Centrifuge is a cross-platform visual scripting tool for authoring story sifting patterns. It is intended to be used with character-based social simulations where a large number of characters interact, form relationships, experience life events, and hold various roles.
Sifting patterns search the state of the simulation, looking for matching characters, events, and scenarios. Our goal with Centrifuge was to develop a tool that enables users to craft sifting patterns without worrying about low-level query language syntax.
Centrifuge is built on DataScript, an open-source in-memory database and datalog-based query language. Users drag-and-drop nodes that correspond to parts of DataScript’s query syntax, and Centrifuge takes care of translating their pattern of nodes into a valid query.
The current version of Centrifuge is configured for use with the Talk of the Town simulationist story generator. Talk of the Town was the simulation back-end for the award-winning augmented reality live-acting game, Bad News. We chose this simulation because in Bad News, one player acts as the Wizard. Their job is to find interesting characters in the simulation by manually coding inside a Python interpreter. This manual story sifting is a tedious process and requires knowledge of python programming and the internal simulation structure. Here is an example video of gameplay.