With the NFL regular season now over, let’s take a look back at the preseason predictions I made for each backfield around the league, review my thought process, grade the outcomes of those predictions (by the letter and by the spirit), and outline the lessons I learned through the course of the year. We’ll go alphabetically by team, starting with the:

Prediction:“Keaontay Ingram adds value in a reserve role behind [James] Conner…, setting himself up to be a sexy handcuff and zero-RB candidate in 2024 redraft leagues.”

Result: Woof. We now have a 62-carry sample of Ingram producing right around two yards per carry in the NFL, and while that’s generally not large enough to base any sound proclamations about a player’s ability off of, we also now have a two-season sample of Ingram getting passed up on the depth chart by guys who we have no reason to believe are anything but complete JAGs. The process for this prediction was based on my stance that Ingram’s 27-for-60 rushing line as a rookie was not enough evidence to move off my initial evaluation of him as a competent three-down player, and the runway for it coming to fruition could not have gone any better: Conner was averaging 17.5 touches per game and gaining more than five yards per carry when he sprained a knee in week five and missed the subsequent four games, opening up the door for Ingram -- who had operated as the team’s RB2 up to that point and received the start in Conner’s first week of absence -- to prove that he could contribute in the league. Instead, he was very quickly relegated in favor of Emari Demercado (though to be fair to all involved, this reshuffling of the pecking order happened on the heels of Ingram going 10-for-40 on the ground, catching both of his two targets, and earning a completely satisfactory pass-blocking grade from Pro Football Focus in that one start, so it’s not as if he Deon Jackson-ed himself out of a job), only touched the ball another 14 times in the next month, and was ultimately cut by the Cardinals. He probably just isn’t that good, and this prediction earns a big, fat F.

Prediction:“Bijan Robinson catches 50+ passes as a rookie.”

Result: Robinson ended up with 58 receptions. I’m sure I’ll dive deeper into his rookie year later this offseason, as I almost certainly did not nail every aspect of my Bijan-in-Atlanta fit analysis, but I’ll take an A on this prediction.

Prediction:Prediction: “JK Dobbins runs for over 1000 yards (or at least paces for it) but doesn’t provide much more than a neutral return on value at top-20 running back prices in seasonal leagues.”

Result: Dobbins got hurt in week one after having rushed for just 22 yards, so this one is basically void.

Prediction: “James Cook gains over 500 rushing yards and over 500 receiving yards.”

Result: Cook ended the year with 1122 rushing and 445 receiving yards, which is obviously not exactly what this prediction entailed, but it’s decently close if we’re going by the spirit of the law (which basically involved Cook being one of the better dual-threat running backs in the league). I’ll give myself a B here, and I think my major takeaway from this situation (and others like it around the league, as we’ll see) is that NFL coaches can really not be trusted to utilize their pass-catching running backs to their full potential. It’s hard to take issue with the Bills deciding to use their most explosive runner as their clear lead back (Cook out-carried the other backs on the team by nearly 100 attempts), and it’s also hard to take issue with them getting an excellent 8.2 yards per target out of that player in the passing game (the fourth-highest mark for any 20+ target runner in the league), but I can’t help but feel like a 281-touch workload -- of which nearly 85% were rushing attempts -- for a guy who weighs no more than 205 pounds and has a history of dynamic and versatile pass-catching does not represent optimal usage. I don’t want to scale back Cook’s contributions (and those of other players like him), I just want to reapportion them into something that is more sustainable and more uniquely tailored to his rare skill-set. I’m already not one to fall for the “he’s actually an offensive weapon” nonsense we hear every offseason, but I will be more jaded going forward in making assumptions about optimal usage for running backs with non-standard tool kits (I also realize that nobody but me is complaining about this Cook usage: he stayed healthy and ran for 1000 yards, the Bills have an effective running game, fantasy drafters mined nearly 14 PPR points per game out of a guy they acquired at mid-RB3 prices, etc.; the point stands, however, that I want my running backs used to the full extent of their unique talents, and does anybody really think that the optimal way to use a guy like Cook is by giving him the same carry/target ratio as 2014 Eddie Lacy?).

Prediction: “Miles Sanders runs for over 1000 yards but isn’t a difference-maker in redraft leagues, and Chuba Hubbard proves to be the more effective runner on a per-carry basis.”

