Inspired by the logical framework laid out by the cool, kind, and very smart Adam Harstad in this thread that describes his process for ranking wide receivers in dynasty, I decided to overhaul the previously very loose methodology by which I form my own dynasty running back rankings. This article won’t reveal what the resulting changes to those rankings were (you can see those for yourself right here), but sweeping changes to the formational mechanism of a product that people pay me for access to seems to warrant transparency, so what this article will do is describe and justify the reasons for those changes to that mechanism. At the end, I will also give a quick run-down of a few players who I identify through this process as being particularly appropriate buys or sells in dynasty leagues given a disparity between how I believe they should be valued and how the market is currently valuing them (if you don’t care about the rankings process and just want to see those recommendations, scroll down to the picture of Breece Hall).

The justification part of this equation is easy: as a companion to the best running back evaluations on the planet, I want this website to also offer the best running back rankings on the planet. My process for putting rankings together (specifically the dynasty rankings, as the fact that the players being ranked for devy formats and rookie drafts are necessarily non-lineup-contributing assets makes the thought process behind those rankings fundamentally different) so far has been to basically fire from the hip my intuitions about how you should value running backs, and while I think those intuitions are and have been good, they could stand to benefit from a structured process that helps guide them in a consistent and (more) objective way. I’ve held this belief for the entire time that I’ve been putting rankings together but have never really landed on a process that felt good and sensible enough to me to actually implement, but then I stumbled upon Harstad’s thread.

I won’t rehash all of it here, but what I took away from his thought process is the great value in devising a way to quantifiably consider the dynamic that exists between players’ current levels of production and expectations about how they might produce in the future. Harstad uses Bob Henry’s rest-of-season projections (which I have never used but Harstad says are demonstrably very good) as a way to quantify the former of those factors, while defaulting to player age as a measuring stick for the latter. Using those resources in tandem enables Harstad to categorize players (especially those beyond the clearly interesting players at the top of the rankings, where things get murkier) as either “producers” or “prospects”, distinctions that are helpful for determining the rosterability of individual players in the context of specific dynasty teams:

I love that categorization framework and have implemented it into the mechanism I’m describing in this article, but with some tweaks given the fundamental differences between how wide receivers and running backs come by their production in the NFL (and, by extension, in fantasy football). Because targets are earned within a competition between all five of the eligible receivers on the field on each play (most of whom are wide receivers), the most productive wide receivers (at least in the context of each offense -- passing volume, quarterback efficiency, and other environmental factors obviously have a large influence on production when we compare players cross-situationally) are those that command the most targets via their own merits (whether that’s route-running skill or catch-point dependability or rapport with their quarterback or whatever). It’s very rare -- almost unheard of, even -- for a wide receiver to simply be given targets without earning them on the field.

Running backs are different. The vast majority of their opportunity comes in the form of carries, which are not earned in the same way that targets are (and even most running back targets are schemed more than they are earned like wide receiver targets). It’s of course correct to say that running backs earn their carries by earning their way onto the field, but once the play is called, the running back is either getting a handoff or he’s not, and there’s nothing he (or the other running backs on the team, even if one or more of them also happens to be on the field) can do about it either way. Because of that, we far more often see wild fluctuations in the week-to-week opportunities given to specific running backs, and we know that those fluctuations don’t mean that the affected players are fundamentally volatile performers in the same way that fluctuations in wide receiver opportunity would indicate. Obviously, those fluctuations simply mean that running backs are periodically thrust into more opportunity or scaled back into less opportunity than they receive in the median circumstance, largely by means of injuries to themselves or other players in the backfield.

