Data says investing in a running back no longer makes sense

Adrian Peterson joins Drew Brees, Saints for fresh start (0:54)

ESPN'S Josina Anderson, Dan Graziano, Mark Schlereth and Herm Edwards examine what led former Vikings RB Adrian Peterson to New Orleans. (0:54)

With the NFL salary cap rising nearly $50 million over the past five years, business is booming for player salaries. Unless, of course, you happen to play running back.

The relatively modest terms of Adrian Peterson's deal with the Saints (one year at $3.5 million guaranteed, plus a one year option), coupled with the fact that the probable Hall of Famer was on the market until later April, are emblematic of a broader trend where RBs are concerned.

According to Spotrac, the average running back salary is $1.4 million, making it the third-lowest-paid position in the game. Even average salaries for kickers ($1.6 million) and punters ($1.7 million) are higher. Only fullbacks and long-snappers represent less lucrative positions.

There are exceptions to every rule. Le'Veon Bell, David Johnson and Ezekiel Elliott served as offensive foundations for their respective teams last season. However, those three were the only players in the league to account for more than 30 percent of their team's yards from scrimmage in 2016. That type of three-down workhorse is rare and doesn't really signify any sort of larger trend at the position. With most teams reluctant to commit significant dollars or a high draft pick to the position, what is the optimal way to acquire a running back? Rookie contract backs are certainly cheaper than proven veterans, but are they consistently more productive? And where do bargain-bin vets, some of whom are cheaper than first-round picks, fit in?

To answer this, we can use a stat called adjusted yards per carry (AY/C), developed by Football Perspective's Chase Stuart, which compares yards per carry to the league average for that season. For example, Mike Gillislee of the Bills had 1.52 AY/C last year, which was the difference between his 5.71 yards per carry mark and the league average of 4.19. That 1.52 mark ranked third among running backs with at least 75 rush attempts. Gillislee, whose teammate LeSean McCoy finished fifth in AY/C, proved to be especially valuable considering that his cap hit in 2016 was just $600,000, according to ESPN's Roster Management System.

Gillislee's 2016 season was one of the 10 best over the past five years. In fact, eight of the top 10 AY/C seasons since 2012 came from players on their rookie deals:

Additionally, four of those had cap hits under $1 million. But that doesn't necessarily mean younger, cheaper running backs produce better on the whole. Over the past five seasons, Roster Management System has salary data for 272 running backs with at least 75 carries in a single season. Using that sample, we can divide them into three groups: rookie deals, prime vets (ages 25-28) and decline vets (age 29+). The average AY/C for each group shows one type of running back is consistently above average every year.

As expected, teams get the best production from the typical pricey veteran. But he's only about two-tenths of a yard better than the average rookie and less than a half-yard better than a bargain vet. For perspective, the 2016 player comps for each group would be Rob Kelley (rookie), DeMarco Murray (prime vet) and Frank Gore (decline vet). Teams pay a premium for prime vets, which we can see based on the average of each group's cap hits since 2012. In this case, the "average" refers to the harmonic mean, which reduces the effect of outliers like Adrian Peterson's $15.6 million hit in 2015.

Even when controlling for outliers, prime vets cost about 72 percent more than decline vets and 318 percent more than rookie deals. It's easy to sell the allure of a pricey free-agent running back as an offensive weapon, but of the top 50 rushing seasons in our data set, only 10 came from prime vets. Jamaal Charles, who leads all running backs since the NFL-AFL merger in yards per carry, was the only prime vet to make the top 50 twice. Fewer running backs are making significant money, but even the ones who do get paid usually fail to deliver multiple seasons of elite production.

Teams should extend that caution to the draft as well. While Leonard Fournette and Dalvin Cook serve as the splashy headliners, recent history suggests the best returns can be found in the middle rounds. Looking strictly at rookie-deal seasons, here's how rookie RBs fared in AY/C based on the round they were drafted in.

The average first-round running back on a rookie contract is actually a below-average player, at least in the past five years. No, this isn't all Trent Richardson's fault; of the 30 individual seasons from first-round rookie-deal running backs, only 13 produced a positive AY/C. Several of those came from one-hit wonders such as Donald Brown and Knowshon Moreno, while Doug Martin and C.J. Spiller put together two efficient years but failed once they signed their second contracts. Mark Ingram stands out as the most successful first-round running back of the past five years, having delivered three seasons of positive AY/C value for the New Orleans Saints during his rookie deal.

Indeed, the best production recently has come from the middle rounds. Last year alone, five of the 10 most efficient rushing seasons came from rookie-deal players picked from Rounds 3-5: Gillislee (Round 5), Ty Montgomery (3), DeAndre Washington (5), Jordan Howard (5) and Jay Ajayi (5).

A few lucky teams enjoy the reliability of elite three-down superstars. But for most teams, the evidence suggests that aggressively shorting the position in the draft and free agency could unearth players who are not only cheaper but also arguably just plain better.

For more from ESPN Analytics, visit the ESPN Analytics Index.