Breakthroughs in Decision Science and Risk Analysis

Chapter 10Outthinking the Terrorists

Jason Merrick¹, Philip Leclerc¹, Hristo Trenkov², and Robert Olsen²

¹ Department of Statistical Sciences and Operations Research, Virginia Commonwealth University, Richmond, VA, USA

² Deloitte Consulting, Center for Risk Modeling and Simulation, Washington, DC, USA

Introduction

Deciding how to defend against terrorist threats requires the best insights that decision and risk analysis methods can offer. Well-established decision analytic tools, such as decision trees, influence diagrams, and Bayesian networks, are commonly used in modeling counterterrorism decisions. These techniques use probabilities to model the uncertainty inherent in the domain, including uncertainty about terrorist attack methods and targets. This chapter reviews three approaches to determining which terrorist plans are most likely: directly eliciting subjective probabilities from intelligence experts, modeling adaptive decision-making and game-theoretic strategizing by intelligent foes, and text mining of intelligence data to see what is actually happening. We focus on two example counterterrorism decisions to better understand the problem: defending commercial aviation against surface-to-air missiles and screening cargo containers for nuclear threats.

von Winterfeldt and O’Sullivan (2006) perform a decision analysis to minimize the risk of attacks on commercial airlines using surface-to-air missiles, called man-portable air defense system (MANPADS). In particular, they assessed ...