went over can be extended to the 270-plus pound EDGE rushers, but there are differences. Of the NFL Combine metrics, the cone drill, broad jump, and agility score are still important, but speed score and the 40-yard dash aren’t as correlated to early NFL success. This makes sense because these heavy EDGE rushers are more reliant on power and hand usage than the speed rushers.

 

 

However, speed does still matter for the big boys, especially when it comes to sacks per game. The six heavy EDGE rushers who averaged more than 0.50 sacks per game in NFL seasons two through four all had a cone drill under 7.26 seconds. That’s the current threshold for elite pass rushers who step on the scale at over 270 pounds.

 

Of the on-field college statistics, there’s one really large surprise -- college pass deflections appear to be important, but that’s solely because of J.J. Watt. If I remove him from the sample, there’s zero correlation. Instead, I’ll be focusing on tackles for loss again. But it’s just less important for the heavier EDGE rushers compared to the sub-270 pounders.

Before you look at all the data below, remember that the sample is looking at NFL seasons two through four. This isn’t analyzing entire NFL careers. This is just designed to find the best running backs for their rookie contracts. It’s up to you if you want to extrapolate the data to either rookie seasons or seasons after the rookie contract. I’m guessing the data would be pretty similar, but the exact correlations would be slightly different.

 

Defensive Tackles

 

 

Defensive tackle is a tough position to analyze because it’s a diverse position with some in to rush the passer and others to stop the run. However, there’s enough evidence in the charts to narrow our focus to just a few things, both on-field and at the NFL Combine.

 

 

College tackles for loss was a better predictor of NFL sacks than college sacks for EDGE rushers, and the same can be said for defensive tackle. Tackles for loss also is the best single pre-draft metric that we have for overall early NFL success. If you want to create a minimum threshold, then I recommend using 0.75 tackles for loss per game.

 

Like you’ll see with linebackers, solo tackles are far more important than assisted tackles. In fact, there’s a moderate, positive correlation to early NFL success with solo tackles and a negative correlation with assisted tackles. If you want to look at production, don’t even bother looking at assisted tackles or total tackles. Just highlight the tackles for loss and solo tackles, then move on.

 

 

If you are looking for sacks, look no further than the broad jump. It’s my favorite measure of explosion, and that’s what’s needed to sack quarterbacks in the NFL, along with technique. As you can see, defensive tackle sack artists all had a broad jump of at least 100 inches. Guess where Aaron Donald is on this chart? You’ll probably get it right.

 

Linebackers

 

 

If you want to play linebacker in the NFL, you better be fast -- specifically straight-line speed -- and you better have a lot of tackles in college -- specifically solo tackles. Everything else is basically worthless. That includes the bench press, side-to-side combine drills, and assisted tackles.

 

 

Elite NFL linebackers have historically made a lot of solo tackles in college. In fact, all six linebackers who have averaged 0.75 AV/G in NFL seasons two through four made at least 4.0 solo tackles per game in their final college season. Linebackers that made the 6.0 solo tackles per game threshold have had a much better hit rate as well.

 

 

You rarely see a trend as strong as this in NFL Draft analytics, unless you’re building models -- you’ll see the results of mine in future columns. If a linebacker doesn’t run the 40-yard dash in under 4.79 seconds, it’s really bad news. Meanwhile, those linebackers that run under 4.6 seconds have had a pretty high hit rate for NFL Draft standards. Speed kills at this position, and it’s arguably only going to get more important with NFL offenses changing.

Before you look at all the data below, remember that the sample is looking at NFL seasons two through four. This isn’t analyzing entire NFL careers. This is just designed to find the best running backs for their rookie contracts. It’s up to you if you want to extrapolate the data to either rookie seasons or seasons after the rookie contract. I’m guessing the data would be pretty similar, but the exact correlations would be slightly different.

 

Corners

Before we look at the correlations, I wanted to note that Approximate Value is nowhere near the best measure for ranking cornerbacks. However, it’s not completely worthless. Here are the top-5 corners in the sample: Richard Sherman, Darrelle Revis, Patrick Peterson, Jalen Ramsey, and Courtland Finnegan.

 

 

 

The NFL Combine has historically been more important for corners than on-field counting stats production. If you’re spending any time looking at any of the tackle stats, your process is flawed. If you’re looking at interceptions or pass deflections, then your process is flawed, but it’s less flawed then looking at the tackles stats. Instead, we need to be heavily using Sports Info Solutions and Pro Football Focus coverage stats. I don’t even need historical data to know that those metrics would outperform any of these counting stats.

 

 

Of the NFL Combine measures, the speed score has been the most correlated to early NFL success historically. However, it still isn’t very strong. My understanding of the position is that it’s so based around instincts -- both reading the receiver and, at times, the quarterback -- and it’s harder to quantify that at the NFL Combine or through counting stats. If forced to use a minimum to find elite corners, I’d set it at 90 since all 10 corners with at least a 0.50 AV/G were above that mark. Unfortunately, that will only eliminate a handful of corners each season.

 

Safeties (Under 210 pounds)

 

 

I didn’t have the free or strong titles in my database, so I opted to separate the two positions by weight. Of course, this is a flawed step, but it’s an improvement over just lumping all safeties into one so deal with it. … For the smaller safeties, a lot of things are somewhat important but no single statistic was very strong. When it comes to the on-field counting stats, total tackles and pass deflections are arguably the two best, but once again, it’s a weak correlation.

 

 

While the trend isn’t strong, we can use a minimum threshold to find out what the elite safeties look like. 14 of the 18 safeties (78%) who had at least a 0.40 AV/G in NFL seasons two through four ran under 4.52 second during the 40-yard dash.

 

Safeties (Over 210 pounds)

 

 

The heavier safeties have slightly different correlation coefficients than the safeties who weighed under 210 pounds. College solo, assisted, and total tackles had a negative relationship with early NFL success for the smaller safeties, but it’s a zero or weakly positive relationship for the big boys. On the flip side, the college interceptions and pass deflection stats have a negative relationship with early NFL success when it was slightly positive for the smaller safeties. These explain the differences in roles between heavy (strong) and skinny (free) safeties.


Source URL: https://www.nbcsportsedge.com/article/numbers/nfl-draft-analytics-def