Because of all this, how a given running back is currently producing in fantasy football -- and how rest-of-season projections indicate we should expect him to continue producing in fantasy football in the short-term future -- is not as dependable an indicator for how we should expect him to continue producing in fantasy football beyond the short-term future as those things would be for wide receivers. A wide receiver who is currently turning a 25% target share into WR2-level production in fantasy can reasonably be called “good” and thus expected to do approximately the same thing next year, but a running back who is currently turning a 65% opportunity share into RB2-level production might just be doing so because the guy who normally plays ahead of him is hurt. Conversely, a wide receiver who is either not getting on the field or playing sparingly and doing little with his opportunities is probably just not good and can therefore be expected to continue not producing (with obvious caveats for rookies getting acclimated to the league, etc.), while a running back who is receiving few opportunities cannot be so readily written off as either not good or incapable of producing in the future (because we know that running backs who are not good can produce in fantasy football given the requisite opportunity).

None of that was profound, but it means that classifying running backs so cleanly as either producers or prospects is not as straightforward or as informative as with wide receivers. Some consideration must be paid to the possibility that a non-producing running back who also isn’t an unknown commodity in the way that a “prospect” is could offer utility to your dynasty team by simply (and somewhat randomly) being thrust into opportunity, and the calculus therefore is not as simple as “prioritize the non-producers that you think are the best players”. We’ll get back to this consideration in a second.

First, though, let’s return to the idea of quantifying current production and expectations for short-term future production. I don’t have a go-to set of rest-of-season projections like Harstad does (maybe I’ll find one in the future, who knows), but I am able to quantify my intuitions for how players can be expected to produce for the remainder of 2023 given how they’ve been producing (and playing) thus far in the season, my impressions of their talent, and how their situation can be expected to either change or stay the same throughout the rest of the year. To those ends, I’ve established a 9-point scale on which to categorize my expectations for player production in the short term, with the following question as my guide: “What sort of fantasy value can this player be expected to offer for the rest of the current season?”

The scale is thus:

Most of these categories are self-explanatory, but let me illuminate those shrouded in greater mystery. The “flex / DFS” category is for players who are not quite weekly starters (in 12-man, 2WR-2RB-1FLEX leagues) but could offer matchup-based utility in lineups. The “contingent” category is for players who need something external to occur (an injury to a teammate, for example) in order to qualify for one of the higher-rated categories. The “speculative” category is for players who probably would not directly benefit from a single external occurrence in the way that a player in the “contingent” category would, but who are on a roster and could conceivably get on the field at some point. The “playing but worthless” category is for players who are currently receiving on-field opportunity but a) are doing so little with it that they might as well not be receiving that opportunity at all, and b) do not stand to meaningfully benefit from an external occurrence in the way that a player in the “contingent” category does.

Here is a (non-comprehensive) sampling of players who I determined fall into each of these categories:

Clearly, categorization for any individual player is subject to change. If I had done this exercise prior to week one, AJ Dillon would have been classified as a 3 (at worst a 2), De’Von Achane would have been classified as a 2 or 1, and Nick Chubb would have been classified as a 6. If Christian McCaffrey suffers a season-ending injury in practice tomorrow (obviously I hope that doesn’t happen), he’d fall to a 0, Elijah Mitchell would jump from a 2 to at least a 4, and Jordan Mason would jump from a 1 to a 2.

You’ll also notice that these classifications are fundamentally constrained in the amount of players eligible for any one of them. While I won’t necessarily always categorize three players as 7s, it wouldn’t make sense to categorize more than three as 7s, and so on based on the seasonal ranking ranges and depth-chart placement that most of the categories refer to.

Now let’s return to the process of quantifying expectations for production in the long-term future. To account for the differences I delineated earlier between running back opportunity and production and those of wide receivers, I’ve decided to take a more talent-in-a-vacuum-based approach to my expectations for future running back production than Harstad did for wide receivers, and strictly speaking the following methodology does not refer directly to either “production” or actual “expectations” at all. Instead, I will use my impressions of a running back’s ability in tandem with his age to answer the following question: “What sort of real-life role and standing in the league could this player’s ability be expected to justifiably maintain for at least two future seasons, assuming he receives the requisite opportunity?”

