Modeling for campaign analysis

the long version

The U.S. military's increasing use of computer modeling has clear benefits, among them the ability to better inform decisionmakers and reduced exercise costs. However, there are also some drawbacks that need to be overcome both in the models (many of which were developed with the Cold War and less-advanced technologies in mind) and the ways they are used (sometimes with unrealistic expectations or with inadequate analysis of the results). The authors discuss some of the significant challenges and offer suggestions for working through them to achieve not only a new generation of models but a new generation of analytic capability; educating analysts and decisionmakers about the needs, methods, and limitations of model-based campaign analysis; balancing the emphasis between the models and the related analysis; improving and sharing databases; developing a set of models with a range of capabilities, rather than attempting to create one supermodel; focusing R & D on the effects and representation of key combat phenomena; and critical peer review of the models and broader disclosure of methods and results.