Result: By a spirit-of-the-law interpretation, I nailed this prediction. Sanders fell far short of the 1000 yards I said he’d run for, but only because he stinks and was beaten out by Hubbard, who did prove to be the more effective runner and chewed up ground at a 1060-yard pace as this team’s starter from week six on. I’ll give myself a C here, and the lesson learned is to defer more to talent evaluations (at least the ones you can be confident in, such as Sanders not being a value-adding runner) than projections (which necessarily involve praying at the altar of median outcomes) when considering how to handle these running backs on a yearly basis (and that doesn’t mean treating talent evaluations as more important than projections themselves, but rather treating talent evaluations as more relatively important in that entire calculus than I treated them this year). It has long been a philosophy of mine to err on the side of your talent evals, but I think I lost the forest for the trees within the process of building out projections this season. Sometimes it’s okay to just say: “this player isn’t good and I don’t want him on my fantasy team even if I can’t tell you exactly how things aren’t going to break for him.”

Prediction: “This backfield is a fantasy mess in which no player is a reliable, weekly starter outside the context of an injury to either [Khalil] Herbert or [Roschon] Johnson.”

Result: I think this turned out to be true. Herbert had only four games all season with double-digit PPR points (and only two of those came in consecutive weeks), Johnson had three such games (none of them consecutive), and D’Onta Foreman had one such game. Gimme an A.

Prediction: “Joe Mixon is more frustrating than he’s worth for the second year in a row in fantasy.”

Result: I’ll let the Mixon managers weigh in on how frustrating he was this season, but I think I was off the mark here. After being drafted as the RB12 (in PPR and according to fantasydata.com), Mixon scored double-digit fantasy points in 13 of his 17 games, he scored at least 15 fantasy points in eight of 11 games following the Bengals’ week seven bye, and he finished the year as the RB11 and at 15.7 points per game. Pretty solid year, pretty bad (and lazy) prediction: I’ll give myself a D on this one.

Prediction: “Nick Chubb flirts with career-highs in the major receiving statistics but isn’t able to break out of the mid-RB1 range that he’s occupied in PPR leagues for his entire career.”

Result: Chubb got hurt in week two, so this one is void (but if we want to get really disgusting with it, he did have four targets for four receptions and 21 yards in just over five quarters of play by the time he went down; those numbers extrapolate out to a season-long receiving line of 54-54-286, which would all represent career-bests for the Chubbster).

Prediction: “Tony Pollard touches the ball at least 300 times.”

Result: Cha-ching. Pollard ended the regular season with 307 touches, and I will also point out that (while I didn’t expect him to experience such a slog on the ground) his efficiency fell off a cliff in very much the same way that I illustrated was possible for previously-dynamic runners stepping into high volume as veterans. I’ll take an A here.

Prediction: “Jaleel McLaughlin has at least one usable week in fantasy this season, even if it benefits DFS sickos more than redraft or dynasty players.”

Result: McLaughlin proved to be a competent receiver and effective runner this year, and he also scored ten or more PPR points on three different occasions (including a two-game stretch of 19.4 and then 17.9 points). I’ll take another A, this one from deep.

Prediction: “Jahmyr Gibbs becomes the sixth rookie running back ever to post 600+ yards as both a runner and receiver.”

Result: Gibbs finished with 945 rushing yards and 316 receiving yards, the second-lowest total for any of the 191 running backs who’ve earned at least 70 targets in a single season since the turn of the century. Gibbs’ abysmally low receiving efficiency is not something I’ve seen anyone talk about, and I’m eager to look further into his deployment and performance to find reasons for such low per-target output as this offseason progresses (for what it’s worth, I don’t currently believe those numbers are any reason to panic). I sort of straddled the line on this projection, but the median receiving yardage total for the nearly 200 players in the aforementioned group is 510, and 28% of them gained more than 600 yards through the air, so I don’t think I was completely off base in predicting the sort of passing-game numbers that Gibbs could have put together this season. I give myself a C here.

Prediction: “Nothing notable happens with these running backs in 2023.”

Result: This was a ridiculous and virtually meaningless prediction, but considering I have just as little to say about this backfield now as I did when I made the prediction, it was also probably correct. I’ll take a B on this one (and Packers’ decision-makers: can we please usher in a new era in this backfield? I love Aaron Jones, but let’s infuse things with some youthful juice here -- fourth-round MarShawn Lloyd would be a godsend).

Prediction: “Dameon Pierce is an offensive centerpiece who touches the ball 300 times in 2023.”

