When to trust your intuition
Lesson 25 of 31 in the course “Brain Hacks”. Free, with animations and a quiz.
The statue that “looked wrong”
In 1983 a major museum in California set out to buy an ancient Greek statue for a huge sum. For fourteen months it was checked: a geologist studied the marble, lawyers examined papers on past owners. Everything matched, and the museum bought it. But several experts in ancient art, seeing it for the first time, instantly felt something was off. One said it looked “fresh”; another felt a chill. They couldn’t explain it right away. Later some of the ownership letters turned out to be fake: a letter dated 1952 had a postcode that didn’t yet exist. The debate continues, and the museum labels it “Greek, about 530 BC, or modern forgery.” This lesson covers when such an instant hunch can be trusted.
Does the body know before you do?
A famous card experiment. Four decks in front of you. The two red ones have big wins but huge losses, and in the end they ruin you. The two blue ones have modest but steady wins. No rules are explained. After about 50 cards people got a hunch that blue was better. After about 80 they could explain why. But sensors on their palms showed a stress response to the red decks by the tenth card. Around then, people also began taking fewer of them. It sounds like proof that the body is smarter than the head. But honestly: other scientists later showed participants understood the game earlier than thought, and the debate continues. The cautious takeaway: experience builds up in autopilot before we can put it into words.
The firefighter’s sixth sense
Another well-known story. Firefighters entered a house where the kitchen was burning and started hosing it down. Suddenly the commander heard himself shout: “Let’s get out of here!” He didn’t know why. As soon as the team ran out, the floor collapsed. The main fire turned out to be not in the kitchen but in the basement right beneath them. Only afterwards did the commander work out what alarmed him: the fire was too quiet for a blaze like that, and his ears were hotter than usual. Together these cues gave a sense of danger. No magic here. Over years of work his autopilot had seen hundreds of fires and learned what a “normal” one looks like and what a dangerous one looks like. Expert intuition is fast recognition of a familiar pattern. It works because huge experience stands behind it.
You’re an expert too, you just don’t notice
Stories about firefighters and chess masters seem like magic. But each of us does similar things many times a day. You hear anger in a friend’s voice from the very first word on the phone. You walk into a room and know they were just talking about you. On the road you notice the driver in the next lane is dangerous before they’ve done anything. You can’t explain how you knew, but you’re usually right. That’s the same expert intuition. Behind it are thousands of conversations and trips where the brain saw patterns and got instant confirmation. So in familiar social situations your gut often works brilliantly. The question is how to tell those situations from ones where it only pretends to know.
Three conditions for trained intuition
Psychologists who long argued about whether to trust intuition eventually agreed on clear conditions. One: the environment must be regular enough, meaning it has stable patterns. Chess, medicine, fires, games, talking to people all follow rules, however complex. Two: long practice. It takes hundreds and thousands of repetitions for autopilot to learn those patterns. Three: fast, clear feedback. You need to find out quickly whether you were right. A chess player sees a move’s result at once, but a doctor making a forecast years ahead learns of a mistake too late. If all three are present, intuition is probably trained and can be trusted. If even one is missing, be careful.
Where confidence means nothing
There are areas where these conditions aren’t met, and there a hunch is nearly useless, even for professionals. Guessing which stocks will rise over years. Making long-range political forecasts. Predicting which number roulette will land on. There’s too much randomness, the patterns are weak, and feedback comes late or teaches nothing. The trickiest part: people in these areas are often very confident. But confidence is just a feeling; it doesn’t prove accuracy. So one of the key pieces of advice sounds harsh: don’t treat anyone’s confidence, including your own, as proof. Look not at how sure someone is, but at whether they have those three conditions.
Danya and two different hunches
Danya has about three thousand hours in his favourite shooter. He senses where an opponent will appear and is almost always right. All three conditions are here: the game has clear rules, practice is huge, and you see a decision’s result in seconds. That hunch can be trusted. Danya has also been trading crypto for two months, and he feels he “senses the market” too. But here the conditions aren’t met: the market is noisy, there’s lots of randomness in the short term, practice is small, and a winning trade may be luck that teaches nothing. Danya concludes: in the game he trusts himself, and in trading he checks every “I feel it’ll go up” against his rules and logs it in a trade journal, so that in six months he can honestly compare the hunch with the results.
Try it: three questions for your hunch
Next time you feel a strong “I just know”, stop for ten seconds and ask three questions. One: does this area have stable patterns, or lots of randomness? Two: have I made decisions like this many times? Three: did I find out quickly and clearly whether I was right? If all three are “yes”, it’s probably experience speaking, and it’s worth listening to. If even one is “no”, treat the hunch as a hypothesis, not a fact. Test it with data, advice from someone who knows, or a small trial. And get into the habit of writing down your strong hunches with a date. In a couple of months you’ll see where your intuition gets it right and where it just talks loudly.
Check yourself
- Under what conditions can intuition most likely be trusted?
Correct answer: Regular environment, long practice, fast feedback. Yes. All three let autopilot learn real patterns. - Where is even a professional’s hunch nearly useless?
Correct answer: In forecasting stock prices years ahead. Yes. Lots of randomness, weak patterns and late feedback. - In the deck study palms “reacted” by card 10. Why is “the body is smarter than the head” marked “disputed”?
Correct answer: Later work showed people understood the game earlier than thought. Yes. The study is real, but its interpretation is still debated. - Danya has traded for two months and “senses the market.” What best follows the lesson?
Correct answer: Treat the hunch as a hypothesis, check with rules, keep a journal. Yes. In a noisy area with little practice, a hunch is a hypothesis; the journal will show if it’s right.