Predicting Combine Measurements for College Running Backs

by Noah Hills · NFL Draft

Almost two years ago, I published an article titled Your Favorite Devy Running Back is Too Small. Now What?. In it, I shared my findings from research into how much weight we can expect running backs to gain from the time they are high school recruits to the time they weigh in at the Combine or a Pro Day as an NFL Draft prospect. What I found then was both interesting and helpful. But I’ve since wanted to continue exploring this problem and fill the holes in the original research by attempting to predict NFL weights of running back prospects. This is done by examining their weights in each year of their college careers, and accounting for their height.

In that original article (and in this one from earlier this year), I started by laying out a case for simply caring about how much running backs weigh when they get to the NFL. Essentially because most of the fantasy-relevant seasons of the past 15 years have come from backs who weigh at least 210-pounds, even controlling for the fact that heavier players represent a larger portion of the overall running back population than do smaller players.

A given running back is more likely to be fantasy-relevant, especially for longer than one year, if he’s relatively big. It thus behooves devy leaguers and forward-looking dynasty gamers to care about how much they weigh before they turn pro; insofar as their weight at a given point in their college career is predictive of what they will eventually weigh as pros. The question then is two-fold:

1) How well can we predict the eventual NFL weights of college backs?

2) How early can we have actionable conviction about those predictions?

If we can’t predict how much college backs will weigh by the time they reach the NFL with any accuracy, then a nihilistic approach to pre-Combine listed weights is appropriate. And if we can predict it with accuracy but not until a player is three or four years into their college career, then not much advantage exists to be gained by paying attention early. Luckily, I’ve found that the answers to the above questions are “pretty well” and “pretty early.”

Average Yearly Weight Gain Patterns

The sample players observed were all running backs drafted between 2013 and 2021 after finishing their careers at FBS schools. I didn’t go further back because yearly listed weights are more difficult to find pre-2013. And I didn’t want to include players for whom I could find weights for only some of their college seasons. I also didn’t include players who spent their entire college careers at non-FBS schools for essentially the same information-availability purposes. This left me with a total of 174 players spanning nine draft classes.

Among that group, players left high school weighing an average of 199-pounds. They eventually weighed in at the Combine at an average of 213.5 pounds. Between high school and the Combine, they weighed an average of 203-pounds as true freshmen, 209-pounds as sophomores, and 211-pounds as both juniors and seniors.

The average yearly weight gain pattern is therefore as follows:

Of immediate interest is finding that players experience the majority of their total eventual weight gain in their first two years of college; or, more accurately, by the time they weigh-in as sophomores. Which could be just 15 or so months from the time they graduate high school. Players gain almost half of the total weight they’ll eventually gain in the year between freshman and sophomore weigh-in. They will have gained almost 70-percent of their eventual total by the time their sophomore year starts. Relatively little weight gain takes place between sophomore weigh-in and junior weigh-in. And then pretty much none after that until a player leaves college and begins preparation for the NFL.

Upon establishing the above as the “normal” weight-gain pattern for eventual NFL running backs, I hypothesized that players who would experience relatively little weight gain or relatively large weight gain would have different weight-gain patterns than players who gain a more “normal” amount of weight.

After binning players into weight-gain groups (those who gained anything less than the 14.5 average, those who gained at least 50-percent less, those who gained anything more, and those who gained at least 50-percent more), I found that, while starting weights for these groups varied relative to the total population average, weight-gain patterns among those groups are relatively the same as they are among the total population.

Those findings are illustrated in these two charts:

Save for those in the Severely Below-Average group who gain less than 7.25 pounds between high school recruiting and the NFL Combine, the weight-gain pattern for each subset generally matches the pattern of the group at large. The Severely Below-Average Gainers are likely to be “finished products” earlier in their careers than others, with an average weight gain of only 3 pounds by Year 2 of college among them.

OK, Now What?

The above analysis didn’t yield a lot of fruitful information. What it did offer is more useful in hindsight than it is in the moment. It’s easy to look back and say that a guy who gained only six pounds between high school and the NFL was done gaining weight as a sophomore. But it’s more difficult to look at a guy who’s only gained four pounds by his sophomore year and say that he’s destined to be a below-average weight-gainer by the time he reaches the NFL.

The most useful tidbits I gleaned here were that guys who come into college at heavier weights tend to gain less than those who enter college at relatively light weights. This is intuitive and also backed up by the findings from the research in my original article. And that players generally gain weight at a rate that resembles the overall average pattern, regardless of whether they are gaining more or less weight relative to overall yearly averages.

The next step was finding how well we can predict a player’s weight in each future year given knowledge of his present weight.

The following table shows the r-squared values of each year’s weight to each subsequent year’s weight:

My first takeaway here is that weight is fairly good at predicting future weight. Naturally, it’s best predicted only one year ahead, no matter what you use as your starting point. And the predictive power of a player’s current weight increases as he ages. Particularly interesting is that by a player’s sophomore season, we can predict his weight as an NFL player. With almost as much accuracy as we can when he eventually becomes a junior or senior. That tracks with the earlier finding that player weight-gain is 70-percent complete by the time sophomore weigh-in comes around.

This is all an important step toward answering the questions I posed earlier. Using only his weight at a given time in his college career, we can predict a running back’s eventual NFL weight with relative accuracy. And as soon as some grad assistant updates his weight on the team website before his sophomore season begins.

Accounting For Height

My next step was accounting for a player’s height.

