Sense and decide
Wanna bet?
Ask someone whether they would stake something on a claim, which exposes the gap between how sure they sound and how sure they are.
Borrowed from former professional poker player Annie Duke, in her book Thinking in Bets. Read the original source.
One session · Minutes to start · Anyone on the team can start it
Try this first
When a confident claim is about to drive a decision, ask whether they would bet on it. The pause does the work.
Use it when
Someone asserts with more bluster than conviction.
Skip it when
Use it on claims that matter, not as needling. Played harshly it makes people defensive and shuts down the honesty you wanted.
How to introduce it
When someone states something with confidence, ask whether they would bet on it. The question is friendly and it does real work. It separates how sure a person sounds from how sure they are, which is the gap that matters before the team acts, including on a confident model output.
How to show up
Keep it light. This is calibration, not a trap. Use it on your own claims so it reads as a shared habit.
How long it takes
Seconds, in the moment.
What makes it hard
Tone carries this one. Kept light it works, kept sharp it wounds. People often hedge once challenged, which is the useful part. The claim was softer than it sounded.
What it looks like when it's working
Confident claims get calibrated before they drive decisions. Bluster still carrying the room means nobody is asking.
How long until it sticks
Expect a few uses before the team gets comfortable with the friendly challenge.
How you know it stuck
People flag their own gap between sounding sure and being sure.
The idea behind it
Sounding sure and being sure come apart all the time. A bet finds the seam.
Where it comes from
Probabilistic decision-making, especially Annie Duke's work. A light challenge that separates conviction from bluster before the team commits. The prompt that surfaces how confident someone is by asking them to stake something on it.