In this paper we investigate the problem of optimal games played over networks and focus our attention on the importance of the topology of communication between the agents. We consider a set of agents which are connected via a directed communication graph: each agent in the network has to optimize a local cost function which depends on the agent's decision and on the decision taken by the set of its neighbors, giving rise to a Network Game. We show that, by condensing the strongly connected components of the control graph into super-nodes, it is possible to give a hierarchical interpretation to the Network Game. Then we apply the proposed architecture to the case of a large scale network which takes inspiration by traffic networks application.