Knowing when to fade stud running backs due to age-based deterioration is not easy, but it can save a fantasy season in redraft leagues and circumvent damaging losses of value in dynasty leagues. The conventional wisdom surrounding the running back age cliff is that it exists somewhere around the 26-year mark, and research from smart folks like Adam Harstad and Tej Seth has further clarified some of the guidelines we should use when evaluating players approaching that point in their careers.

Harstad’s research revealed that it’s probably better to think of running back career arcs as having a greater degree of flatness than the typical “age curve” language would imply, and that -- after experiencing a “mini-decline” in the season prior -- runners typically see a sharp and relatively sudden decrease in their performance at a point that becomes increasingly likely the older they become. In other words, we often get a canary-in-the-coal-mine season right before backs completely fall off the age cliff.

Seth’s research revealed some benchmarks to look out for as we try to identify when those age cliffs might occur, with two major points revealed to be the end of a running back’s rookie contract (typically in year four) and their crossing the 1500-carry threshold in the NFL.

Using those insights, I want to explore the degree of possibility that exists for each of the most relevant older runners currently in the league to fall off the age cliff in 2023. The players in question will be each of the runners who a) will be 26 or older when the NFL season starts this fall, and b) are currently ranked in the top-50 of dynasty running backs over at KeepTradeCut, with the following players satisfying both conditions (in order of KTC ranking): Christian McCaffrey, Saquon Barkley, Austin Ekeler, Nick Chubb, Tony Pollard, Derrick Henry, Miles Sanders, Aaron Jones, Joe Mixon, David Montgomery, Dalvin Cook, Alvin Kamara, James Conner, Damien Harris, Rashaad Penny, Jamaal Williams, and Samaje Perine.

My methodology was pretty simple. I first looked at the performance of each player in each season of their career in each of various rushing metrics: Box-Adjusted Efficiency Rating, Relative Success Rate, Rushing Yards Over Expected per Attempt, missed tackles forced per attempt, yards after contact per attempt, and raw yards per carry. I then assessed the level of alarm that could reasonably be assigned to those respective performances based on recent results and apparent trends, and then quantified that alarm on a subjective, 0-3 scale, where 0 = no alarm, 1 = slight alarm, 2 = reasonable alarm, and 3 = considerable alarm. I did the same exercise based on recent history of injury and missed games for each player, and then totaled the results for each player in each metric to create an overall alarm “rating” that was then subject to small modifiers based on Seth’s rookie contract and 1500-carry thresholds (which really only affected Henry, as all the players considered in this study are beyond their rookie contracts while Henry is the only one who has eclipsed the 1500-carry mark). Players with the highest total scores are those I’d consider to be the most likely to fall off the age cliff in 2023, but let’s start at the bottom.

Two players in this study received a 0 in every metric I looked at and have also not suffered worrisome injuries or missed an alarming number of games in recent seasons: Aaron Jones and Nick Chubb.

Jones will turn 29 in December, but he’s coming off his fifth-straight season of at least 14 PPR points per game and has not experienced any significant decline in his performance on a play-to-play level. He averaged over five yards per carry for the fourth time last season while posting the highest rate of missed tackles forced per attempt (0.25, which ranks as the fifth-highest single-season mark out of 92 combined seasons from runners in this group) of his entire career, and his marks in both BAE Rating and RYOE were above the 80th percentile among league-wide runners since 2016. The only recent indication of decline we’ve seen from him was a 28th-percentile RSR of -5.1% from the 2021 season, but he returned to the positives in that area with a 54th-percentile mark of 0.4% last year after his 2019 and 2020 marks also landed near the league mean, so it seems likely that the down performance there in 2021 was a blip. He’s not immune to falling off, but Jones hasn’t given us any reason to suspect he’ll be a significantly lesser player in 2023 than he has been in recent seasons.

Aaron Jones is a former fifth-round pick who will finish the upcoming season at 29-years old, but he still one of the best running backs in the league.

