The paper is a critical analysis of Hubert Schwyzer’s idea of meta-institutional concepts. First, I isolate a presupposition in Schwyzer’s example of chess as ritual. I then show how Schwyzer’s idea of meta-institutional concepts is far from being the endgame in the research on levels of institutionality. In fact, we can iterate on meta-institutional concepts. Schwyzer’s idea has to face an infinite regress. I try to avoid such a regress by introducing the concept of technical end of game. A game defines its own terminal status. People playing the game can then attach different meanings to this norm-dependent terminal status. Hence, meta-institutional concepts are not conditions of possibility: they belong to pragmatics, not to an extra ontological layer. I conclude the paper sketching a classification of games based on different kinds of technical ends and I advocate the need to adopt a pluralistic conception on games and practices.