Abstract: In order to study the influence of agents’ action strategies on the evolution of social economic system, the basic majority game and the evolutionary majority game in which agents prefer to stay in the majority are put forward on the basis of minority game. We build agent-based models,all the agents in the model are given bounded rationality, and they use their best strategy or modify their strategy probility each round in order to compete for finite resources. When evolving to stable state, phenomena which are unique to human society also occur. In the basic minority game, if agents in the majority are ruled winners, we get the basic majority game,and the game has higher resources utilization when the dimension of the strategy space, m, is smaller. As m gets larger, the utilization becomes lower, approaching that of the random choice game finally. While the utilization of evolutionary majority game isn’t affected by m, so,when m in the game is larger,the evolution is so perferable to improve the utilization of the system. With the same parameters in the game,different random initial conditions may cause different side of distribution of P(p) at p=0.5.The average number of winners each round is also found to be related to the distribution of P(p) at the same time. If we use different boundary conditions to reset agents’ strategy probabilities, we will also get different distribution of P(p). For further analysis of evolutionary majority game’s utilization, we find that the lager the parameter r which means the relation between agents’ old and new p,the flatter distribution of P(p) is,and the utilization also gets larger.At last, if we improve R representing the value of the prize-to-fine ratio, the distribution of P(p) will also be affected,resulted to a lager utilization of the game.

Key words: action strategies, minority game, basic majority game, evolutionary majority game, bounded rationality