Ezekiel Elliott and the Myth of the Replaceable Running Back

Mike Tanier@@miketanierNFL National Lead WriterApril 18, 2016

Ohio State running back Ezekiel Elliott (15) runs the ball against Michigan State during the first quarter of an NCAA college football game Saturday, Nov. 21, 2015, in Columbus, Ohio. Michigan State beat Ohio State 17-14. (AP Photo/Jay LaPrete)
Jay LaPrete/Associated Press

Ezekiel Elliott is one of the five to 10 best players in this draft class. He's one of the special guys, the immediate difference-makers, the "can't-miss" Pro Bowlers after you knock on wood for injuries. 

But Elliott is also a running back. NFL teams have shied away from drafting running backs in the first round in recent years. Only six running backs have been selected in the first rounds of the last five drafts. Only one, Trent Richardson, has been selected in the top eight since 2008, and that did not turn out so well.

There's a theory that bubbled forth from analytics circles years ago that running backs are "fungible." Fungible means easily and cheaply replaceable and essentially interchangeable. Your cellphone is not fungible, but its charger is. Tom Brady is not fungible, but the Patriots go through three or four running backs per year without worrying that none of them are Adrian Peterson.

I also bubbled forth from the analytics circles, and I always thought that the "running backs are fungible" concept took the results of the research way too far. (I secretly thought analysts just really loved writing the word fungible.)

Sure, every year a Charcandrick West or Thomas Rawls appears from nowhere to replace a superstar like Jamaal Charles or Marshawn Lynch. But does anyone think West or Rawls was equal to Charles or Lynch? Would anyone spend the hours/days/weeks after Super Bowl XLIX wondering, Why on earth didn't the Seahawks just give the ball to Thomas Rawls at the goal line?

Does it really make sense for a team to avoid drafting a player as good as Ezekiel Elliott as a matter of analytic principle?

Video Play Button
Videos you might like

A quick reexamination of the evidence suggests teams were drafting too many running backs in the first round back when football analytics first became a thing. But draft strategies have changed, as have salary structures. And while many running backs are interchangeable, the truly special ones like Elliott often require a first-round investment.

Rare Commodities

First, let's tackle the idea that "a team can get a good running back anywhere." It's bupkis.

The chart below shows all of the seasons in which a running back gained 1,200 or more yards from scrimmage since 2006, according to Pro Football Reference's tracking, broken down by draft order. Why 1,200 scrimmage yards instead of 1,000 rushing yards? Because we want to include more high-quality seasons by running backs like Brian Westbrook and Matt Forte who might have rushed for just 800 yards but added another 700 receiving yards.

As the chart shows, the number of quality seasons by high draft picks dwarfs the quality seasons by lower draft picks.

Granted, there are many duplicate seasons from the high-round selections. There are seven seasons of Adrian Peterson among the 37 good years by top 10 picks, for example. But that further illustrates the point that quality running backs—the type that produce multiple Pro Bowl-type seasons—are more plentiful in early rounds. The point of picking in the top 10 is to get a perennial difference-maker, right?

Looking at it another way, 34 running backs have gained 5,000 yards from scrimmage from 2006 to 2016. That's a low bar for finding a "good" running back: 5,000 yards from scrimmage means the player started for a few years or had a fairly long career as a committee back. Newcomers like Todd Gurley have not had a chance to reach the 5,000-yard barrier. But the list gives us a good sense of when quality running backs are drafted, while focusing on career totals keeps us from multi-counting Adrian Peterson and other superstars.

This chart shows where these "good" running backs were drafted, again broken down by draft order:

As you can see, the list overwhelmingly favors early-round picks. The only fourth-round rushers to crack the list were Marion Barber, Brandon Jacobs and Darren Sproles. Michael Turner represents the fifth round, Alfred Morris the sixth and Ahmad Bradshaw the seventh. The three undrafted free agents are Arian Foster, Pierre Thomas and Fred Jackson. Justin Forsett (seventh-rounder), Rashad Jennings (seventh-rounder) and Lamar Miller (fourth-rounder) will likely join the list, as may Devonta Freeman (fourth-rounder).

So where the heck did the idea come from that good running backs are growing on trees?

