Consult an outside observer to counteract narrative pull

Describe the scenario to someone who doesn’t share your narrative context and listen to their probability estimate.

Why it works

The conjunction fallacy is driven by the "inside view": you’ve built a coherent story and all its elements feel probable because they cohere with each other. An outside observer hasn’t assembled that narrative, so they’re more likely to assess probability from base rates and less likely to be inflated by the coherence of the story you’ve told. This is the same mechanism as Kahneman’s outside-view forecasting: another person’s estimate counteracts the narrative immersion that makes your own estimate feel solid.

How to do it

1. Describe the scenario briefly (30 seconds) to someone who hasn’t been involved in assembling it.
2. Ask: "How likely does this seem to you — what percentage chance?" before sharing your own estimate.
3. Compare their number to yours; investigate any gap larger than 20 percentage points.
4. Weight their estimate more heavily if your gap is in the "narrative makes it feel more likely" direction.

Evidence

The outside-view correction in forecasting research (Lovallo & Kahneman, 2003) shows that observers uninvolved in a project produce more accurate predictions than the project team. Direct evidence for this as a conjunction-fallacy correction is mechanistic, but the transfer is well-motivated. (mechanistic)

Outside observers can carry their own biases; the value comes from their lack of narrative investment, not from superior knowledge.

Sources

• Lovallo & Kahneman (2003), Delusions of success: How optimism undermines executives’ decisions, Harvard Business Review

Common mistake

Briefing the outside observer too thoroughly before asking — the moment you’ve shared your full narrative, you’ve imported your conjunction inflation into their assessment.