First, an article on punters and now running backs – the two positions the analytics community hates most. I’m definitely not part of the “running backs don’t matter” crowd and with the Eagles truly committing to the run this past season, I wanted to take a deeper look at running backs. Here I am taking a look at what college traits translate to NFL success for RBs and who could be later round finds with the following profiled:
Why EPA is misleading
First, some comments on EPA as I am not using it here. Last week I was in an interesting Twitter exchange about what RB metrics may matter that Connor LaPlante (@cplant_) started. We know traditional metrics like yards-per-attempt aren’t useful and you must be careful using EPA, the go-to analytical metric for almost everything, for evaluating individual players. EPA measures the value a play generated vs. what was expected based on the field position, down, and distance. Given this, the game situation in which a running play is called has a huge effect on a running back’s opportunity.
To show this, below is the per-rush EPA broken down for all rushes, “bad rushes” (rushes on 1st-and-10), and “good rushes” (short yardage, less than 3 yards-to-go). I included the league average as well as stats for Jonathan Taylor (one of the top backs last year and one of the highest use short-yardage backs in the league at almost 18% of rushes), Miles Sanders, and Dalvin Cook (two above average RBs that have much lower short yardage usage).
| Running Back | EPA Overall | EPA 1st-and-10 | EPA Short Yardage (<3 yds to go) |
| Average for all RBs | -0.060 | -0.095 | 0.049 |
| Jonathan Taylor | 0.069 | -0.097 | 0.218 |
| Miles Sanders | -0.038 | -0.093 | 0.016 |
| Dalvin Cook | -0.027 | -0.099 | 0.054 |
First, rushes on 1st-and-10 have a depressed EPA because the chance of picking up more yards than a pass is really low. But short-yardage runs have a much higher EPA because so many of these runs either result in a touchdown if near the goal line or a new set of downs, both of which EPA values highly. A back could run for 4 in both scenarios and have vastly different EPAs.
While Jonathan Taylor had the 3rd best EPA per rush in the league, he looks no different than other RBs on less favorable rushing situations (1st and 10) – his high short-yardage usage skews his overall EPA stats. Still a good back, but you see how this makes EPA problematic for evaluating a RB individually and against other RBs.
EPA is not a bad metric, it is just better at aggregate (season or game level) and evaluating game situation / decisions. Looking at individual players with EPA, especially on smaller sample sizes, just requires digging in for needed context.
The metrics that do matter
The two metrics that have been shown to be (1) most stable from college to pro and (2) most accurate in isolating a running back’s performance are:
- Elusiveness – PFF’s rating based on yards generated after contact and the number of tackles avoided or broken
- Explosiveness – the percentage of rushes that gained 15 or more yards
Elusiveness
By focusing on what a running back generates on their own – blocks avoided and yards after a defender makes contact – elusiveness attempts to isolate a back from the blocking they receive. And it has been shown to be a stable metric from college to pros, meaning if a back was elusive in college, you can expect them to be elusive in the NFL as well.
While not perfect (there is no perfect metric here), elusiveness is very helpful. The below shows RBs drafted over the past 5 years with their college elusiveness rating on the x-axis and their NFL rushing grade on the y-axis. While there are exceptions (Pollard and Jones have high ratings despite average elusiveness scores), a back’s elusive rating absolutely correlates with NFL success.
Explosiveness
I’ve written before about the value of explosive plays – a single explosive play triples the chance of a drive ending in a score (Gannon’s defense, while infuriating at times to watch, is rightly focused on limiting explosive plays). And explosiveness and elusiveness are obviously related – a back that can avoid or break tackles and gain more yards after contact has a better chance of creating an explosive rush.
Below shows the same drafted RBs over the past five years with their college explosive rush rate vs. their NFL rushing grade – again, their college performance is predictive of NFL success, although a bit less strongly correlated than elusiveness (explosive rush rate has an R2 of 0.24 vs a better 0.33 for elusiveness).
A quick look at the 2021 draft class
This is my ongoing “what were the Steelers thinking?” part of the post. While I am not a run game hater as I said, I think it is insanity to draft backs high or give them a second contract. Here is a look at the 2021 drafted running backs with their college elusiveness rating on the y-axis and college explosive rush rate on the x-axis.
When you look at the 2021 backs, the top four elusive backs also happen to be the top 4 run graded and 4 of the top 5 top EPA/rush backs in their rookie season. Kenny Gainwell was actually the highest EPA/rush rookie because he had 5 rushing touchdowns, one from the 1-yard line and the other four from 7 yards out or more, all of which are very highly valued and pushed his aggregate EPA up.
Again, any single number does not tell the whole story on a back, but it is interesting to look at where RBs were drafted vs. the rookie season performance. You do not need to get a starting-caliber back high which is a consistent story each year. Javonte Williams (pick 2-35), Michael Carter (R4-108), Rhamondre Stevenson (R4-121), Kenny Gainwell (R5-151), and Khalil Herbert (R6-218) all stand out among the rookies but are all over the draft board. I’ve posted data in the past that shows RBs are one of the best value positions (along with IOL, IDL, and kickers) to take in rounds 5-7.
