摘要：We introduce a novel beauty contest experiment to study the gap between individual preferences and beliefs about collective preferences of physical attractiveness. In the first round, participants vote their 3 favorite ladies in the second round they vote the 3 ladies they believe were the most voted in the first round.Unlike other beauty contest experiments, our setup does not investigate depth of reasoning in a cognitively intense task. Instead, it is meant to investigate the existence of shared definitions of physical attractiveness, and whether these may be successfully employed as focal points.Our results show that most participants hold mistaken beliefs about collective preferences and overestimate and underestimate how well liked certain ladies are. Regardless of these mistakes, the winning portraits win by a wide margin in both rounds. Moreover, our participants are better at predicting the portraits which will not be the most voted than those which will.
