Input Volatility, Week 16
Plus what is wrong with fantasy analysis
One of the great joys of starting this newsletter in 2020 was how it was a clean slate in some ways, and I could start from ground zero to hit on concepts and ideas I thought were important, in the ways I thought they should be discussed. But I realize there are a lot of you joining every year, and I’ve started to feel a bit of a push and pull as we move toward the end of our third regular season here, where I want to keep everyone in the loop on the foundational stuff, but I’m also most passionate about continuing to find new things, and building on past concepts.
That’s not a major issue, but I’ve been reminded a couple of times recently about how not all of you have been around since the beginning of this thing. I actually got a tweet the other day that asked me to define signal and noise, and I couldn’t tell if it was earnest or potentially sort of mocking the liberties I take with the column, which is probably just me being poisoned by the internet (in the sense that it probably had no ill intent).
For anyone unsure about that, signal would be observable stuff we believe is meaningful and will help predict the future, while noise would be randomness we should not expect to, which can be equally important, and maybe even more so if we think the fantasy football market is going to put weight on something that isn’t actually useful. Of course, after writing a column for seven years where I note “signal” and “noise” for every game all season, I take some liberties with those definitions at times to make them fit a fantasy football column, and in the interest of just not having the brain power to give every one of those summary points the proper analytical focus it might need to defend the “signal” and “noise” monikers I throw on them, every week. A little peek behind the curtain here, but I spend quite a lot of time on those bullet point summaries every week, sometimes just sitting there stumped like, “How do I want to frame the Signal/Noise section here?”
Another thing a newer subscriber mentioned in a recent email was he really enjoyed going back through the archives, so I will note that if you want to do that, you always can at bengretch.substack.com. In terms of where to start with that, I’d recommend focusing on the offseason posts from past years, as well as any one-off in-season posts that weren’t Stealing Signals posts for those weeks. That grouping will tend to be more evergreen in terms of how applicable the content is, and then of course the Signals articles do have intros that have some evergreen thoughts as well. There’s a “like” feature on Substack, and it’s probably the case that the posts with the highest number of likes are the very best ones to start on, as well.
I mention this push and pull for a couple reasons. The first is I’ve talked for some time about a goal to better catalog the archives, which I didn’t get done last offseason, but is one of those failed goals I’m only becoming more encouraging to put together. It feels pretty necessary to have something that is a pretty good guide to get people up to speed, where a lot of the best thoughts and ideas are in one place.
The second reason I have more new thoughts that build on past stuff I’ve written. And one of the ways my thoughts have evolved this year have been to comment more on how we do fantasy analysis as a whole industry, which it’s probably no secret I think is pretty flawed, in the way that things sometimes just flow a certain direction because that’s how they were always done, and no one really stops and says, “Hey maybe this isn’t the right way to do this.” Or perhaps more accurately, maybe no single entity could stop the flow anyway.
The following tweet got retweeted into my timeline yesterday, and it immediately made me think about the ways we just look at fantasy points against a position or talk about start/sits, as a whole industry.
Before I dive into this, I want to be clear I am not intending to criticize any content creators specifically, who are trying to fill a need because generating clicks and views is a necessary part of this job. I’m also in no way criticizing anyone who is subscribed here, because you’re all part of a subset of the fantasy marketplace that I’d call more engaged. There’s a fine line to discussing this, and it’s probably part of why it doesn’t get discussed that much, in a “the customer is always right” sort of way.
I am going to talk about how a ton of consumers treat fantasy analysts like they are their comanagers, and just how absurd that dynamic is when the whole idea of this hobby is it ostensibly gives people the power to manage their own teams. The fine line is that it’s not me being against asking questions of fantasy analysts about your team.
I’ve been pretty open over the years with you guys that I do hate the part of the job where I sometimes feel responsible for other people’s teams. My passion is in analyzing football, and it genuinely bugs like hell when I give bad advice. Answering direct start/sit lineup questions every week just ratchets up the potential for me to feel like I let down tons of people. As much as that sounds like a complaint, I’d argue that’s just describing being passionate about something, particularly in an industry where predictions are more or less the whole game. If you’re a fantasy analyst who cares about their work, you’re going to feel that way; I’m not alone in that feeling among my peers. What I’ve more or less said to myself after doing fantasy analysis for seven-plus years now is those direct answers on coinflip decisions, while on one hand being very time-consuming, can also increase the potential that I will feel like shit when I “clock out” and step away from the job. So that’s what I’ve touched on before — I’ve made a conscious decision to try to avoid that stuff.
