Qualifying Dissertation Directed Emergent Drama

Substantial research effort has been put into creating a truly interactive drama that is not only entertaining but also gives the player a sense of coherency and that she is an active participant in the unfolding drama. A fully interactive drama has the potential of revolutionising entertainment, gaming, education, therapy and news coverage. If pre-authored stories are directly implemented into an interactive virtual environment then there is a call for authoring all possible story twists which is an intractable problem for the authors. One philosophy for tackling this problem is to use planning and narrative structures to reduce the authoring. The problem with this approach is that due to the complexity of planning it is infeasible for real-time application, the games do not scale well because the computation quickly becomes intractable. Additionally, this approach has not managed to sufficiently reduce the authoring problem. A simpler and more computationally feasible way is to have the drama emerge from the play of autonomous actors. This means that the computation has been distributed amongst the actors and that the interaction of the player is responded to dynamically and in context by the actors. The main problem with autonomous actors in emergent drama is that they are unlikely to play out the preferred plot structure, for example a murder mystery without some direction. To tackle these problems we propose to divide the drama into narrative episodes which we call schemas. Schemas are used by a director and a set of actors to structure the drama so that it emerges into a fully developed drama that adheres to a specific genre. The schemas are pre-authored in an abstract way such that they can be deployed multiple times in the same drama, which removes the authoring bottleneck. The schemas are not dependent on events having taken place in a specific pre-determined way. This means that the schemas greatly reduce the planning complexity. The director oversees the dramatic structure of the emergent drama and chooses schemas to deploy in the emergent drama based on the needs of the characters in the drama, the plot progress and observed interaction of the player. The actors use the schemas to guide their play and to coordinate their actions with other actors. We will define and implement our solution in the Second Life virtual environment to play out typical English murder mystery dramas. We will use Second Life players to test the implementation. We will also examine whether the implementation adheres to the English murder mystery genre and whether the system execution time is acceptable.