Fantasy, in Theory: It's Okay Not to Know Things

A lot of knowledge can get you in trouble where a little knowledge would go a long way.

I honestly believe it is better to know nothing than to know what ain’t so.
-Josh Billings

For those that don't know, one of my jobs around here at Footballguys is projecting player upside. I think it's one of the more important aspects of fantasy football; in the later rounds of your draft or on the last few spots on your bench, you're selecting between players who by all expectations are going to be useless for fantasy purposes. Your goal shouldn't be getting the guy who is likely to be slightly less useless, it should be getting the guy who has the best chance of actually being useful.

Similarly, during the season upside can play a large role when setting your lineup. If you're favored during a matchup, you should be favoring players on your roster with higher floors to try to avoid catastrophically bad showings. If you're an underdog, you shouldn't care so much about floors as you do about your chances of players overcoming long odds to have a massive game.

These upside projections are then fed through the Draft Dominator and Lineup Dominator to give you the best, most specific, most tailored advice possible, to maximize your chances of ending the season with a trophy on your mantle.

I like to think it's fitting that I'm the guy who does the upside projections because upside is really about uncertainty. Players with upside are really nothing more than players we're wrong about. Which is fortunate, because I consider myself Footballguys' resident expert on not knowing things.

And because uncertainty matters so much when setting the end of your roster, I wanted to talk a bit about it today to hopefully help make you an expert in not knowing things, too.

There Are Two Types of Uncertainty

(But I'm Not Sure Which is Which)

(I actually know quite well which is which, but I can never resist a joke.)

When discussing potential sources, uncertainty is often broken down into two categories. The first is something called "aleatory uncertainty", while the second is "epistemic uncertainty".

Aleatory means "dependent on chance", while epistemic means "relating to knowledge", and those definitions get you most of the way to understanding the difference between the two types of uncertainty. Epistemic uncertainty reflects things that we theoretically could know but don't. Aleatory uncertainty is irreducible uncertainty, the uncertainty of randomness.

Consider: I hold up a coin and ask you what are the chances it will come up heads when I flip it. Your first concern should be whether or not this is a "true", or unweighted, coin. The fact that you don't know represents epistemic uncertainty.

But even if you know whether the coin is true or weighted, you still don't know how the flip is going to come up. The best you can do is provide an estimate of how frequently one side or the other will result. That's because coinflips are random processes and therefore subject to aleatory uncertainty.

These types of uncertainty are everywhere. What are the chances my favorite song comes up next on the radio? That's aleatory uncertainty. What are the chances I like the song that comes up next on the radio? That's epistemic uncertainty.

Most forms of uncertainty are actually a combination of the two. What are the chances I get cancer in my life? There's a lot of epistemic uncertainty there, things that you could theoretically know that would allow you to get a better estimate. Am I a smoker? (I am not.) Do I have a family history of cancer? (I do.)

But at the end of the day, there's a lot of aleatory uncertainty there, too. Sometimes smokers with extensive family histories of cancer remain cancer-free. Sometimes non-smokers with clean family trees get cancer. We can't ask enough questions and get enough answers to reduce the uncertainty here all the way to zero.

Known Unknowns

Okay, so this might seem a little bit obtuse even for Fantasy, in Theory, a column whose entire charter is writing up things that are interesting and relevant to fantasy but with little-to-no practical application. But I actually think the distinction between aleatory and epistemic uncertainty is a meaningful one for how we manage our teams.

Over short timelines, (such as when choosing who to start and who to sit this week), aleatory uncertainty is pretty important. If we define a "50/50 ball" as a pass where a receiver has a 50% chance to come down with it, (aleatory uncertainty), then if you're an underdog you want to start a guy who will see a lot of 50/50 balls and then pray that he somehow manages to snag 80-100% of them.

Marquise Goodwin is the ideal example of this; on any given target he can get you an 83-yard touchdown. Now, most of the time he's not going to, but if you're a big underdog there are worse strategies than starting him and hoping that this week is the week the passes fall.

Over long timelines, though, aleatory uncertainty works itself out of the system. Again, by definition, a receiver is only going to catch 50% of 50/50 balls. That's tautology. In the long run, betting on a 50/50 ball specialist doesn't get you any farther along.

Instead, when making decisions about what to do with the end of your roster, you should focus more on epistemic uncertainty, or players for whom we simply might not have the whole story.

Adam Thielen is my favorite current example of this. Adam Thielen is third in the NFL in receiving yards and has basically been killing it in fantasy all year. Why? Because it turns out Adam Thielen is actually a really good receiver and we just didn't know it. We thought he was kind of mediocre, which was why we were drafting him so late.

In theory, the fact that Adam Thielen was actually a good player is knowable. Had we known it, we would have drafted him higher. There are other players out there who have this same epistemic uncertainty around them, players who we think are bad but who are actually good and we just don't know it yet.

Smart teams devote their #3 receiver spot on gameday to Marquise Goodwins, (provided they're underdogs and need to chase upside), and the end of their benches to Adam Thielens. They prize aleatory uncertainty in the short term and epistemic uncertainty in the long term.

But most importantly, smart teams remain respectful of the fact that there's a lot they don't know, and they use that lack of knowledge to their advantage. It's okay not to know things. In fact, it's encouraged.

Wisdom don’t consist in knowing more that is new, but in knowing less that is false
-Josh Billings


More articles from Adam Harstad

See all

More articles on: Timeless

See all