Most fantasy managers do not recognize the subconscious threat that the NFL’s first game of the week poses to the chances of winning their matchup. 

About a decade ago in a small town on the coast of Australia, researchers came across a startling series of statistics. Poring over years of the city’s data, they noticed that on days where ice cream sales were highest, so too were reported shark attacks. They checked and rechecked, but the numbers were irrefutable: the sales figures of ice cream across town appeared to rise and fall at an eerily identical rate with shark-related injuries or fatalities. Was there something about ice cream that drew the sharks to these swimmers? Perhaps carnivorous fish have a sweet tooth, just like we humans do? The researchers were flummoxed, but they advanced their findings: ice cream sales are linked to shark attacks. And, perhaps, eating ice cream causes your chances of being attacked by a shark to rise. 

Import your team to My Playbook for instant Lineup & Trade advice

But of course, this insight was fallacious. Ice cream intake was not the cause of rising or falling instances of shark attacks. The researchers fell victim to something that mathematicians refer to as a spurious relationship. Said differently: correlation does not prove causation. 

Correlation Does Not Prove Causation 

Before we go further, I should note that this ice cream/shark attack story should be regarded more as a teaching tool than peer-reviewed research. It’s been told and retold across academia for years, but it isn’t clear that this study was ever meant to be taken seriously (or that it was even published). 

You see, there was a confounding variable that these supposed researchers failed to consider when assessing the data: temperature. Warmer temperatures cause ice cream sales to rise; they also cause more people to head to the beaches. But the two disparate events didn’t have a “cause-and-effect” relationship; one wasn’t causing the other.

But whether real or dramatized, this study isn’t an anomaly. Stories like these proliferate our news media landscape. Like a study that suggested that dogs that are walked by men are more aggressive. Or one that said that candy consumption predicts future criminal behavior. These types of headline-grabbing studies claim a cause-and-effect relationship when the truth is that there isn’t causation – only correlation. And these correlations are often caused by confounding variables – or in some instances, are merely a coincidence. 

Researchers and authors have tried to fight back against spurious correlations by demonstrating just how dangerously misleading linking two random events in a “cause-and-effect” manner can be. Like how divorce rates are linked to margarine consumption. Or how eating organic foods causes autism. Or that the number of films Nicolas Cage appeared in is linked to the amount of people who drowned by falling in a pool in a given year. These are all not-so-subtle reminders of how we can make data tell all sorts of stories – including fictional ones. 

And yet, time and time again, we seem to quickly forget that correlation does not prove causation. So, why are our minds particularly susceptible to fall for these faulty conclusions? Nobel Prize-winning psychologist Daniel Kahneman said it best: “the world makes much less sense than you think. The coherence comes from the way your mind works.” 

Our minds love organization: we relentlessly strive to bring order to what is actually a chaotic, disorganized world. Often, this cognitive trait is helpful; but sometimes, it isn’t. Sometimes it causes us to find patterns in randomness. Sometimes it fuels our willingness to believe conspiracy theories. We want everything to “make sense” – so we’re quick to support insights that simplify, assign motive, or explain our world. 

The Fallacy of Thursday Night Football 

Now, here’s where this gets us into trouble in fantasy football. If I were to ask you whether Patrick Mahomes’ performance on Thursday Night Football will influence Cam Akers’ fantasy output against the Cowboys on Sunday, you would say: absolutely not. Just because both players might be on your fantasy team, Mahomes and Akers are not in a “cause-and-effect” relationship. 

But even though we understand that intellectually, we don’t act that way. We see our fantasy lineups as cohesive, and worse: we see them as linear. In other words, when players on our fantasy team generate big outputs on Thursday Night Football, we tend to believe that “we’re off to a good start.” Similarly, if one of our players underperforms on TNF – or if a player on our opponent’s lineup goes off – we believe that “we’re off to a bad start.” And then we’re tempted to let this one game influence our lineup decisions for the rest of the week. “Slow” TNF starts will increase the likelihood that we’ll choose to start a “high ceiling/low floor” player on Sunday because we (erroneously) view our fantasy lineups through a real-game lens; in other words, we’re trying to play catchup. 

But remember: correlation does not prove causation. If every NFL game was played on the same day, and at the same time, we’d realize that the data points in our lineup are completely random. Go ahead and look back at the biggest weekly score from your fantasy team in 2019 – scroll down that week’s lineup, and I bet that even in a big week, you’ll still spot one or two clunkers. The same is true for your bad weeks: even on your worst fantasy week from 2019, I bet you had one or two players perform above projection. 

This is the fallacy of Thursday Night Football – it’s extrapolating (and highlighting) one data point, so it feels like it’s giving us a “taste” of what’s in store for our fantasy week – but it isn’t. It’s merely giving you one – or two, or three – completely random outputs: data points that won’t result in some linear progression, or influence the rest of your players’ outputs whatsoever. Your Thursday Night totals could be the best tallies you’ll generate from your fantasy team all week – or they may be the worst. But there’s no way of knowing, and thus, the results should hold precisely no weight in your start/sit decisions for the rest of the week. The moral of this story: don’t chase your Thursday Night results on Sunday, or Monday. 

Look, by season’s end you may indeed find that your biggest Thursday Night Football outputs were correlated to your biggest fantasy weeks of the season – but it’s not cause-and-effect. Don’t let a monster week from Patrick Mahomes – or a suboptimal total from one of your opponent’s players – trick you into believing you should play it safe with your lineup on Sunday. Conversely, don’t chase a “bad start” to your week by haphazardly starting boom-or-bust options over more reliable ones. 

Remember, eating ice cream does not increase your risk of a shark attack. 

And this revelation is directional: because we mistakenly perceive Thursday Night Football games to be a window into our upcoming fantasy week. 

But in truth, these games are more mirror than window: a reflection of how our mind works. 

Import your team to My Playbook for instant Lineup & Trade advice

Subscribe: Apple Podcasts | Spotify | Google Podcasts | Stitcher | SoundCloud | iHeartRadio

Beyond our fantasy football content, be sure to check out our award-winning slate of Fantasy Football Tools as you prepare for your draft this season. From our free mock Draft Simulator – which allows you to mock draft against realistic opponents – to our Draft Assistant – that optimizes your picks with expert advice – we’ve got you covered this fantasy football draft season.