A First Look at the 2026 RB's
A lot of backs to like, but only one to Love
Last year, I wrote about the Mitchell line, a threshold I came up with based on former 49ers RB Elijah Mitchell. Its purpose was to differentiate real-deal, fantasy-startable RB prospects from the rest, using a cutoff of 250 PPR points across the first three seasons of a player’s career.
I don’t want to go into too much more detail; you can look at the piece here, but in short, it divides prospects pretty cleanly. Our namesake back, Elijah Mitchell, resides just over that line; more importantly, though, he’s in his own zip code, with barely any other players living in the fifty-point range below him. Those are guys like Benny Snell, useful players who never became surefire fantasy starters.
And that’s why I think the Mitchell line works so well: if you’re spending a pick on a running back, you want guys who have the capability to be a starter. Even if Mitchell’s career was fraught with injury, I think his level of early-career production is a good benchmark for what we should shoot for when drafting dudes.
The 2025 Class
How, then, can we predict which backs will surpass the Mitchell line, and which will fall below it? I’ve been a fan of Next Gen Stats’ NGS Score in the past, which blends a running back’s on-field performance with his traits to get a holistic view of him as a prospect. It’s no surprise, then, that among the countless metrics I tried, it was best at sifting through RB’s.
Let’s first apply NGS Score, then, to the 2025 class and see how we do. I should note that this is in-sample for our data, meaning that it was part of what we used to determine our cutoffs. Yet it’s still instructive of what makes a great player (or, conversely, something less than that).
In the graph below, you’ll see two areas, one blue, and one orange. (Note: hover over the dots to see more backs). The bottom zone is what I’d call the ‘dead zone’, where backs who received an NGS Score below 67 live. In the past decade, only one out of the 22 backs below that threshold in the past decade, only one of them cleared the Mitchell line. That’s a pretty astounding result, and seems like a death sentence for any back.
The first thing that pops about the 2025 class is the number of clear-cut, high-impact starters in it. These are the guys in the blue zone (above 81 NGS Score); players in this region typically hit. In 2025, most of these guys went early in dynasty drafts, and have already justified their cost.
What’s equally exceptional, however, is how none of the top 20 backs by dynasty ADP fell into the dead zone. The closest is Woody Marks, who, after seeing newfound competition from David Montgomery, feels like the perfect example of a borderline case.
Marks succeeded where others failed for two reasons. First, situation: despite going far later than, say, Kaleb Johnson, Marks had extremely weak (and injured) competition ahead of him, whereas Johnson was forced to sit between two healthy, productive players. Second, draft capital: the fourth round is a relatively high price to pay for an RB, and you’d like that investment to give you some decent snaps. It’s why a lot of nominally superior prospects, like Damien Martinez, en up getting cut just because they were taken later.
The 2026 Class
If the 2025 class was stellar, then what can we make of the 2026 class? In a way, this year’s group is the mirror image of last year’s: similar amount of mid-tier guys, except with way more duds than studs on the periphery. The only sure thing this year is Jeremiyah Love, who has a real chance to go top 10 in the upcoming NFL Draft. He’s so good, in fact, that I had to adjust the chart below around him; if I included him, his 94 NGS Score would’ve greatly distorted the rest of our data.
As for the rest, the name of the game is, as always, value. If we go by the mock draft consensus, guys like Adam Randall and Desmond Reid (more on him later) seem like pretty solid bets relative to dudes like Singleton or Washington, who you’re going to be paying a far higher premium in fantasy for.
Now, there is, of course, a caveat, which is that based on big boards, Randall and Reid are likely to be mere seventh-rounders, if not UFA’s. While NGS Score is nice, the biggest indicator of prospect success is the real-life draft capital spent on them. If they go in the fifth or sixth round, take notice; otherwise, it might be wise to pass on these potential August cuts.
