Digital games are essentially about fun, challenge and entertainment. With the increasing number of games-literate people, society is slowly learning to harness this technology for other than mainstream market exploitation, thus, driving game design into unexplored territories. Nevertheless, when observing the selection of contemporary commercial games, there seems to be little diversity in the type and forms of gameplay. There is an oversupply of competition, destruction, solo-heroism and personal sandboxes. One of the less ventured areas is games that are designed for collaboration. This paper examines the issues of puzzle or challenge design (Rollings & Adams, 2003) in the context of collaborative gaming. Questions like ‘How to create meaningful puzzles requiring collaboration?’, ‘Is there life beyond trial and error?’, and "How far game developers can go with collaboration design?’ are of relevance. The questions are approached from the game design point-of-view, while the application domain is within the thematic setting of learning and group dynamics. Games are intrinsically instruments for learning, which can be, and have been, adopted for collaboration. In many computer games players learn lessons that can be applied to other aspects of their life. Many players thrive on and long for the challenges games provide, and are enriched by the learning that follows (Rouse, 2000). Games do not just entail having participants in virtual surroundings. Instead, they offer meaningful and motivated actions for the players, which, in turn, enhance the potential for collaboration (Baker et al., 2002). The case study illustrated in this paper includes an experimental game design, eScape (2003), which has been constructed as a total conversion modification for Unreal Tournament 2003. In order to harness the potential of multi-disciplinary expertise, the design of eScape was supported by collaboration between the Universities of Oulu (LudoCraft Game Design and Research Unit and Educational Technology Research Unit) and Jyväskylä (Institute for Educational Research). Educational experts provided a set of pedagogical concepts, such as, negotiation, conflict, shared understanding, common goals, and group forming. These acted as objective phenomena to be pursued by game design. Game design and development knowledge was applied into learning context without compromising either the factor of fun or the requirement for purposeful learning. Traditional pitfalls of edutainment were avoided by finding a fruitful combination of a game system and a learning environment. eScape is a 4-player collaborative game, which can be defined as social action adventure especially targeted for novice players. The high concept of the game is an escape story where the group of players has to solve a set of problems, or puzzles, in order to flee from the ancient prison colony. The puzzles are designed in a way that the effort, commitment and actions of every participant are required for successful completion. In addition to visual and non-verbal in-game interaction, the player-to-player communication is supported by voice-over-IP speech system, which allows free dialogue. Rich interaction is enabled in a way which is as intuitive and non-intrusive for the players as possible. The choices, manoeuvres and other interaction features are simple enough to be used by non-gaming community. A special laboratory environment was constructed so as to capture all the required data during the experimental game sessions. The multiplayer nature of eScape required extensive data collection arrangements because every player’s actions had to be recorded. In order not to compromise the research setting, the players were physically isolated from each other by wall sections. These cubicles were arranged in a way that the players were not disturbed from outside the game world. This made it possible to have a multitude of data recording devices and personnel within the set, while the players were experiencing realistic computer-mediated collaboration. The eScape empirical experiment was organised with the participation of six groups of four test players chosen from the non-gaming community. On the first day the players were given a brief training session in the game tutorial. On the second day they played the game, immediately followed by a stimulated recall interview. Data were gathered using several methods: background information questionnaires, video feed from each of the players (over-the-shoulder view), combined views from all the four players (over-the-shoulder views), video feed from a virtual camera (in-game), audio recording of spoken dialogue, demo recording within the game platform (enables free camera movements during playback), stimulated recall interviews, and the players’ personal notes. The analysis of player collaboration in puzzle solving was conducted by studying the perceivable interaction forms (Manninen, 2003) that were evident in dialogue transcripts and in-game video recordings. Based on this analysis the implications for puzzle design were further evaluated against the proposed puzzle design framework. The main challenge was the design of motivationally guided, logical and challenging puzzles that would require true collaboration. In order to avoid ending up with trivial, non-motivational and/or frustrating activities, the theoretical and practical expertise of game design was utilised. However, since there are relatively few occurrences of truly collaborative games in the entertainment sector, the theories behind eScape involved a combination of game design, virtual environments, collaborative work and education. The study revealed encouraging results on the possibilities of designing puzzles for collaboration but some dangers were also identified. The design produced relatively simplistic and secure puzzles, which enabled safe trial-and-error procedures. Higher level of collaboration could be supported by increasing the pressure, risk levels and/or creativity in the design for collaboration. When players acquire a concrete feel for beneficial, or purely enjoyable, collaboration, they will naturally engage such strategies. The main limitation of the eScape, and one possible reason for the lack of higher level collaboration, was the shortage of supported collaborative construction. The explicit construction process provides endless representational variations for the participants to express themselves. Combined with collaborative challenge, the participants can creatively enrich each other's experiences. For example, the findings indicate that the participants found highly innovative ways in overcoming the obstacles – sometimes even exceeding the boundaries set by the designers. In comparison to destructive features evident in contemporary games, the constructive features, while difficult to design, offer a higher range of possibilities. Sample References Baker, K., Greenberg, S. and Gutwin, C. (2002). Empirical development of a heuristic evaluation methodology for shared workspace groupware. Proceedings of the ACM Conference on Computer Supported Cooperative Work, ACM Press. eScape – Electronically Shared Collaborative and Pedagogical Experiment (2003). LudoCraft Game Design and Research Unit, University of Oulu. [online] [cited 30 November 2004]. Available from: http://ludocraft.oulu.fi/escape/ Manninen, T. (2003). Interaction Forms and Communicative Actions in Multiplayer Games. In Game Studies, The International Journal of Computer Game Research 3(1). Rollings, Andrew & Adams, Ernest (2003) Andrew Rollings and Ernest Adams on Game Design. New Riders, p. 500 Rouse, R. (2000). Game Design: Theory & Practice. Wordware Publishing, Inc., Plano, Texas. Salem K. & Zimmerman E. (2004). Rules of Play – Game Design Fundamentals. Massachusetts Institute of Technology.