Chubb’s resumé is even cleaner. He’s never averaged less than five yards per carry, he’s never posted a BAE Rating below the 65th percentile (and 2022 saw him produce a 137.0% mark that lands in the 90th percentile and ranks as the second-best of his career), he’s never posted RYOE numbers below the 75th percentile (he finished in the 87th a year ago), and his missed tackles forced and yards after contact figures are always near the top of the league (his forced 0.27 missed tackles per attempt last season, the second-highest mark of his career). Even his RSR numbers, which had settled in the -2.4% to -4.1% range from 2019 to 2021, saw a strong resurgence in 2022, as his 4.3% mark landed in the 72nd percentile. Chubb is playing as well as he ever has entering 2023.

The other guy whose recent numbers earn him a place in the “peak performance” category is Tony Pollard, whose most alarming single-metric performance from last season was the career-low RSR of -1.1% (46th percentile) that he posted after hovering right around the 1.0% mark in his previous three seasons. That’s not a very significant drop-off, and considering that it came under a substantial volume increase and alongside improvements in each of BAE Rating, RYOE, MTF per attempt, and YAC per attempt from 2021 to 2022, I’m not worried at all about the strangely-old Pollard experiencing an age-based decline in 2023.

While Pollard maxed out our previous group with an alarm rating of 1, this group of five likely-not-washed runners earned between 2 -- Austin Ekeler -- and 3 points -- Jamaal Williams, David Montgomery, Damien Harris, and Rashaad Penny -- on our scale.

The 28-year old Ekeler doesn’t show many signs of slowing down, as 2022 saw him post his highest marks since becoming a regular starter in 2019 in both BAE Rating and RYOE (with numbers in the 91st and 84th percentiles, respectively), and his MTF, raw YPC, and RSR numbers from last season were all right in line with his career averages (which is especially impressive in the case of RSR, where he’s been above the 74th percentile in each of his last three seasons). Slight alarm does come in the form of a five-year slide in yards after contact per attempt, however, as his most recent mark of 2.92 ranked 22nd among runners with at least 100 carries last season. Still, Ekeler has missed only one game in the last two years, is nowhere near the 1500-carry threshold, and has performed up to his own high standards in pretty much every other measure of play in recent seasons. The canary in his coal mine is still kicking.

The only reason Penny finds himself in the “probably fine” group rather than the “peak performance” group is his troublesome injury history, because he certainly hasn’t started playing at anything below an elite level when active. He’s averaged over six yards per carry, gained more than four yards after contact per carry (he was first in the league in this area last season), gained more than a full yard above expected per carry, and forced 0.23 missed tackles per attempt in both of the last two years, and he posted the highest and second-highest marks of his career in RSR and BAE Rating, respectively, in 2022 (numbers in the 72nd and 92nd percentiles). Penny is always one of the few best ball-carriers in the league on a per-touch basis, but it’s also true that the injuries that make him difficult to trust in fantasy football also make him a candidate for a sudden decline with little on-field notice.

New team, same story: Rashaad Penny’s injury history poses a lot of risk, but he could smash if he manages to stay healthy for a legitimate stretch of time this season.

Harris is the first player in this study who earns ratings representing at least “slight alarm” in more than one of the metrics in question: he’s now strung together two straight seasons of declining marks in raw yards per carry (though his 4.40 average from 2022 was still above the league mean), and his rate of missed tackles forced per attempt dropped off a cliff last season, going from 0.14 in 2020 and 0.18 in 2021 to 0.09 in 2022 (the lowest among any player in this group in the last three years). His marks in BAE Rating and RYOE have been remarkably steady in his three seasons of legitimate work, and his RSR and after-contact numbers both were higher in 2022 than in 2021, but his recent struggles to remain healthy (he missed six games in both 2022 and 2020) combined with some performance dips in key areas should keep us attentive to further signs of decline.

