Tennis You have been invited to bet on one of three tennis games. In game 1, two extraordinarily good tennis players are up against each other. In game 2, two extraordinarily poor tennis players are up against each other. In game 3, one very good and one very bad player are up against each other, but you do not know which is good and which is bad. As a result, as far as you are concerned, the probability that any given player will win is 50 per cent. Suppose that you are ambiguity averse. Which of the three games would you be least likely to bet on? Why?

There is no principled reason why people cannot be ambiguity prone rather than ambiguity averse. In fact, evidence suggests that people’s behavior in the face of ambiguous probabilities depends on the context. According to the competence hypothesis, for example, people are less averse to ambiguity in contexts where they consider themselves particularly knowledgeable. Thus, a football fan may be ambiguity averse in the Ellsberg case (where outcomes are completely random) but ambiguity prone when predicting the outcomes of football games (where he or she feels like an expert).