ABSTRACT
Normal situations sometimes yield abnormal behavior. We know there are serial killers, pathological liars, paranoid psychotics. But they are uncommon enough that unless your line of work is psychiatry or criminal investigations, they would not be prominent in your understanding of people’s behavior. But in data from experiments it is not so unusual to see a major fraction of choices violate what both common experience and essentially universal social norms would lead us to expect. The problem is not that players may behave differently in the lab. Something of that has to be expected, since no lab experiment can capture exactly the conditions of choice outside the laboratory. But the sort of problem I want to explore here arises when choices in an experiment look so completely different from what we might expect that there is a challenge in conjecturing what kind of context might prompt such abnormal responses from apparently normal people.
