My colleague and friend Denny Carter has gone on record saying running backs aren’t people or something like that. I’m not entirely sure. I didn’t read the article. From what I can gather via the headline, we want to be drafting running backs who get the ball a lot, talent be damned. In this article, I’ll argue Denny is right, but only sometimes, and the cases when talent plays an outsized role are extremely important. We’ll get to the part with some player takes eventually too.
A Thing I Recently Read
While endlessly consuming content, I came across a Substack post from Ben Gretch in which he paraphrased fantasy football living legend Shawn Siegele with the line, “The elite RB seasons almost require 3-4 FPOE per game from the RBs.”
FPOE stands for fantasy points over expectation. RotoViz defines the expectation part of their metric like this:
Based on the average Fantasy Point value of plays in similar down/distance/field position instances.
A first-and-10 carry at an opponent’s one-yard line is going to have an extremely high fantasy point expectation while a third-and-two dive from a running back’s own 12-yard line is virtually useless for fantasy purposes. If Derrick Henry breaks loose on that third-and-two for an 88-yard score, he massively outperformed the average expected fantasy points of that play. You can do the same calculations for targets or pass attempts and RotoViz has done just that. Total FPOE per game is a mix of a volume and efficiency stat. Because it’s cumulative within a game, a player needs to get a lot of touches to rack up FPOE. However, they have to exceed the expectation of that volume, which is a measure of efficiency. Devin McIntyre has a good explanation of how backs get over their expectation.
If you consider that if you have HOF-level efficiency and get 20 touches per game, you are only generating 0.1 FPOE per carry, or 2 FPOE per game.
— devinmcintyre.bsky.social (@devinmci) July 31, 2023
But TD expectation drops to basically 0 FP outside the 5 yard line or whatever. Those long TDs are where you can stack FPOE.
Finding Elite Seasons
Digging into the veracity of Siegele’s claim…it checks out to a stunning degree. Over the past decade, 38 running backs have exceeded 18 half-PPR points per game on a sample of at least 10 games within a season. The average fantasy points over expected per game for those backs was 3.3. That number rose to 3.9 for all running backs who topped 20 points per game. Even more incredible is just how mandatory topping that expectation was in order to achieve a strong fantasy season. Of those 38 backs, only one rusher finished the season with a negative FPOE. Dropping the threshold to 16 points increases the total number of players to 57 and only adds two more negative FPOE runners. All three players who fell below expectation did so by less than a point per game.
This tells us a few things. The first is very obvious. An NFL running back can only get the ball so many times in a season. Looking at the best fantasy seasons separates the players who got the ball a lot and did even more with it—Josh Jacobs breaking the fantasy landscape in 2022—while removing the backs who turned 300 touches into useful but not particularly exciting seasons—Najee Harris in 2021.
This matrix from RotoViz’s Blair Andrews is useful for understanding how the metrics I won’t shut up about are interacting with each other. A running back’s previous success (and to some degree their prior displays of talent) can be used to predict future opportunities. FPOE isn’t extremely stable from year to year, but it does have some staying power. It’s even better at predicting future expected points. Pat Kerrane recently showed a similar example when he pointed out that success rate—using NFL’s Next Gen Stats—helps predict players who will see more high-value touches in future years.
I don’t want to imply that the stability of FPOE is exclusively measuring running back talent. It is absolutely measuring the yearly carryover of:
- Offensive line play
- Scheme
- Play-calling
- Quarterback play
- Dozens (at least) of other things
However, if we can use it and other metrics like success rate to get small insights into running back talent, we should run with them.
When This Stuff Matters
Pat Kerrane previously pointed out that in the early rounds, running back is a volatile position. There are running backs that break the game and running backs that kill your roster. The middle ground between these outcomes is less pronounced than it is at receiver. As we move later into the draft, solid but not game-breaking seasons can be massive wins. Jamaal Williams averaged .3 FPOE last year, good for 36th among all running backs. He ranked sixth in advance rate in Best Ball Mania III. The final component of winning any fantasy format is the market.
