Being a productive rusher in the NFL takes a rare mix of skills and talents, such as speed, elusiveness, vision and anticipation. Those who have excelled at it have historically been rewarded with team-defining roles, league accolades, furious media attention and nice contracts to boot.
Of late, however, top rushers have seen their roles diminished and their pay stagnate. In the modern NFL, teams appear reluctant to commit resources to ball carriers like they used to.
Perhaps this reflects the new offensive landscape in the NFL, in which teams pass more and better than ever before. But it may also reflect a growing recognition that, for all their talent, traditionally great running backs probably don’t actually contribute that much to their teams’ chances of winning.
Consider the case of Adrian Peterson — the 2012 NFL MVP and the only running back to win that award in the past 10 years. Peterson has made by far the most money for a RB in NFL history after the Vikings paid him more than $12 million each year for the past six years — during which the Vikings averaged just over seven wins per season. He’s 32 years old and spent much of 2016 injured but led the league in rushing as recently as 2015. As an unrestricted free agent this offseason, Peterson signed with the New Orleans Saints for a modest $7 million over two years, with only $3.5 million guaranteed. In 2017, he’s slated to cost less than Bengals backup RB Giovani Bernard (who will cost $3.7 million against the cap) — not to mention 10 different kickers.1
While Peterson’s situation has its own contours and meniscus tears, it’s something of an emblem for the NFL’s approach to RB talent right now. With top players at most positions getting paid more and more as the league’s revenues and salary cap have grown (and in some cases, exploded), the average pay for top running backs has stalled and even declined in recent years:
In 20172 the top 16 running backs (the league’s above-average starters, give or take a few) are slated to make less than any offensive or defensive positions — even tight ends and safeties, who were the bottom-feeders of the nonkicking game in 2000.3 Additionally, running backs and six other positions – WR, DE, CB, LB, DT and OL — averaged between $3 and $4 million in 2000. In 2017, top RBs will average around $6 million — the other six positions will each average more than $10 million. And even this picture may be slightly rosier than it appears, because for several years Peterson bent the RB curve upward.4
If we look at each figure as a percentage of all money spent on the top 16 players at each position, things look even worse for running backs:
Relative to other positions, top running backs are slated to make around half as much in 2017 as they did at the start of the 2000s.
So why is this happening? Perhaps NFL teams have consciously devalued the running back, or perhaps it’s just a multitude of decisions in many different circumstances that have happened to lead us here. But while the RBs’ new situation almost certainly results directly from the league’s shift toward pass-centric offenses, it may also stem from the league ever so slowly wising up to the fundamental math of its own game. Much like with the 3-point shot in the NBA, passing the football in the NFL has virtually always looked better on paper. And, like the midrange jumper, the NFL seems to be (perhaps more slowly) creeping in the direction of running just about as rarely as the situation and game theory require. Indeed, the history of offense in the NFL is largely a story of running less and getting more efficient:
Over time teams have been passing more and more, and they have been picking up more and more yards as a result. Part of that is that teams are getting better at passing, but it’s also that they have gotten better at knowing when to pass. And, of course, those two trends play well together: As teams have gotten better at short, high-percentage passes, the rationale for running in situations where you need less-explosive but high-percentage plays has declined.
That said, the NFL still has a long and easily demonstrable history of running way, way too much — even recently. We can tell because we know how every play in the NFL since 2000 has affected a team’s likely win percentage. The stat is called WPA, or Win Percentage Added. WPA allows us to look at the results of plays beyond just yards gained5, which is helpful since a successful run should positively impact a team’s chances of winning even if it picks up fewer average yards, etc. Indeed, if every team played perfectly and everything was in game-theoretical balance, we wouldn’t expect to see much difference between runs and passes at all. This is not the case.
For starters, let’s cut out a bunch of special circumstances and look at the most vanilla run/pass decisions possible: first and second downs, with 5-10 yards to go, outside the Red Zone, outside the last two minutes of either half — giving us about 278,000 plays to work with. This is where the traditional, workhorse, MVP-type running back butters his bread. Now, since the decision to run or pass is largely a function of how far ahead or behind a team is and how late it is in the game, let’s break our results down by quarter and score margin before the play:
Basically, there is pretty much no ordinary situation in which running produces better results than passing. If a team is more than 10 points ahead in the second quarter, running has seemed to do OK. And that’s about it. Even situations where running a lot is pretty standard — like up fewer than 10 points in the third or fourth quarters — passing has done substantially better. Of course, some amount of run/pass balance is necessary, or defenses would completely tee off on the pass every time. But this issue is likely overblown: As a pretty straightforward application of introductory game theory, if one option keeps producing substantially better results than the other, you should do it more often.6
Of course, running the football has ancillary benefits, such as burning time off the clock, avoiding turnovers, gaining positive yards more consistently, picking up shorter yardage a higher percentage of the time, keeping the defenses honest, and so on. (There may even be situations in which teams pass too often, such as with 2-point attempts.) That sounds like a lot of good uses for the run! But note that, when it comes to these things, the quality of your running back — at least by conventional measures like how many yards they gain — is of secondary importance.
This is because even a great rushing attack is still worse at picking up yards than even a mediocre passing attack. The all-pro running back may gain a lot of yards as his team funnels its offense through him, but many (or even most) of those yards are picked up in spots — like when a team is slightly up or down in the third quarter — where passing would have been better (or at the very least, where teams should be passing more often). Indeed, much like with having a good punter, there’s a danger that a great running back could hurt his team, if he entices them to run too often. (Conversely, a potential problem with having a great run defense is that opponents may be bullied into passing!)
None of which is to say that the running back position will die out, or that the league’s unwillingness to pay a lot for them will continue indefinitely. Running backs and rushing may still be an important part of the game, so long as you aren’t trying to use it to pick up a bunch of yards on the ground. There are better ways to do that, and better things you could be doing with that slot.
For example, running backs who excel in short-yardage situations — such as Marshawn Lynch, Jerome Bettis or Marcus Allen7 — or “third down” or pass-catching RBs who can be legitimate multiway threats in a spread offense — may actually be more valuable than they seem. As the athletes who play “running back” get better at things like opening up the passing game and helping pick up first downs, the position may be leveraged more efficiently and see its value increase commensurately. Note the sole runner scheduled to make eight figures this year is Pittsburgh’s extremely versatile LeVeon Bell (whose one-year franchise player contract will earn him a little over $12 million in 2017).
But committing money to “workhorse” running backs who provide little outside of their ability to grind out a large number of yards inefficiently — a description that arguably fits Peterson as well as any great RB — is like doubling down on buggy whips when everyone else is scrambling to make flying cars.
CORRECTION (May 16, 1:32 p.m.): The “Why running sucks” chart in an earlier version of this article misattributed the source of its data. It was from Pro-Football-Reference.com, not ESPN Stats & Information Group.