Assessment of groups in a network organization based on the Shapley group value

• We extend the individual game-centrality of Gomez et al. (2003) to the framework of groups.
• We define the Myerson group value (based on the Shapley Group value), for the assessment of groups in a network organization.
• We offer several decomposition of the Myerson group value to obtain a precise interpretation of this complex group measure.
• We illustrate how to use empirically the proposed methodology in a well-known advice-seeking-network (Borgatti, 2006)

The focus of this paper is the assessment of groups of agents or units in a network organization. Given a social network, the relations between agents are modeled by means of a graph, and its functionality will be codified by means of a cooperative game. Building on previous work of Gomez et al. (2003) for the individual case, we propose a Myerson group value to evaluate the ability of each group of agents inside the social network to achieve the organization's goals. We analyze this centrality measure, and in particular we offer several decompositions that facilitate obtaining a precise interpretation of it.