Rewind: Ambiguous Backfields

00:00:00
Speaker 1: Hey everyone, it's JJ. Things are a little bit crazy over here. I'm working on getting the late Round draft guide finished for the end of the month, and my kids are home from school today, so dad life is really hitting. And yes, that's kids plural. Some of you likely saw, but I posted on Twitter last week announcing that my wife and I adopted a baby boy all the way back in early February. He was six weeks old when we met him. He's now six months old, so he's been around for a while now. He's also amazingly awesome. We love him to death. But this offseason has been wild, to say the least. With all that being said, I wanted to throw a rewind episode your way this week. Last season, I did an episode solely on the concept of ambiguous backfields. People tend to avoid running backs when they've got competition in their backfield, but oftentimes those are the types of backfields the target and fantasy football. One of the things I talked about on this episode actually led me to being very high on Tony Pollard last season. I'm not sure we'll have a situation as clear cut as that this year, but looking at market trends can really be helpful and the Pollard Love from last year is a good example of that. This was done last offseason, so it does not include twenty twenty two data. The Late Round Draft Guide has that update and it talks in more detail about these concepts. So if you're interested and you like what you hear today, make sure you're checking that out on lateround dot com. Thanks for listening, thanks for supporting, and I hope you enjoy this rewind on ambiguous backfields. I've done a lot of episodes of this podcast through the years. I've done studies on running back handcuffs, on quarterback rushing, on breakout wide receivers. There's been a wide variety of topics covered. Plenty of those topics quickly come and go. You listen, you take it in, and you move on. But others over the years have really stuck. And one of the things that has stuck most since I started doing this podcast is the concept of ambiguous backfields. Years and years ago, I did a study on how to spot breakout running backs and fantasy football. I revisit the study every year, but the gist of it is that I looked at non early round running backs and found which running backs outperform their average draft position. Most From there, I found traits that these players seemed to share. One of those traits was that they came from, as I dubbed it, ambiguous backfields. That these breakout running backs weren't emerging from backfields with any sort of established starter. Initially, that's all. The ambiguous backfield idea meant that the running backs that we should be drafting or coming from teams where there's no established starter, where the first running back drafted from that team is going in round five, around six, around seven, a round eight. Then last year happened, and that's where this story gets a lot more complicated. The market would draft these high end handcuffs relatively early with a single digit round pick, but those players needed that injury to happen to their starter to really have league winning value, and those injuries don't happen all that often. Not only that, but when they do, the backup running back isn't always who he thought he was. He either isn't very effective or he himself splits the backfield with another player. Regardless, the data told me last year that team RB ones are better than Team RB two's in the middle rounds now Historically, I've done this breakout running back study by looking at season long numbers, but I want to switch it up and look at points per game numbers. It basically shows the same thing, but I think there's a little bit more accuracy with points per game. If you look at my Fantasy League redraft ADP from twenty eleven through last year and analyze running backs who number one played at least eight games in the season at number two were drafted between rounds six through nine, so pick sixty one through one to eight would all call the middle rounds. You get the following demographic. You get one hundred and twenty eight running backs in total. Of those one hundred and twenty eight, seventy five of them were the first running backs drafted from their NFL team. They were the team RB ones. That means fifty three players were not Team RB ones. Now pretend you had a scatterplot. Were on the X axis you have average draft position for every running back since twenty eleven, and then the Y axis you have points scored or how many points that player scored in that season. So if you had a running back with an overall ADP of fifty and he ended up scoring one hundred points that year. That's a data point, that's a point in the scatterplot. Now do that hundreds of times and you can create a trend line that shows expectation, or how many points we'd expect a running back to score given a particular average draft position. Like I said, historically, I've done this by looking at overall cumulative season long numbers, but I switched it up this year to look at points per game among running backs with eight or more games played. That way, things aren't skewed by random injuries that sideline a player for a few weeks. The overall point is still the same, since there's still a strong correlation between overall points and points per game, but this to me makes things a little bit more precise. So, like I said, we have seventy five team RB ones and fifty three team RB twos that have been drafted in a round six through nine since twenty eleven, and they've played at least eight games, and this is redraft data, not best ball data. Of those seventy five team RB ones, twenty two of them were able to exceed EIGHTYP expectation by more than three PPR points per game. That's a rate of over twenty nine percent. Now that three number, it's arbitrary, but it's a threshold that's not really easy to hit, and it's a number that I can keep consistent throughout today's episode. So twenty nine percent of RB ones drafted in the middle rounds exceeded eightyp expectation by three or more PPR points. Of the fifty three RB twos, that number was nine, which equates to seventeen percent. That's a good bit lower than the team RB one's. Hence, RB ones have been better val using the middle rounds historically than RB twos. That showed up when looking at season long numbers, and it's showing up with points per game numbers two. And when you bump that up to five points per game versus expectation, the RB ones still have an edge over the RB twos by about five percentage points. The edge still exists. I took this all a step further last year. Data seemed to show that when there were multiple running backs from the same team getting drafted in these middle rounds, the RB ones on those teams hit at a substantially higher rate. And that's true when looking at point per game numbers as well. So again, we had one hundred and twenty eight running backs drafted in the middle round since twenty eleven who played at least eight games. Seventy five of those players were the first running backs selected from their team. Fifty three of them where RB twos or lower. Of those seventy five mid round RB ones, twelve of them had teammates who were also drafted in the middle rounds with them. Of the fifty three team RB twos, seventeen had a teammate drafted with them. Now you may be wondering why those numbers are different, and it's because a team are RB two could be drafted after a team RB one, and that team RB one may not have played at least eight games. That and there was one team, the twenty nineteen Eagles, who had three running backs getting drafted in the middle rounds. So, for instance, last year we had Ronald Jones and Leonard Fournette as an example of this, they were both drafted in round six through nine. Now, of those twelve team RB ons, five of them exceeded expectation by three or more points per game. That was a rate of almost forty two percent, which is bonkers compared to the other hit rates that I'm talking through on today's show. It was the highest hit rate subset in this entire study. Meanwhile, just one of the seventeen RB twos hit, meaning less than six percent of team RB twos who were drafted in the middle rounds with their teammate RB one have exceeded expectation at a high rate. This is what I dubbed the ambiguous RB one theory last season. The ambiguous RB one theory states that when you're looking for ambiguous backfields to target, you should be attacking middle round team ones who have teammates who are also being selected in the middle rounds. The logic behind this and why I thought it wasn't totally irrelevant was at number one, the market just hasn't been that bad at spotting which running back is better in one of these muddy backfields. But number two, think about why two teammates would both be middle round picks. It's likely because either player is good, or that both players are in a good situation. Maybe if one emerges, He'd be said, in hindsight, I wish I would have approached all this in a much different way, because realistically there are flaws to this approach. The top flaw being that ADP is going to fluctuate from one source to another, and it's going to even fluctuate during draft season or while you're drafting. For example, last year, Chase Edmonds and James Connor seemed to have this type of pairing in relationship that they were both middle round picks. With how I'm defining middle round technically, though in hindsight, Chase Edmonds ADP ended up being fifty seven overall according to the source that I'm using, So James Connor didn't really fall into this category of being a team RB two drafted with his team RB one in the middle rounds. Even Leonard Fournette last year, there were a number of people who would message me and say, but Ronald Jones is a higher ADP than Leonard Fournette does. Why isn't he the obvious target? It was poor communication on my end. As an analyst, I have to constantly be looking at what went right and what went wrong, and I have to make sure I'm getting the right information across. So I want to set the record straight with data and how to apply this stuff. The idea shouldn't be so strict mathematically. It doesn't make that much sense for that to be the case, since there's a lot of variance with what we're looking at that variants being that one ADP source is going to be different than another. That makes it really hard to apply to your drafts. So let's look at this at a high level. First number one, you should prefer team RB ones to team RB two's in the middle rounds overall. And again this is focusing on normal redraft leagues. There's a difference in how you approach your best ball drafts, and I'm not talking about that here. I'm speaking to managed leagues. So get RB ones over RB twos generally speaking. But here's where things should change. If you do get an RB two, which is still fine to do. Yes, RB ones have better hit rates, but RB twos are still hitting. If you get an RB two, you're better off finding an RB two that doesn't have an early round running back associated with them. In other words, what we saw with James Connor last year, and data really backs this up. Of our one hundred and twenty eight running backs who played eight or more games and who were drafted in the middle rounds since twenty eleven, thirteen of them had a teammate running back get drafted in the first two rounds on average, so thirteen middle round running backs had a running back teammate with a top two round ADP, meaning all of these thirteen running backs were not the RB ones from their team. Of those thirteen, none of them exceeded ADP expectation by three or more points per game, not one. A lot of them were fine, they can still exceed expectation, but none of them were typically league winning type players. When looking across an entire season, that makes a lot of sense and really backs up the work that I've done historically on running back handcuffs. It's not that these handcuffs can't have big weeks when they're called upon. It's at number one. We don't know when they're going to get an opportunity. If a top running back gets hurt. We can't assume that injury is going to happen at the start of the season. So in turn, a handcuffs point per game rate looks pretty bad. That doesn't excuse things though. Not only could an injury just never happen to a top running back, but handcuffs in most leagues, especially non deep leagues, those handcuffs are the first to go on your roster when you start picking up players off the waiver wire. The easy examples historically include Ben Tate and Tevin Coleman. Coleman's an interesting case because he'd exceed expectation by a little bit, but never by enough to be super valuable on a points per game basis because Devonte Freeman in that backfield was in the way. These types of players can be values, but they're rarely league winning players across an entire season, which makes a good bit of sense since they're sharing a backfield with a really good running back. Now, once you hit round three for a running back teammate, and maybe mid round three, I would argue the ambiguity at least starts to begin. I've gotten this question a lot, actually, what defines an ambiguous backfield? Since I focus so much on players after round five, I've generally defined it as that as running backs who are part of backfields where the RB one isn't drafted before round five. But I haven't put that much thought into it because I never thought that it mattered that much. But I think it does when communicating this idea, So I looked at the data truthfully, we should feel a lot more comfortable with RB twos when their RB one teammates are drafted in round three or beyond. We've had some big hits from middle round running backs who had running back teammates drafted in rounds three through five, and interestingly enough, that's large the running back dead zone, or a big chunk of the dead zone. Since twenty eleven, we've had twenty three running backs picked in round six through nine with their running back teammate going in rounds three through five. In other words, their middle round running backs team RB twos, but their running back teammates are basically dead zone running backs. You ready for this, Almost thirty five percent of those players ended up exceeding expectation by three or more points. Among all the different running back buckets within this data, that one had the second highest hit rate. The only one that was better were the RB ones with RB two teammates drafted in the middle rounds with them the ambiguous RB one theory. Now, your first thought is probably that this has to do with injury. That the guys drafted in the dead zone they got hurt and so their backups were able to step in and perform well, but that's only partially true. It's only a little bit true. Running backs miss games, it's just part of the position. But of the team RB twos in this group who exceeded ADP expectation by three or more points, there areb one teammates missed on average just three games. I don't think we can solely blame injury for these running backs performing above expectation. I think instead we recognize why these running backs were drafted where they were drafted in the first place. This is another knock on running backs who get drafted in the running back dead zone. My reasoning as to why the running back dead zone exists in the first place, which, by the way, I did a rewind episode on the dead Zone just a couple of weeks ago in the podcast feed, so listening to that will allow this to make a lot more sense. But my reasoning as to why the running back dead zone exists is because of a supply shortage combined with teams wanting to fill their rosters. Most fantasy managers want multiple running backs early on in their draft, but true projectable belcow running backs don't really exist past like RB fifteen or RB sixteen each season. Instead, of recognizing that fact and going with a wide receiver, managers force the issue in they reach consistently. That creates ADP where these running backs are being over they're inferior players going too high in drafts. And what does that allow for? It allows for competition, It allows for another running back in that backfield to step in and do something. Because the RB ones aren't very good, or at least they're not round three, round four, round five good. We know that these dead zone running backs don't give us league winners at a much better rate than running backs that you get a couple of rounds later, and it's probably because they're not as good. So if they're not as good, their teammates are able to make noise and be relevant in fantasy football. Not only that, but we're talking about middle round team RB twos. In this scenario, if a player is the second running back on his own team but he still has that type