But the other side of that fine line is it’s also genuinely one of my favorite parts of the job when you guys ask questions that understand the basics of what I write and think about, and it’s clear you value my opinion and want clarification a couple of layers deep. I freakin’ love those conversations, and many of you know that, because I will send you private theses that are hundreds of words long sometimes when your questions really get me thinking. They help me by advancing the discussion. Hell, I’ve carried a bunch of those questions and answers into the newsletter itself, and if you’ve been reading along since the offseason you almost certainly think I don’t have to defend this point, despite feeling obligated to. I can’t even imagine how many thousands of words I’ve written in this newsletter that were in direction response to your questions about your teams. Point made.
So what I’m talking about is the random consumer, and by that I just mean your typical hobbyist, who very well might be a rocket scientist or brain surgeon or whatever, and is plenty capable of understanding football but just doesn’t have the time to. There’s an interaction between fantasy analysts and those types of consumers, who make up the vast majority of the fantasy consumer base (not necessarily scientists and doctors, but people who aren’t engaged at a super deep level). And that interaction often carries this tacit expectation, from my past experience and what I still see all the time, that the analyst is always available to be essentially a comanager for that individual. There’s almost this “Do your job!” element, and most fantasy analysts have probably gotten messages like that before.
And I’m not complaining about assholes. There are assholes in every industry. Some of you have to deal with them every day. I also fully understand not everyone has the time to consume fantasy analysis every week, and of course there are those who just need the answers. But I mean step back and think about how silly that consumer/analyst dynamic is. Fantasy football is about you managing your very own team of football players! Like, the entire concept is that an individual gets to make their own decisions. And then there are the people who ask these questions and then if the advice they get doesn’t work, they abuse the person they asked! You see the, “You’re terrible at your job, you cost me my championship!” stuff around the industry, and even setting aside the part where the fantasy analyst went out of their way to help, that was still your team to make the decision on. You chose to ask that analyst. And then ultimately, you chose to listen to them. If the analysis didn’t work, that’s a bummer, but that’s not on the analyst for offering a thought!
I think most good analysts want to analyze football and this game, and then to empower consumers to manage their teams how they will, based on whatever they may take from that analysis. And judging by tons of interactions I’ve had, that’s the dynamic we have here, and you guys really get that, and it’s what I absolutely love about this newsletter. It’s also why I guess I felt the need to write this, because I feel lucky among my peers and in a position where I can comment on this without it really impacting my audience.
And what I want to say is it’s BS that analysts trying to make a name for themselves, who actually care about this, have to go through an initiation period of trying to get their name out there by attracting as much consumer attention as possible, and because of how the whole thing operates, that requires playing this game of being consumers’ comanager, and then getting blamed for being wrong. What also happens is the analysts that do care a ton about being right — they are going to have a harder time with it, as I described having a hard time myself. Meanwhile, to get back to that tweet that I shared, the analysts that can tell you the what but not the how — those analysts aren’t actually as stimulated by all the actual analysis of this job. But because they don’t mind the person-to-person interaction element as much, they are going to paint with a broad brush, offer cursory stuff like “I like the matchup for Position Group A of Team X better than Position Group A of Team Y so I’d go with them,” and then let the natural variance of things dictate that they’ll be right on enough calls to start building a reputation among those consumers.
The logical conclusion of this point is to say that not only is this dumb for the types of stuff that’s rewarded in the industry, it’s dumb for the consumers! It’s this short-term gain at the expense of long-term success, where consumers who have the expectation of analysts answering all their weekly lineup questions so they don’t have to — those consumers wind up gravitating toward worse analysts.
For those analysts, it’s a game of volume — answer as many questions as humanly possible without giving any real reasoning, and don’t actually sweat being wrong. That’s the key! Not giving a shit what people think when you give them information that doesn’t pan out, which over a longer timeline quickly becomes a detachment from any accountability over your analysis at all. And therein also lies the irony, as I just laid out. The entitled consumer in the fantasy marketplace who seeks that type of quick advice and feels empowered to then blame the analyst giving it, because they weren’t capable of predicting the future — that consumer winds up attracted to whichever snake oil salesman took on the responsibility over their team and got lucky with their calls. On the analyst side, the game of volume dictates there will be consumers where variance breaks right over multiple calls, and they suddenly think you know stuff other people don’t.
And to be clear, I am referring specifically to analysts who recite superficial stats the consumer could look up for him- or herself, but can’t actually detail how those matchups compare, or how the “Fantasy Points Against Position Group A” data was compiled and why it should be trustworthy. They just don’t actually have a lot to say at all about the “how.” They can say the “what,” and they can do it the way fantasy analysts have been doing it for years before them.
