1082: Mailbag - Transfer Data, Luther Burden, and More

00:00:02
Speaker 1: This is the Late Podcast with your host, JJ Zacherthis JJ zacherfon. What's up everyone. It's JJ Zachariesen in this episode ten eighty two of the Late Round Fantasy Football podcast. Thanks for tuning in. Before getting to the mailbag questions today, I just want to send another reminder that the Late Round Prospect Guide is currently available for pre order. This year's guide lets you know everything that goes in the ZAP model. It's got profiles for every running back, wide receiver, and tight end going to the NFL combine. And there's also the Year two profiles. I have my Year two model for second year NFL players. I've got a little over thirty profiles for running backs and wide receivers in this guide too. Basically, there's a ton in this guide and you can pre order it right now for fourteen ninety nine over on Lateround dot com. That's gonna go up to nineteen ninety nine when the guide launches in mid March. Again, just head a Late Round and you can learn more there. The first question this week is from Twitter. It's from at SF Underscore David Geo. It says, as you dig into the prospects, have you noticed any more impact of this transfer era now that there's a larger sample, or does teammates score account for it enough. So modeling was definitely a little bit more straightforward before players started transferring the way they are right now, mostly in that things were a little bit more contained and a little bit more controlled. You could more easily do things like account for teammate competition. Teammates score, though, is a complicated thing to really nail down from a statistical perspective. I've been thinking about this a lot with Judaian Price in particular this year, with him having played behind Jeremiah Love. On one hand, you're gonna hear people say, of course, his production profiles sucked he played behind Jeremiah Love. I actually got an email about that this week, and it's an incredibly valid thing to say for the record Tuesday show, So we'll just me talking about production only. But on the other hand, yeah, we may not have many instances where legit running back played behind a stud like this, but it's not like we haven't seen split backfields in college in the past with multiple NFL backs. You can look at Ohio State just last year, and we can give those players a little bit of a boost, it's just a question of how much, and that's where you can really get yourself into trouble. I mean, think about Roshan Johnson. He plays behind Bijon Robinson in college and so we excuse away some of his lack of production, but he hasn't been usable in the NFL. So if we make too many excuses, it can lead to misses in our evaluation. Now, I think I've built a reasonable teammate scoer that can properly boost a player like Price. He'll be right there from a Zapp score perspective, with Jonah Coleman and Emmitt Johnson more than likely, which I think is fair right now here in February. But what does make teammates score a little difficult is at number one what I just talked to, which is knowing the degree in which you should boost a player's profile. But then there's number two, which is underclassmen. For my teammate score, I'm essentially looking at the final two years of a player's career to see if he played with anyone good and I helped define good by draft capital, so Generian Price, he gets a boost for playing with a guy who's likely going to be a top ten, top fifteen pick, but think Jackson Smith and Jigba on the wide receiver side. He played alongside some really good wide receivers in college. His sophomore season was insane, and he was playing with Garrett Wilson and Chris A. Lave who were early first round picks. So the model can recognize that because they came out a year before he did. But Chris Olave and Garrett Wilson their teammates scores it was boosted because they played with each other. Yes, and they were coming out in the same draft class, but their teammates score didn't recognize how good Jackson Smith and Jigbo was because JSN came out a year later. Like look at Carnell Tis this year. Tate has a good production profile, but he's playing alongside Jeremiah Smith. Now we know that Smith is going to be looked at favorably when he eventually comes out, But you can probably start to see where things become kind of an issue from a modeling perspective, I can make that claim pretty confidently for Smith, But where do you draw the line? All of a sudden, if you're projecting the underclassmen, you're injecting a lot of subjectivity that can easily do more harm than good with a model, So teammate score is not perfect to begin with the transfer stuff. It's not as big of a deal though, because I can model things to recognize teammates at different schools. The bigger problem for teammate scorer is that it's just hard to capture properly. It does get signal in the model, but I try to find the right balance. What the transfer era has done the bigger impact that has nothing to do with teammate score. It's made classes generally older and we have fewer early declares too. Not to be clear, I don't think if this is only transfer related. It's just a piece of the pie. Something that I think is playing a huge role as the COVID season that shifted eligibility and kept players in school a lot longer. So the short term data it's going to be skewed higher than what we'll probably see in twenty twenty seven, twenty twenty eight and beyond. So it's the transfer era, it's the COVID era, and there's also nil At least some players are going to stick around school longer because they can still make money doing so, all while hoping to improve their draft stock. Well, let's focus on the transfer stuff. Transferring comes into play with age because it gives players an out. Rather than having to sit out a year, they can now immediately go somewhere else and play, Which means if a player's contemplating going to the NFL, where he might be looked at slightly unfavorably, maybe as like a fringe day two to Day three pick, he might say, screw it. I want to find a better situation somewhere else in college, and I want to improve my draft stock. I'm going to stay in college, I tweeted to the other But just look at recent running back draft classes. If you hit the last three draft classes at running back twenty twenty four, twenty twenty five, and then this year twenty twenty six, when we