Finding The Superior Athlete: Impact Running Backs In The 2016 NFL Draft

In two previous posts (on Defensive Ends and Defensive Tackles) we re-introduced you to a metric called SPARQ, which is a single number designed to summarize a player's athleticism. The number is calculated with a proprietary formula that incorporates player weight, bench press, broad jump, vertical jump, forty-yard dash, ten-yard split, short shuttle and 3-cone drill (details here).

In those two posts, we combined the SPARQ metric with a metric for the college production of defensive linemen called the 'Production Ratio' to see which draft prospects would emerge as the most productive AND most athletic.

Today, we're turning our attention to running backs. Unfortunately, we don't have a simple metric like the Production Ratio for running backs. Instead, as a measure for the college production of running backs, we'll use a metric introduced by Kevin Byrne of ColdHardFootballfacts in 2011 called the "Rusher Rating."

The Rusher Rating is calculated in a similar fashion to the better know passer rating, but factors in three efficiency metrics instead of four:

• Yards per attempt
• Touchdown percentage
• Fumble percentage

The underlying premise is that more often than not, good running backs achieve a high YPA, have a high TD percentage, and ideally don't fumble away the ball too much. Hard to argue with that logic, even if we've been conditioned all our lives to look at volume stats for running backs.

Using the CHFF formula, here's a summary of the Rusher Rating for 26 draft-eligible running backs with a draftable grade. The data is limited to the last two college seasons of each running back (click on the blue column headers to sort):

If you sort the table by Rusher Rating, you'll see Ezekiel Elliott and Derrick Henry near the top of the table, as you'd expect from two players who are expected to picked in the first and second round. You'll also see some other names at the top of the chart, players who've been highly productive in college for various reasons, but may not have garnered as much attention as some other players at the spot.

And while we have the Rusher Rating for all the backs in this draft, we don't yet have a SPARQ rating for all of them. Here's an overview of the SPARQ scores for this year's RB draft class, courtesy of Zach Whitman of 3sigmaathlete.com (click on the blue column headers to sort):

Unfortunately, we don't yet have SPARQ numbers for every back in this year's draft class. Notable omissions include Devontae Booker, Jordan Howard, Jonathan Williams, and Aaron Green.

Nevertheless, we can take the above data one step further and graphically visualize who the top running backs in this draft are (if you're going by college production and athletic potential):

For a detailed discussion of what each quadrant means, please refer to one of the earlier posts on the topic.

What stands out here is that there's a cluster of four A-quadrant players that are pretty tightly bunched, all with above average athleticism and above average production. Not a big surprise to see Elliott and Henry show up here, but C.J. Prosise and Kenneth Dixon show up very well in this approach.

Prosise of course is a relative newcomer at the position, only converting to RB in 2015 after never having played the position before. His 7.0 career YPA (6.6 in 2015) compares favorably to Elliott, and his athleticism is ranks him in the 64th percentile among his NFL peers. He has a fairly high fumble rate of 3% (five fumbles on 166 carried from 2014-15), and perhaps that's a result of his late transition to the position. It is a watchout though.

Dixon shows up as the third running back on some big boards, and his markers suggests that's a pretty good assessment. Dixon has drawn comparisons with Thomas Rawls, and given that the Cowboys were very interested in Dixon last year, he might be high on their list this year.

If your A-quadrant players are gone, and there's a good chance they'll all be gone after day two of the draft, you may want to look at prospects in the B quadrant. They all have above average athleticism, even if their college production was slightly below average, though there could be a perfectly acceptable reason for that. Teams just need to understand what's behind each individual's production number.

As you look into the C and D quadrants, the choices become increasingly difficult. Running backs who did well against college opposition will find that once they enter the NFL, their previously elite skill set is often par for the course at best on an NFL team. As a matter of principle, NFL players are bigger, faster, stronger, and more talented than college players, and if a running back doesn't have good athletic markers, his chances of success in the NFL dwindle rapidly.

The mandatory caveat: There are a multitude of factors that determine how well a prospect will do in the NFL. College production and athletic markers are just some of them, but at the very least, they provide some interesting input into the evaluation process.

Almost every draft has produced stand-out running backs from the later rounds. But if a team had known upfront that a guy was going to be a stand-out running back, he'd have been drafted much higher in the first place. As such, hitting on a running back late in the draft is often nothing more than luck; great if you're the team that made the pick, but no way you can plan on that.

This year's draft class, as measured by SPARQ, is a little short on elite-level athletes. Only three guys are in the top 70 percentile compared to their NFL peers (Daniel Lasco: 94th percentile, Derrick Henry: 82nd, Tyler Ervin, 81st), and there a few more backs with above average athleticism. But that doesn't mean a team like the Cowboys won't find a guy they like. It's just that that particular guy may not be waiting in every round.

By popular request, I'm adding some historic data for standout running backs. Unfortunately, there isn't a lot of historic SPARQ data around, so we'll have to make do with the following eight players, who should make us feel much more confident about the model and the data in this post.

Translated into our four-quadrant matrix, six of those eight players present themselves as A-quadrant players: