Base rate neglect when we think about possible scenarios or answers

One of the consistent attacks on the rational homo economicus is that we so often fail to let our judgments of probability stay close to an informative statistic, referred to by cognitive researchers as a “base rate.” If you were asked the average number of lawyers in the AmLaw 200 firms, you might quickly say 400, or some number. You retrieved that number so quickly that you didn’t stop to think about what you might know that could give you a decent baseline for your estimate: the sizes of large firms retained by your department, your sense of firm size when you last interviewed, or the list you reviewed of global large firms or any other source of a base rate.

We should start with base rate clues, even if they are partial. More generally, and the source of this post’s idea, Daniel Kahneman, Thinking, Fast and Slow (Farrar, Straus & Giroux 2011) in Chapter 14, at 154, urges us to question the diagnosticity of our evidence – the degree to which it favors the hypothesis over the alternative.