The Running Back Age Cliff Is Real, and It Starts a Year Earlier Than You Think
The claim everyone repeats
“Never draft a running back over 30.” “The cliff is at 28.”
You have heard both, usually from someone who has not looked at the numbers and occasionally from someone who has looked at the wrong ones.
We pulled eleven seasons from 2015 through 2025 using nflverse weekly scoring joined to NFL rosters for birth dates, then measured running back aging two different ways. The two methods tell very different stories, and understanding why is the whole point.
Method one, points per game by age
This chart lies.
Take every running back season with at least eight games played and average full-PPR points per game by age.
| Age | RB seasons | Mean PPG | Index vs. peak |
|---|---|---|---|
| 22 | 78 | 9.88 | 1.00 |
| 24 | 147 | 9.27 | 0.94 |
| 26 | 128 | 8.61 | 0.87 |
| 27 | 97 | 9.72 | 0.98 |
| 29 | 52 | 8.98 | 0.91 |
| 31 | 21 | 9.09 | 0.92 |
Read that literally and there is no cliff at all. Twenty-seven-year-olds score about as much as twenty-three-year-olds. Thirty-one-year-olds score more than twenty-six-year-olds. Startable rate tells the same story. The share of running back seasons finishing RB24 or better is actually highest at age 27, at 27.7%.
The problem is survivorship bias, and it is brutal at running back.
A 31-year-old back only gets eight games if he is still good enough to be on the field. The bad ones have already disappeared from the sample because they are no longer getting meaningful work or are no longer on rosters.
That is why an age chart built from raw points per game can make it look like running backs barely age at all.
Method two, follow the same player
Here is where the cliff shows up.
The better question is not “how do 28-year-old running backs score?” It is “what happens to this running back next year?”
So we followed the same player from one season to the next, required at least eight games in both seasons, and measured the change in his points per game.
| Age in year one | n | Median change | Share who declined |
| 22 | 55 | +3.9% | 45.5% |
| 23 | 87 | -2.3% | 50.6% |
| 24 | 97 | -2.7% | 51.5% |
| 25 | 93 | -7.1% | 63.4% |
| 26 | 62 | +0.4% | 50.0% |
| 27 | 52 | -22.3% | 71.2% |
| 28 | 43 | -16.1% | 60.5% |
| 29 | 24 | -13.8% | 62.5% |
| 30 | 17 | -30.4% | 70.6% |
| 31 | 10 | -29.8% | 90.0% |
Three things stand out.
The real break is the 27-to-28 transition. Not 28 to 29, which is where a lot of rankings seem to put it. A running back playing his age-27 season loses a median 22% of his per-game production the following year. From 23 through 26, the median change stays within roughly plus or minus 7%, which for fantasy purposes is pretty close to flat.
Age 25 is a mild speed bump, not a warning. A median decline of 7% with 63% of players getting worse is worth a small adjustment. It is not a reason to pass on a good running back.
Past 30, the floor falls out. The median decline is around 30%, and nine out of ten backs entering their age-31 season got worse. The handful who survive and keep producing are exactly why the first chart looks much healthier than reality.
What this means on draft day
Do not fade a 26-year-old. He is still inside the flat part of the curve. If your board says he is the best value available, take him and stop apologizing for it.
Apply the real caution to the age-27 season and what comes next. The back who is 27 this year is the one to be careful about paying full price for in dynasty, especially if his value is based heavily on last year’s production. In redraft, the age-27 season itself is usually still fine. The bigger concern is what happens after it.
Do not turn this into a hard rule. A 10% projection haircut on a back who is 60 points clear of the next tier still leaves him well ahead. Age should help break ties inside a tier or separate two similar players. It should not make you jump two tiers.
Ignore the 30-year-old at market price. Not because he cannot produce. Some of them do. The problem is that the market is usually pricing him off last year’s production while the data says the odds of decline are overwhelming. If he falls three rounds past that price, he is a perfectly reasonable bench swing.
Who this is about in 2026
Ages below are as of September 10, 2026, from nflverse roster birth dates. Prices are twelve-team PPR ADP. This is a price question, not a talent question.
Targets – inside the flat band and priced behind an older back
| Player | Team | Age | ADP | The read |
|---|---|---|---|---|
| Javonte Williams | DAL | 26.4 | 37 | Inside the flat band and going after Josh Jacobs (28.6, ADP 32). Same range of outcomes, one of them has the year-over-year data on his side. |
| Tyjae Spears | TEN | 25.2 | 138 | Tony Pollard (29.4) costs 75. Spears is the cheapest real path to a Tennessee workload and the younger side of the same backfield. |
| Bucky Irving | TB | 24.1 | 52 | Cheapest lead back who is nowhere near the break. Nothing in the aging data argues against paying this. |
| Jordan Mason | MIN | 27.3 | 114 | Priced identically to Aaron Jones (31.8, ADP 113) in the same backfield. If the two are coin flips on role, the aging split is not a coin flip. |
| Tank Bigsby | PHI | 24.0 | 155 | Contingency price behind Saquon Barkley (29.6). You are buying the age gap, not projecting a starter. |
| RJ Harvey | DEN | 25.6 | 90 | Level price with J.K. Dobbins (27.7, ADP 93). Flat band versus the season right before the break is the tiebreaker. |
Careful at market price – 30 and up, priced on last year
| Player | Team | Age | ADP | The read |
|---|---|---|---|---|
| Christian McCaffrey | SF | 30.3 | 6 | The most expensive place on any board to bet against a nine-in-ten decline rate. He may produce. The price gives you no room if he does not. |
| Derrick Henry | BAL | 32.7 | 25 | The exception the aggregate chart is built out of, which is exactly why the market keeps paying. Fine if he falls, not at RB12 money. |
| Saquon Barkley | PHI | 29.6 | 18 | Entering the range where the median same-player change turns sharply negative. First-round price, no discount for it. |
| Tony Pollard | TEN | 29.4 | 75 | Priced on volume that a 25-year-old on the same roster can take. |
| Aaron Jones | MIN | 31.8 | 113 | Past 31 the decline rate is 90%. There is a younger back at the same price on the same team. |
Fair, not a trap — Josh Jacobs (28.6, ADP 32), Rhamondre Stevenson (28.5, ADP 78), and Rico Dowdle (28.2, ADP 82) all sit in the age-28 group, median -16%. That is a haircut on the projection, not a reason to move them down a tier.
What this data does not say
It does not say old running backs are bad players. It does not support a “cash out at 26” dynasty rule, because the flat portion of the curve still runs through 26 and the biggest drop begins after 27.
It also does not tell us whether an individual player’s decline came from aging or from losing opportunity, because points per game cannot fully separate the two.
What it does say pretty clearly is that the year-over-year drop starts to show up at the 27-to-28 transition, and that raw points-per-game age charts can hide the decline because the older backs who remain in the sample are the survivors.
Responses