look at running backs who were in the ZAP model, which is every back who was either drafted or at the NFL combine since twenty eleven. From twenty eleven through twenty twenty three, we had eleven running backs who were twenty four years or older by the time they're drafted. Over the last three years twenty twenty four, twenty twenty five and twenty twenty six, just three draft classes. We've had fifteen of those backs, so we far surpassed that mark of twenty four years old and just three seasons. Meanwhile, early declares are dropping now I care more about declare status at wide receivers. Let's just look there. From twenty eleven through twenty twenty three, we frequently saw ten or more early declare wide receivers with each draft class. Over the last three years, we've had fewer than twenty total. What's happening is typically only the best wide receivers are not declaring early because, like I said, college players now have more outs. It's not only due to transferring, but that's definitely playing a role. Now. I know the common response in this type of thing is, well, your model isn't capturing those things properly now because the models tested off of data that goes back to twenty eleven and the landscape has changed. Yes, that's true, the landscape is different. With that being said, younger players are still generally better bets than older players are, and you can still look at this through the lens of age adjusted production. That hasn't really changed. Maybe it's a little bit more difficult for players to break out early because now there are older, more capable guys that are sticking around college. But let's not pretend to that sample is massive. We only have twenty running backs of the combine this year. There are fewer than fifty wide receivers, and a lot of them they won't be relevant in the NFL. We're just talking about a handful of cases for the most part in the grand scheme of college football. And as usual, guys, the model isn't there to just give you answers. It's to keep you level headed and to guide you to help you know when to deviate from draft capital. And I think it still does a great job of doing that despite the changes across college football. This has question is from Joey from Patreon. It says, Hey, JJ, wanted to pick your brain on Luther Burden. I know you mentioned him as a bye and the general fantasy community is projecting a breakout. What's your guess as to why he isn't getting on the field the injury in training camp, limited deployment a lah Jaden Reid. I have him on a bad dynasty team and wondering if trading him for a player within your tears makes sense. Thanks. I will say if consensus becomes very, very very bullish on Burden, then there's a chance I'd be out on him. Remember this is all related to cost. I think Burden is an awesome player for many reasons, but as you're noting here, there are still some red flags being overall usage. Burden didn't have a seventy percent rout share in any game last year. DJ Moore was at an eighty six percent routeshair all season long. Roma duneesday eighty eight percent. Burden, even when he started to get more run was still in that like sixty to seventy percent range. That's why anything we do with Burden is based on projection, not something that we've consistently seen. Now, I will say I don't think this is a Jayden Reid situation at all for a couple of reasons. First, their deployment wasn't the same as rookies. Reid ran seventy six percent of his routes out of the slot as a first year player. Burden gets pegged as a slot guy, and he definitely profiles as someone who can do that given what he did in college. But in twenty twenty five, he ran out of the slot on just forty five percent of his routes run. He played plenty on the outside. Burden also had one of the best yards per route run rates that we've ever seen from a rookie wide receiver, at least over the last fifteen years. And he even had a better targets per route run rate than Jayden Reid did as a rookie. And Burden was a better prospect by a wide margin. That's not a hindsight take either. And look, another thing to remember when we point out these singular outcomes is that circumstances just won't be the same from one team to the next. Reid has been good, not great, at least as a fantasy asset. But in saying that, we're also assuming the Packers are using him optimally, which in my opinion, they should allow him to do more on the perimeter. We can't assume the Bears are gonna do that exact same thing with Luther Burden. But here's what I'll say about Burden. The raw volume was not there. That's the red flag to this profile, Like we didn't see him as a full time player. Why that didn't happen, which is what your questions asking. We're never gonna know for sure. We're not in that front office, we're not in that locker room, we're not in the coaches meetings. My guest though, is that he was a little bit raw coming out as a prospect. He was injured early on. It's something easy offense, just immediate learn and there's decent competition within that offense. So I think it's a multitude of factors. That doesn't mean that things will definitely change. The reason to be bullish on Burden is because based on him seeing the work that he saw, he was great what he was asked to do. He looked the part. You can visually see that when you watch him play, and you can look at the data too. We're trying to project that trajectory. It wouldn't be the first time that a player took a step forward after having good peripherals and not a ton of volume in year one. But that lack of volume is also why we're not ranking him close to like a Tetoroh McMillan. There's more downside risk. I think the upside is not that far off from someone like TMAC, especially when you consider team environment. But what's different is the downside. What if Luther Burden's route share doesn't grow? What if this is all small sample sized noise. So when we evaluate, we don't need to say that this is definitely gonna happen or definitely won't happen. That's not the way that I look at this stuff. That's not the way that anyone should look at this stuff. Don't think in absolutes, and when I do that, when I weigh the upside with the downside, I think that Luthor Burden is a little bit undervalued by the Dynasty market. I don't think it's like super egregious, but I have him ranked higher than the market does, And admittedly I think a lot of that has to do