Coordination and Emergence in the Cellular Automated Fashion Game

Authors:
  • ZCZhigang Cao
    Zhigang Cao
  • HGHaoyu Gao
    Haoyu Gao
  • XQXinglong Qu
    Xinglong Qu
  • MYMingmin Yang
    Mingmin Yang
  • XYXiaoguang Yang
    Xiaoguang Yang

Abstract
We investigate a heterogeneous cellular automaton, where there are two types of agents, conformists and rebels. Each agent has to choose between two actions, 0 and 1. A conformist likes to choose an action that most of her neighbors choose, while in contrast a rebel wants to be different with most of her neighbors. Theoretically, this model is equivalent to the matching pennies game on regular networks. We study the dynamical process by assuming that each agent takes a myopic updating rule. An uniform updating probability is also introduced for each agent to study the whole spectrum from synchronous updating to asynchronous updating. Our model characterizes the phenomenon of fashion very well and has a great potential in the study of the finance and stock markets. A large number of simulations show that in most case agents can reach extraordinarily high degree of coordination. This process is also quite fast and steady. Considering that these dynamics are really simple, agents are selfish, myopic, and have very limited information, our findings are rather surprising and exciting. Phase transition, as well as emergence of many interesting patterns, such as continents, strips, mazes, and wheels, are also observed.