Another Look at Search-Based Drama Management

Abstract

A drama manager (DM) is a system that monitors an interactive experience, such as a computer game, and intervenes to keep the global experience in line with the author's goals without decreasing a player's interactive agency. In declarative optimization-based drama management (DODM), an author declaratively specifies desired properties of the experience; the DM intervenes in a way that optimizes the specified metric. The initial DODM approach used online search to optimize an experience-quality function. Later work questioned both online search as a technical approach and the experience-quality optimization framework. Recent work on targeted trajectory distribution Markov decision processes (TTD-MDPs) replaced the experience-quality metric with a metric and associated algorithm based on targeting experience distributions. We show that, though apparently quite different on the surface, the original optimization formulation and TTD-MDPs are actually variants of the same underlying search algorithm, and that offline cached search, as is done by the TTD-MDP algorithm, allows the original search-based systems to achieve similar results to TTD-MDPs. Furthermore, we argue that the original idea of optimizing an experience-quality function does not destroy interactive agency, as had previously been argued, and that in fact it can capture that goal directly.