The other two guys in this category are here not because they’ve been especially good in recent years, but because they’ve been largely underwhelming for most of their careers and then experienced sharp declines in one efficiency metric each last season. Montgomery’s 2022 marked the second straight year with a steep drop-off in BAE Rating, as he went from 120.3% in 2020 to 93.0% in 2021 to 85.9% last year (bringing him from the 77th to the 27th percentile), and he hasn’t been impressive according to either raw yards per carry or RYOE since he entered the league. The things he did well in 2022 are break tackles -- he forced 0.23 missed tackles per attempt last season -- and churn out positive yardage on a consistent basis -- his 6.2% RSR landed in the 79th percentile -- and it’s in that sort of reliable grinder vein that I anticipate he’ll add value on the downward slope of his career. He might not be washed, per se, but he’s now a decidedly low-ceiling version (at least on a down-to-down basis) of the already low-ceiling player he’s always been.

The other player who fits that general description is Williams, who has seen his RYOE decline in each of the last two years and was most recently a 26th-percentile performer in that area. Unlike Montgomery, Williams’ MTF and YAC numbers have always been pretty low, and with a career’s worth of BAE Ratings that have never even sniffed the 100% mark, Williams will have to rely on his always-solid RSR numbers to add value. Just like Montgomery, Williams’ age has turned him into a lower-ceiling version of a player who already had a low-ceiling style of play.

Neither of the players in this category scored more than a 1 (indicating “slight alarm”) in any one category, but they each earned that score in three different metrics while also having some mildly concerning recent injury history.

The first of these elite talents is Saquon Barkley, who had a stretch of health struggles in 2020 and 2021 that are hopefully in the rear view after a relatively uneventful 2022 season. Barkley’s return to health saw him bounce back from atrocious efficiency numbers to post solid marks in each of BAE Rating, RSR, and raw yards per carry last season (all between the 63rd and 66th percentiles), and while they didn’t quite reach positive levels, his YAC and RYOE numbers also jumped up a bit from where they were prior. It’s a lot to ask of an aging, athleticism-reliant player to completely return to the pre-injury heights we saw early on in his career, but it seems likely that Saquon will be at least a competent runner for the immediate future.

The second of these guys is Christian McCaffrey, who famously missed most of both 2020 and 2021 before playing all 17 games last season. There aren’t any massive on-field red flags for him either, but some slight concern is raised by the fact that his BAE Rating, RSR, and RYOE marks all saw significant drop-offs from 2021 to 2022. His -2.8% RSR lands in the 36th percentile and is the lowest mark of his career, while his 101.3% BAE Rating and 0.01 RYOE per attempt numbers are both barely above their respective baselines and indicate replacement-level output on a per-carry basis. He was solid from a through-contact perspective, though, with YAC and MTF numbers that were right in line with how he performed during his heyday. I ultimately don’t think he’ll have a hard time producing efficiently in the San Francisco offense (and much of his value comes from the versatility he lends to the unit as a whole anyway), but it’s worth keeping an eye on CMC’s team-relative and expectation-derived numbers in case a downward trend begins to develop.

The two players in this category -- Joe Mixon and James Conner -- have each had their share of injury issues in the past few seasons, but the real issue is the substantial decline that they’ve both seen in multiple rushing metrics.

For Mixon, those declines have come in BAE Rating and RYOE, the best context-laden measures we have of relative rushing performance. While his raw numbers haven’t been good in recent seasons (he’s maxed out at a 49th-percentile per-carry average of 4.13 yards going back to 2019), he’s been below the 100% mark in BAE Rating in each of the last three years and in the negatives in RYOE in two of them (2022 and 2020). He’s also been a relatively ineffective tackle-breaker recently, as his YAC and MTF numbers ranked worst and second-worst, respectively, in this group of aging backs last season. Really, Mixon’s biggest value-add of late has been as a chain-mover, as he’s posted positive RSRs in every season of his career and managed marks above the 66th percentile in both of the last two years. If he remains with the Bengals in 2023, he’ll be in line for solid, volume-fueled production on a great offense, but you should not be counting on Mixon for situationally-agnostic fantasy utility at this point in his career.

Somehow, the dual-threat alien that Joe Mixon was at Oklahoma has become a plodder in the NFL.

