Ratio Bias

What is ratio bias?

Ratio bias is a tendency to let the number of favorable outcomes influence a probability judgment or choice without adequately considering the total number of possible outcomes. The numerator can become more compelling than the proportion.

Ratio bias example: red beans

In Denes-Raj and Epstein’s 1994 study, participants sometimes preferred drawing from a bowl with more winning red beans but a lower winning proportion, such as 7 red beans out of 100 instead of 1 out of 10. The corresponding probabilities are 7% and 10%. Some participants reported this preference despite knowing which odds were better.

Ratio bias and equal ratios

Ten out of 1,000 and one out of 100 are both 1%. Treating them as different chances because one contains more winners would be another illustration; both express the same chance of winning.

Limits

The bean experiment demonstrates a particular probability conflict. It does not establish how often this pattern determines decisions in every real-world setting or whether a given communication change corrects it.