Confirmation Bias

Category: Probability & Belief

The tendency to search for, interpret, favor, and recall information in a way that confirms or supports one's prior beliefs or values.

How it works

Once you hold a belief, your mind stops acting like a neutral investigator and starts acting like a defense attorney for that belief. It shows up in three places at once: in search (you go looking for supporting evidence), in interpretation (ambiguous facts get read as confirmation), and in recall (you remember the hits and forget the misses). The belief shapes the very data you let in.

The driver is the discomfort of cognitive dissonance, contradiction feels bad, especially when a belief is tied to your identity or to a position you've stated publicly. Agreeing evidence is pleasant and effortless to process; disagreeing evidence is unpleasant and forces costly mental work, so we unconsciously route around it. We ask questions whose likely answers we already favor, and we hold contrary evidence to a far higher standard of proof than friendly evidence.

This makes confirmation bias uniquely dangerous, because it's self-reinforcing: the more you 'research' within a closed loop, the more certain you become, and certainty feels exactly like being right. It's the mechanism that turns echo chambers, conspiracy theories, and entrenched grudges into stable, evidence-resistant structures.

Where you'll see it

  • Someone convinced a coworker dislikes them notices every curt email and ignored hallway nod, while filing the friendly moments under 'they were just being polite', and ends up certain of a hostility that was never there.
  • A day-trader who 'knows' a stock is a winner reads only the bullish analyst notes, dismisses the bearish ones as 'fear-mongering,' and remembers their good calls while quietly forgetting the bad ones.
  • A voter scrolls a feed algorithmically tuned to their views, sees the other side only at its worst, and concludes that everyone reasonable agrees with them, because every piece of evidence they encounter is pre-filtered to agree.

Where it comes from

The systematic study of confirmation bias is often credited to British psychologist Peter Wason, who in the 1960s designed the famous 2-4-6 task. He gave people a number triple that followed a hidden rule and asked them to discover the rule by proposing their own triples. Most generated only examples they expected to fit (like 8-10-12), receiving 'yes' after 'yes,' and confidently announced the wrong rule, because they never tried to falsify their hypothesis. Wason's work, alongside his Wason selection task, showed that people naturally seek confirmation rather than disconfirmation, and the broader bias was later synthesized in influential reviews by Raymond Nickerson.

How to counter it

Deliberately try to falsify, not confirm. Borrow the scientist's habit: ask 'what evidence would prove me wrong, and where would I find it?' Actively going hunting for the disconfirming case is the one move the bias can't survive.

Steelman the other side. Force yourself to state the strongest version of the opposing view well enough that someone who holds it would say 'yes, that's exactly right.' If you can't, you don't understand the issue yet, you only understand your side of it.

Diversify your inputs and track your record honestly. Follow thoughtful people you disagree with, and keep a log of your predictions so the misses can't quietly evaporate. When a belief is load-bearing, ask a trusted contrarian to poke holes, outside eyes see the evidence you've trained yourself not to.

The tell

You're doing it when you go looking for evidence that you're right instead of evidence that you're wrong.

Related biases

Featured in