maxims, however, can be supported from historical evidence. Others emerge from relevant social science theory, decision analysis, and/or game theory. Finally, some are suggested by anecdotal material and, though plausible, are untested and of unknown generality.
An Aside: Remarks on the Data
Elsewhere in this paper, reference will be made to analyses based upon an historical data base. This data base was prepared by Dr. Barton Whaley, then of M.I.T., as part of an ongoing research effort on deception. From Alam el Halfa to Yugoslavia, the data base currently consists of over fifty quantitative and qualitative attributes of 232 military engagements over the period 1914 to 1973. Data elements include categorical attributes (e.g., was deception employed, was surprise achieved, did the attack plan reflect the opponent's preconceptions, etc.) as well as quantitative variables (e.q., strengths, casualties, etc.). For many entries in this data base, there is general agreement among the various source materials consulted. For some, however, the data were more ambiguous or even contradictory. Finally, there are cases for which some data are missing entirely and reasonable estimates