I hypothesized that taller players would gain more weight between high school and the NFL than shorter players, given a greater “frame” to which to add. Though I quickly detoured to calculate our ability to predict a player’s NFL height using his listed height as a high school recruit or in each given year of college. I found that NFL height can be predicted from any yearly listed height just as accurately as weight can from a player’s junior season listed weight, and that the average player measures in half-an-inch shorter at the Combine than he does in any season from high school recruitment to junior year in college. The actual calculations are a bit more involved. But as a general rule, whatever year it is and whatever height a guy is listed at, just subtract a half inch and you’ll come pretty close to nailing his eventual NFL.

Back to our regularly scheduled programming.

Accounting for player height is indeed helpful in projecting eventual player weight. In order to do this analysis, I started by once again binning players into groups. This time based on high school recruiting weight “classes.”

Players who weighed between 161 and 165-pounds as high school recruits were placed in the 165-pound weight class. Those who weighed between 201 and 205 were placed in the 205-pound weight class, and so on. I then found the average height of players within each of those weight classes to establish a baseline for comparison of relatively tall vs. relatively short players among those who weigh similar amounts. It didn’t make sense to me to compare the weight gain of a 6-1 player and a 5-8 player if one of them started out at 215-pounds and the other at 180-pounds.

From there, I found the average eventual weight gain within each weight class, as well as the average eventual weight gain of relatively tall players and relatively short players within each weight class. Finally, I was able to use the differences in those average eventual weight gains to find how many (positive or negative) pounds it’s worth to be tall or short relative to weight. And, using a (confusing unavoidable pun alert) weighted average, how many pounds it’s worth to be tall or short in general.

Phew. All those findings are below:

On average, it’s worth over two pounds of eventual added weight to enter college as a relatively tall player. Shorter players can expect to add approximately three fewer pounds than players of average height in their respective weight classes.

Accounting for height also increases the predictive power of our weight projections.

Below are the r-squared values of yearly projections of eventual NFL weight using three sets of independent variables: weight alone, weight and height together, and then weight and height in a given year combined with weight from prior years:

You can see here that accounting for how heavy a player was at past yearly checkpoints in addition to their current weight and height makes for predictions of future weight that are relatively strong starting in a player’s freshman year in college. And they’re stronger than predictions of NFL weight using player weight from a year in the future. In other words: if I look at a player’s weight and height as a sophomore while also accounting for his weight-gain pattern given his yearly weights as a freshman and high school recruit, I can predict his NFL weight with virtually identical accuracy as I could if I had a time machine and knew what he was going to weigh as a junior.

Putting It Into Action

This finding is potentially useful for identifying early on in college careers which players are likely to weigh-in at ideal sizes by the time they turn pro. I should also mention that all of this analysis was done within that sample of 174 post-2012, FBS-school drafted running backs. But it was tested out of sample on a group of 55 undrafted and/or pre-2013 players for whom I had full-career yearly weight data. The r-squared values among this out-of-sample group were consistently at or above what they were among the original sample group.

As an example based on this analysis, here are projected Combine measurements for every running back taken in the first two rounds of the NFL Draft since 2017 based on information available to us while they were underclassmen or earlier (projected weights are calculated yearly, projected height is calculated as of year two in college), next to their actual Combine measurements:

Cells highlighted in red represent final heights that proved to be more than an inch off of projected height and final weights that proved to be more than five pounds off of any projected weight. Saquon Barkley‘s dark red shows that his final weight was more than 10 pounds off of any projected weight.

Cells highlighted in green represent accurate projections;. Those in light green show height projections that were within half-an-inch of actual eventual height and weight projections that were within five pounds of actual eventual weight. Those in dark green show height projections were within a quarter-inch of actual eventual height and weight projections that were within three pounds of actual eventual weight. There’s much more green here than red. And as early as we have a player’s final height and weight listings as a high school recruit, we’re batting .591 on weight projections within five pounds of eventual Combine weight.

Let’s take our ability to project NFL measurable sand apply it to the most prominent running backs in the current college football player pool. Here are the projected Combine heights and weights of 20 of the best, calculated using their listed measurements from 2021 rosters:

With these projections, I’d be tentatively worried about the prospects of Blake Corum, Jahmyr Gibbs, TreVeyon Henderson, Will Shipley, and Sean Tucker being built like traditional three-down backs by the time they get drafted. And I’d be almost certain that Devon Achane and Deuce Vaughn won’t get there. That’s not to say I’d be disqualifying these guys from my devy ranks or fading them specifically. But relative to the respective styles that dudes play, it’s certainly worth having a good idea of how they might weigh in as pros. It’s easier to be excited about a slightly-undersized Gibbs at 5-10, 208-pounds when he catches 30 passes a year in college than it would be if he was the same size but wasn’t a great receiver (*cough* Kenneth Walker *cough*).

Last Word

I suppose that’s that.

Any future research I do in this area would likely focus around exploring whether there are certain schools who are more or less trustworthy or accurate in their yearly listed measurements than average. If Kansas State chronically overshoots their guys’ weights or if Michigan consistently lists dudes at two inches taller than they eventually measure in at at the Combine, I’d want to know. It’s also worth noting that all this analysis was done using a sample of players who have already turned pro. I would assume that weight-gain and height-change patterns of players who weren’t good enough to reach the NFL might be slightly different than the patterns of the players looked at here. And it’s probably worth keeping an eye on yearly weight-gain for that reason.

Pooka Williams and Chris Thompson were both 4-star talents who weighed 175-pounds as high school recruits. Williams was the better college player. But it’s probably not a coincidence that he went undrafted after measuring in at 170 at his Pro Day. Thompson was selected in the fifth round after being 192 pounds at the Combine. It’s unfortunate, but face it, fellas: size matters.