在酒吧模型的基础上，本文提出了集市模型A和集市模型B。在三个模型中，主体采用不同的行动策略：在酒吧模型中，他们采取少数者行为；在集市模型A中，他们采取多数者行为；在集市模型B中，他们采取少数者与多数者混合行为。本文讨论了主体的预期规则、学习和优化行为，并通过参与人数均值的收敛情况考察了不同行动策略下模型演化的均衡状态。发现对于不同的行动策略方式，会出现预期的自我实现和自我毁灭的两种现象，主体需求得到满足的程度——即集市的最终参与情况不同。在集市模型A中，集市出现两种极端状态：拥挤和冷清，集市的初始参与人数和历史参与数据对于主体行为选择的正反馈作用决定了模型演化的最终状态；在集市模型B中，主体行为选择受到合意人数范围限制时，集市的最终参与状态受到主体要求的市场合意参与人数的值和范围大小的影响。

Inspired by the El Farol Model, this paper puts forward fair model A and fair model B. Agents take different action strategies in these three models: they prefer to stay in the minority side in the bar problem, while they prefer to stay in the majority side or they play mix-game in the fair models. This paper discusses the prediction rules、the learning process and the optimization strategies, and studies the long-time behavior of the three models as indicated by the convergence of the mean attendance in different parameters setting. The emergent phenomena are different in these three models: 'self-fulfillment of expectation' and ' self-destroying of expectation'. And it shows that the satisfaction degree of the agents' requirements of exchanging goods, which can also be measured by the fair participation level, is different when they take different action strategies. The country fair with agents who choose to stay in the majority side seems to be either crowded or empty, and the initial condition and the positive feedback of the historical data determine the equilibium state in the fair model A. The requirements of the agents affect the market's emergent phenomenon when the agents choose to join the market with consensus size range in the fair model B.