Conner shocked me by putting together a really strong fantasy campaign in 2022, because he was downright bad by many measures of running back play the year prior: he averaged just 3.72 yards per carry, he ranked 43rd among qualifying runners in YAC per attempt, and his 21st-percentile BAE Rating of 81.8% suggested that his per-carry struggles were not just a product of the poor offensive situation he found himself in. Last season, though, he bounced back in a big way, averaging a solid 4.24 yards per carry and jumping back over the 100% mark in BAE Rating while posting a career-best and 82nd-percentile figure of 7.0% in RSR. The washed risk with Conner comes from the litany of minor injuries he’s dealt with in recent years on top of a two-year decline in MTF numbers (from 0.21 to 0.19 to 0.16) and the possibility that his bounce-back 2022 season was a temporary blip of solid play in the midst of a larger and more legitimate falling off process that started in 2021. Unless he completely deteriorates, however, Conner should continue adding value as a Mixon-, Montgomery-, and Williams-style grinder who can reliably keep his offense on schedule.

Now we come to the red alert portion of the programming, in which four players scored a total of 6 points on our alarm rating scale, giving them each an average of 1 point per rushing efficiency category -- not good! Each of Derrick Henry, Samaje Perine, Dalvin Cook, and Miles Sanders came by their high-risk status by slightly different means, though, so let’s examine them one-by-one.

I sounded the alarm on Henry last offseason, when he was a 28-year old coming off an injury-shortened season that saw him post the lowest marks of his career in BAE Rating and RSR, as well as the second-lowest marks of his career in RYOE per attempt, raw yards per carry, and missed tackles forced per attempt, in addition to the third-lowest mark of his career in yards after contact per attempt, with each of those non-career-lows representing the worst he’d performed since he became a full-time starter in 2018. Instead of fall off, though, Henry looked quite a bit like the beast who we came to appreciate as the best runner in the NFL in 2019 and 2020, as his BAE Rating, RSR, MTF, and YAC numbers each returned to pre-injury levels. From some perspectives, however, Henry’s 2022 mirrored his 2021, as his 4.41-yard per-carry average looked more the previous season’s 4.28 than it did the 5.16 that he averaged between 2018 and 2020, and his 0.06 mark in RYOE -- while still solid -- was even lower than 2021’s 0.09 and a far cry from the 0.45 that he posted in his prime. Overall, though, last season’s bounce-back would be enough for me to look past a down 2021, except for the fact that Henry crossed the dangerous 1500-carry threshold early last year and is now 250 attempts beyond it. He still carries high-end upside, and maybe he’s one of these guys who just continues to defy our expectations for what normal, human athletes are supposed to be capable of, but you’re not allowed to be surprised if King Henry’s reign comes to an abrupt end in 2023.

Perine was a surprise to me this high up the list, as I generally think he’s a quality, reliable, jack-of-all-trades back who could be a much bigger factor in the Denver than his current ADPs suggest, but he regressed in four different categories last season: BAE Rating, raw yards per carry, MTF per attempt, and YAC per attempt. Basically, he wasn’t dealing with contact well, and his overall efficiency suffered as a result. These haven’t all been minor dips in performance, either, as Perine’s BAE Rating dropped from the 89th to the 63rd percentile (even as Joe Mixon was less efficient), his YAC number dropped more than three-quarters, from 3.86 to 3.07, and his raw yards per carry dropped by 0.32 (from the 65th to the 50th percentile) and for the third straight season. His MTF number technically went up by .01, but last season’s 0.15 marked Perine’s second season into a substantial fall-off from three years’ worth of 0.28-level play from 2018 to 2020 (and the 0.29 rate that Perine put together in 2020 is the second-highest single-season mark among backs in this study, behind only Chubb’s 0.30 from the same year). Still, these numbers aren’t actively bad, and it’s clear that Perine is still an effective player. But, as Harstad said, running backs “tend to experience a “mini-decline before they fall off the cliff entirely,” which is exactly what happened with Perine in 2022. Tread carefully.

