The Lab Is Not The World

For decades, we’ve been told the human mind is riddled with biases, and now there are more than a hundred on the list. But did you know that most of them were “proven” with only hypothetical puzzles? Take base rate neglect, for example. A hypothetical school is 90% jocks and 10% nerds. Twenty percent of jocks wear glasses. Sixty percent of nerds do. You bump into someone wearing glasses. Jock or nerd? Most people will say nerd only because it feels right, because glasses point to nerds. But the math says the person is three times more likely to be a jock. Ninety jocks at 20% is 18 pairs of glasses. Ten nerds at 60% is 6. Here we have the crowd outweighing the signal. But if you rewrite the same problem with counts instead of percentages: how many jocks, how many nerds, how many of each wear glasses? We see far more people get the answer correct. It is the same math presented in a different format that yields a different result.

That suggests the mistake is sometimes in the translation and not always in the mind. Why? Because our intuitions were shaped by messy, real-world experience, where information arrives as things we count and see, not as abstract probabilities. If you put a good tool in an unfamiliar setting, it can look broken. Ergo, before you call a behavior irrational, ask what environment it was built for.

When presented with these moments, don’t simply articulate one plan needing improvement. Instead, find the time to create different plans that don’t overlap at all. Can you refine and improve each one enough that you’d be happy randomly choosing any one of them? Because by the time you’ve built out independent plans, it’s quite likely you’ll have discovered the right path.