with the fact that I liked him a lot as a prospect. His ZAP score was amazing. It'd be one thing if he did this while the model didn't like him coming out. Instead, it's showing that he's been really effective at every stop so far. His year two score it's not that much different from a Mecca. Abukas and Abukahs ranked significantly higher within the Dynasty market. That's not a knock in Abuka either, that's just me noting that Luthor Burden looks like a great upside play, but when you're looking to trade for him, just make sure you're not ignoring that downside risk because yes, the volume, the U usage, the route share, that's still definitely a red flag for Luther Burden. This s's questions from Fat gold Chain on Patreon. It says looking good today, JJ. Thanks. In a recent episode, you mentioned that a rookie running backs year one ZAP score was actually more predictive than year one fantasy points per game. With that in mind, what do we do about Caleb Johnson? Yeah, I talked through that on the Late Round Fantasy Football show on Wednesday too, about how a player's ZAP score is more predictive for year two through four fantasy production than is year one pp your points per game rate, or I should say it helps explain that production more. Now, when I say things like that, I don't want anyone to think that it just completely overrides everything else. I just mean that if you were to isolate those specific metrics, you'd be better off using a player's prospect scorer. Competitor was year one pp your point per game production? With Caleb Johnson. Unfortunately, even with the updates that I made to the Zapp model this offseason, it still thinks that he was a good prospect. It's just the model miss. But I've written up Caleb Johnson's Year two profile on the Prospect Guide, and when I was doing the research, I looked at running backs picked on Day two who failed to reach fifty rushing attempts as a rookie since twenty eleven. There actually had been some spike seasons for those players in that group in years two through four. The problem is that that group is not very inspiring. Just hear the names Shane Vereen, Kenyan Drake, Ronald Jones, Damien Harris, Darryl Henderson. Johnson lands in the flex play category in that year two model with a bunch of other guys from last year's draft class. But I don't see how we can be very bowlish year. It feels like the Jalen Polk situation last year. But at running back. Fortunately we can anchor a little bit more to that running back Prospect score. But I'd be pretty surprised if Caleb Johnson ends up being something that were desperately wanting on our fantasy rosters. It's a model miss though. I'm just glad I avoided him in redraft this past year. The siness questions from at bricks forty two, It says you mentioned using only rate stats yards per team pass attempt, rather than raw stats total receiving yards. Have you ever tried any kind of blend of both, So I'm assuming you're referring to the prospecting work since you're referencing yards per team pass attempt. I've tested a ton of statistics through the years. What you guys hear me talk about is the best combination of factors that I've found to go along with things like draft capital size, teammate score, and so on. But I wanted to talk about this because I think it's really important. When you hear the word per as in per attempt, or per rush or per route run, it's very easy to automatically think of a rate stat that those metrics are rate stats. Most of them are rushing yards per rush. That's a rate stat. It tells you nothing about volume, just about what happens per individual rush. But some of these metrics disguise themselves as volume based metrics. Receiving yards per team pass attempt is actually capturing volume. Think about it. It's total receiving yards gained divided by the number of pass attempts a team has. When we refer to volume, we're often talking about a player being on the field so snaps and then running routes, routes run, and then seeing targets targets and then getting yards. The more of those things that a player has, the better he is. Generally in fantasy football, that's volume. Receiving yards is a volume based number. A player could see zero yards in a game, which we'd assume meant that he didn't see that much volume, and then team pass attempts is just a way to sort of smooth things out for the type of offense that that player might be playing in. So in order for a player to be really good in receiving yards per team pass attempt, he's going to probably need to be seeing a good bit of volume or he's just insanely efficient. But over a larger sample that ends up normalizing unless you're in an offense like navies or something where you just need more context. The same is true with total yards per team play. Total yards is capturing both rushing and receiving, and then we're just normalizing for the amount of plays run by an offense. But a player who's ranking high there within that metric, he can get there by either being a volume hog on the ground or through the air. But it's capturing a lot of things that we inherently want to capture. Volume means a team is intending to use a player in a certain way, or that the player is probably pretty talented in a metric like receiving yards per team pass attempt, it's capturing that to some degree. But then also it's still going to capture some efficiency, because players can still rank high enough if they're insanely efficient, and then it's equalizing things by dividing that by some team total. So yes, I've tested all sorts of metrics, but I do think it's common for people to view some of these numbers as rate stats when they're actually capturing volume. When you look at per rush or per reception, you're only looking at a player's volume itself plays when the player is involved. When you change that denominator to be the entire team environment, you're all of a sudden shifting the meaning of that metric, which is why those numbers can actually be pretty powerful both in college and NFL analysis. That's it for today's show, though thanks to all of you for listening. If you had subscribed to the Late Round Fantasy Football podcast, make sure you are bestworting for it pretty much anywhere podcast can be found in effort to follow me on Twitter and on Blue Sky at Late Round QB. Thanks for listening this week. Everyone greatly appreciate you. Have a great weekend, and I'll catch you next week.