But I can tell you that I’ve played this game for more than 20 years and I write about a ton of concepts, particularly about the chaos and uncertainty that defines it. And as I’ve told you before, the truth about start/sit calls is absolutely that there are huge overlaps in the ranges of outcomes of basically any close call you might have an urge to ask about. I spend so much time talking with my comanagers, many of whom are the sharpest people I know in this industry, about what is ultimately the dumbest stuff, as we just try to find a tiebreaker that makes some sense, to split decisions we all know very well are basically just coinflips. Basically everything I’ve seen from the most advanced players I know at the highest stakes suggests almost all of them labor over their start/sit decisions longer than the average social media fantasy analyst who is answering start/sits on the timeline does. Those analysts are just firing off answers in 30 seconds or less, while the most knowledgeable players in the world spend considerably longer, with more of knowledge base to build off, trying to find marginal reasons to separate two options in a tight call.
Again, pardon my language, but how fucking stupid is this dynamic? That was all my big point was today. It’s just the stupidest shit in the world. No one benefits from this. The whole industry, and the superficial analysis that drives it because the majority of consumers want the quick answers — it’s all batshit bonkers that this is what fantasy analysis is for so many people.
Every Sunday on social media you’ll find a market of consumers wandering around the virtual town square looking for the psychic whose tarot cards tell them what they want to hear, as they shout out “Pick A or B” to anyone they can then pin the fault on when things don’t go the way they need them to. “You cost me my fantasy championship!” No, your unwillingness to manage your own damn team — which you signed up for when you decided to play fantasy football — is what cost you. Don’t pin your desire to pass the buck — which as I’ve explained often gravitates you toward worse analysts — on the analyst. It’s on you, Gary.
Sometimes it’s just on variance or bad luck! Shit happens in football. It’s often multi-faceted why a guy succeeded or failed, and a lot deeper than what we pin it on so our pea brains can blame some easy answer. There are so many damn examples of us being like, “Ah, we were dumb to not see X coming,” and stopping there, when a fair analysis of X would point out it required a teammate injury to open opportunity, another lucky outcome like a coach whose scheme smashed or an offensive line that overperformed expectations, and then maybe still some other stuff on the weekly level of this small sample game we play where big-play variance ran in one direction for the year. And if you knew all that stuff in advance, X happening would have been a much easier call!
I want to again emphasize this is a “Don’t hate the player, hate the game” post from me. I don’t begrudge anyone doing their thing within this dynamic I’ve described. Of course, I also wrote it to help explain why I do fantasy analysis a little differently here, which I’ve probably explained before, and also to give you some feedback about what types of stuff you see out there and whether to listen or not to it. I think discerning what is actionable from the sea of information is the whole game (hence a post centered on signal and noise!).
Let’s talk about some input volatility spots for this semifinals week, and I’ll offer my surface-level thoughts and leans, and then you can make your own lineup decisions. Also, apologies to anyone named Gary. Just felt like the right name.
I’m going to start by plowing through a bunch of weather-related discussions, as those are huge this week. The big concern for me is there is tail risk in some of these games that might not even be captured by things like the over/under, where perhaps in the most extreme 10% of outcomes, some games can get pretty brutal. That often makes for some downward input volatility; we saw a game on Thursday night with some wind where the overall stats were negatively impacted, although also due to the Jets and Zach Wilson being who they are.
Ultimately, there was just one touchdown, and among all the skill position players on either team, only Evan Engram and Travis Etienne scored more than 10 PPR points. I played Garrett Wilson in some spots, and that sucked. Christian Kirk and Zay Jones and Zonovan Knight did nothing. All I’m trying to argue is weather increases the likelihood of a really poor overall environment for fantasy production, and yet, in most cases someone is still going to have some production, like Engram and Etienne did in that one.
I’m going to write a bit about current expectations in these games, as well as what to look for in the forecasts, but my advice is to keep an eye on these spots if you have a tough call, and take your decision up to the start of the game.
Saints-Browns
This is the big headline, with winds well into the area of concern (typically anything over 15 mph is concerning, and winds during the game are expected to be about twice that speed with gusts even higher). There are also expected to be snow showers throughout the day.
High-level takeaway for me is this is bad for everyone, with the over/under plummeting and expectations of a low-scoring game. So as I noted, the input volatility here is downward. I think the thought process shouldn’t be about who you can’t play, but more about who you can.
Nick Chubb is reportedly good to go after popping up on the practice report, and you have to feel pretty good about how he fits into this type of game environment. I’m a little more sketched out about Alvin Kamara because I don’t love when I’m banking on him to get there as mostly a pure rusher, but Cleveland’s run defense is bad. Taysom Hill might be a viable TE option if desperate, but his floor is always very low.
I’m pretty skeptical of the pass catchers on both sides, though David Njoku is one pass-catcher you can kind of understand, again because the TE floor is so low.