There are three factors at work:

First, it's true that pretty good committee backs are readily available in mid- to late rounds. If you are in the market for a Roy Helu Jr., Bilal Powell, Jacquizz Rodgers, Alfred Blue, Andre Ellington or Boom Herron to play a role, you can get one.

Among the backs with or approaching 5,000 yards, many of the late-rounders (Forsett, Jennings, Bradshaw, Thomas, Jackson) were role players who stuck for years. Rushers like that are available in the late rounds or in discount free agency, though they aren't exactly lying around on the sidewalk.

Second, great late-round sleepers were more plentiful 20 years ago, when Stephen Davis, Dorsey Levens, Terrell Davis and Jamal Anderson were playing in Super Bowls and having MVP-type seasons. Such diamonds in the rough are scarcer now.

Finally, many of the first-round "successes" among the 34 running backs are considered by most to be disappointments or busts: Ronnie Brown, Reggie Bush, Cedric Benson, Darren McFadden, Thomas Jones, Knowshon Moreno. They were all talented backs who had some very good seasons, often after rocky starts to their careers, but none of them lived up to their expectations.

Those first-round disappointments were cautionary tales against drafting "replaceable" running backs in the first round. Why draft Bush when you can get equal or better production from Sproles? Why draft Jones only to see him lose the starting job to Marcel Shipp, or Benson to see him lose out to the aging Jones?

Back in the 2000s, when the "running backs are fungible" theory became popular, powerful forces were negatively impacting the relative value of running backs. In the last five years, those forces have either been blunted or reversed.

The New Market

Let's look back for a moment at how running backs were valued a decade ago, before the Great Fungibility Revolution:

Running back salaries were ballooning. Bush's rookie contract as the second player taken in the 2006 draft was reportedly $62 million over six years. McFadden's contract as the fourth player selected in the 2008 draft was a reported six-year, $60 million deal. Both players received about $26 million in guaranteed money.

Frank Franklin II/Associated Press

Rookie salary inflation was a major issue from 2005 to 2010 and was one of the driving forces behind the 2011 lockout. But high-round rookies weren't the only running backs cashing in. Shaun Alexander signed his infamous eight-year, $62 million contract in the 2006 offseason. Larry Johnson signed for six years and $45 million after a preseason holdout in 2007. Both Johnson and Alexander were veterans who signed deals that would theoretically pay them millions into their mid-30s, an age at which even Hall of Fame running backs are worn to the nub.

Keep in mind that a decade ago, the salary cap was about 40 percent lower than it is now. Rookies and aging veterans were taking up a shocking percentage of cap space at a position where careers were short (frequently shorter than a six-year contract) and committee systems had proved successful.

Teams were drafting three running backs in the first round per year. From 1993 through 2010, an average of 3.2 running backs were drafted in the first round every year. At least two running backs were selected in the first round each year, though no more than five were selected in any given year.

A deep dive through the data shows that diminishing returns set in once teams reached for those third, fourth and fifth first-round running backs.

Here's the typical Approximate Value (according to Pro Football Reference) of running backs selected in the first round, by the order in which they were selected:

(As a point of reference, a perennial 1,000-yard rusher usually has a career Approximate Value in the 60s.)

The first running back off the board typically had an excellent career, despite some famous busts in the mix. The third or fourth guys were a hit-or-miss bunch. The three running backs who were the fifth first-round selections of their drafts were Rashaan Salaam, Trung Canidate and Chris Johnson—two busts and one difference-maker.

Teams picking the first running back off the board from 1993 to 2010 had a good chance of landing the likes of Marshall Faulk, Edgerrin James, Jamaal Lewis, LaDainian Tomlinson or Adrian Peterson. The second or third back drafted might be Eddie George or Marshawn Lynch but had a better chance of being Tim Biakabutuka, Chris Perry or Felix Jones. The deeper into each class teams reached, the more likely they were to come away with John Avery or Beanie Wells.

Scouting hadn't caught up with strategic trends. The first-round bust rate for Big Ten rushers drafted in the 1990s and 2000s was extremely high. Fourteen of those backs were selected from 1993 to 2010. Robert Smith, Eddie George and Larry Johnson became stars; Tyrone Wheatley and Rashard Mendenhall had some solid years; but the rest of the list is full of Ki-Jana Carter, Curtis Enis and Ron Dayne types.