Diving Deeper into the Dead Zone
OK, so it looks pretty obvious that unless we go over this with a fine-tooth comb, there aren’t many stars in this year’s draft class. But if we’re looking at a bunch of JAG’s, then what moves a dude from ‘meh’ to ‘do not draft’? Should we add more criteria for success?
The answer, it seems—after a stupid amount of trial-and-error—is a pretty resounding ‘yes’. But what are these magical features that will lead us to the promised land? My goal was to find indicators that do two things. First, they have to be aspects of running back play that complement each other, i.e., they aren’t redundant. Second, they have to say something that NGS Score hasn’t already said.
This rules out most rote production metrics—rushing share, receiving efficiency, etc.—which are already rolled up more or less in the NGS score itself. Where this ultimately leads us, then, is to weight and PFF’s elusiveness rating. From a mathematical standpoint, it basically doubles our effectiveness in finding busts.
NGS score was truly a killer on its own, to be sure. Of the 22 players with an NGS Score below 67 in the last decade, all but one fell below the Mitchell line. Yet when you add in the cross-section of weight and elusiveness rating, that number grows to 38 (i.e., 56% of total busts)
Even if that number doesn’t sound like much, the important thing is, we have an incredibly low false positive rate. Some of the guys the system flags as busts might do OK in the NFL, but very few of them will become outright fantasy starters. Of the 42 potential busts we flagged using our method, only four of them turned out to be real-deal, productive backs, and some of those, like Isiah Pacheco, were massive outliers.
But more importantly, I chose these metrics because they were both clean and intuitive. It’s easy to understand how, if a back is both elusive and big, they’ve got a good shot at success. Again, few players in this class fit that bill, and are probably more one than the other (i.e., we have a lot of bruisers and slippery little guys). Only Jeremiyah Love clears both bars resoundingly, and we all know how good he’s gonna be.
2026 Class
With all that out of the way, what does this practically mean? Well, the bad news is, we don’t pick up much new information from this. It mostly confirms what we were seeing with the NGS score, and if I had to use one word to describe the ‘26 running back class, it would be borderline.
Just as we saw all of our guys jumping out of the NGS dead zone, most of our guys are also just elusive and big enough to meet reasonable NFL thresholds. Everyone else—the dudes who are too scrawny and slow—we already flagged as potential busts via NGS Score.
The two notable exceptions whom NGS Score failed to flag share similar names. Demond Claiborne and Desmond Reid both sit well under where you’d like your backs to be in terms of size, with Reid specifically standing at a very tiny 5-foot-6. Despite not measuring at the combine, Reid is quite possibly a sub-4.4 back, yet his low elusiveness rating suggests his speed didn’t translate to the field.
Claiborne instills more optimism: despite being similarly undersized, he actually ran at the combine, giving us concrete proof of his elite speed. It’s telling, too, that Claiborne is projected to go about 75 picks above Reid—around pick 150—where Reid is almost in priority free agent range.
But I do think these two players are instructive of where we’re at with this class. Even including the more traits-y guys, it’s still a class full of box-checkers, and not outright studs. Perhaps an even better word to describe it would be flat, though some of these guys are definitely better than others. Yet there’s still not a whole lot separating them, and I wouldn’t be surprised if teams leaned more on preference than pedigree come draft time.
Summary
Overall, I think we’ve done a good job at confirming the consensus about the 2026 RB class. There’s still plenty of time between today and the draft, of course, and what looks like an extremely strong correlation between NGS Score and success rate is bound to get noisier once we know where these prospects are headed. (To my understanding, NGS Score incorporates scout grades, which means it should be as in-line with big-board rank as it currently is).
Yet I don’t think the overall picture will change too much from now ‘til then. Maybe we’ll have a better idea of which fringe guys got a boost in the draft, and who had an unfortunate draft-day fall. When that happens, I’ll be back with an improved RB model of my own. Until then, we can still make mental notes of early red flags we saw, warning signs to heed lest we fall prey to some runaway hype train.