364: Mailbag 4/10/20

00:00:02
Speaker 1: This is the Label podcast with your host JJ zacher Retha JJ zacher reson What's up everyone. It's JJ Zacharyesen, the editor in chief at fandel and at number five dot com. This is episode three hundred and sixty four of the Late Round podcast and the many shows that are part of the FanDuel podcast network. Thanks for tuning in. We've gotten a lot better at analyzing fantasy football through the years, like a lot better. If you go back and read the things that I was putting together when I first started writing and analyzing fantasy It's legitimately not very good. I'm not ashamed of the content necessarily, but I've gotten better at analyzing, and the whole community has too. As we've gotten better, there's been a push to find really predictive metrics, to ignore this stuff that doesn't matter as much and focus on stuff that's only predictive. But I don't really agree with that. I don't disagree with the idea that finding predictive metrics is great, but football statistics that aren't predictive are still very helpful when analyzing fantasy football because it helps us understand regression and it helps us figure out what matters and what doesn't when our competition may not totally get it. I'm bringing this up because this week I got a lot of questions about regression. One was from at ff Zinger that said touchdown regression is the most commonly discussed, and rightfully so regression and fantasy football, But are there any other types that you value? For example, Mike Williams twenty point four yards perception is bound to drop as well as perhaps Hollywood Brown's twelve point seven will rise. Now, I won't dig into yards per reception specifically here, but I want to dig into the idea of non touchdown statistics regressing. You can think of this in a really basic way to get your mind going. If a player has a really high yards perception rate, or if a player has a really high yards per target rate or a really high yards per carry rate, it's probably going to be pretty hard for that player to repeat that performance over and over and over again. But the thing is a lot of casual fantasy manag and I'd actually say like eighty percent or more fantasy managers they don't think of data this way. For example, we rarely see players rush for five yards per carrier more in consecutive seasons. But when Johnny from down the Street saw Philip Lindsay's five point four yards per carry rate with one hundred and ninety two attempts during his rookie year, he might have said, Wow, imagine how many yards Philip Lindsay would have if he had more carries. And Lindsay did see more carries during his sophomore season, but he still rushed for fewer yards because situations change and things regress, and maybe that big run one year didn't happen the next year. So if you see someone say yards perception doesn't matter, I'd say that it does since you're playing against other humans in fantasy football, and if other humans are going to overstate what a touchdown total might say or what a yards per carry rate says, then you're at an advantage when you analyze not just the statistics that matter, but the ones that don't matter too. So I guess to answer this question from ff Zinger more direction, Yes, there are a ton of statistics that I look at when analyzing regression, even though those statistics don't matter for predictiveness. Now, before get into the rest of the questions today, I want to let you guys know about Fandel's MBA simulations. Each night, Fandel will offer a pool of players from a given night's scheduled NBA games. You'll be able to draft a full lineup using the standard roster limits and salary cap, just like you would for a normal NBA multi game format. Players will be randomly assigned stats from a game they played during the season, and scores will be slowly revealed throughout the night. The contests are one hundred percent free, but winners will be awarded site credit and potentially other prizes each night. For more details, was it Fandle dot com or download the Fandel app Today. Eligibility restrictions apply this first question or I guess next question on today's show says hey, JJ, I have a regression question for you. The PFF Fantasy Football projections for twenty twenty have Alvin Kamara projected for eleven point one touchdowns on the ground and three three point nine touchdowns through the air. I had a lot of Kamara last year and his lack of touchdown production was very frustrating given his yardage totals. But watching his film, it appears at Taysom Hill's usage around the goal line really hurt Kamara last year. He didn't vulture a ton of touchdowns, but there were a lot more formations where it looks like Hill replaced Kamara. And if you aren't on the field near the goal line, you aren't going to score very many touchdowns all off season. Camara is going to be the regression candidate poster boy. I appreciate the random chance of football and the importance of understanding regression, but if we don't dig deeper and look at why a guy like Kamara goes from fourteen rushing touchdowns to five, isn't that lazy analysis? As the best numbers driven analysts in the business, thank you, How do you factor in the nuances of play calling, in personnel or just red zone usage when you look at regression candidates? Sorry for the long question, Love the show, Nick, And then in a PS he listed Kamara's red zone rushing in twenty eighteen and then his red zone rushing in twenty nineteen, and at the bottom of it it says, I'm just saying, so there's a lot to tackle here. Let's start with the first part, where PFF