Rewind: Mailbag 7/7/23

00:00:00
Speaker 1: Hey everyone, it's JJ. Of course, who else would it be, But it's JJ. It's Fourth of July week, and this is usually a time of year a lot of you start coming back and getting into fantasy football again, which is great for people hopping back in. Hopefully you're able to listen back to an off season's worth of episodes and catch up on what's going on in fantasy football right now. If not, no big deal, You've got plenty of time. There are plenty of things to help, whether it's this podcast feed or the Late Round Draft Guide which you can purchase right now over on lateround dot com. Those wall help, and this episode should help too. I went back and picked out five mailbag questions from this offseason to feature on this mailbag rewind. So, if you've been around, you've heard these answers before. But it's a good sort of recap of my process and how I'm thinking about this stuff at a higher level, at least I think. Anyway, I hope you dig it, Hope you had a great Fourth of July holiday, and I'm excited to get back to it early next week with a brand new episode. The first question is from at mikeyed it says, where would you have ranked ja Jackson, Smith and Jigba among wide receivers if he was able to come out last year. I feel like I've been in this JSN bubble for the last week and a half of my life and I just can't get out. I can't escape. But I still have more to say about him because I keep seeing more and more takes about him. As I've said in the past, draft capital is part of my wide receiver model, meaning if a player ends up going higher in the draft than expected, his prospect score is going to change. It's going to get better, and then the reverse is true too. Right now, JSN is projected to go twentieth overall on NFL mockdraft database dot Com, and that places him in the ninety six percentile on my prospect model. Last season, Drake London, Chris olav and Garrett Wilson had better prospect scores than that. Trey Lon Berks and Jamison Williams and Johan Dotson did not. Now how he would have ranked if he was in that class, I don't know, because this past year did hurt his draft stock more than likely, and each draft is unique with different team needs and such so I could just comment on how my views all of this and have views London, Alave, and Wilson as better prospects than Jackson, Smith and Jigba. Not by a ton, but they're better. I feel like this has made me some JSN hater. At the end of the day, he's gonna be my wide receiver one in this class more than likely. I think he's got an extremely good profile, and he's more than likely someone I'm gonna be drafting this year. I just have seen this taken way too far. At least in my opinion, JSN is not a Jamar Chase prospect. He doesn't have projected draft capital in the top ten, which is really important. I sent a newsletter out last week to subscribers talking about this, and I talked about it when comparing Drake London to JSN on last week's Mailbag episode Draft capital Matters. There's an inherent evaluation with draft capital. So if JSN does go twentieth overall strictly looking at draft capital, his odds of being elite aren't nearly as good as someone who is drafted in the top ten if you're only looking at draft capital. So given that we know that He's not a Jamar Chase level prospect or. It appears that way right now. I can't understand the way people are viewing the one O two and rookie drafts right now. I actually continue this conversation about JSN, and I sent another newsletter on Wednesday about it. That one was a little bit more ranty and more qualitative. But there was a tweet from at McNamara Dynasty, who's a really strong follow on Twitter, by the way, and the tweet was referencing a Dynasty trade. It asked which side would you prefer, Chris Olave, Kenneth Walker and the one to eleven or AJ Brown Elijah Moore and the one O two. And this is a single quarterback league that's very important. Now, I have a preference for what side I would take here, but I don't think it's like a bad trade or anything. What really got me going with this, because I'm such a nerd about this stuff, was how people were commenting saying that the best two assets in that trade were on the AJ Brown side. In other words, AJ Brown and the one oh two were the two best parts of that deal. And I'm just baffled. I don't understand you're telling me that you would trade Chris A Lave for the one oh two. I mean, you could argue that you shouldn't be trading Kenneth Walker for the one oh two. I'm lower on Kenneth Walker, so I can understand that a little bit more. But Chris A Lave. And yes, some people responded and say that they would that they would trade Chris A Lave for the one oh two, specifically because they think that Jackson Smith and Jigba is a better prospect than Chris Alave was. And this is where I want to talk a little game theory and logic and how I think about this stuff, because I think it's important to peel back that layer. It is one hundred percent cool and fine and understandable. If you think that Jackson Smith and Jigba is a better prospect this year than Chris A Lave was last year, there's like a one to two percentile difference in my model. It's not like I see things so drastically different. This isn't like Johan Dotson or Sky Moore versus Chris A Lave. So you might hear that and you might think, Okay, then why is this such a big deal. If we like JSN more than Chris Alave straight up right now, if we like JSN more as a prospect, why not trade for him? And the answer is because we've seen Chris Alave dominate in the NFL. Among fifty plus target rookie wide receivers since twenty eleven, a Lave's yards per route run rate was fifth best, his target's per route run was fourth best. His target share per game rate was also fifth best. His yards per team pass attempt was sixth best. Even if you don't care about those peripheral numbers, despite them being predictive, he was still top twenty and PPR points per game among all rookie wide receivers since twenty eleven. A Lave had an amazingly good rookie season and it's been amazingly undervalued. Just look at the way that people have viewed Garrett Wilson this past year versus Chris a Lave. Garrett Wilson won Rookie of the Year, and