You’ll notice a few key points here. First, I’m not giving any consideration to fantasy football at all. Because (especially) running back production is so closely tied to the environment in which players operate, and because those environments are so difficult to project from year-to-year, it’s almost futile to predict how a given player might produce in fantasy even a year out without making substantial assumptions about his situation either changing or not. I can more effectively reason through my beliefs regarding a player’s potential to produce in fantasy by referring directly to his inherent capabilities.

Second, the phrase “justifiably maintain for at least two future seasons” is operative here. I can’t predict which running back will randomly receive opportunity beyond the level that his ability warrants, and I can’t predict how long the particular roster state that allows him to persist at such a level of opportunity will exist, but (I’m fairly confident) I can identify the level of opportunity that a running back’s ability would warrant him maintaining in a league-average situation for a few years at a time. While not explicitly accounted for, player age is wrapped up in the answering of this question in that I’m considering player ability out to at least two years in the future.

Third, while this element of the process is also not directly reflected in the wording of the question, I’m hypothesizing in the answering of it something closer to a sustainable ceiling case for each player than a median outcome.

Finally, as with the short-term future production expectations, my answer to the above question in each player’s case will involve categorizing them on a numerical scale:

I think these categories are all self-explanatory, but let’s do another (non-comprehensive) sampling of players who I determined fall into each of them.

Remember that these categorizations refer to players’ speculative roles and standing in real life, not in fantasy, and it’s entirely possible for players to over- or underproduce their real-life, in-a-vacuum, in-a-league-average-situation ability for seasons at a time, which is how someone like Alexander Mattison might be classified as a 4 in terms of rest-of-season production expectations but as a 3 in terms of justifiable, long-term role maintenance (and given his current level of play, it’d probably be reasonable to classify him as a 2 in the latter area).

It’s also worth pointing out that these categorizations are less subject to change than those concerned with short-term production, especially in-season. They aren’t completely static, as age, injury, and increased information about player ability gained via their performance on the field could all influence how a player’s hypothetical ceiling looks, but external factors far outweigh internal ones, and the former simply won’t affect my view of player potential nearly as much as it does my view of player production.

Also unlike the short-term production classifications is the fact that these categories are more flexible in the amount of players they can logically accommodate. It would be silly to consider a large number of players to have a justifiably sustainable top-3 running back ceiling, but we’re not limited to only three players in that category if there happens to be a lot of high-end running back talent in the league at a given time. Similarly, the number of running backs who might categorize as a 4, 3, 2, or 1 on this scale is nearly unlimited, especially when compared to the short-term production scale that is constrained by the limited number of league-wide roster spots that are actually dedicated to running backs at a given time.

By generating a weighted (and variably so based on where each player is at in their individual career arc) average between those two ratings for each player, I can produce an overall “score” that I’ve serendipitously found to provide a pretty good and intuitive baseline for arranging their respective dynasty values. I will tweak from there either based on the wider market or by using my intuitions to fill in the gaps of where this process looks past apparent next-year-or-so opportunity for players whose hypothesized ceilings clearly don’t match -- in whichever direction -- that level of opportunity, but the main ordering mechanism going forward (until I decide that it sucks, which I don’t foresee but is certainly possible) will be that dynamic between rest-of-season production expectations and sustainable ceiling hypotheses. Now formulated via this process, find the revamped version of the dynasty rankings (which are now certainly the best on the planet) right here.

Now that that’s over with, let’s look at a few players who I believe should either be targeted or traded away in dynasty leagues given a substantial difference in how I rank them versus how the market values them (as measured by KeepTradeCut values and FootballGuys rankings). As a quick note, I’m going to stay away from value disparities that are clearly the result of a difference in win-now-or-rebuild approaches between ranking services. I trust that you are intelligent enough to navigate the varying utility and value to your individual rosters of a 27-year old RB2-level producer versus a 22-year old third-round rookie who we think could be good, and I therefore think pointing out that KTC skews toward production or that my rankings skew toward youth (or whatever) is not very instructive. But I digress. Here are four players who I think look like solid buys:

Najee Harris is my RB21 but is the RB33 at KTC (he’s the RB19 at FBG). He’s obviously not offering much of anything in your lineup right now, but Harris is playing well in the context of his awful circumstances. “Talent over situation” is a widely-accepted axiom in the long-term format that is dynasty, which would make Najee an obvious buy across the industry if everyone wasn’t more interested in memeing about how bad he supposedly sucks than legitimately evaluating his profile, but that phenomenon makes him all the easier to trade for. There’s no guarantee that his circumstances improve enough to make him useful in fantasy any time soon, but acquiring players at prices that under-represent their inherent abilities is a winning strategy in the long-term.