has Camara with fifteen total touchdown My projections have Camara at about thirteen touchdowns, so I'm a little bit lower, but I still think that your question would still stand. So let's just keep this going. Let's look at did Taysom Hill actually hurt Alvin Kamara last year? I don't think the evidence shows that he did to the degree in which you might think that he did. The Saints ran the ball a lot less in the red zone and at the goal line last season than they did in twenty eighteen, so the statistics that you mentioned at the end of the email aren't really that relevant. By the way, the statistics that he did mention were rankings for Alvin Kamara and red zone touches in twenty eighteen and twenty nineteen. What we need to look at is the percentage of those types of touches that a player is seeing, rather than the raw number and where that player ranked, because in the end, goal line rushes come down to offensive efficiency and effectiveness, not the individual player's effectiveness. Players in bad offenses don't see as much red zone work because the offense aren't in the red zone as much. In twenty nineteen, Kamara actually had fifty three percent in New Orleans goal line touches and twenty eighteen that number was forty percent. He did see a fewer percentage of red zone looks last year, but it was a drop by less than eight percent, going from forty six percent to almost thirty nine percent. And keep in mind he played fewer games in twenty nineteen than he did in twenty eighteen. Meanwhile, Taysom Hill's numbers in the red zone and goal line on the ground actually dropped year over year. He didn't have a single goal line rush last year, and he saw fewer than nine percent of the team's overall red zone attempts. In twenty eighteen, he saw ten percent of New Orleans goal line rushes and almost fourteen percent of the team's red zone attempts. Now, Kamara did go from seeing three targets of the goal line to none last year, whereas Taysom Hill went from zero to four, So they basically flip flopped. But that's not going to explain the difference in touchdowns for Kamara one year to the next, not even close. That might explain like one or two touchdowns. To me, there's more important evidence out there that points to Kamara's lack of touchdown production. Like I said, the Saints didn't produce even close to as many goal line plays and red zone plays last year, but they also saw their starting quarterback get hurt and Kamara himself, as we know, he was also hurt throughout the season. If you look at how the Saints did in twenty nineteen versus twenty eighteen and twenty seventeen, there's a pretty clear difference. We saw a higher pass to rush attemp ratio last year in neutral game scripts, and a higher pass to rush ratio in general. That can explain a higher ratio of passing touchdowns to rushing touchdowns, which is what they had last year. The Saints threw three passing touchdowns for every one rushing touchdown. The previous two seasons saw that ratio at one point twenty seven and one. That's a significant difference. Most importantly, the Saints just weren't as good offensively ats scoring touchdowns. They scored a total of forty eight. The previous season saw them at fifty nine. So basically what happened in twenty seventeen is that the Saints had a very low pass to rush touchdown ratio. That helps touchdowns a little bit in the favor of their running backs. In twenty eighteen, they were able to maintain those touchdown totals because they scored a lot of touchdowns, so the ratio changed, but it didn't matter to the running backs bottom line because there were more touchdowns to go around. But in twenty nineteen they likely threw for touchdowns at too high of a rate. They had the fourth highest pass to rush touchdown ratio in the league last year. That's likely to not remain the same. I know I just dumped a bunch of stats onto you, guys, but this is why Kamara is very unlikely to come close to repeating his touchdown totals from last year as long as he's healthy. Maybe thirteen total touchdowns seems high if you think that Taysom Hill takes some, but based on what we've seen, that's likely to only be a touchdown or two. Kamara should at least be projected for double digit scores. This last question is another regression related questions from at ike Azar. With touchdown regression coming and only having a nine percent target share in games where both Jamal Williams and DeVante Adams play year, is Aaron Jones a top ten running back? My projections actually have Jones at RB ten, so I don't think he's a lock to be a top ten guy. No. Now, for the record, the players ahead of him are Christian McCaffrey, Saquon Kamara, Dalvin Cook, Zeke Chubb, Jacobs, Fournette, and Derrick Henry, and there are arguments to be made for guys like Miles Sanders, Austin Eckler, Joe Mixon, or even Le'Veon Bell. I mean, the running back competition this year is actually pretty fierce. And two things to keep in mind with this. I don't draft based off of projections. They're really just there to help me guide conversation and learn more about a player situation. The other thing, though, is that projections capture regression because it's showing what a player is projected to do in a given season. So with regression baked in, Jones is still pretty damn good. I think there are worries, but there are a lot of worries for pretty much any running back outside of CMC and maybe Saquon Barkley. So to me, he's top ten, but barely, and you can make arguments for him probably being overdrafted this year. Chris Carson or the twenty twenty one oh four. This one's pretty easy for