Result: This one is worse than the Keaontay Ingram prediction. After a four-year career at Florida in which he was perennially both effective and underutilized, and after a rookie season with Houston in which he posted quality efficiency numbers on high volume and despite abysmal offensive infrastructure, Pierce was maybe the worst running back in the NFL in 2023 (at least among guys getting regular and sustained playing time). I figured he’d continue to grow into a three-down role and that Devin Singletary would not amount to much more than a competent breather back, both of which proved false. Singletary basically turned into rookie-year Pierce (no really, look at their numbers), and as one just seven backs to gain fewer than 600 yards from scrimmage on more than 150 touches in a single season in the last decade (Miles Sanders and Kareem Hunt also qualified for such a group this year), Pierce basically turned into late-stage Peyton Barber. This prediction deserves an F-, but I will quickly point out that I was pretty dialed in on the non-Pierce factors that went into making it: given a re-tooled offensive line, a revamped receiver group, and the addition of CJ Stroud, I was all over the Texans’ offense being a much-improved unit in 2023.

Prediction: “Evan Hull supplants Deon Jackson as the team’s primary backfield steward before [Jonathan] Taylor returns to the lineup.”

Result: Hull got hurt on his second touch of week one, but Jackson was awful enough to lose his job by the time the Colts’ next game rolled around anyway. Getting beat out by a suddenly-effective veteran in Zack Moss isn’t the same as getting beat out by a speculative rookie in Hull, but I think I was on track enough here to earn a B.

Prediction: “Travis Etienne leads the backfield in goal-line carries and routes run but doesn’t finish as a top-ten running back (on a per-game basis).”

Result: Etienne ran more than 200 routes than D’Ernest Johnson and Tank Bigsby did combined, and he handled nine of the team’s 15 total rushing attempts from inside the five-yard line, but he also finished as the RB7 in PPR. I get a D on this prediction. I have already atoned for my sins against Etienne, but the greater lesson here is that I need to trust my evaluations. 2023 was the first offseason in which I incorporated film study into my scouting process, and while I believed at the time that I had put together a sound film-charting methodology, it was hard to be completely confident that I was going about things in a way that made sense -- both in the film study itself and in how much weight I was giving it relative to the weight given to numbers-based evaluative elements -- when I had no existing results and couldn’t really backtest my process. That lack of certainty resulted in some episodes of capitulation, one of which involved giving Bigsby a lot of hypothetical credit for maybe getting to do things in the Jacksonville offense that I didn’t really think he was going to be good at. With results for the 2023 running back class’ rookie season now in the books -- and pointing to pretty accurate evaluations of Bigsby, Gibbs, Robinson, De’Von Achane, Kendre Miller, Tyjae Spears, and Zach Charbonnet (as well as plausible deniability with guys like Zach Evans, Tiyon Evans, DeWayne McBride, and Evan Hull) -- I better understand how to balance my process and have much greater confidence in the film-charting aspect of it than I did a year ago, and my treatment of the 2024 class will reflect as much.

Prediction: “Isiah Pacheco finishes the season as a per-game RB3.”

Result: Pacheco finished the season as the RB14 in PPR, fueled largely by a late-season surge in which he averaged 21.8 points per game in four healthy appearances between weeks 12 and 17 (though the 12.7 points he averaged prior to that stretch would still have been good enough for an RB24 finish). He’s a funky runner, but he was legitimately good this season, and I think the lesson for me here is to be less confident in the irrelevance of any one or more players in an ambiguous backfield. Such a lesson is really an extension of the one I outlined in the Panthers section earlier: don’t get lost in the weeds of projections. It might be hard to configure an RB14 outcome from Pacheco (and more specifically one that makes workable sense as something resembling a median outcome) with the information available to you in the preseason, but that obviously doesn’t mean it’s not possible (I also don’t want to act like I would have thought such an outcome was impossible or that nobody else looking at this situation would have been justified in being high on Pacheco). Perhaps the lesson learned is to tinker more with projections as a way of exploring a situation’s range of possibilities: no need to marry yourself to the vanilla universe of median projections when you can tweak the inputs based on differing hypotheticals and consider the likelihood of a variety of resulting possible worlds. This prediction earns an F.

Prediction: “Zamir White carries the ball nearly twice as many times this season as all the non-[Josh] Jacobs backs in Las Vegas did in 2022 combined (that mark was 38, so let’s say averages at least four attempts per game and paces for a season-long total in the 70 range).”

