Confirmation Bias: The Mind's Lawyer

In 1960, a London psychologist named Peter Wason sat people down with a tiny puzzle. “I have a rule in mind that some sequences of three numbers obey,” he told them. “The triple 2, 4, 6 obeys it. Propose your own triples, and I’ll tell you whether each one fits. When you’re sure you know the rule, say so.” People got to work. They tried 8, 10, 12 — fits! They tried 20, 22, 24 — fits! — and 100, 102, 104 — fits! Confident now, they announced the rule: “even numbers going up by two.” Wrong. The real rule was almost insultingly broad: any three numbers in increasing order. 1, 2, 3 would have fit. So would 5, 50, 500. The reason almost nobody found it is the reason this entire course exists: every triple they tested was one they expected to pass. They went looking for yes, got yes every time, and mistook a pile of confirmations for a proof. Not one of them tried a triple designed to come back no — the only kind of test that could have revealed the rule.

That is confirmation bias: the deep, automatic tendency to seek, interpret, and remember evidence in ways that confirm what we already believe — and to skate right past whatever would prove us wrong. It is not stupidity, and it is not (only) dishonesty. The sharpest, most educated minds do it; in fact, as we’ll see, a sharper mind often does it better, because it can build a more airtight case for the conclusion it already wanted. The 18th-century essayist Francis Bacon diagnosed it four centuries ago: “The human understanding when it has once adopted an opinion draws all things else to support and agree with it.” The mind, in other words, behaves like a lawyer defending a client it has already decided is innocent — not like a scientist trying to find out what’s true.

This course turns that lawyer back into a scientist. You’ll learn the precise definition and the three places the bias operates — in what we search for, how we interpret what we find, and what we later remember — and you’ll watch the Wason task happen to you in an interactive demonstration, so the trap stops being a story and becomes an experience. You’ll see why the same study can leave two opposed camps each more convinced they were right, why echo chambers and algorithmic feeds are confirmation-bias machines, and why the bias is so universal that it must have been useful once. Then we build the antidote — a small set of deliberate moves that don’t take willpower so much as a habit: consider the opposite, steelman the other side, seek the disconfirming case, and pre-register what would change your mind. It assumes you’ve met one earlier idea — that incentives quietly bend belief — and builds the cognitive machinery underneath it. By the end you’ll have the rarest reflex in all of thinking: when you feel most certain you’re right, you’ll have learned to ask the one question the lawyer never does — “What would prove me wrong, and have I actually looked?”