Military, diplomatic, and intelligence analysts are increasingly interested in having a valid system of models that span the social sciences and interoperate so that one can determine the effects that may arise from alternative operations (courses of action) in different lands. Part I of this article concentrated on internal validity of the components of such a synthetic framework – a world diplomacy game as well as the agent architecture for modeling leaders and followers in different conflicts. But how valid are such model collections once they are integrated together and used out-of-sample (see Section 1)? Section 2 compares these realistic, descriptive agents to normative rational actor theory and offers equilibria insights for conflict games. Sections 3 and 4 offer two real world cases (Iraq and SE Asia) where the agent models are subjected to validity tests and an EBO experiment is then run for each case. We conclude by arguing that substantial effort on game realism, best-of-breed social science models, and agent validation efforts is essential if analytic experiments are to effectively explore conflicts and alternative ways to influence outcomes. Such efforts are likely to improve behavioral game theory as well.
Date of this Version
political simulation, agent-based models, game theory, validation, policy analysis tools
Date Posted: 07 September 2007
This document has been peer reviewed.