757: Mailbag 4/21/23

00:00:02
Speaker 1: This is the Late Podcast with your host JJ zacher Retha j J zacher reson. What's up everyone. It's JJ Zacharyesen in this episode seven hundred and fifty seven of the Late Round Fantasy Football podcast sponsored by FanDuel. Thanks for tuning in. Before we get into the mail back questions this week, I want to send you another reminder about this year's the Late Round Prospect Guide for fourteen ninety nine. You can get profiles of every running back and wide receiver who were at the combine, as well as my model's prospect scores, and there's a big breakdown of the model itself and the guide too. There's also an entire section dedicated to the year two model. That one can help you understand what to do with second year players like Damian Pierce, Chris Olave, Wandel Robinson, and more. Once the draft is over next week, you'll get an updated version of the guide to your email inbox. That way you're all set for your rookie drafts right away. Head to Lateround dot com to purchase yours. The first question this week is from Patreon. It's from rev Up those Friar Mooths. It says Throughout the twenty twenty one and twenty twenty two season, everybody was hyping up this twenty twenty three draft class. Now that we're almost at the draft, it seems that this class is not going to meet the hype, and then it regressed as a class during the twenty twenty two college football season. Do you agree with this take? Should we put less weight into the hype of future draft classes because they are hard to predict prior to a college football season. I remember during rookie drafts last year, people were valuing twenty twenty three picks higher than their equivalent twenty twenty two picks, and now it seems to me that the twenty twenty two class was stronger. So look to be transparent with everyone. I don't start prospecting until January each year. I don't analyze college football like I do the NFL. I have a general idea of some of the better prospects before January, but in January I really dive in. As a result, I'm not an analyst who has these super strong opinions of each draft class as a whole. A believer that it's really easy to overrate or underrate these situations. When we're a year and a half out, things will inevitably change. Not every prospect as b John Robinson or Marvin Harrison junior. So because of this, I agree with the take that the twenty twenty three class seem to get worse in the eyes of people who follow those things closer than I do. It started out a ten, and then it became an eight, and then it became like a six. I think this class is fine, but it's not elite. And I thought last year's class looked really good, especially a wide receiver, but that wasn't the consensus take really during the college season. It's sort of nice to be able to focus on these classes in January each year because I'm coming at it without any sort of overall perception. Some may view that as a bad thing, but I think it's more good than bad. As we've seen many times historically, the masses will believe one thing about a class and it ends up being wrong. So I'm at least avoiding any of that for the most part at least. But overall, yes, this class is not as good as everyone once thought, including myself. But I also think that we're going to see plenty of players emerge. I've seen people one out of all of their rookie picks this year, but there could easily be more clarity post draft. I don't think some of these wide receivers were as bad as other people think. There's something wrong with Jordan Addison. There's upside with Quentin Johnston that still exists. They're not perfect prospects. It's not as good of a class as we saw last year, but they can still emerge. And if everyone's thinking about trading their picks, you'll end up devaluing them in a trade, and that's never a good thing. The last question is from Twitter. It says thoughts on Alan Robinson's fit in Pittsburgh. So the Steelers traded for Alan Robinson this week and it was a low risk trade overall, as it should be because Robinson has been dusty. Two years ago, his yards per route run rate was one point one to three. Last year it was point nine to three. To give you some context of what that means. Over the last two years, among wide receivers with fifty or more targets, those yards per run rates ranked in the fourteenth and the sixth percentiles. But when I saw this trade go down, my thought was that Robinson was going to become Pittsburgh's main slot guy. They had a big slot with Chase Claypool, and it's a natural transition for an older wide receiver who used to win with his physicality, and the Steelers have used those physical slot guys in the past. George Pickens and Deontay Johnson rarely lined up in the slot. Last year, they were there on thirteen percent of their routes. Now, Robinson's ever run more than forty two percent of his routes from that area of the field in any season of his career, but he's been in the thirty three to thirty seven percent range over his last two seasons. Now, I don't think he's going to see some career rejuvenation in Pittsburgh. I think instead the Steelers see that their wide receiver depth chart is lacking experience. This allows them to be more flexible during the draft because there are slot guys in this class that they can develop and even draft on Day three. Overall, this just allows for more flexibility. The next question is from at cam Shats. It says, listen to the running back episode last week with Derek Brown and it was excellent. Thank you, but I was disappointed