Result: White finished with 104 carries, but he had only 20 before Jacobs went down with a quad strain in week 13 -- he definitely wasn’t “pacing” for a season-long total in the 70 range. I assumed the Raiders would want to lessen Jacobs’ workload a bit and that White would prove capable of providing that relief, and while the latter’s 84-for-397 stretch as the starter during the final month of the season gives me some vindication with regard to his competence, the idea that this team would scale back the duties of a guy who was a) coming off a 1600-yard rushing season and b) playing on the franchise tag probably wasn’t tenable. I’ll take my F.

Prediction: “Austin Ekeler beats his previous career-high in season-long rush attempts by at least 20.”

Result: Trying to thread the needle on the age cliff for each of these running backs is brutal. Ekeler looked almost completely toast at times this season, while other aged runners (like James Conner and Christian McCaffrey, even with their more extensive injury histories) put together career years. The former UDFA did pace for 217 carries, which would have represented an 11-carry improvement over his previous high-water mark, but some missed games on top of a sharp decline in effectiveness (that surely encouraged Joshua Kelley’s increased early-season workload) was enough to derail this prediction. I probably deserve a C here.

Prediction: “Cam Akers runs for over 1000 yards.”

Result: I’ll take the F, but let the record show that I was only 834 yards off the mark. The lesson learned here is similar to the Miles Sanders one about not getting lost in the weeds of median projections, but there’s also a less universal takeaway about allowing for maximum shenanigans to happen with coaches we know to be broody and unpredictable. If I was defending myself in bad faith (though I would never), I would point out that my assumption of there being good enough infrastructure in this offense to allow for a 1000-yard rusher was clearly true given the fantastic season Kyren Williams put together. I need to revisit his performance now that the year is over, and I’m pretty open to becoming Kyren-pilled after having pushed against the bandwagon for most of the last few months (see 1, 2, 3).

Prediction: “De’Von Achane will lead this backfield in opportunity share across the eight games following the Dolphins’ week 10 bye.”

Result: This one didn’t quite hit, as Achane’s 83 post-bye touches were fewer than Raheem Mostert’s 106 (each posted in six partially-overlapping games), but I think I did well by the spirit of the law here. The subtext of the prediction was that Achane would play well enough to establish himself as the backfield’s best player at least by the time the second half of the season rolled around, and his per-carry performance as perhaps the single best runner in the league certainly fulfills much of what I was getting at there even if Mike McDaniel continued favoring Mostert in the midst of a very healthy and very effective season. I’m taking a B on this one.

Prediction: “Alexander Mattison doesn’t crack the 12 points-per-game threshold in PPR leagues.”

Result: Here we see an example of me properly weighing player talent versus the rigidity of median projected outcomes in a way that I did not with players like Miles Sanders, Tank Bigsby, and Cam Akers. Sometimes you have to defer to the fact that a guy just isn’t good at football even when there doesn’t seem to be an alternate route through which to funnel fantasy points. Mattison averaged 8.3 points per game, so I’ll claim an A.

Prediction: “Rhamondre Stevenson earns over 80 targets and catches more than 60 passes for the second season in a row.”

Result: Stevenson finished with 38 receptions on 51 targets but paced for 17-game totals of 53 and 72, respectively: I was close, but a bit long. It’s clear in hindsight that I should have anticipated more of a role for Ezekiel Elliott, but Stevenson’s individual play falling off a cliff was not something I was predicting. I’ll take a C here.

Prediction: “Alvin Kamara scores fewer PPR points per game than Rachaad White does.”

Result: I’ve been annoyed at myself for having made this prediction all season, and not just because it turned out wrong (and it did, as Kamara averaged 17.9 points per game while White averaged 15.8). Most of these predictions are rightfully self-contained, but this one stupidly reached across teams and made itself unnecessarily dependent on multiple scenarios going the way I thought they would. We’ll get to White in a bit, but the Kamara part of this prediction fell apart not because he played better than I figured he would -- on the contrary, his per-carry average fell below the four-yard threshold for the second time in three seasons while his per-target average was the lowest of his career despite a career-best 87.2% catch rate -- but because I made incorrect assumptions about everything else going on in this backfield. That would also be fine on its own, but to center this process around a learnable lesson, it’s clear that those assumptions were based on capitulations on my own player evaluations: while I didn’t think Jamaal Williams was very good at this point in his career, I projected him for a sizable role in this backfield anyway, and while I didn’t have much confidence in Kendre Miller’s abilities, I projected him for an even greater share of the backfield pie. It was perhaps reasonable to be tentatively enthusiastic about Miller, but my default position should have been to regard the Saints as more likely than not to continue riding a diminished Kamara until the wheels completely fall off. I did correctly predict that White would be a useful fantasy asset, so I’ll take a D instead of an F on this overly convoluted prognostication.