he deserved to. He balled out. He's great, but his numbers weren't any different than Chris in fact, on a per game basis, on a per target basis, on a per reception basis, Chris Alave looked better. You can't just ignore what Chris Alave did as a rookie. Data points in the NFL are so much more valuable than data points from college. It's why if you look at my year two model, it's far more accurate in predicting future success than the prospect model is. And of course it is the NFL is different. Some players dominate in college, but they don't dominate in the NFL. When you have actual evidence of a player doing well at the NFL level at a young age as a rookie, you have to weigh that heavily. Like think of it this way, what's more predictive of future fantasy football success in the NFL? Chris Olive is incredibly good rookie season in the NFL, or Jackson Smith and Jigba's great sophomore season in college. I think we all know the answer to that, and as all of you know, I think about this stuff in a non black and white way. Of course, there's a possibility that jasn is better than Chris A. Lave. What if he ends up being the best wide receiver of all time? You have to assign some sort of probability to those types of outcomes. What's possible isn't necessarily probable. It's more probable for Jasn to have that high of a ceiling than some random wide receiver in this class. Yeah, but it's also more probable for Chris Alave to get to that upside given what we saw from him as a rookie, and how if that's aligned with other stud wide receivers in fantasy football. I said this in my email blast earlier this week, But there's a scene from Family Guy where Peter and Lois Griffin are given this opportunity to get a boat, but then they're interrupted by the choice to take a mystery box instead. So they could either have a boat or they could have the mystery box. And there's a very famous quote from Peter that gets thrown out there all the time where he says, a boat's a boat, but the mystery box could be anything. It could even be a boat. I'm all about chasing upside and gobbling up rookie picks to build value on My Dynasty. I've talked about that at length the last couple of years on this show. But why are we ignoring the potential upside of Chris Alave himself, who flashed that upside as a rookie. To me, it feels like, if you're gonna trade Chris Alave for the one oh two, you're taking the mystery box on the chance that you could get a boat. Could you explain why your prospect in year two? Models generally don't account for things like injury or bad quarterback play. So this is an important question, and it's one that I touch on in the Prospect Guide. But even if you're not reading the guide, which is totally fine, I think the answer here still might be helpful. The first thing that keep in mind with my prospect model and my Year two model is that they're meant as a guide more than anything else. To me, they're there to keep you in check. There's no reason you need to feel some strong conviction. When a player has a ninety second percentile score versus ninety first percentile score, the model sees one player as better than the other. But the model also isn't going to be perfect. If you like that ninety first percentile player, more have at it. If it's a ninety second percentile guy versus a seventy fifth percentile one, yeah, maybe you should think twice. That's what I mean by the model keeping you in check and being a guide. So what that means is that you can inject your own subjectivity into things. I wouldn't inject that subjectivity when scores are dramatically different. But if you notice something about a player's profile that the model might miss, go for it. The issue is that we can often overstate what those things are. Like you're asking why the models don't account for injury or bad quarterback play, Well, how are you defining that? Why should you assume that a player not getting hurt would have meant he performs really well, or why assume that quarterback play was the result of the poor production that a player had. This question seems more geared towards wide receivers, so I'll answer with that in mind. The prospect model uses some per game numbers, but the biggest production related metric in the model is yards per team pass attempt and that metric is looking at a player's best season, and it's not a pro rated number or a per game number. It's the raw receiving yards a player had, divided by his team's total pass attempts. So clearly, if a player got hurt, he'll look worse in a given season. Within that statistic, the numerator is going to be a lot smaller. So how is that fair? Well? With questions like this, you have to ask yourself about the alternative. In this case, the alternative is should I use some sort of games cut off for yards per team pass attempt or should I use per game yards per team pass attempt numbers? You can start to come up with different iterations of the metric then, and I did that, I tested that, and within the model that I'm working with, the raw number worked best. Or I should at least say there wasn't any significant difference in using another method. And you might say, how is that possible? Well, a few things. Number one, production is just one of the many inputs in the model. A player dropping in yards per team pass attempt doesn't make them irrelevant. Look at Drake London last year. Drake London was my wide receiver one in the model, and Drake London was hurt throughout college. His yards per team pass attemp number was not very good. Number two, though, you can still run into issues doing this a different way. For example, if you did a per game yards per team pass attempt number, which is totally fine to do for the record, but if you did it per game, you run into sample size problems for players who didn't play a full season. A player who played six games could have a really good number there, but because it's across so few games, it's not necessarily indicative of the type of player that he is, the type of receiver that he is. Nothing is going to be perfect, but that's a good reason why across a large sample there's not a meaningful difference within this model when looking at raw numbers versus per game once. And then obviously when