Models, Data, and War
A Critique of the Study of Conventional Forces
Many combat models exist which simulate phases of conventional war. These models must employ data inputs to provide quantitative estimates of effectiveness useful for decisionmaking. This report surveys the quality of both the data and the combat models. Attention is given to the concepts of firepower scores and indexes, estimates of terminal ordnance effects, and estimates of ammunition expenditure which are used in most models. It is shown that major deficiencies exist in both the quality and kinds of empirical data necessary for adequate analysis of combat operations. There is inadequate testing of most of the behavior relationships embedded in models. Combat modeling appears structured to accommodate the inadequate database, with much of the available data being of either unknown relevance or obscure empirical foundation. The structural inadequacies of combat models and the poor data quality appear to be mutually reinforcing. Attention is given to how this situation can be improved by appropriate operational testing or field experimentation.