Whodunit: A Generative Model for Murder Mysteries asan Information Game

This research focuses on creating a model to generate information for a narrative-based game of murder mysteries. The literature review aspect of this research will review previous related research works in the subject of procedural narrative generation, explains the concept information game in
the context of murder mysteries as well as on how the narrative structure of murder mystery stories in fiction can assist us in constructing a model that can generate information for a game experience.
I developed a proof of concept using the existing knowledge about the structure of murder mysteries and procedural narrative generation in a way that it can provide information for a murder mystery narrative-based game. After developing this proof of concept, I tested the prototype using a human-computer interaction method called Wizard of Oz. I hypothesized that this prototype is able to provide narrative information coherently for a compelling narrative experience with a believable cast of characters that can evoke feelings of suspense and surprise. By surveying participants after the play-tests, we concluded that the proof of concept can create a coherent narrative experience with a believable cast of characters in a way that it can create the feeling of suspense in players, but it was not able to create the feeling of surprise on the revelation of the culprit for most users. Based on this study, the prototype was also able to create a generally compelling narrative experience for the users.
You can read the full thesis at: