<p>In flocking systems, social relationships structure the interactive framework among individuals and guide collective group behavior. Individuals tend to synchronize their movements with those to whom they are more closely connected. Within such social-relationship-based interaction networks, some individuals exhibit heterogeneity in terms of interaction strength or node attributes. These heterogeneous individuals play a pivotal role in the transition of the group from disorder to order or in the evolution of collective behavior. Although they may not differ in outward appearance, they exert an asymmetric influence on collective dynamics. This study focuses on recognizing the critical individuals within social structures who drive the transition of collective behavior. We find that individual differences mainly lie in the strength of received interactions. Individuals who can interact effectively with more members tend to have more stable movement and influence a wider range of neighbors. To address this, we propose two new metrics based on observable swarm’s trajectories: Surrounding Centrality, which measures how much an individual is surrounded, and Fluctuation Centrality, which assesses the variability of an individual’s trajectory. These metrics integrate spatial and temporal features and include an adaptive selection mechanism based on trajectory characteristics, thus improving the method’s adaptability to different scenarios. Our approach offers a new perspective for uncovering the mechanisms behind collective behavioral transitions and provides a quantitative analytical framework for identifying key individuals within the group.</p>