A laboratory experiment on the minority games with aggregate information and partial information.

Abstract

The impacts of different information treatments on the Minority Games have attracted wide attention from researchers. Further to preceding research on impacts of aggregate information and full information treatments in the Minority Games experiment, this paper investigates the effects of partial information on both efficiency level and payoffs distribution through laboratory experiments. The experiment contains 12 sessions with 20 participants and manipulates two variables, namely, the length of information history provided and the amount of publicized action history of individual players. The research testifies the previously studied relationship between allocative efficiency and the length of information history. Meanwhile, the research discovers a positive correlation between allocative efficiency in the Minority Game and the amount of partial information provided. However, analysis of payoff outcomes indicates that partial information does not significantly influence the payoffs distribution. The paper concludes that partial information has positive effects on the efficiency of the Minority Game and ambiguous effects on its payoffs distribution, and the payoffs distribution is more related to number of strategy changes.