One of the things that’s become sort of common to hear leading up to draft season is that strength of schedule (SOS) doesn’t matter and you shouldn’t pay attention to it. It’s an idea that’s been tested a decent amount by various people, so rejecting the importance of SOS might even be prudent, since if you test for SOS effects they’re either very small or non-existent.
There’s just one problem though: if you listen to people who play fantasy football at high levels, it’s not uncommon at all to hear about SOS. In fact, the reason I’m writing this post is because I was listening to the High Stakes Fantasy Football Hour and heard Blake Pyle, the recent winner of $300k, mention that he took Antonio Gates in a draft because Gates has the easiest schedule among TEs.
I’m not trying to make a backdoor appeal to authority here by implying that if someone wins $300k, every idea they have must be a good one. In fact I will concede that it would be very possible to be really good at player evaluation, or working the waiver wire, and that could essentially win you some prize, while paying attention to SOS might just be noise. Except that I regularly hear guests on HSFF Hour talk about SOS. You’ll hear them mention that they might look at playoff schedules, since many high stakes prizes are top heavy prizes. Or you might hear them talk about SOS as if it’s sort of a tie breaker for them.
If you pay attention to sports analytics at all you might also know that the long held belief among some analysts that the hot hand in basketball doesn’t exist has mostly been debunked recently. When SportVU data became available it showed that players that make a string of shots are more likely to make the next shot, although they also start taking more difficult shots. Then it also turned out the original paper debunking the hot hand may have relied on flawed math in reaching its conclusions. Keep in mind that the idea that the hot hand was a fallacy persisted for some 25 years. The reason this shares some parallels to the issue of SOS is that if you asked any player whether there’s such a thing as a hot hand they would almost invariably say that the hot hand does exist. These are people who have a lot of domain knowledge, even if they never stopped to study whether the hot hand exists. It’s an imperfect analogy to high stakes fantasy players using SOS, but I think there’s a key similarity.
In both cases the finding of the analysis is that something doesn’t matter, or doesn’t exist. When I’m doing research I’m almost always more comfortable to say that something does matter, than to reject the existence of some effect. The reason is that there’s a good chance that I just haven’t thought to test something in the right way. So when I hear high stakes players putting their money where their mouth is and saying that they use SOS information in their drafts, it totally changes the threshold to say that SOS doesn’t matter. When I hear someone with skin in the game saying they use it, I would need to see that using that information is actually harmful, rather than just seeing that it’s potentially irrelevant. There are lots of reasons to take this position, including the potential that the effects aren’t linear, and the only method used to test the existence of the effects has been linear regression. RotoDoc wrote about this recently when looking at QB types.
None of this is to say that SOS is, or isn’t, a meaningful thing to look at in fantasy drafts. But it is to say that knee jerk attempts to dismiss it, when people with skin in the game are relying on it, are probably not meeting the burden of proof that should apply.