Confirmation Bias: Why You Only See the Stats That Prove You Right
Confirmation Bias
Confirmation bias is the tendency to seek, notice, and remember evidence that supports what you already believe, while ignoring or explaining away everything that contradicts it. In fantasy it turns a hunch about a player into an unshakable conviction — you collect every stat that proves “your guy” is good and wave off every red flag. Beating it means deliberately hunting for the case against your own opinion.
| Confirmation bias in action | What you do | What you should do |
|---|---|---|
| Researching “your guy” | Read only the bullish takes | Seek the strongest bearish case |
| A good game | “See, I was right” | Ask if it was role or variance |
| A bad game | “Just a fluke / bad matchup” | Ask if it’s a real warning sign |
| A rival’s player you fade | Notice only his busts | Weigh his good games honestly |
Confirmation bias is quiet and insidious because it feels like doing research. You look up stats, read analysis, watch film — but you unconsciously filter all of it through the opinion you already hold, keeping what agrees and discarding what doesn’t. The result is a manager who feels informed and is actually just entrenched. It’s how you end up defending “your guy” deep into a season he’s clearly failing.
How It Works
Once you form a belief about a player — he’s a breakout, he’s a bust, he’s undervalued — your brain stops evaluating evidence neutrally and starts recruiting it. You:
- Seek sources and stats likely to agree with you (the bullish article, the one good metric).
- Notice confirming evidence more readily than contradicting evidence (his 100-yard game sticks; his three quiet weeks blur).
- Interpret ambiguous information in your favor (a bad game is “a fluke,” a good one is “the real him”).
- Remember the hits and forget the misses, so your track record feels better than it was.
Every one of these feels rational in the moment. Together they build a wall around your original opinion that new information can’t get through — which is exactly the opposite of what research is supposed to do.
Where It Bites in Fantasy
“My guys.” The clearest symptom. Every manager has players they’re high on, and confirmation bias turns fondness into blindness — you draft them too early, hold them too long, and defend them past all reason, because you’ve spent months collecting evidence they’re good and none that they’re not.
Selective stat-shopping. Modern fantasy has a metric for everything, which means you can always find one number that supports your take. Confirmation bias makes you cite the flattering stat (his air yards!) and ignore the damning one (his target share is cratering). Cherry-picking metrics feels analytical and is actually just bias with a spreadsheet.
Explaining away bad games. When a player you believe in busts, confirmation bias supplies an excuse — bad matchup, game script, a fluke — so the belief survives. Sometimes those excuses are valid; the bias is in always reaching for them instead of asking whether the bad game is a real signal.
Fading rivals’ players unfairly. It runs the other way too: once you’ve decided a player is overrated, you notice only his failures and dismiss his good games, which leads you to sell or pass on genuinely good players.
How to Fight It
The antidote is to deliberately argue against yourself:
- Actively seek the bear case. Before committing to a player, go find the strongest argument against him — the sharpest skeptic, the ugliest metric. If you can’t state the case against your own take, you don’t understand it yet.
- Ask “what would change my mind?” Decide in advance what evidence would prove you wrong, then watch for it honestly. A belief that no evidence could shake isn’t a read; it’s faith.
- Steelman the other side of every trade. When you’re buying a player, force yourself to articulate why the seller is happy to move him. When you’re passing on one, argue why you might be wrong.
- Track your calls honestly. Write down your takes and revisit them. Confirmation bias survives on selective memory; a written record exposes the misses you’d otherwise forget.
- Treat a stat that contradicts you as more important, not less. The metric that disagrees with your take is the one you most need to explain, not dismiss.
Worked Example: Defending “Your Guy”
You targeted a breakout receiver all offseason and drafted him a round early — he’s your guy. Through six weeks he’s disappointed, but you’ve stayed confident, because every week you notice the encouraging signs: a couple of deep targets here, a “he was open” there, a favorable upcoming matchup. You dismiss the bad weeks as game script and bad luck.
What confirmation bias hid from you: his snap share has quietly dropped, a rookie is eating into his routes, and his target share has fallen for three straight weeks. Those are the exact signals you’d have jumped on for a player you didn’t love — but because he’s your guy, you filtered them out and kept collecting reasons to hold. A manager evaluating him neutrally would have sold weeks ago; you’re still defending him because you went looking for agreement instead of truth.
Common mistake: researching a player to confirm what you already think instead of to test it. Once you love or hate a player, you unconsciously collect the evidence that agrees and explain away the rest — so more “research” just entrenches you. Before you commit, go find the strongest case against your opinion, decide what would change your mind, and treat the stat that contradicts you as the most important one. If you’re only finding reasons you’re right, you’re not researching — you’re rationalizing.
The Bottom Line
Confirmation bias disguises itself as diligence: you gather information, but you filter it through the opinion you already hold, keeping the agreeable and discarding the inconvenient. That’s how “my guys” get defended into oblivion and good rival players get unfairly faded. The fix is to argue against yourself on purpose — seek the bear case, define what would change your mind, and weight contradicting evidence more heavily, not less. Research to find the truth, not to prove you were right, and you’ll drop the losing convictions the rest of your league clings to.
For more on research, evaluation, and biases, continue with overconfidence bias, sticky vs. unsticky stats, and player valuation.