Do players correctly estimate what others do?: Evidence of conservatism in beliefs

n a simple experimental environment a group of subjects was asked to give estimates of a second group’s choice frequencies in a set of lottery-choice tasks. The results show that subjects in the first group are on average able to correctly predict the option that is chosen with higher frequency by the second group, but the predictions are systematically inaccurate in that they are distorted toward the uniform prior. Two mechanisms to elicit the expectations were used in the experiment, a quadratic scoring rule and a bidding mechanism. Aggregate results being similar under both mechanisms, the use of the former mechanism consistently yields more accurate predictions.