As we head full steam ahead into summer, we’re in the heart of projections season and about to enter full-blown hype season in fantasy football. It all seems so easy and predictable at this time of the year. Plug in some numbers, adjust for rookies and free agents, make a few bold predictions, and boom; you’ve got a perfect set of shiny new projections or rankings. The best-of-the-best (see: Ulitmate Draft Kit) has gotten extremely good at this process. Even so, fantasy football is a game of numbers based on a real-life sport with so many unpredictable variables that no projection or model can possibly be perfect.
Before I go any further let me make something clear. I’m not knocking projections or rankings. I actually got the idea to write this article when l heard Mike refer to Sony Michel’s 2020 fantasy outlook as a “mystery box” in an episode earlier this month. The purpose of this article is to highlight how we can take advantage of the indisputable math concept at play behind the scenes. We get bombarded with phrases like “high ceiling”, “upside”, and “boom or bust”, but what do these phrases actually mean? They all center around a mathematical concept. Variance.
There are people in this industry with much stronger math backgrounds and understanding than I possess, so forgive me if this is a bit dumbed down. Variance defines the spread from an average. As we should all know, fantasy football regularly produces a large separation from an average, also known as outliers. I didn’t do any hard variance calculations for this article (a project for another day perhaps) but I think we can all agree that it exists.
Variance in the Draft
As the redraft season approaches, let’s first look at this concept from a draft perspective. Our industry has gotten extremely efficient at forecasting top-end performances. For the first two or three rounds, it makes a lot of sense to play it safe. Hammer running backs and wideouts, avoid oft-injured players, take the “sure thing”.
But as the rounds progress, predictability breaks down. We see it every year. In 2018 Robert Woods was a ninth-round pick that finished as the WR10. In 2019 Austin Ekeler was a late-fifth-round draft pick that finished as the RB6. Undrafted wideouts A.J. Brown and D.J. Chark both finished as top-16 fantasy wideouts in 2019.
Looking ahead to 2020, who are some hight variance players you could snag in the mid-to-late rounds? Sony “Mystery Box” Michel could end up as a workhorse running back on a run-first offense, behind an elite offensive line, that tries to run out leads in the 4th quarter. Is it likely? Not really. It would be an outlier, which means that unlikely as it is, it could happen. Maybe Ronald Jones dominates snaps in Tampa Bay while rookie Ke’Shawn Vaughn rides the bench in Bruce Arians’ doghouse. Could N’Keal Harry rise from the ashes of his lost rookie season to become Jarrett Stidham’s top target? Will Teddy Bridgewater utilize his plethora of offensive weapons to become a QB1? Only time will tell.
Without diving too deep, variance is also one of the driving factors behind the late-round quarterback strategy popularized by JJ Zacharision and championed by The Footballers. Four of the top-six quarterbacks in 2020 were drafted in the eighth round or later.
Simply put, the later you get in the draft, the less predictable player outcomes become.
Variance applies just as well, if not better, to in-season performances. Every fantasy football platform shows weekly projections. It’s enough to drive a fantasy manager crazy as you struggle to set your optimal lineup. The problem is that most of these projections are based on averages. You can, however, take advantage of this whenever you find yourself as an underdog in a given week. Based on average projections, you want to use a few of your starting slots on players that can go well above their projections due to their wide range of outcomes (AKA high variance) when you find yourself as an underdog. Of course, you don’t want to go overboard. You’re not starting Sammy Watkins over Michael Thomas. But over Danny Amendola? Absolutely.
Take the 2019 season for example. As noted in my 25 WR Statistics from 2019 article, five of the top-ten single-game wideout performances of the season came from receivers that finished outside of the top twenty on the season. The top two came from WR54 Will Fuller (Week 5: 46.7 points) and WR51 Sammy Watkins (Week 1: 42.3 points). Outliers can win you a week.
It should be noted that variance works both ways. While it wasn’t the focus of this article, busts also exist. Your best bet is to mitigate risk early on and lean into variance more and more as your fantasy draft progresses or if you ever find yourself as an underdog in a weekly matchup.
The whole point of this article is that, in the end, weird things will happen and math will back it up. Every year there will be surprises. Expert analysts will pour countless hours into rankings and projections. Some will be better than others (once again, see the deservedly touted Ultimate Draft Kit) but variance will always rear its head, and outliers will exist. Math doesn’t care. It’s what makes this game about a game so fun.
Remember way back in February, back before I’d even started on my 2020 fantasy footb…