We studied the content of the team mental model with a case study about a successful online game team. This team was formed by high-ranking managers in a large raid guild to conduct a raid in War of Worldcraft. They were interviewed with a set of 5-W questions developed according to the conceptualizations of team mental model (Levine & Moreland, 1991). Toward the interview protocol, a content analysis was conducted with two-cycle coding method from exploratory to explanatory and a hierarchical framework from code, category, theme, to theory were generated to describe the team mental model. In the first cycle, 17 codes were extracted from participants' own wording for concepts that were shared by whom controlled various gaming characters. These codes are knowledge convergence about the problem at-hand, actors, events, affects and outcomes - what the core gamers have experienced through the collaborative gaming process. In the second cycle, coding was guided by learning theories. Six categories, mingled from 17 codes, showed collective knowledge of co-work process, leaders' works, work under supervision, seeking joint fun, relationship oriented, and balance between extrinsic-intrinsic motivations. From 6 categories, two themes were synthesized: the team (1) performed ''joint hard work'' for (2) seeking ''joint hard fun.'' The first theme comprised declarative and procedural knowledge representations and we consider it to be the commonality between this game team and ordinary work/learning teams. The second theme was composed of affective and cognitive evaluation components about intrinsic motivation which is in accordance with Self-determination theory (SDT, Deci & Ryan, 2000). In general, ''Jointly hard work for hard fun'' is the gist content of the team mental model. We found ample evidences that members explicitly recall shared motivational beliefs of team mates and emotional-motivational events in gaming. Based on the results of this study, several implications have been addressed for teachers to enhance students' intrinsic motivation in conducting quality Computer-Supported Collaborative Learning. Future studies are needed to further explore the relationship between the quality of team mental models and team-level performance.