to not hear anything about Sean Tucker. His stock's fallen a lot lately. What gives It feels like a case where he gets round two draft capital and he's everyone's RB three or RB four again, what do you think? Yeah, Look, my prospect model loves Sean Tucker. He has one of the best production profiles in the class. He's got pretty good size, and he was used as a pass catcher in college. He's not as useful as a receiver like Jamior Gibbs or anything, but his best season reception share was almost sixteen percent, and when you're able to get this kind of size with that kind of number, it increases your hit rate for a running back. The issue with Tucker isn't how he looks in that model. It's his medicals. He's evidently been medically cleared, but he hasn't tested yet. He's supposed to test on the twenty fourth. If that goes well, hopefully it improves his draft capital and everyone can become bullish on him. Once again. I've been someone who likes Tucker more than the consensus throughout this process, so I'm hoping things go well. This last question is an email and it says, dear JJ, huge fan of your work. Thank you. I was looking through the Prospect Guide because I was trying to separate the profiles of Taj Spears, Kendre Miller, and Izzy Abanacanda, who are all really close in your rankings. One thing that stood out to me is that while all of them have a best season reception share below seven percent, they vary pretty widely and their total career receptions. Spears has forty seven, Izzy as thirty eight, and Miller has just twenty nine. I've heard you say before that receptions show versatility and being used in multiple ways transfers to the NFL. If that's the case, though, why not use career receptions in the model instead of reception share? Or why not use targets? If a team wants to use a player in several ways, they target that player a lot, regardless of how often they throw the ball in general. Right, I guess I'm just wondering how to reconcile the argument that teams throw to talented running backs with the idea that we should adjust for the amount that other players catch around them. I'm hoping you can explain a bit more. Thanks again, for all you do. Aiden So, just to correct this question a little bit, Spears actually had a best season nine percent reception share. Now, with that being said, this question is really a question about why we use market share statistics and not raw volume in the first place. Like it really doesn't have anything to do with just receiving. It has to do with market share in general. And the reason for that is pretty straightforward. You kind of answered it in the question. Every offense is different. We see differences in the NFL, but those differences are even more drastic and more dramatic in college. Some college offenses are insanely pass heavy. Others barely throw the ball at all. So it's not really fair to judge a player on a team with two hundred pass attempts versus one with seven hundred of them in a given season. And you may say, if this running back is so good, they would just keep dumping the ball off, But philosophy matters. Coaches have an idea in their head of what they want their offense to be. Sometimes these players are stuck in these offenses that just don't throw the ball that much, So we're not really adjusting for the amount that other players catch around them. I mean, we kind of are inherently, but what we're really doing is adjusting for the type of offenses that these guys play in. This is why there's this completely false narrative out there that Jonathan Taylor wasn't a pass catcher in college, therefore he must have looked terrible in my model. Now, he didn't have high raw reception totals, but Wisconsin was consistently throwing the ball less than three hundred and fifty times per season. His best season reception share was ten and a half percent. That's not some elite number, but it's totally fine. And the list goes on and on. Leonard Fournette never had a twenty reception season in college, but that season resulted in almost a thirteen percent reception share because LSU was throwing it fewer than three hundred times per season. In college, you can be very, very efficient on the ground, and if you have a player like Jonathan Taylor or Leonard Fournette, you're gonna give them the ball on the ground. But if they're getting the ball on the ground and through the air, that's what shows the versatility and that's why you're giving them context within their offense, and of course some offenses are going to use their running backs more than others, and all that additional stuff that this metric just won't capture. One metric will not capture everything. Many metrics can get you closer to your goal, though. Reception share is just one of those inputs, and the data shows this. All of these metrics are tested, they're more strongly associated and correlated to success, and sometimes raw numbers do correlate more to success, especially when you pair them up with other metrics. But we're using these share related statistics because it gives us a better picture. Whether we're talking target share, reception share, backfield share, it's more predictive than the raw number. This best christ is from Patreon. It's from Garrett. It says, it feels like everyone is subjectively down on Deuce Vaughan because of his size, but your model obviously likes him because of his production profile. That being said, if Deuce is a later round pick, isn't he almost guaranteed to have a positive draft capital delta? That's to say, is the market undervaluing