speaking about quarterback play, you have to remember that both models are utilizing market share statistics, which means a player is being compared to his teammates within those statistics, which means those guys are playing under the same circumstances. They're playing with the same quarterback. Now, I will say the year two model, it's a little bit different with this because something like yards per route run is in that model, and that can be influenced by better quarterback play. But we have to remember that if there's an exception to the rule, it doesn't mean the rule is invalid. We can sit here and say, look at this wide receiver who missed games as a rookie who ended up doing well. But that isn't happening nearly as much as you'd probably think. For example, if we focus on the wide receivers who are in my prospect database, so guys who would have a year two model score in the first place, a whole lot of them played fewer than half of their games as rookies, so fewer than eight games. I'm talking like three hundred and fifty four hundred players here, because a lot of guys are undrafted and just couldn't find the field in year one. Guess how many of those players. All of those players gave us a single season above thirteen pp euro points per game, and either year two or year three one of them one, and it was Alan Hearns, a guy my model wasn't going to really pick up on. Anyway. We have to remember that the year two model is trying to predict what's gonna happen in year two and year three of a player's career. So when a player gets off to a slow start, if he misses games, it's gonna be harder for him to capitalize in the second or third year and that's why Jamison Williams, while he can absolutely hit, because his prospect score was great, we have to be level headed about a situation. In my opinion. Of course, every once in a while we're gonna have a situation like Traves etn But even then, since he plays the running back position, I told people to just ignore what happened in year one. He was a player to target in my draft guide. So we can use subjectivity to help us out with these very rare instances. But mathematically, the reason we don't need to overdo it just for these types of players is because number one, those types of players who miss an entire season don't exist in large numbers. But then number two, there actually might be reason to fade those players within the context of looking at year two and year three, we don't see hits from those guys very often at all, So you'd then be changing a model for the sake of just a handful of guys, and to me, that doesn't make that much sense to do. What impact will Sharboney's draft have on Seattle's backfield, particularly Kenneth Walker. One of the top topics this week in fantasy circles has been the Zach Sharboney pick and what it all means for Kenneth Walker. I talked about the pick itself on Monday's episode, giving reason as to why the Seahawks made that pick. I really think it comes down to two things, the first being that Kenneth Walker had the worst success rate among one hundred plus attempt running backs last season, and his rate of gaining positive yardage per rush was also worse in the league. He was a home run hitting running back, so they paired him with Sharbonay, who was good at picking up tough yards without that same type of big playability. But there's also pass catching, which I think is just as important. Kenneth Walker was not a good pass catcher in college. You guys got sick of me bringing that up last year. His five point four percent best season reception share was really really low, and his yards per route running college never exceeded point seven to seven in a single season. Sharbonnay's, meanwhile, was never below that mark, and during the two players' final seasons, Sharbonnay had basically triple the yards per route run compared to Walker, and then all the while Sharbonnay had more than double Walker's best season reception share, and this is all coming after a season where Walker didn't do that much as a pass catcher. He ranked low in PFF's receiving greade among running backs. He was one of the worst in the league per route run, and he averaged about a seven and a half percent target share per game when he took over that backfield. That's nothing special. I think it would be silly for anyone to defend Kenneth Walker the receiver. The pass catcher. Defending Walker the rusher, that makes a lot more sense. He can be electric with the ball in his hands. He was the best pure runner in last year's class. But the Sharbonay pick hurts him. Not because that early down work is just gonna vanish, but because the receiving ceiling is now going to be tougher to achieve. I think there's at least some chance, some chance that Sharbonay becomes a favorite for that coaching staff if he ends up hitting, if he ends up being a good player. But I want to be clear here. Just because I'm explaining why the Seahawks may have made the Sharbona pick, it doesn't mean I think he's gonna outperform Kenneth Walker in fantasy football. I still think that Kenneth Walker is the favorite there. My point is more so there are a lot of people out there who seem to think that this is not that big of a deal. That it's a deal, but it's not a big one. That's where I disagree. Before that pick, Walker had hoped to be in every down back. Now I've had him as a cell all off season, but I still recognized that it could go in that direction. With the Sharboney investment, those odds have dropped considerably. He's going to need Sharbona to bus as a player in order to maintain that value or get hurt. My early projection see Walker as more of a low end RB two with upside. That's for twenty twenty three. That's off of a fifty two percent rush share in that offense and about a five percent target share. Maybe you feel more optimistic than those numbers, but I think the numbers are logical when you figure the player that they just drafted in the investment they made to get him. And yeah, it's the Seahawks. They love their running backs and they value the position more than the rest of the league. But it's not like Sharboney himself is a bad running back. He was projected to go in Round two, maybe early round three. And I will say this, the