Anchor on base rates before adding details

Start with how common the outcome is in the relevant population, then update — don’t start with the story.

Why it works

The conjunction fallacy is a symptom of inverting the correct process: a vivid description is assembled first, and probability is reverse-engineered to match it. Anchoring on base rates first — how often does this kind of outcome actually occur? — gives deliberate reasoning a numerical floor that resists narrative inflation. Even a rough base rate drastically reduces sensitivity to representative-but-irrelevant details because the anchor shifts the reference frame from "how well does the story fit?" to "how often does this occur?"

How to do it

1. Before engaging with the specific details of a claim, ask: "Among cases in this category, what fraction have the outcome I’m evaluating?"
2. Write that base rate down as your starting probability.
3. Update for specific details only if those details are actually diagnostic, not just coherent.
4. Watch for cases where details feel relevant but don’t change the base rate.

Evidence

Base-rate neglect is a documented judgment failure, and instruction to attend to base rates partially corrects conjunction errors in trained subjects. Kahneman and Tversky’s work on base-rate neglect directly underpins the conjunction fallacy literature. (observational)

Anchoring on base rates helps but does not fully eliminate the fallacy; the representativeness pull is strong even when subjects know about base rates.

Sources

• Kahneman & Tversky (1973), On the psychology of prediction, Psychological Review

Common mistake

Treating every additional detail as diagnostic information and updating upward on each one — details can make a story feel complete without changing its underlying frequency.