The era of the first-round running back is largely behind us. Every once in a while, a generational talent is picked early and proves he was worth it—Adrian Peterson, taken number seven overall by the Minnesota Vikings in 2007, is probably a choice most teams would make again (ignoring, of course, the season he missed after he pled no-contest on child abuse charges). The Los Angeles Rams are probably pretty happy with their selection of NFL Offensive Rookie of the Year Todd Gurley with the number ten overall pick last year, given how much of a phenom he’s turned out to be. The Titans got a few very good years out of 2008 first rounder Chris Johnson (number 24 overall), and the Panthers enjoyed several seasons with Deangelo Williams running the ball (number 27 overall in 2006). Marshawn Lynch, the twelfth overall pick in 2007, didn’t exactly reap huge rewards for the Buffalo Bills, who drafted him, but he had a heck of a career after he landed in Seattle.

But generally speaking, the first round isn’t a great place to get a star running back anymore. Between 2011 and 2016, only four running backs were taken in the first round—and with the exception of Gurley, who sure looks like the real deal (but who’ll have to maintain last year’s pace for a few years to prove it), they haven’t been great calls. The Saints took Mark Ingram with the twenty-eighth pick in 2011, and though Ingram has matured into a good runner, it took him a long time to get there. The Browns drafted legendary bust Trent Richardson at number three the following year, and no running backs went in the first round in the two drafts that followed. It wasn’t until the Rams picked Gurley (and the Chargers took Melvin Gordon five picks later) that we saw another running back get selected on the draft’s first day.

But this year, Jerry Jones and the Dallas Cowboys selected Ohio State’s Ezekiel Elliott with the fourth overall pick. It was weird. Elliott is a fine running back—in fact, he might turn out to be a very good one—but for the Cowboys, who have a whole host of needs on defense, ignoring the cornerbacks and pass rushers available to take a running back was a strange decision. At the very least, it puts Elliott in an awkward position. As just the fifth running back drafted in the first round in the last five drafts, there are astronomical expectations for what he’ll have to do to warrant the opportunity cost of not taking Florida State corner Jalen Ramsey or Oregon defensive end DeForest Buckner.

Elliott will probably put up good numbers with the Cowboys. The Cowboys’ offensive line is great, and even an aging back like last year’s free agent pickup Darren McFadden averaged 4.5 yards per carry and put up over a thousand yards in ten starts. But that also means that the bar for Elliott is high. If he comes in and posts 4.8 yards per carry and 1,100 yards on the season—the same numbers that Todd Gurley put up in his ROY campaign—he’ll end up being a marginal improvement over a free agent they picked up in his eighth year in the league. (Gurley did score ten touchdowns last year, which is significantly better than the three McFadden scored.) So how good, exactly, does Ezekiel Elliott have to be to warrant his draft position?

A good place to start looking might be the 2014 performance of former Cowboys running back Demarco Murray. He rushed for 1,845 yards with an impressive 4.7 yards per carry, scoring thirteen touchdowns. But Jones let Murray walk after posting one of the best rushing seasons any player has ever had, leaving him free to sign with the Philadelphia Eagles—where, behind a lesser offensive line, his per-carry average dropped more than a full yard. That proved a decent decision, as McFadden turned out to be a capable replacement once he took over starting duties. But it means that we can expect, at this point, that any running back worth a job in the NFL should come pretty close to 4.5 yards per attempt rushing behind the Cowboys’ offensive line. (Last year, midseason free agent pick-up Robert Turbin averaged 4.3 yards per attempt, while Joseph Randle—who was cut by the team and is no longer in the league, managed 4.1 yards on each carry.) Elliott will need to exceed the bar set by McFadden last year. He might have to surpass Demarco Murray too—otherwise, the Cowboys could have just kept Murray around. He will have to be more than just good. He’ll have to be great.

There’s a strong feeling among observers that Jones drafted Elliott in an attempt to recreate the “triplets” era of Cowboys football, with a star quarterback in Tony Romo, a star receiver in Dez Bryant, and—with Elliott—a star running back to serve as this generation’s Emmitt Smith. But football is different now than it was in Smith’s era. Smith’s per-rush average was rarely eye-popping—his skill was at scoring touchdowns. But even the best running backs in the contemporary era of the NFL don’t score touchdowns the way that Smith did. He had multiple seasons where he rushed for twenty-plus touchdowns—if Ezekiel Elliott scores 20 or more touchdowns on the ground in 2016, he’ll be the first player to do so in the past decade. (Only four backs—Peterson, Williams, Arian Foster, and LeSean McCoy—have managed more than 15.)

So let’s look at what a basis for comparison would be. Among rookie running backs, the best seasons since the turn of the millennium belonged to Adrian Peterson in 2007 and former Broncos back Clinton Portis in 2002. And that’s really the level that Elliott would have to reach in order to be a marked improvement in the backfield over McFadden. Both Portis and Peterson averaged at least 5.5 yards per attempt, and each of them managed to rush for at least 1,340 yards. They each found the end zone a lot, too—Portis scored 15 rushing touchdowns, while Peterson had 14.

If Elliott puts up numbers like that, then Jerry Jones will have earned his smug smile. It’d still be hard to say that the Cowboys don’t need a corner like Jalen Ramsey or a pass rusher like DeForest Buckner, but if the team has a ground game that can surpass last year’s by a full yard per carry and score touchdowns instead of field goals, a lot of the problems the team had on defense (namely, the fact that the defense was on the field so much) will get covered up.

So that’s the bar, roughly, for Ezekiel Elliott. It’s a high bar—literally two players this century have hit it—but he was taken with the fourth overall draft pick to be a once-in-a-generation player. We’ll know that he’s earned that status if he rushes for five-and-a-half yards a carry, manages somewhere in the neighborhood of 1,500 yards, and scores at least a dozen touchdowns. Good luck!