Williams’ expected points were drastically undersold by his ADP. His points over expected were functionally irrelevant because we didn’t need an efficient season from the 14th-round pick who had a high-volume, goal line role on a strong offense. He was nearly locked into a great season the moment he put D’Andre Swift on the back burner. Najee Harris, on the other hand, had to outperform his volume in order to pay off an ADP that saw him come off boards and the one-two turn. He did not and turned in a pitiful advance rate of eight percent. That’s under half of what the average advance rate (2-of-12) is.
Early in drafts, we want elite runners with three-down roles, touches near the end zone, and targets. Later in drafts, it’s fine to take players who don’t check all of these boxes. Given how stable roles are compared to efficiency, we want to be targeting some of the boring runners Denny has become obsessed with in his old age. Brian Robinson, a player he highlighted in his article (which I did actually read), is one of my favorite players to draft this year. Now let’s get to the part you’re all here for:
The Player Takes
Jonathan Taylor
Taylor averaged 2.3 FPOE as a rookie and then boosted that number to five in 2021, his RB1 overall campaign. He recorded the second-most yards after contact in a single season, trailing only Derrick Henry’s legendary 2020 campaign. Taylor faltered last year, but his offensive line play took a hit, his quarterbacks tanked the offense to 31st in EPA per dropback, and Taylor himself dealt with an ankle issue throughout the year. Pro Football Focus ranked the Colts’ line at 10th heading into the season and his quarterback is going from a collection of statues to the most athletic passer to ever enter the league. Taylor’s standoff with Jim Irsay and the Indianapolis front office is a concern, but in talking with Denny, Kerrane, and Patrick Daughtery on the Rotoworld Football Show, no one believed he was at risk of not playing in 2023.
Taylor is the perfect bounce-back candidate and we’re getting an absurd discount right now because of an unlikely holdout.
Tony Pollard
Pollard led all running backs in FPOE per game last year and has surpassed his expectation in all four of his pro seasons. Since entering the league, Pollard ranks first in yards after contact per carry, fourth in missed tackles forced per carry, and fifth in explosive run rate. His backfield combined for the fourth-most expected points in 2022 and Ezekiel Elliott is currently out of the picture. Pollard strikes me as a back who we’ll be arguing about as the 1.01 this time next year.
Breece Hall
Still on the PUP list and Dalvin Cook rumors swirling, Hall’s ADP is in a tailspin as the season approaches. Looking at our ADP Trend Report, Hall’s draft position has dropped half a round this summer and Taylor is the only bigger faller in the first five rounds of drafts. While the rumors of Cook joining the team are not ideal for his fantasy output, Hall looked like the type of back who could force his way into 20 touches a game whether or not there was a veteran presence on the roster. He led all backs in rushing yards over expected per attempt while also excelling as a receiver, ranking second in yards per route run as a rookie. He averaged 2.7 FPOE and that number jumped to 5.5 in his final four games. Now going in the fourth round, the risk of him taking a few weeks to get up to full speed is mitigated by the reduced cost.
Kenneth Walker
Walker ended the year at 12th in rush yards over expected on the back of three carries of 60 or more yards. As Devin pointed out, long touchdowns are the easiest path to a high FPOE. Walker earned a respectable 2.2 points over expected per game as a rookie. He also dominated the Seahawks’ high-value carries, taking all nine of the team’s rush attempts inside the five-yard line. With Zach Charbonnet in the mix, Walker is definitely more of a long-shot to post a league-winning season, but his price in drafts also reflects that.
D’Andre Swift
I haven’t been drafting Swift much because I think his odds of getting the expected points part of the equation high enough to be an elite back are low. His previous team saw him as a role-player and his current team has at least two other backs they plan to platoon with him. However, he is a back that scares me to enter the league with so little exposure to. Swift has seasons of 2.1 and 1.5 FPOE per game and neither came on a massive weekly workload, meaning he has shown the ability to greatly outperform what is asked of him. He now gets to run behind the league’s best offensive line and has a quarterback who can freeze opposing linebackers at will. Swift needs to supplant Kenneth Gainwell for work on passing downs, but there’s a path to an Alvin Kamara-like season if he proves to be as electric as the advanced stats claim he is.