MModern Wisdom
← All frameworks
MindsetAnnie Duke

Make Beliefs Explicit and Close the Loop

Turn gut calls into examinable forecasts you can check and learn from

Difficulty
Advanced
Time to result
~ongoing to results
Steps
6
Confidence
82%

Duke argues great decision-making requires two intentional acts: seeing the luck accurately and examining the beliefs that inform your choices. The gut fails not because it's always wrong but because it's a black box where bias and noise live; you can't examine it, teach it, or check it. Her fix is to make the implicit explicit. Break a decision into its component beliefs and state precisely what must be true of the world for the choice to win, for a stock, exactly what you think you know that the market doesn't. Express those beliefs as probabilities or bounded ranges, then record the forecast. Because it's now explicit, you can close the feedback loop later: compare what happened to what you predicted and measure how calibrated you were. That accountability is what turns experience into genuine learning instead of pin-the-tail-on-the-donkey.

Origin

Annie Duke built this from cognitive science and 18 years of professional poker, where noisy, high-luck environments make implicit gut reads impossible to learn from without explicit, checkable forecasts.

Core principles

  • 01See the luck accurately: map the spread of outcomes and their probabilities
  • 02The gut is where cognitive bias and noise live and can't be examined
  • 03A real decision tool is repeatable, teachable, and accountable
  • 04Explicit forecasts let you close the feedback loop and actually learn

How to run it

  1. 1

    See the luck

    Map the spread of outcomes that could occur and how likely each is, accepting you can't control luck but can perceive it.

  2. 2

    Break the decision into component beliefs

    Instead of 'it looks like a good stock,' identify the specific beliefs the choice depends on.

    Pro tip Ask what you think you know that the market doesn't; that's the belief you must defend.

  3. 3

    State what must be true to win

    Make explicit the conditions of the world that have to hold for this to be a winning decision.

  4. 4

    Forecast precisely

    Express each belief as a probability or a lower-and-upper bound rather than vague language.

    Pro tip Precision is what lets you later tell whether you were right, so weasel words can't rescue a bad call.

  5. 5

    Record the forecast

    Write down your predicted states of the world and their probabilities at the moment you decide.

    Watch out We misremember our past state of knowledge; without a record you can't honestly reconstruct it.

  6. 6

    Close the loop and check calibration

    After outcomes arrive, compare them to your recorded forecast to see how well calibrated you were and update.

In the wild

Justifying a stock purchase

Rather than buying a stock because you 'have a nose for value,' you pull the decision apart: you believe the market is underestimating how many widgets the company will sell. You forecast a specific unit range and the probability the market is wrong, and you write it down. Months later you compare actual sales and the stock's move against your recorded forecast, learning whether your edge was real or luck, and recalibrating for the next call.

A checkable track record that converts each investment into a genuine learning loop.

Common mistakes

Trusting an unexaminable gut

Because a gut call is implicit, you can't inspect its reasoning, teach it, or check it, so you can't tell a good decision from a lucky one.

Skipping the written forecast

Without a recorded prediction you'll misremember what you believed, and hindsight will corrupt any lesson you try to extract.

Is it for you?

Best for

Consequential, repeated decisions like investing or strategy where calibration and learning compound over time.

Not ideal for

Low-stakes, high-frequency choices where the overhead of explicit forecasting isn't worth it.

From the transcript

in order to be a great decision maker you have to do two things ... you have to understand and see the luck ... but…

Annie Duke · 00:00

i have no way to go back and close the feedback loop into in order to create a really good learning loop because i can't…

Annie Duke · 20:30

From the episode

How To Make Better Decisions - Annie Duke - #233

Annie Duke