Which day 3 backs in the 2022 class could be interesting?
The Eagles re-discovered the run game this past season, ending the season with the 3rd highest EPA per rush and top rushing success rate in the league. Hurts was a large part of this, but Sanders, Howard, and Scott all had above 70.0 run grades running behind the number two rated Run Block Win Rate (RBWR) offensive line.
Looking forward, I am assuming they will continue to prioritize the run and while Sanders is still the number 1 back, he is in the last year of his rookie deal and only played 12 games each of the past two seasons. There is a price that makes sense for everybody, but assuming what he will likely get, it makes no sense to re-sign him. Scott and Howard may also both be gone, and while Gainwell was drafted last year, the Eagles should look for depth this April. They won’t (shouldn’t) take a back high in this draft but like last year, finding a potential gem like Gainwell late in the draft, particularly a larger back, to prepare for the post Sanders era is smart.
Below shows the same graph I used above but for the 2022 running back draft class showing college elusiveness and explosive rush rate.
I am not going to spend time on the top backs – Breece Hall, Isaiah Spiller, Kenneth Walker, Kyren Williams – that will most likely go early but instead look at projected day 3 RBs that could shine.
Hitting on day 3 is tough, no matter the position, but since the Eagles are unlikely (or not) going to use a top pick on a back, I’ll highlight guys that have a better than average chance.
Dameon Pierce – Florida
| Measurables: | 5’10”, 215 lbs |
| ADP: | R4-131 |
| Elusive Rating: | 138.0 (4th) |
| Explosive Rush Rate: | 9.0% (9th) |
The top graded back in the NCAA, Dameon is a physical back that runs hard and caused missed tackles at the 2nd highest rate in college this year. One that is viewed as a power back but he is much more than that – he had the 9th highest explosive rush rate, 4th best elusive rush rate, and was the fastest player on the field at the Senior Bowl, hitting 20.66 mph (source @JimNagy_SB tweet here). He went for 69 yards on 9 rushes against Georgia and is an effective blocker. One of my favorites in this list and somebody that would be a great fit on the Eagles.
Tyler Allgeier – BYU
| Measurables: | 5’11”, 220 lbs |
| ADP: | R4-134 |
| Elusive Rating: | 111.4 (8th) |
| Explosive Rush Rate: | 7.3% (18th) |
Another big back in the class. Allgeier won’t quite give the same explosive rush rate as others on the list but he makes up for it in breaking tackles.
Zonovan Knight – NC State
| Measurables: | 5’11”, 210 lbs |
| ADP: | R6-180 |
| Elusive Rating: | 110.8 (9th of 29) |
| Explosive Rush Rate: | 7.9% (16th of 29) |
Zonovan is 3rd in the draft class, behind Walker and Spiller, in tackles avoided per attempt and is 2nd in the NCAA in yards per attempt when contacted behind the line of scrimmage. His fumble rate (7 fumbles in 3 years over 439 attempts) scares me away and his pass blocking needs to improve.
Jerrion Ealy – Ole Miss
| Measurables: | 5’8″, 190 lbs |
| ADP: | R6-197 |
| Elusive Rating: | 131.4 (5th) |
| Explosive Rush Rate: | 9.1% (8th) |
The smallest back in the list, Ealy is 4th in the draft class behind Zonovan Knight, Spiller, and Walker in tackles avoided per attempt and 7th in college in yards after contact per attempt. Like White above, Ealy needs to improve his pass-blocking but provides some kick return value.
ZaQuandre White – South Carolina
| Measurables: | 6’1″, 215 lbs |
| ADP: | R6-204 |
| Elusive Rating: | 153.6 (1st) |
| Explosive Rush Rate: | 11.2% (4th) |
White is another back I really like, especially where he is likely to be drafted. He only had 215 college rushing attempts but led the FBS in both yards after contact per attempt (4.61 yards) and yards per route run among running backs (2.17 yards) while playing in the SEC. He needs to greatly improve his pass blocking to get consistent time in the NFL.
Kennedy Brooks – Oklahoma
| Measurables: | 5’11”, 215 lbs |
| ADP: | R6-214 |
| Elusive Rating: | 118.9 (7th) |
| Explosive Rush Rate: | 12.7% (2nd) |
Another bigger back, I often see Kennedy projected as a short-yardage back but his traits say he is much more than that – as a comparison, his missed tackle rate generated is twice Jordan Howard‘s with the 2nd highest explosive rush rate among prospects. Brooks is my personal favorite on the list – he has the size the Eagles could use, had no fumbles over the past two seasons and only one in his college career. He showed up consistently all year, including 139 yards on 22 rushes and four that went 15 yards or more against Oklahoma State, the top college run defense last year in EPA allowed. He overlapped with Hurts at Oklahoma in 2019, which doesn’t matter but is interesting if nothing else.