In 1960, British psychologist Peter Wason sat a group of participants down and told them: “I have a rule in mind that generates the number triple (2, 4, 6). You need to figure out my rule by generating your own triples — I’ll tell you whether each one fits.” Sounds simple, right? Here’s the catch: the actual rule was laughably broad — any three numbers in ascending order. But Wason didn’t want to test if people could figure out easy rules. He wanted to watch how people think when they think they already know the answer. What he found became one of the most replicated and consequential findings in cognitive psychology: almost nobody tests the hypothesis they disagree with. They generate (11, 13, 15) to confirm “each number is +2” — never (11, 12, 19) to test if that’s really the only possibility. When told the rule is wrong, they don’t question their strategy; they just tweak the rule slightly and generate more confirming triples. It took most participants an absurd number of tries to arrive at “any ascending sequence,” and some never got there at all.
🧠 The Classic That Made the Point Unforgettable
A few years later, at Stanford, researchers Charles Lord, Lee Ross, and Mark Lepper designed an even more devilish experiment. They recruited people with strong opinions about the death penalty — half for, half against. Each participant read two scientific studies: one claiming capital punishment deters crime, one claiming it doesn’t. The twist? Both studies were completely fictional, and participants read exactly the same pair of studies — just with the conclusions swapped between groups. The result was devastating: supporters of the death penalty rated the pro-deterrence study as “better designed and more convincing,” while opponents said the opposite about the same study when it appeared to support their side. After reading identical evidence, both groups became more extreme in their original positions. The evidence didn’t reduce polarization — it deepened it.
🤔 The Counterintuitive Twist
The logical thing to do when encountering mixed evidence is to update your beliefs — move slightly toward the middle, incorporate the new information. That’s not what happens. What happens is that you unconsciously apply a double standard: evidence that agrees with you is subjected to a quick and generous review (“looks solid, makes sense”), while evidence that disagrees is put through a grueling cross-examination (“the sample size was too small, the methodology is questionable, this doesn’t prove anything”). You don’t notice you’re doing this, because your brain isn’t trying to find the truth — it’s trying to protect your existing identity. And the cruel irony? The smarter you are, the better you are at rationalizing. Higher intelligence doesn’t protect you from confirmation bias — it just makes you a more skilled lawyer for your own side.
🔗 Why It Matters
If you run a product like 渡心阁 where people receive deeply personal readings (八字, tarot, personality analysis), confirmation bias is working in your favor and against you. On the good side: users who resonate with a reading will remember the hits and forget the misses, increasing satisfaction. On the tricky side: a user who comes in skeptical will find reasons to dismiss even accurate observations. The real opportunity is in designing the experience so the bias works for both sides — making the reading specific enough that confirming evidence is plentiful, but honest enough that skeptical users don’t feel manipulated. This is the same dynamic behind why people stay in echo chambers on social media, why investors hold onto losing stocks, and why it’s nearly impossible to change someone’s mind with facts alone.
🎲 Bonus
Wason’s 2-4-6 experiment has a darkly funny consequence: even when participants were explicitly taught the concept of falsification — that they should try to disprove their hypothesis instead of confirming it — they still failed. The bias runs deeper than conscious strategy. Try this yourself today: pick one belief you hold (anything — a political opinion, a food preference, a personality judgment about someone). Then ask yourself: “What evidence would genuinely convince me I’m wrong?” If nothing comes to mind… congratulations, you’ve just met your confirmation bias face to face.