Political opinion formation: Opinion distribution and human heterogeneity

Political party competition has always been a hot topic both in social science and self-organized system. Previous agent-based simulations mainly focus on construction of new models or analysis of network structures. However, even under the same network structures and with the same simple mechanism, the results of opinion formation can still be various. On the basis of a classical opinion model, we consider the coevolution of initial distribution of opinions as well as the human heterogeneity, and study the critical phenomenon of fragmentation, polarization and consensus. We show that the initial distribution of opinions leads to different evolution results on similar random networks. Besides, the existence of stubborn nodes plays a special role, only nodes that are extremely stubborn can cause the change of final opinion distribution while others only delay the time of reaching stable. We further introduce the average ratio of near neighbours to characterize the stage of opinion formation. This theoretical work goes deeper on an existing framework, it is an early attempt on qualitative and quantitative study of party competition and it provides substantial inspiration for decision-making in business.