probability-base-rate-anchoring

Base rates improve probability forecasting by anchoring estimates in an appropriate reference class and its historical frequency before applying case-specific adjustments. This outside-view approach produces defensible calibrated probability ranges instead of overconfident guesses.

The outside view in risk estimation grounds predictions in historical base rates for comparable situations rather than case-specific details. You need it when forecasting outcomes or evaluating whether your initial probability estimates are overly optimistic.

Adjust base rate probabilities by identifying verifiable differentiators that explicitly push the likelihood up or down from the historical frequency. Combining these structured adjustment factors with the initial base rate yields a final calibrated probability estimate.

Selecting a reference class for base rate anchoring requires defining a genuinely similar category of past cases large enough to yield stable historical frequencies. A well-chosen reference class ensures the outside-view base rate accurately reflects the target outcome.

Ignoring historical base rates leads to forecasting mistakes by bypassing the outside view, causing predictions to rely solely on case-specific details. This often results in overconfident estimates that fail to account for the actual historical frequency of comparable outcomes.