Generative Experimentation and Social Simulation: Exploring Gaming for Model Verification and Validation

Abstract

The agent-based modeling paradigm has been actively applied to address social normative issues such as: values, cognition, morality and behaviors. The abstraction of human and human-like social processes and mechanisms result in misalignment of computational model with existing (old) verification and validation techniques and expose significant challenges. We argue that human sources represent a sound approach for verification and validation of agent-based social simulation models. We propose a novel conceptual gaming framework that extracts required information from relevant sources as part of game play.