me. I'm taking the one oh four all day. I don't think it's really that close. Carson's gonna be twenty six this year. He hasn't been able to stay healthy, and it's not like he's got this crazy profile to back him up. With the one oh four, you're guaranteed one of the best backs in this class or one of the two elite wide receivers. So to me, I'd take the one oh four and not even second guess of the decision. This last question is from at J. L. Woods ninety eight. Ninety eight. You see a bump up for Mike Williams with Tyrod or Tua at quarterback? Tyrod has an underrated deep ball. No, maybe some positive regression in the touchdown department in store for Williams. I feel like he could be a by Lowan Dynasty slash keeper right now, but would love your thoughts. Williams should have scored more touchdowns than he did last year, Yes, but I think it's hard to make the argument that anyone in that Chargers offense is going to see a bump in twenty twenty maybe longer term. If the quarterback that they pick ends up being really good. But one of the pluses with Rivers was that the Chargers were actually pretty pass heavy in neutral game scripts when Rivers was under center. When games have been within six points, They've ranked in the top five and passed to rush ratio in four of the last five seasons. And Rivers isn't necessarily a conservative quarterback. He ranked sixth in deep ball rate last year, which I'm defining as passes that travel fifteen or more air yards. He was six last year among two hundred plus attempt passers, and he's generally been in the middle of the pack in recent history. And that's all while ranking really well in completion percentage on those throws. In fact, since twenty eleven, his worst season in completion percentage on those throws on those fifteen plus air yard throws is thirty nine percent. The average completion percentage on those throws over this time period is forty one percent, So his worst season in completion percentage since twenty eleven was actually pretty close to the average across the league. So with Tyrod, maybe he'll be able to keep up the frequency and throwing it deep. The Chargers are likely not going to be as pass heavy as a team with Tyrod Taylor under center if they can help it. Maybe a rookie comes in in surprises and is able to outperform expectations, but Mike Williams is not in the most desirable spot for twenty twenty. This NXT question says, how do you factor in bad offensive play when it comes to incoming rookies a la cam Akers? Okay, so every player should be looked at individually to understand context fully. But what I don't understand is people trashing cam Akers for his lack of production given the offense that he was in, when his market share numbers actually tell a pretty decent story. That's why market share is so important. Instead of comparing the raw production of Jonathan Taylor to Cam Akers, you're looking at the production compared to the rest of that player's offense. You're looking at how frequently he was used, because if he's used a lot and he's constantly getting the ball, that signals that that player might be pretty good. Using final season numbers, cam Akers has the fourth best rushing yard share in this class, meaning if you look at a team's total rushing yards, cam Akers percentage of rushing yards for his team was fourth best in this class. The only players who were better were James Robinson, Eno Benjamin and Keishan Vaughan. Jonathan Taylor was fifth. That's not an input in my model, but it gives his rushing total some context. One of the inputs in my model those reception share, or the percentage of receptions that a player had in his offense. Cam Akers ranked in the top ten in reception share. Now, I understand the offense might have been bad, but the offense was bad for everyone in that offense. That's why we're looking at market share numbers. And just to reiterate, I don't just look at these numbers and take them at face value. We should be investigating all of these guys in their situations. But the numbers are predictive, They can, and they do help us make better decisions. That's especially true if we're modeling this last question says, who's a running back currently going outside the top twenty that you could see being top five according to Fantasy Pro's rankings right now, Damian Williams is the RB twenty two, and I think that he makes a lot of sense for this question. In regular season games over the last two years where he's seen ten or more carries or when he's been the most heavily utilized back in that backfield, Damian Williams has averaged twenty one point four PPR points per game. You know how many running backs had a better points per game average in that last year? One his name was Christian McCaffrey. And that even continued into the playoffs and the Chiefs five playoff games. Over the last two years, Williams has scored at least one touchdown per game while averaging sixteen carries and almost seven targets per contest. What if the Chiefs in the draft decided to not spend at the running back position again this year. Damian Williams burned a lot of people last year. I totally get it, but he also wasn't totally healthy. When he's healthy and active, he's a legit option in that offense with a ton of upside. That's going to do it for today's show, though. Thanks to everyone for listening. If you've yet subscribed to the Late Round podcast, make sure you are by searching for it pretty much anywhere pop cast can be found, and don't forget to follow me on Twitter at late round QB. Thanks everyone, hope you're all staying safe and I'll talk to you on Monday.