This paper is a case study of a particular kind of gaming arrangement in an internet café, in which two people play a Warcraft III mod called Angel Arena, seated next to each other, while a third person is seated beside them, providing an ongoing commentary and fomulation of the gaming. This session is interesting as a case study because while the two players are controlling avatars in the game, the -third man- is present as a commentator and instructor for the other two, and is therefor not a necessary part of gaming in any -game technical- sense. The question that this gives rise to is if and how -the third man- manages to become part of the gaming interaction, how he gets involved in the situation and how he influences the way in which the game gets played by the controlling players. Interaction in gaming has often been glossed over and could, using the ethnomethodological terminology, be considered a -missing what- in research on games and gaming. Even though ethnographic studies of computer gaming can say something about the contexts and activities of gaming, it will very rarely say anything about the detailed, embodied interactions, on- and offscreen, that make up that gaming. Such details of interaction are only recoverable through the use of video recordnings, and careful and detailed transcription and analysis of the ongoing interaction. Interaction analysis has been conducted under the auspices of ethnomethodology in a wide variety of lay and professional setting, but have only rarely been applied to sessions of computer gaming. This paper is based on a video recording showing the present players and their screens, and has been transcribed and analyzed using conventions of conversation analysis and ethnomethodology. The recording is part of a larger collection of video recordings of gaming sessions at an internet café, and were collected during a short-term fieldwork, where young people-s (aged 11-21) gaming practices were studied. The -third man- tries to ensure that the two controlling players are constantly orienting to his presence as a necessary part of the gaming, and how this orientation is mutually constructed in interaction is interesting as it allows for an analysis of how collaborativity is manifest as a concrete phenomenon in gaming interaction. The -third man- uses a variety of embodied interactional resources in order to secure his place in the gaming session, and the orientation towards him as a bona-fide collaborator to the gaming is observable in a multitude of ways. The gaming as an intersubjective, mutual and concerted activity is accomplished using embodied resources such as pointing, gesturing, posture and spatial positioning, as well as talk and on-screen activities. In negotiating his status as a participant to the activity, the -third man- uses these embodied resources for producing himself as a more competent gamer (in this particular game) than the other two players, who rely on him for tactial and strategical advice, as well as for input on aesthetics and general commentary on the state of the game. By orienting to the assymetirical distribution of knowledge in the gaming session, the -third man- is constructed by the participants as a necessary part of the interaction. Also, by carefully negotiating between the two players, he ensures that he is seen as a neutral party in the gaming activity . This serves as an additional resource by which the -third man- necessitates his presence in the gaming. The ways in which competence gets done and understood by the participants in naturally occuring gaming is related to a discussion of peer cultures and young people-s ways of self-organizing in relation to computer gaming in specific contexts. By demonstrating how gaming competence is produced as an interactional phenomenon, this paper discusses peer culture not as an abstraction or thro