How Rich is the Rich-Club? Super Rich-Club Organization in Complex Networks

Abstract

The rich-club coefficient is used in complex networks to quantify the tendency of progressively higher-degree nodes to be more densely connected than expected by chance. However, it does not capture whether high degree nodes towards the core are actually more interconnected than the original network. Here, we jointly analyze the rich-club coefficient and an enrichment coefficient, which quantifies whether the internal average degree of progressively higher-degree subgraphs is greater than that of the original network, despite the decreasing number of nodes. We apply this framework to real and synthetic networks. Our results reveal three distinct regimes of core organization: rich-club organization without enrichment, enrichment without rich-club organization, and a combination of both, which we call Super Rich-Club organization. In the latter regime, high-degree nodes form cohesive and unusually dense cores that support efficient communication. Our results show that characterizing core organization across structural scales requires considering both relative density and internal connectivity, revealing a stronger form of core organization in real systems with implications for network structure, spectral properties that govern dynamical processes, and communicability.