摘要：We analyze competitions where the contestants evaluate each other and find the first contestant to be disadvantaged. We suspect that this is due to information diffusion, Bayesian belief updating taking place in course of the contest and initial uncertainty about a contestant's relative quality
