Concept
Probabilistic Beliefs
Probabilistic beliefs are beliefs held as degrees of confidence, such as being 70% sure, rather than as true-or-false verdicts, so that new evidence adjusts how sure you are instead of declaring you simply right or wrong.
Most of us sort beliefs into two drawers: true and false. The project will ship on time. That candidate won't work out. The split feels decisive, but it hides how much you actually know. It also turns every correction into a defeat: you weren't slightly off, you were wrong. Holding a belief as a probability fixes both problems. Saying "I'm about 60% sure the project ships on time" makes your uncertainty visible to you and to others. It also invites the information that would move the number. When evidence arrives, you update from 60 to 45 rather than defend a verdict. Annie Duke's test in Thinking in Bets is blunt: imagine someone asks, "Wanna bet?" Once money is on the table, most confident claims shrink. What's left after the shrinking is your real confidence level.
A real-life example
A product lead tells her director the launch will definitely happen in March. He asks whether she'd bet a month's salary on it. She hesitates and settles on "maybe 65%." Saying the number out loud prompts him to ask about the other 35%. Two risks surface: a vendor contract and an untested payment flow. They fix one early. When the launch slips to April anyway, nobody calls her wrong. Her 65% had left room for it.
How to use it
- 1Pick one belief you're acting on this week and put a number on it: how sure are you, from 0 to 100%?
- 2Run the bet test: would you stake real money at those odds? If you'd hesitate, lower the number until you wouldn't.
- 3Write down what evidence would move the number up or down by 10 points, and revise it when that evidence arrives.