The key player problem (KPP) identifies a set of key nodes that have a central role in a network. In this paper, we propose a generalized KPP (GKPP) that extends existing work on KPP-Pos and KPP-Neg in such a way that it can consider network structure, node attributes, and the characteristics of edges. We also articulate a novel concept called the key player problem for exclusion (KPP-E), which selects a set of nodes to enforce the centrality of a given set of nodes of interest. To solve this problem efficiently, we propose a sequential greedy algorithm that significantly reduces computational complexity. To corroborate the conceptual meaning and effectiveness of the proposed sequential greedy algorithm, we apply GKPP and KPP-E to several real and random networks.