The collective wisdom of behavioral game theory
研究发现，将七种行为博弈理论模型加权平均，能比任何单一模型更准确地预测一次性同时行动博弈中的策略选择，不同模型通过捕捉不同行为模式实现互补。
Abstract We apply an algorithm from the wisdom-of-crowds literature to optimally combine behavioral game theory models to more accurately predict strategic choice in one-shot, simultaneous-move games. We find that the optimal weighted average of seven behavioral game theory models predicts out-of-sample choice behavior significantly better than any of the individual models. The crowd of behavioral game theory models is wiser than any single one of them. Different strategic choice models complement each other by capturing distinct patterns of behavior. The field of behavioral game theory is enriched by having this diversity of models.