Probabilistic Analysis of Complex Combat Scenarios

The United States Department of Defense is constructing the Bayesian Enterprise Analysis Model (BEAM) to evaluate potential military strategies and force compositions in future wars. BEAM has less resolution than campaign models in that, rather than modeling individual weapon systems or units, BEAM applies a probabilistic approach to represent the statistical distribution of quantities of assets in the combat theater. Additionally, BEAM uses the probabilities of mission success along with their associated uncertainty and does not model the details of mission execution. BEAM implements military strategy through allocating assets and their required support to missions. The dynamic evolution is evaluated by threads, which are based on combinations of cases of the marginal asset distributions from each of two sides. The threads are reset after each time period to address dispersion of outcomes. We detail the various aggregation necessary to accomplish a rapid scenario evaluation that enables a search of the military strategy and force composition space for each of the warring sides.