Thinking in Bets: Write the Decision Down First

It's a Sunday in February 2015, and the Seattle Seahawks have the ball on the 1 yard line with 26 seconds left in the Super Bowl. Pete Carroll calls a pass. It's intercepted, New England wins, and the call becomes shorthand for coaching malpractice.

Annie Duke watched the reaction and saw something familiar. She'd spent years as a professional poker player, and early on the veterans warned her about a specific error. They called it resulting, and it means grading a decision by its outcome.

The interception told you almost nothing about whether the call was sound. Duke's book Thinking in Bets opens with a harder claim: two forces determine how your life turns out, the quality of your decisions and luck, and we're bad at keeping them apart. A good process can lose the hand, and a bad one can win it.

The reason is hindsight bias. Once the ending is known, it feels like it was always going to happen. The sentence that gives you away is, I should have known that would happen.

Say it and you've already rewritten the record.

Duke's fix is small and unglamorous. Attach a percentage to your beliefs. She says "I'm 70% sure" instead of "I'm sure."

That phrasing is more accurate, and it does something else: it invites the person across the table to hand you the information you're missing. Admitting uncertainty is how you get corrected.

Then she turns to the part that matters if you run a company. Deciding whether an outcome came from skill or from luck is itself a bet, and she calls it a consequential one. Every piece of feedback passes through that sort.

Most of us do it in a self-serving direction, taking credit for the wins and writing off the losses as variance. Get the sort wrong often enough and experience stops teaching you anything.

So the prescription is to record your state of knowledge before the outcome arrives: what you believed, how confident you were, what you expected to see if you were right.

For a founder, this is the case for a decision log. The record has to be written while you're still ignorant of the ending, because the moment you learn it your memory starts bending toward the result.

Six months later, the log is the only account of your thinking your memory never got to edit.

That's the part a book can argue for and can't do for you. Thinking in Bets makes the case for writing the bet down. The writing is the work.

Grade the decision. Note the luck. Run the next experiment.