Chuba Hubbard is my RB35 but is the RB47 over at KTC and the RB64 at FBG. He’s simply a better running back than Miles Sanders is, and we shouldn’t be surprised if he surpasses the former Eagle on the Carolina depth chart at some point this season (he already played more snaps and touched the ball more than Sanders did in week four). Even if he doesn’t, he’s a young running back who is averaging nearly 10 opportunities per game and producing efficiently on them, so he should at the very least be valued next to the Tank Bigsby and Jaylen Warren types who we wish were doing the same thing.

Jaleel McLaughlin is my RB36 but is the RB43 at KTC and the RB72 at FBG. I think this guy is legit. He’s done nothing but be dynamic on limited work this season after having finished his career at Youngstown State as the most productive college running back of all time, and he may have already surpassed Samaje Perine as the RB2 in Denver. He’s not going to supplant Javonte Williams, but there’s room in a Sean Payton offense for two running backs to eat. Hop on the bandwagon before you get priced out of your shot at the next Phillip Lindsay.

Zach Evans is my RB51 but is the RB65 at KTC and the RB66 at FBG. There are less-talented players who are also not currently producing but should be ranked higher than Evans in dynasty given a clearer path to work than he seems to have, but of the guys that I’d rate as “speculative” in terms of short-term opportunity, Evans might be the only one who I’d say has the potential to justifiably sustain himself as something better than a “functional starter” in the league. In other words, he has the highest ceiling of any longshot running back in the player pool, and he should be prioritized ahead of the Deuce Vaughn, Kenneth Gainwell, Chase Brown, and Jerick McKinnon types because of that upside.

And here are three players I think you should be looking to trade away:

Kenneth Walker is my RB10 but is the RB4 at KTC and the RB6 at FBG. While I certainly like Walker as both a player and long-term dynasty asset, I think the fact that he’s currently producing as the RB8 in PPR leagues (aided by five touchdowns in the first four games) is allowing those who didn’t care about the lack of receiving profile (he currently has a 7.8% target share, lower than the mark that Dontrell Hilliard earned last season) or the addition of a second-round runner in Zach Charbonnet to this backfield to get away with their cock-eyed optimism for the time being. Walker is a good asset who is currently valued as a near-elite asset, which represents a perfect opportunity to flip and accrue value.

Kyren Williams is my RB27 but is the RB15 at KTC (he’s the RB44 at FBG, which is too low given his current level of production). Nobody at any position in the entire player pool is currently outproducing their talent profile to the level that Williams is right now. He might keep this up all season for all we know, but it’s also possible that his opportunity share decreases to the point that he can no longer be counted on as a top-24 option, which could happen for any variety of reasons (an injury to himself or Matthew Stafford, an untimely fumble or ill-advised joke that offends his wishy-washy head coach, etc.). Basically, he fell into opportunity and doesn’t have the requisite ability to give himself staying power if things go south, so unless you absolutely need him in your lineup, every solid fantasy outing he puts together is a trade window-extending gift from the dynasty gods.

Miles Sanders is my RB39 but is the RB32 at KTC and the RB28 at FBG. Sanders is currently producing as the RB30 in PPR and, as I mentioned above, is getting thoroughly outplayed in his own backfield by Chuba Hubbard. He’s not providing current value in your lineup and doesn’t profile as a guy who we should expect to garner difference-making work in the seasons to come, so I’m not sure what he’s doing on dynasty rosters other than taking up space that could be occupied by cheaper players who might also not be providing current value in your lineup but who do offer much higher long-term upside.