The Workhorse Cliff Is a Myth. The Birthday Isn't.

Every August somebody tells you a running back is cooked because of last year's workload. Four hundred touches. Three hundred and fifty. Pick your number. The back is used up, the tread is gone, fade him.

I believed a version of this. So I measured it: every top-24 running back season from 2008 through 2024, and what happened to each of those backs the following year. I counted a season as trouble if the back missed three or more games, or his half-PPR total fell by 20% or more, or he never played another NFL season at all.

The workload story fell apart the minute I ran the numbers. What actually predicts trouble has been sitting in plain sight on every player page in the world.

First, the number nobody bothers to look up

Here's the question I should have asked before anything else: what happens to a normal top-24 back?

251 of 408 of them. 61.5%. Three out of every five top-24 running back seasons since 2008 are followed by missed games, a big scoring drop, or no season at all.

That is the number every scary workload stat has to beat. It almost never gets quoted, because on its own it isn't a take. It's just the weather. Running backs get hurt and running backs decline, and if you draft one in the first four rounds you are buying a 61.5% chance of being annoyed next season no matter whose name is on the jersey.

Keep that number in your head for the rest of this article.

The workhorse cliff, measured

Now the claim. Backs who cross 320 touches in a season: how often does trouble follow?

64.6%. (53 of 82 seasons.)

And everybody else in the top 24? 60.7%. (198 of 326.)

That's a gap of four percentage points, and when I put an honest margin of error on it, the range runs from negative eight to positive sixteen. It straddles zero. In plain English, I cannot tell a 320-touch back apart from any other high-end back. The heavy workload effect isn't small in this data. It's absent.

The famous version of the claim doesn't survive either. Four backs have crossed 400 touches since 2008, and yes, all four ran into trouble the next year: Chris Johnson, DeMarco Murray, Le'Veon Bell, Christian McCaffrey. Four for four sounds like a law of nature. Run the arithmetic against that 61.5% base rate and four in a row happens 14% of the time by chance alone. You would need seven straight before it meant anything. There are four, and one of them is a contract holdout, not an injury.

Bottom line: the number of touches a back had last year tells you close to nothing about whether he survives this year.

What actually separates: the birthday

Now set those 82 workhorse seasons aside completely and look at the other 326, the ordinary top-24 backs nobody writes scary August articles about. Same seventeen years, same definition of trouble, split at 27 years old on September 1:

| | followed by trouble | | --- | --- | | 27 and older | 74.8% (77 of 103) | | Under 27 | 54.3% (121 of 223) |

Twenty points. The margin of error runs from +10 to +31, nowhere near zero, and the odds of a gap that size showing up by chance are under one in a thousand. Out of every comparison I ran, this is the only one that clears the noise, and it isn't close.

Twenty points is the difference between a coin flip and a bad bet. A top-24 back under 27 gives you roughly even money on a clean season. At 27 and up you're paying the same price for a three-in-four chance that it goes wrong, and "goes wrong" here means missed time, a big scoring drop, or a career quietly ending. Not all of it is injury.

And the best part? You don't need me for it. Touch counts require a spreadsheet. Age is printed on every player page on the internet, for free, and the market still doesn't discount it nearly enough.

The part where I got it wrong first

I have to own this one, because I nearly published the opposite.

My first pass tested age inside the workhorse group. Nineteen backs 27 or older with a 320-touch season, against 63 under 27. The older group came out at 58%, the younger at 67%, and I wrote down that age didn't matter and moved on.

Nineteen seasons cannot answer that question. The margin of error on that comparison is 25 points wide in both directions. A real effect could hide in there all day and never show up. When I ran the same split across the 326 regular top-24 seasons instead of the 19, age separated decisively.

I had read a null result from a sample too small to produce any other kind of result, and mistaken it for evidence that nothing was there. That is a mistake I will make again if I'm not careful, so I'm writing it down: a small group that shows no difference has not told you there's no difference. It has told you nothing.

So what do I do with a 30-year-old back?

Here's my live position, 2026 draft season, and it makes me look inconsistent on purpose.

I took Christian McCaffrey. Late first round, coming off 413 touches, going into his age-30 season. He is the only back on that 400-touch list twice, and both of his previous workhorse years ended with him hurt. Everything above says the age half of that is a real risk and the touch half is noise. I bought him anyway, because a top-five back in the late first is a price that pays for the risk. That's a value argument, not a health argument, and I want to be clear about which one I'm making.

And I'm out on Derrick Henry. Same age bracket, similar profile, opposite call. Which is worth explaining, because when I checked my own reasoning against the data, most of it fell apart.

The three reasons I fade Henry, graded

"Lamar takes his scoring chances." Wrong, and not a little wrong. Inside the ten-yard line last season, Lamar Jackson's share of Baltimore's opportunities fell from 10.8% to 6.5%. Henry's went up, from 45.9% to 51.9%. He handled more goal-line work in 2025 than in 2024, and only one back in the entire league saw more of it than he did.

"He kept getting stopped at the goal line." I can picture two of these clearly: the Detroit game, three cracks inside the three and nothing; the Rams game, stuffed from the one. Both are real. I checked them.

They are also two of the seven drives where he got multiple carries inside the five, and he scored on the other five. His failure rate on those drives was 28.6%. The league's was 50.0%. He converted repeated goal-line chances better than the average back, in the exact season I remember him failing at it.

Zoom out and it gets worse for my case. Inside the five in 2025, Henry scored on 37.0% of his carries. The league average was 37.4%. From inside the two he scored on 61.5%, against a league mark of 48.9%. Average in one spot, better than average in the other. And a full season of goal-line work is only twenty-odd carries, so at that size a thirteen-point swing is a coin landing differently anyway.

"The offense isn't what it was." This one holds up, and it's the big one. Baltimore scored 518 points in 2024 and 424 in 2025. That's 94 points gone, an 18% drop, and it is the biggest, most solid fact in this whole article.

So the fade survives, on half the reasons I thought I had. Henry plays this season at 32, on the wrong side of the only signal in this article that held up, in an offense that just lost 94 points. That's enough. The goal-line stuff was a story I told myself with two vivid plays and no denominator.

How to actually use this

Don't hear what I'm not saying. None of this means you should never draft an older back. I just did. Here's what it does mean.

1. Stop paying the workload tax. Last year's touch count is not a health forecast. If a heavy-usage back is falling in your draft because everyone read the same August article, that's your discount to collect.
2. Start paying attention to the birthday. Twenty points of risk, free, on every player page. It's the closest thing to an edge in this article.
3. Price the risk instead of avoiding it. Three in five top-24 backs hit trouble. If you're waiting for the safe one, you're going to be waiting in the eighth round. I took a 30-year-old in the late first because the price paid for the risk, and I'd do it again.
4. Don't trust the play you remember. I watched a full NFL season and confidently drew the wrong conclusion from two goal-line stands. If your case against a player is built from things you can picture, go count how many times the thing you're picturing didn't happen.

Every number here comes from 2008 to 2025 nflverse data. Two honest caveats. The same backs show up more than once, so my margins of error are slightly optimistic, which never rescues the results that came back null and makes the age finding strong rather than certain. And a 20% scoring drop partly measures regression from a career year rather than damage, which is why I say trouble "followed" a season rather than was "caused by" it. Both arms of every comparison were graded by the identical rule, so the gaps are fair even where the level reads harsh.