Prediction: “Saquon Barkley eclipses 60 receptions for the first time since he was a rookie.”

Result: He caught 41 and paced for 50. Meh. Give me a D.

Prediction: “Breece Hall performs well enough in the latter part of the year that Bijan-versus-Breece is a legitimate debate in dynasty startups next spring.”

Result: There was a nine-game stretch in the middle of the year in which Hall averaged 2.53 yards on 99 carries, but after the bye in week nine (though this time period overlaps significantly with that midseason slog), he gained 893 yards from scrimmage, caught 53 passes, scored six touchdowns, and averaged over 20 PPR points per game. We’ll see where public opinion ends up, but I think Hall-versus-Robinson is a legitimate conversation (and I might defer to Breece). I’ll take an A here.

Prediction: “D’Andre Swift produces like a fringe-RB1 in the second half of the season.”

Result: This one is pretty similar to the Achane prediction in that I nailed the spirit of it but missed on the specifics of its manifestation. Swift did end up ascending to the top of this depth chart and producing like a fringe-RB1, he just did so in week two and through midseason (he averaged 14.4 PPR points per game in his first eight) rather than at midseason and through the end of the year (he averaged 9.9 points in his final eight games). Gimme a C+.

Prediction: “Najee Harris averages over four yards per carry and catches close to 50 balls (45+) this season.”

Result: Harris averaged 4.06 yards per carry but caught just 29 passes (and was only targeted 38 times). I missed here given that Jaylen Warren ended up settling in as this offense’s preferred option on passing downs, but I’ll take the vindication of seeing Najee run the ball well this year (though the down-to-down experience can still be frustrating at times!). He gained 537 yards at 4.22 per carry and averaged over 14 PPR points in the seven games after Matt Canada got fired on November 21st, and he finished near the top of the league in both RYOE per attempt and positive RYOE rate. Najee does not suck, and I’ll take a C on this prediction.

Prediction: “Christian McCaffrey gains 700+ yards both on the ground and through the air.”

Result: McCaffrey ended the year with 1459 rushing yards and 564 receiving yards, so as with my James Cook and Jahmyr Gibbs predictions, I lose out for having trusted too much in a coaching staff’s ability to fully utilize a versatile running back in the passing game (though Kyle Shanahan can hardly be criticized for the way he ran the offense this year). I will take another C for splitting the difference.

Prediction: “Kenneth Walker pulls a JK Dobbins by not providing value on top-20 prices in seasonal leagues despite rushing for over 1000 yards.”

Result: Walker finished as the RB20 in PPR points per game while rushing for 905 yards in 15 games and after having been drafted as the RB16 (according to fantasydata.com). This one nets me an A.

Prediction: “Rachaad White slogs his way to more PPR points per game than JK Dobbins, Miles Sanders, Cam Akers, and Kenneth Walker.”

Result: I don’t know why I was so drawn to involving other players in my White-centric predictions, but as Dobbins got hurt, Sanders lost his job, Akers got cut, and Walker averaged nearly two and a half points fewer than White’s 15.8 (posted despite a 3.64-yard per-carry average), this one ended up correct. Give me another A.

Prediction: “Derrick Henry finishes outside the top ten running backs in PPR points per game for the first time since 2018.”

Result: Give me yet another A, because Henry finished as the RB16.

Prediction: “Antonio Gibson is the anti-Kenneth Walker, catching over 60 passes but not cracking the top-24 running backs in PPR points per game.”

Result: Gibson did indeed not crack the top-24 running backs in PPR points per game, but he also didn’t catch 60 passes. I completely misread the tea leaves with this team, as I was content to plug Gibson into the “JD McKissic role” and also wrote that “[Brian] Robinson is a good player who doesn’t seem set up to produce like a fantasy starter while Gibson is also in the backfield.” Gibson fell short of the typical McKissic receiving contributions, Robinson was a solid RB2 for much of the year, and I probably deserve a D on this prediction.