In the late 1990s and early 2000s, many Big Ten running backs still lined up in the I-formation and ran downhill 20 times per game throughout their college careers. Meanwhile, the NFL was changing to the single-back, pass-intensive strategies we see today.

JOHN BAZEMORE/Associated Press

Many first-round misses of a decade ago were big, overused college backs who couldn't contribute to the passing game. T.J. Duckett caught 20 passes in two years as a college starter, but the Falcons made him the 18th pick in the 2002 draft. Dayne, a four-year starter and Heisman Trophy winner, caught just 31 passes while rushing 1,220 times in college; the Giants drafted him 11th overall in 2000 but lacked a time machine to take him back to 1977, when he would have fit an NFL offense. Biakabutuka caught a whopping 12 college passes; the Panthers selected him eighth overall in 1996.

It's genuinely shocking that these one-dimensional power rushers were waved into the first round in an era when the shotgun was becoming a base offensive set and teams were beginning to pass over 60 percent of the time. It was like buying the best eight-track tape or Betamax player on the market in the era of CDs and DVDs.

Three major course corrections have occurred since the days when drafting a running back early was a poor investment:

• No running back has a contract with a reported value anywhere near $60 million. Only Adrian Peterson and LeSean McCoy have contracts with reported values over $40 million. Todd Gurley, the 10th player taken in last year's draft, will make $13.8 million over four years. That's less than Bush's guaranteed money a decade ago, when the cap was much lower.

• The third- to fifth-best running back in a draft class is unlikely to be drafted in the first round.

• One-dimensional collegiate power backs are now generally valued as mid-round picks, though there are exceptions (Melvin Gordon caught 22 passes in college, for example). Teams are more wary of collegiate workloads than they were in the past. Also, few power-conference teams line up in the I-formation and play Cro-Magnon football anymore. For example, Elliott caught 58 passes in Urban Meyer's offense and had pass-protection responsibilities.

The course corrections of the last five years have changed the value of running backs. Carlos Hyde, Elliott's predecessor at Ohio State, might have been a first-round pick in the late 1990s: a Big Ten bruiser who could catch a screen pass if you placed the ball right on his hands. As a pricey first-round pick, his early career would be considered a disappointment. As a late second-rounder making less than $1 million per year, he has a chance to grow into his role.

Steve Nesius/Associated Press

Here's another example: What happens if you take Doug Martin and saddle him with a Bush or McFadden contract? The Buccaneers would have moved on after 2014, when he rushed for 494 yards. Maybe Martin's 2015 season would have taken place elsewhere, or maybe changing locations would have further hampered his value. Martin would have been a bust for Tampa Bay. Instead, he's an affordably priced veteran.

Changing the salary structure and selecting first-round picks more judiciously altered the equation for running backs in the last five years. The theory of "replaceable" running backs must be updated to reflect the new market.

Wise Investments

After performing science-y research and drawing conclusions, it's important to be careful when applying a general principle to a specific case. The average April daily high temperature in your region may be 55 degrees, but that doesn't mean you throw away snow shovels or wait until June to try on cargo shorts. Exceptions to the general principle occur all the time.

A decade ago, teams valued running backs incorrectly. They were overpaying untested rookies and aging veterans. They were digging too deep into draft classes too early in the draft. Sometimes, they selected the wrong type of player. It created a situation in which Reggie Bush was making millions of dollars to back up Pierre Thomas.

Because committee-type running backs were (and are) plentiful, and two or three of them might take up less cap space than an early first-round pick, it didn't take much analysis to conclude that drafting a running back early was an unwise investment. Stat-heads noticed, and so did teams.

But there are still special running backs who can do things a committee cannot replicate, guys who can run like CJ2K at his peak or enter Beast Mode or just be Adrian Peterson. There may only be three or four of them in the draft every five years, but when they do arrive, they are now available at a fraction of the old Bush price. It makes no sense, analytically, economically or old-school football-wise, to pass on a player like that.

Ezekiel Elliott is one of those special players. His game film reminds me of LaDainian Tomlinson. His skill package, from power and speed and big-play ability to blocking, places him a thick notch above "interchangeable."

Teams may pass on Elliott because they rank Laremy Tunsil, Jalen Ramsey or someone else higher. Fair enough. But if you think teams should pass on Elliott because of the "research," then you had better take a closer look at the research.

Mike Tanier covers the NFL for Bleacher Report.