Update incrementally as evidence arrives rather than waiting for certainty

State your current best-guess probability, identify what would shift it, and update when that evidence arrives.

Why it works

Ambiguity aversion often produces binary thinking: “I don’t know enough” vs. “I know enough.” Bayesian updating reframes this as a continuous process: each piece of evidence shifts your probability estimate, and you don’t need certainty to act — only sufficient confidence relative to decision stakes. Making the update process explicit — stating your prior, the evidence, and your posterior — reduces the emotional weight of ambiguity and replaces it with a tractable question: how much evidence do I need for this specific decision?

How to do it

1. State your current best-guess probability for the key outcome (your prior), even if it’s rough.
2. Identify what evidence would shift that estimate and by how much.
3. As evidence arrives, explicitly revise your estimate and record the revision.
4. Set a confidence threshold for action in advance: “I’ll act when I reach X%.”

Evidence

Superforecasting research (Tetlock & Gardner, 2015) shows that explicit probability estimation and systematic updating outperforms gut-feel judgment for ambiguous outcomes. Good Judgment Project studies documented significant improvements in forecasting accuracy through probabilistic thinking training. (observational)

Bayesian updating requires good signal; in highly novel situations, early evidence may be unrepresentative and updating on it too aggressively can lead to premature closure.

Sources

• Tetlock, P.E., & Gardner, D. (2015). Superforecasting: The Art and Science of Prediction. Crown.

Common mistake

Stating a prior but never actually updating it — the practice requires revisiting the estimate when real evidence arrives, not just setting a number and ignoring subsequent information.