backfield's a little bit different in best ball leagues, where you won't have to worry about how things will go down each week. But I did want to at least get this out there because it was easy to take what I said on Monday and think that I just liked Sharbonay more. Kenneth Walker is still projected to be the better fantasy asset, but I do think I'll be lower than consensus on Walker, but that hasn't really changed. I'd been lower on Walker than the consensus, and I do think that Sharbona is at least somewhat of a threat, somewhat of a threat to be the one A in that offense down the line. He's a good running back who does things that Walker doesn't do, and that in turn limits Walker's ceiling. This says question is from Brandon on Patreon. It says, how do you go about projecting a player like Calvin Ridley who missed significant non injury related time and is also on a new team. You cross your fingers and you hope for the best In all seriousness. Though projections are not an exact science. There's a lot of subjectivity that goes into them. I've done shows in the past on the process of building projections. I think the most recent was episode three hundred and seventy three with rich Reebar that was back in twenty twenty. But we're making educated guesses when we're building them. I mean, maybe educated guessing isn't the right way to phrase it. If you have a projection system and it's thoughtful and it works off of the right things, that's going to be better than some when building projections manually, Like with my projections, I start at the team level and I work my way down to the quarterback level, and then I get to dipping things up between running back, wide receiver, and tight end. And to get the team level numbers, I've got some regression related formulas that I use to help with quarterback numbers. It's the same thing. And then the big thing to figure out with the skill position guys is market share, statistics, target share, and backfield share. That's what drives a player's projection most because volume drives production, so some players are easier to project than others. We have a pretty good idea of the type of target share and efficiency a player like Justin Jefferson's gonna see this year. His touchdown total might fluctuate a little bit because that's less stable. But from a volume perspective, he's been in the twenty eight to twenty nine percent target share range in each of his last two seasons. Projecting him for a twenty eight percent target chair isn't that much of a stretch. That's what we've seen over the last two years. Calvin Ridley's projection is going to have more volatiletility, and this goes for any player who's in a new environment or who's coming off an injury. You're looking at what they did historically on the volume front. For Calvin Ridley, he saw target share per game rates of fifteen percent and seventeen percent as a first and second year player, and then he blew up in year three and he had a twenty five percent rate. He started to see good volume in his fourth year, but he didn't play that much and then last year he was out. So chances are he's not going to be under a fifteen percent target chair, right. We know that's his absolute floor. He did that as a rookie, he's never been below that number during his entire career. He may not get to a twenty five percent share, though, when you consider that he's only done that once across an entire season, he hasn't played in like two years, and he's on a new team, and then you factor in the competition on said team. So when you play it all out, Ridley's likely in that twenty one maybe twenty two percent target share group. And then from there you can use some regression analysis to figure out his yards per target, which feeds into his total yards, which feeds into his projected touchdowns. It does a very generalized way of talking about projections, but hopefully it helps a little bit with a player like Calvin Ridley. This next question is from at Bulldozer John. It says, when you roster one of the backs in an uncertain backfield eg. Khalil Herbert, what do you think about trading for all of the backs and said backfield versus more running back diversity on your roster? It seems like at least one of them fits this criteria for me each offseason. So this is something that I've answered before in the podcast, but it's worth repeating. Do not go into your draft with the intent to capture an entire backfield, especially when that backfield is Chicago's. And this is not a best ball take for the record, but the idea of ambiguous backfields is that players ADPs drop in season long drafts due to them not having high projected volume. They're splitting a back to it with someone else, so their price drops. But oftentimes those are the types of running backs who end up emerging and becoming huge values in fantasy football. I have tons of episodes through the years on the topic. So people then come to me and say, well, JJ, then why not just draft every single running back from a particular backfield. That way you can get the one who emerges the huge value in fantasy football. And the very simple answer to that is not all of these running backs end up emerging. These ambiguous backfields often stay ambiguous, so you can't really do this, And I say especially Chicago's backfield, because that means you're gonna draft three running backs from the same team just hoping that one emerges, and then you also have to hope that that one player is effective in fantasy football, that his environment is even good enough for him to be a difference maker, and I'm not sure Chicago even has that kind of environment given what Justin Field is going to do with his legs. Players will emerge from these ambiguous backfields. They do every single year, but capturing the backfield is a bad idea in season long leagues. You're making the assumption that a player is going to emerge from the single backfield that you targeted. More often than not, the backfield stays ambiguous. So instead you should target players in these specific backfields who fit certain criteria that'll increase your chances of hitting on players who end up emerging. And there is data that can help you out with this. I cover it all in the draft guide.