Centered Low-Intelligence Subjects with Appropriate Noise Enhance Coordination in Networks

In the era of intelligent interconnection of all things, society gradually forms a system composed of numerous in-telligent entities that collaborate and evolve through division of labor. However, in reality, high-intelligence individuals often fail to achieve high system efficiency after interacting to form intelligent groups. The relationship between interactions among intelligent individuals and the emergence of system efficiency remains to be explored and answered. In this study, we use color coordination game as the basic scenario to investigate the impact of decision error rates of different types of intelligent individuals, the proportion of the same type of intelligent individuals, and the differences in decision error rates within structured groups on problem-solving. Through simulation experiments, we find that in high-intelligence groups, adding an appropriate number of relatively low-intelligence individuals can optimize and stabilize problem-solving. Conversely, adding lower-intelligence individuals to a low-intelligence group will decrease overall performance. Notably, when the added lower-intelligence indi-viduals are at the center of the group, their impact on problem-solving efficiency becomes more pronounced. Additionally, we discover that having significant intelligence differences within high-intelligence groups and moderate intelligence differences within low-intelligence groups can also enhance overall group efficiency. This research provides an explanation for the typical phenomenon of mismatch between individual intelligence and system efficiency in collaborative processes and offers guidance for improving system efficiency in real-world social scenarios.