Using Agent-Based Models for Prediction in Complex and Wicked Systems

This paper uses two thought experiments to argue that the complexity of the systems to which agent-based models (ABMs) are often applied is not the central source of difficulties ABMs have with prediction. We define various levels of predictability, and argue that insofar as path-dependency is a necessary attribute of a complex system, ruling out states of the system means that there is at least the potential to say something useful. ‘Wickedness’ is argued to be a more significant challenge to prediction than complexity. Critically, however, neither complexity nor wickedness makes prediction theoretically impossible in the sense of being formally undecidable computationally-speaking: intractable being the more apt term given the exponential sizes of the spaces being searched. However, endogenous ontological novelty in wicked systems is shown to render prediction futile beyond the immediately short term.