Skip to main content

Ambiguity Aversion

Preferring a known risk to an unknown one, even at a cost

ProbabilityDecision-making

What is it?

Ambiguity aversion is the tendency to prefer an option whose odds are known over one whose odds are unclear, even when the unclear option may be just as good or better. Daniel Ellsberg illustrated it with urns: many people would rather bet on drawing a colour from an urn with a known fifty-fifty mix than from an urn whose mix is unknown, and they keep that preference whichever colour they bet on. Missing information about the odds feels like a risk in itself. Craig Fox and Amos Tversky later found that the aversion is strongest when a clear option sits right next to a vague one, or when you feel less knowledgeable than others. In decisions, it pushes people toward the familiar employer, the known investment, the treatment with published statistics or the market they already serve, and away from new options simply because they are harder to quantify. It tends to get worse when you will be judged by others and when you feel out of your depth.

Example

Turning down a promising startup offer because the company's future is harder to estimate than your current employer's. Keeping savings in a familiar fund instead of one you have not researched. Choosing the well-documented supplier over a new one whose track record is simply unknown.

References

Ellsberg, D. (1961). Risk, Ambiguity, and the Savage Axioms. The Quarterly Journal of Economics, 75(4), 643-669.

Fox, C. R., & Tversky, A. (1995). Ambiguity Aversion and Comparative Ignorance. The Quarterly Journal of Economics, 110(3), 585-603.

How to Prevent It

Doxa uses AI and can make mistakes. How it's built

Question

Am I rejecting this because it is worse, or because it is harder to estimate?

Question

What would a reasonable range of outcomes look like for the unclear option?

Question

Is the known option truly safe, or just better documented?

Question

Do I feel less competent here than the people I am comparing myself to?

Question

What could I learn cheaply to make the unknown less unknown?

Technique

Write a best, worst and likely case for each option, including the vague one.

Technique

Look for base rates from similar situations to put rough numbers on the unknown.

Technique

Evaluate each option on its own before comparing them side by side.

Technique

Run a small, reversible trial of the unfamiliar option.

Technique

Ask someone who knows the unfamiliar domain to fill in what you are missing.