Fantasy football success: The running back stats that matter – and those that don’t

Fantasy football managers often rightly ask themselves…which stats matter and which don’t when it comes to predicting success? I’ve always pushed people to use a 50/50 balance between watching matches and rating players, while using metrics to back up your analysis, or take a second look if the metrics differ wildly from your opinion.

But I also have access to TruMediathe huge database. And while I was in a Zoom meeting with the fantasy editors (Brandon and Nando), we were digging around and playing around with some numbers and came across a cool idea for a column – this column. TruMedia, you see, lets you plot graphs on the x and y axes and it will spit out an R-squared number that shows the correlation between any two stats you can think of. So I took a bunch of them, ran metrics on the y-axis and half-point FPPG along the x-axis… and was able to find out which elements actually help predict the fantasy success. There was a lot of trial and error – and several surprises along the way.

I’m going to do these positions position by position using common (and some uncommon) metrics often mixed with the perception that they’re important to fantasy football. For this study, I went back eight years, using a threshold of at least 100 rushing attempts for running backs – my goal was to get around 200+ players. If you are unfamiliar with R-squared (aka rSq, r²), it shows the variation/correlation between two variables where 0 is no correlation and 1 is a perfect/exact relationship. In much simpler terms – the larger the r², the more important this metric (and the closer the points – which are real players – to the drawn line). The lower the r², the less they count and the points will be more scattered.

As a visual example, here are two charts of data points from high and low correlation statistics.

Higher correlation (predictive)

Lower correlation (irrelevant)

We mainly focus on per-game stats because they give a better picture of a metric’s correlation – but I’ve included career-wide numbers for reference. Obviously, no more snaps/keys/etc. will result in more fantasy points. And, yes, we go as far as “NFL Draft Pick Round” to try to find stats that will help you win.

The results (and what it all means below):

So let’s go !

The stats you thought were important, but they’re not!

NFL Draft Capital / Round… we do not care? Whether it’s a first-round or undrafted running back, there’s no correlation to fantasy success.

Even in Half-PPR scoring (and it doesn’t increase much for Full-PPR), the field play percentage for passing plays (Pass Play %) does not increase the value of a running back.

“That running back faced a ton of 8-man fronts.” Guess what? Vs 8 man box has no importance! As Emory Hunt once pointed out in these (virtual) pages, an 8-man front sometimes gets a talented running back to the second tier and past all the guys who are now turning around and were stacked on the line.

back to the draft, 40 yard dash times have always been overstated, but now we have proof of data correlation so you can stop worrying about them altogether.

Yards per carriage the rate doesn’t say much about the image, and oddly enough, Yards after contact actually a little less to predict the success of the fantasy.

Total number of plays equals more chance for fantasy points…kinda. It’s a decent correlation, but there are better options.

Statistics that predict success

Weighted opportunity! What is that? It’s a statistic created by Scott Barrett at FantasyPoints (and the next column I’ll have the one I made for wide receivers and tight ends). You can read the full story in the link, but the quick summary is that not all snaps and dabs are created equal. We know that, but this metric also carries the strongest correlation at a fantastic mark of 0.918.

Readings per game stand out hits per gamewhich seems counter-intuitive, but being on the pitch and touching the ball frequently is as easy as it gets.

Hasty attempts, yards and affected all see high correlation marks. Again, this seems obvious.

Goal to achieve (inside the 10 yard line) and Goal line attempts per game have significant value while Goal-to-Go Rush % (percentage of rush attempts inside the 10 meter line) is not relevant. As long as the ball carrier gets the attempts frequently every game, the percentage that equals (of his total carries) doesn’t matter. Here it is in graphic form! :

What it all means

Looking at my projections for the 2022 season:

Back to the 2021 season:

  • Derrick Henry, Alvin Kamara and Najee Harris were the three running backs with 18+ Rush Weighted Opportunity (Half-PPR).
    • I will reference this calculation throughout the year with waivers, reports of concern and more.
  • Jonathan Taylor, James Conner and Damien Harris were the Top 3 and the only ones with 1.0+ touchdowns per game (fourth was Joe Mixon at 0.81).
  • Every running back in the Top 18 for games per game (all 15.7 or higher) has finished in the Top 18 for Half FPPG, except Barkley.
  • The only running back in the Top 18 FPPG but not Plays per Game was Aaron Jones, but he was 21st.
  • As for G2G attempts, the Top 10 everything had at least 13.7 FPPG (Top 16 in FPPG).

All in all, we have a clear picture of what predicts running backs’ fantastic success – and what doesn’t – headlined by Weighted Opportunity. We’ll dive into wide receivers, tight ends, and quarterbacks in the coming weeks, giving you a better chance to predict your fantasy success.

(Photo by Wesley Hitt/Getty Images)

You may also like...