Bombing Problems—A Statistical Approach II

A continuation of “Bombing Problems, A Statistical Approach,” Opns Res, 5, 75–89 (1957), this paper studies the probability of a hit within a circle centered at a moving target (kill probability) from the viewpoint of statistical inference. The best estimate of such a probability based on results obtained in past experiments, is given when the precision of the bombing apparel is unknown and bombing is performed under already observed experimental conditions. A “predictive” probability is derived in case bombing is performed under new experimental conditions. “Dangerous” regions around the target are discussed. A method is proposed to obtain the best estimate of a hit within a fixed area when the apparent target differs from the unknown true target. Generalizations to three dimensions are outlined.