Probabilistic Thinking

Abstract

While games of chance have been played for thousands of years, and while experienced players often had good intuitive ideas of the relative frequencies of various outcomes, the concept of probability as referring to the ratio of favorable outcomes to total outcomes of a ‘fair’ gambling device emerged only 500 or so years ago in Western societies. That allowed evaluation of new games not yet played and later led to a more abstract conception of probability as a measure satisfying certain conditions (‘axioms’). Only in the past 150 years or so has probability had a major role in science and only in the last 50 years or so in everyday life—as for example in evaluating medical outcomes or technological risks or manufacturer liability. Thus, especially in attempting to assess probability in everyday situations, our intuitions are often deficient—more so than those concerning quantity, space, and time. The major problem is not, however, one of making random errors in probability judgments but of making systematic ones that result from a number of well-known and research systematic cognitive biases and heuristics, which are described in this article. These systematic departures from rational assessment do not imply that coherent and accurate probabilistic thinking is (anywhere near) impossible, but that when departures from coherence and accuracy occur, they tend to follow a predictable pattern.