It’s about time the remaining NFL GMs who still think paying top dollar for a running back is a good idea to take a hard look in the mirror and come to grips with the fact that they are egregiously inept at their job.

It seems like most GMs understand paying running backs is a bad idea as evidenced with Melvin Gordon shooting his air-ball shot this past offseason. With quants preaching their 400-level course metrics such as EP, EPA, and DVOA to showcase the relative unimportance of a top-tier running back, many GMs have indeed studied up, delved into the data themselves and re-tooled their approach to team salary allocation.

But what about the remaining GM’s who don’t get it yet? Maybe some of them still don’t believe in all these fancy voodoo statistics and are more comfortable with football strategies suitable for success back in the Reagan Administration. Not to name names but how is your favorite team ever going to learn this fact when this is the guy sitting in the back of the classroom you’re forced to rely upon?:

Don’t you worry. I won’t leave these “unique-paced” GM learners behind. That’s what this class is all about. This article does not require that you know what EP, EPA, DVOA or any other 400-level course football sabermetric is. I basically only mentioned those from the start as a humble brag that I even know what they are. Instead, I’m presenting a more 100-level “No GM Left Behind”-like course. In such, to make sure everyone understands the lesson ahead, we will be using convoluted data visualization techniques data scientists who have passed the fourth-grade refer to as “tables” and “bar graphs” displaying a statistical metric known to both theoretical physicists and students held back from the third grade as “average”. This way, no matter what pace of learning your team’s GM requires, they will be sure to comprehend.

Here’s your first and final worksheet of this class. As a matter of fact, it’s even all filled in for you. You just have to reference it. It’s simply a list of NFL running backs since 2009 grouped by average annual salary I pulled from OverTheCap.com:

To properly evaluate if these high contracts have been beneficial, I used these somewhat arbitrary groupings of salary to compare how well each group of highly paid backs performed during the tenure of their contract vs their collective backup running backs in simple yards per carry (YPC). The idea here isn’t to compare 13M backs to the 8M or 5M varieties, but instead to see if the difference in YPC between a 13M back and his backups is meaningfully more than the difference between lesser paid starters and theirs. If there’s no real difference, why the hell are they being paid on par with starting quarterbacks?

Quick boring z-copy on the analysis: to make this a fair comparison, only rush plays that met the following criteria were considered:

• Only rushing plays by a running back were used (no QB, FB, WR, etc.)
• Only rushes between both ten yard lines were used as these plays disproportionately bring down YPC averages for starting backs who are depended on for goal line carries
• No garbage time rushes were used. This was defined as plays in which there was either less than a 20% or more than an 80% win probability as determined by the wizards at NFLScrapR
• Only rushes with ten or less “yards to go” for a first down were considered (limits long downs where defenses are likely to allow more rush yards)

Sorry for the boring disclaimer, but sometimes the boring stuff in school is actually important. Skeptics will claim the situations of backup rushes are different so doing all this filtering puts things on an even playing field. I’ll reiterate one key point too: only rushes during the tenure of each player’s contract were considered. No GM cares what happened before one earned his contract or how bad he did after getting cut or moving on.

One last consideration to think about is box size. Many people have this idea that when a stud running back is in the game, defenses stack the box against them. This is a pretty good theory except for the fact that it’s dead wrong. But to my credit, I gave it a fair chance. The only data I am aware of that comprehensively logs men in the box comes from the 2019 #BigDataBowl provided by the NFL and Next Gen Stats for the 2017 and 2018 season. Here is how the combined top-five highest rushing backs of both years compared vs their own team’s backup running backs:

Summary: in 2018, the top-five running backs faced .01 (but actually .005 before rounding) more men in the box on average than their backups. In 2017 that number was exponentially higher at .09 more men. That’s another .26 MIB per game for a 20-carry workhorse and four entire men in the box more per 16, 20-carry per game, season. Call me crazy, but if your 13M thoroughbred can’t handle that heat, he’s probably better off served as glue.

With all the haters concerns addressed, let’s get to business. Let’s directly compare each of the three salary ranges provided in your pre-filled worksheet above and find out if those with the higher contracts have a greater difference in YPC than their backup running backs:

Results:

I realize these “tables” have no pictures and might be confusing to some NFL GMs. Without shaming them in any way, let’s convert these tables into graphs so these statistical stragglers have a picture book version without words they can read better:

Everybody get out your rulers because ain’t a damn fool going to tell any of these three sets of bar graphs are any different at first blush.

There we go. If you look very closely and squint really hard, you still won’t even see the entire .01 YPC difference from 13M and 5M-8M backs. And sure, the middle group did a touch better but remember, we’re not really interested in comparing production of top-tier backs with one another because they play on different teams with many other factors that could explain the insignificant difference in YPC. What we really care about is the differential between the “Contract” backs and the combined backup running backs on their respective teams. And as you can see in the tables and graphs, there is literally, LITERALLY no difference in the production of 13M+ running backs over their backups than 5M-8M running backs over theirs. Both have a differential of .32 YPC.

POP QUIZ! What does all of this mean from a general management perspective? (HINT: it means, it’s probably a bad idea to pay a lot of money to running backs).

Did you all get that right? Jerry, even you?

The data here is suggesting that the attributes that make a “good” running back (burst, speed, shiftiness, power) provide about .3 YPC over their team’s replacement-level running backs. And regardless of how much “better” players like Ezekiel Elliott and Todd Gurley seem to the eye, performance above replacement basically plateaus around that .3 YPC range. This doesn’t mean all running backs are as replaceable as the next. Otherwise that gap between starter and backup wouldn’t exist. Perhaps a running back in that 5M range is worth the extra .32 YPC in the long run. What all this means is that you don’t need an Ezekiel Elliott, or Todd Gurley, or Le’Veon Bell to accomplish this and they are not giving you anything extra for your dollar.

If this is all still too much numbers and statisticals, think about it historically. If you take all 26 5M+ salaried running backs of the past decade from your pre-filled out worksheet, exactly Marshawn Lynch of them won a Super Bowl during the tenure of their contract. Meanwhile, below is a comprehensive list of Super Bowl winning running backs rostered from the past five years (2014 – 2018):

James White, Sony Michel, Rex Burkhead, LaGarrette Blount, Wendell Smallwood, Kenjon Barner, Jay Ajayi, James White, Dion Lewis, LaGarrete Blount (again), Ronnie Hillman, C.J. Anderson, Shane Vereen, Steven Ridley, Jonas Gray, Brandon Bolden and for crying out loud LaGarrette Blount yet again.

Just look at that heaping pile of mediocre. Kudos to you if your brain hasn’t defragged half of these names from working memory. Pop Quiz! Can you point out which of these running backs are starters? Would the starters even know themselves? But this is exactly what NFL GMs need in a running back: capable, and forgettable. Forget about “excellence”. That’s so 1990s anyway.

This concludes the 100-level course on elite running backs don’t matter. Now comes the test. I dream of a society in which all GMs, regardless of natural ability, are able to bend their minds and absorb this third-grade appropriate research and straighten course. All they really need to do is pass this one question long “No GM Left Behind” test:

Which of the following is better?

A) Paying 5-8M instead of 13M or more for a running back that also provides your team .32 YPC more than your backups

B) A

C) B

D) Any of the above

-El Jefe