The illusion of understanding
Lesson 10 of 31 in the course “Brain Hacks”. Free, with animations and a quiz.
“I knew it all along!”
Before the US president’s 1972 trips to China and the USSR, psychologists asked people to rate the chances of various outcomes: would he meet Mao, would there be an agreement? After the trips, the same people were asked to recall their estimates. For events that happened, people “remembered” giving them higher odds than they really had. This is hindsight bias. Once you know the outcome, the brain rewrites the past, and it all seems obvious. That means we overrate how well we understand the world.
Success stories lie by omission
Every unicorn startup later gets a beautiful story: a genius idea, grit, a bold move at the right moment. It sounds like success was inevitable. But thousands of teams with similar ideas, grit and boldness shut down, and nobody writes about them. Luck is almost invisible in success stories because it’s hard to narrate. When you read “how he did it”, remember: the story is assembled in hindsight, and there’s less of a recipe in it than it seems.
Experts and forecasts
One researcher spent nearly 20 years collecting forecasts from 284 experts who made a living advising on politics and economics. Over 80,000 forecasts of the form: the status quo stays, there’s more of something, or less. The result was devastating: the experts did worse than if they’d simply given all three options equal odds. And the most knowledgeable were often the most overconfident. Knowledge helps explain the past, but predicting the future in complex systems is far harder than it seems.
The illusion of skill in investing
A researcher studied nearly 163,000 trades by individual investors: cases where someone sold one stock and immediately bought another. Each was sure they were swapping worse for better. On average it was the reverse: the stocks they sold later did better than the ones they bought, by about 3.2 percentage points a year. And the most active traders had the worst results on average. The feeling “I understand the market” is strong, but over short stretches there’s a lot of randomness.
The planning fallacy
In 1997 the new Scottish Parliament building was estimated to cost up to £40 million. Two years later the budget was £109 million, then £241 million, then £376 million. The building was finished in 2004 for about £431 million, more than ten times the plan. That’s the planning fallacy: we plan for the best case and forget how much usually goes wrong. Danya does the same with his term paper: “I’ll write it over the weekend” turns into three weeks.
The outside view
A well-known psychologist and his team were writing a school textbook. After a year, everyone estimated how much was left: about two years on average. Then he asked the team’s expert: how long did similar teams take? The expert was embarrassed: about 40% never finished, and the rest took 7 to 10 years. He had that knowledge in his head, but hadn’t used it in his own estimate. The textbook was finished 8 years later. The outside view is the question “how does this usually go for others in a similar situation?”.
Hack: the premortem
Before an important plan, imagine: a year has passed and everything failed. Now spend five minutes writing the story of that failure. Why did it happen? Danya tested his plan “learn Python over the summer and find a side gig” this way: ran out of money for the course, trips ate the summer, quit after a month out of boredom. Strangely, once the failure has “already happened”, reasons come much more easily than when you ask “what could go wrong?”. Now you can fix the plan in advance: a free course, a schedule, a study buddy.
Try it: a forecast journal
For two weeks, write down your predictions with a confidence percentage: “ace chemistry — 90%”, “the coin will rise this week — 80%”. Then mark what came true. Danya was surprised: average confidence 84%, yet one prediction in four came true. Without notes, he’d remember only the hits and think he understood everything. The journal treats two things at once: hindsight bias, because you can’t rewrite an old forecast, and overconfidence, because you see your real hit rate.
Check yourself
- A team won a tournament and a friend says: “I knew it.” Why should you take that with a pinch of salt?
Correct answer: Knowing the outcome, memory rewrites past estimates. Yes. In the 1972 trip study, people “remembered” giving events that happened higher odds than they did. That’s hindsight bias, not lying. - Individual investors sold one stock and bought another. What did the data show?
Correct answer: The stocks they sold later did better than the ones they bought. Yes. On average the gap was about 3.2 percentage points a year in favour of the stocks sold. Confidence in understanding the market wasn’t backed by results. - Danya plans: “I’ll build my portfolio site over the weekend.” Which question gives a more honest estimate?
Correct answer: How long does a site like this usually take other beginners?. Right. That’s the outside view: other people’s experience includes everything that usually goes wrong, unlike a best-case plan. - What is a premortem?
Correct answer: Imagine the plan already failed and write down why. Yes. Once the failure has “already happened”, reasons come more easily, and you can fix the plan before starting, not after.