his collegiate production. It seems as if everyone is giving no value to it. With him being listed as RB fifteen plus on some analyst rookie rankings. Okay, So for everyone who's not aware, draft capital delta is just the difference between a player's draft capital percentile and his prospect score percentile. In other words, a player going first overall in the NFL Draft, his draft capital percentile be in the one hundredth percentile, whereas that player may not have a one hundred percentile prospect score. So the difference between those two numbers is draft capital delta. You could think of draft capital delta as how my model views a player without draft capital in mind, a player's prospect score is what's most important. But as I've talked about on the podcast in the past, a positive draft capital delta does lead to better hit rates than negative ones. And Deuce Vaughan in the first iteration of the Prospect Guide had a decent prospect score because his production in college was otherworldly and the model definitely uses size as an input for running backs, and he still looked good now. That guide came out over a month ago, and since then, duce Van ran a four five six point forty at his pro day, which is absolutely horrendous at his weight. He fell well below that ninety speed score threshold that the model wants guys to hit, which means his prospect score fell a whole lot after that forty time. Now, funny enough, he could still have a positive draft capital delta because his production profile is that good, which would be a rarity for a player with that kind of speed score. But that shouldn't be the focus. The focus should be the prospect score. Draft Capital Delta is only there to get a general feel about how the model views a player versus the actual NFL draft, and while there's better odds for positive ones, we have to recognize that it's the score itself that matters. The prospect score itself. Draft capital Delta also has nothing to do with how the model is viewing a player versus the fantasy mark. It's against the NFL draft the NFL market, and those are two very different things. But look, Vaughn's prospect score was in the eighties when the Prospect Guide launched. Now it's in the sixties, so his potential to hit is much much lower than it was before, and it's because we got that new piece of data. And it may sound unfair, but a ninety speed score not only is pretty easy, for a running back to hit. It's also a mark where players just don't end up hitting when they don't get to it. I talked about that on episode seven hundred and forty seven. And since there's volatility in that range within the model, it can really tank a player's score. That doesn't mean I'm completely out on Vaughn and rookie drafts because everyone is out on him. If you want to throw a late round dart, totally reasonable. When he has the production profile that he has, he's a very unique prospect. When you're in those dart throw rounds, you want to target unique prospects because they're players that we don't often see, and when we don't often see them, that means that there's a lot of volatility there. There's a lot of volatility to do Zvon. That's why, at the end of the day, whether you draft a player or not comes down to opportunity cost. The last question this week is from at Pit to the SB. It says you talked about comps a little bit last mailbag. How are comps done? How do you choose a metric overall profiles? And doesn't the way different stats are weighted lead to bias? Now as I said on that Mailbag episode, comps for me or to help tell a story and to just see if we've seen something happen before. When we see a lot of hits with a certain profile, that's obviously a good thing, we feel more comfortable. And then the reverse is true too for comps. In the model I use size, athleticism, and production metrics. Any method of finding comps is going to have some sort of bias. But the metrics that I use for comps are the ones that are in my model, the ones that seem to correlate best to NFL success, like something like receiving yards per team pass attempt. I have a process for it all, but some comps are better matches than others. Sometimes it's Josh Downs to Elijah Moore, which is a very very close comp. Sometimes it's a little bit more difficult, or I've got to loosen parameters a bit. But I think any process for comps is gonna have bias because there's no uniform way to create them. Whether you're looking at fillmore data, there's no single way to create comps. That's why comps are all over the place all the time. But in general that's the way that I approach them. That's it for today's show, though, thanks to all of you for listening. If you get subscribe to the Late Round Fantasy Football podcast, make sure you are by searching for it pretty much anywhere. Podcasts can be found in doff it to follow me on Twitter at Late Round QB. Before I let you go, I want to give a shout out to Bruce Gibson. He donated ten Prospect Guys to be given away to people who may not be able to afford them this year. So if you're in that boat, and please be honest about that, head tolateround dot com right now and purchase the twenty twenty three Prospect Guide. When you do that, use promo code Gibson seven. That's Gibson and the number seven, and do that at checkout and the cost of that guide should drop to zero dollars as usual, make sure the price is zero before purchasing. So thank you so much, Bruce. It is greatly greatly appreciated. But thanks for listening everyone. Enjoy your weekend and I'll talk to you next week.