of ADP, then that player is probably pretty good himself. For example, in twenty eighteen, Tarik Cohen was drafted in the eighth round. Jordan Howard, his teammate, was getting picked up in the middle of the third round. Cohen was coming off a rookie season where he was electric with the ball in his hands. He caught fifty three passes. Jordan Howard was this typical dead zone esque running back to avoid, given he had two seasons under his belt and he hadn't hit thirty receptions in either season. Tarik Cohen was talented and he clearly had a role. Jordan Howard was clearly being forced up draft boards because the position dried up and there's a natural demand for running backs. If you have a middle round running back who isn't sharing a backfield with an obvious stud like this scenario, it could signal that Number one the market thinks that running back is good since he's got a decent enough ADP, and then number two, if he's good and if the player ahead of him is a little bit overrated, then there's a higher chance that RB two can carve out a role. We have to remember that I'm defining the worth of these players by comparing their output to their expected output. The later in the draft you get, the lower the expectation. So if you're raising inferior players up draft boards into that dead zone area, they're going to have a tougher time exceeding expectation. Meanwhile, their teammates are going to be able to do well themselves. So your question then might be, well, then why are team RB ones who get drafted with their running back teammates in the middle rounds so much better? Why wouldn't we see that even out a little bit more like we do with these dead zone running backs. And I think the answer to that is because we're not terrible at picking out who the RB one is in a given backfield. When you look at the dead zone analysis, the dead zone running backs may be playing lower than expectation, but they're still generally getting to ten, eleven, twelve, thirteen PPR points per game. It's just that where they get drafted, they're expected to score a lot more than that with two middle round picks, with two middle round running backs, expectations are generally the same. So the guy who people think is going to emerge is probably going to have an easier time exceeding expectation. Let me also be clear that the middle round running back data is small. There's variance to all of this, but I do think it makes more sense to get these RB twos who share backfields with a dead zone running back since something is probably going right talent wise for that RB two to still have some sort of higher ADP despite having a teammate getting drafted so high. Now I know I focus on the middle round running backs primarily so far, but it's because those are higher leverage picks. Later round running backs they're mostly dart throws with later round running back picks, though a lot of the same ideas hold true. Since twenty eleven, we've had two hundred and seventy eight running backs get drafted between round ten and round sixteen, or pick one oh nine to one ninety two. That's at least according to my data. Most drafts are about sixteen rounds, which is why I stop there. Of those two hundred and seventy eight, only fifteen of them were their team's RB one, which is crazy to think about that you could get a team RB one that late, but it's happened before. Of those fifteen players, four of them ended up exceeding expectation by three or more PPR points per game. That's a rate of almost twenty seven percent, which is slightly less than the RB one rate that we saw from middle round running backs. The remaining two hundred and sixty three running backs drafted in the later rounds were non RB ones, surprising, and of those players, fewer than twenty three percent ended up exceeding ADP expectation by three or more points. So, once again, RB ones have been better than RB twos, but by a lesser amount compared to the middle round selections. Looking more at these non RB ones who were drafted late, eighty one of them had an RB one teammate drafted in rounds one to two, ninety two of them had an RB one teammate drafted in rounds three to five, and ninety of them had one drafted in the middle rounds around six through nine. The hit rate for these three buckets were as follows, sixteen percent, twenty one percent, twenty three percent. To put this another way, as the backfield became more ambiguous, running backs in the later rounds exceeded expectation at a more significant rate. So what does this all mean? How do we put a bow on all this? Well, let's look at what I said at the start. Team RB ones are better bets than team RB twos across the board. That's the case whether you're looking at the middle rounds or the late rounds, and it seems like they've really hit when they've got teammate RB twos in the middle rounds with them, though there is some variance there if you're getting a team RB two, though, I'd focus on RB twos who are obviously talented who could have a role in their offense, and I wouldn't shy away from RB twos who have RB one teammates getting drafted in the dead zone. And generally speaking, team RB twos who have teammates getting drafted in rounds one or two, they're not hitting at a super great rate. Again, ADP is going to shift, and this type of analysis, which is ADP focused, isn't going to work for all league types and structures, but it's definitely good to know. It's good to think about because I do think utilizing this information can help you increase your probability of winning at the running back position in your league. Target those ambiguous backfields.