Cook’s high alarm rating is both unsurprising and strangely sobering: we know he’s not the guy he was, but I hadn’t really grasped the direness (to be a bit melodramatic) of the situation, a factor that adds an interesting dynamic to Cook’s free agency (I’m writing this before I go on vacation, it’s currently June 22nd and this publishes on July 2nd, so Cook may not be a free agent anymore). Of course we can’t be positive that he’ll turn into a shell of himself after what then would be a swan-song 2022, when he did experience some stark drop-offs in a couple metrics. His raw yards per carry followed up a 4.99-to-4.65 dip the year prior with a 4.65-to-4.44 dip this year, and while that’s still a 64th-percentile mark, but some of the more contextualized and metrics indicate an underlying dip in performance. Cook’s RSR was negative for the first time in his career, dropping from a 68th-percentile 3.7% to a 39th-percentile -2.3%, and his RYOE per attempt dropped like a rock, going from -0.06 in 2021 to -0.75 in 2022, landing in the 21st-percentile last season. His BAE Rating stayed afloat because Alexander Mattison wasn’t very good, but it went down a bit too. Seeing severe drop-offs in both rate- and average-based stats strikes me as a bad sign, and Cook’s still-good marks in the MTF and YAC categories indicate either that he’s fighting for every yard and just not getting as far as he used to, or that he’s still got it physically and has seen regression in other areas due to some factor other than approaching the age cliff. That feels like a bold bet to make regarding any soon-to-be 28-year old runner, so I think we have to be prepared for the possibility that Cook doesn’t pose quite as big a threat to whoever the incumbent new teammates he ends with as a prime or near-prime Cook would, and -- more importantly -- to accurately calibrate expectations for the various effects that such a phenomenon could have.

Perhaps even more surprising than Cook’s placement on this list is the placement of Sanders, who had easily his most productive season in 2022 while also falling mostly in line with the raw efficiency numbers he’d posted thus far in his career (he dropped from 5.08 to 4.96 yards per carry). However, his BAE Rating and RYOE numbers both fell, the former just slightly from the 77th to 69th percentile, but the latter precipitously, falling a full yard from 0.88 to -0.12. Such a combination raises my suspicions of a subtle Sanders regression being masked by still-high raw efficiency marks kept afloat by a still-dominant Philadelphia offensive line. A two-year drop in RSR (he’s now in the 38th percentile) suggests that Sanders is having an increasingly difficult time churning out positive outcomes, something that shouldn’t be an issue behind a run-blocking unit that ranked fifth and third, respectively, in the NFL in PFF’s grading system in the the last two years. That downward trend in play-to-play consistency has coincided with a stark drop-off in YAC per attempt in the last two seasons, as Sanders went from numbers of 3.28 and 3.38 as a rookie and sophomore, respectively, to a collective 2.90 in 2021 and 2022. He’s still breaking as many tackles as he ever has, he’s just not making as much happen afterwards, with what really amounts to an across-the-board decline in measures of play (his yards per target also dropped from 4.6 to 3.0 last year). Sanders is set up for a nice season in Carolina, but no longer having the juice is a possibility as well.

He’s one of my favorite players, but I think Alvin Kamara is already washed. The only of these six metrics that he did not regress in last season was raw yards per carry, where he posted a 3.98-yard mark that stands as his second-straight below the 4.0 threshold after a 4.97-yard career average prior to that. Otherwise, he posted respectable numbers while seeing moderate dips in BAE Rating and RSR (from the 72nd to 65th percentile, and from the 71st to the 66th percentile, respectively), but experienced significant drop-offs into actively-bad territory in the remaining metrics. He forced only 0.15 missed tackles per attempt last season (which is below the seasonal average of 0.17 among this group) after doing so 0.21 times per run from 2019 to 2021, his YAC per attempt numbers continued a now-three-year slide and ended up ranked 49th out of 69 qualifying backs last season, and his RYOE followed up its 0.80-yard drop to -0.25 from 2021 with a 0.24-yard drop to -0.49 in 2022, a mark that lands in the 33rd percentile. His biggest strength as a runner when he was in his prime was his contact balance, and he’s now no longer dealing with contact well, so it makes sense that Kamara’s overall efficiency and consistency as on the ground have also suffered. He’s still been productive in fantasy, but his underlying metrics have looked like those of a washed up running back in recent seasons, and he now has a potential suspension to deal with. I wouldn’t be shocked if we’ve seen the last of Kamara’s fantasy utility.