Page:CIAdeceptionMaximsFactFolklore 1980.pdf/17

C00036554 than not been keyed to enemy preconceptions — according to the data in 110 out of 131 (or 84 percent) of the cases. This supports the assertion that deception planners subscribe to the principle. Second, these data support the conclusion that when deception is keyed to enemy preconceptions, the probability of surprise is greater. Though the overall degree of success (measured by the fraction of cases resulting in surprise) using deception is large, 123 out of 131 (or 94 percent) of the cases, a more disaggregated analysis is possible. Specifically, when deception was keyed to existing belief, surprise resulted in 106 out of 110 (or 96 percent) of the battles, whereas, when this was not the case, surprise resulted in "only" 17 out of 21 (or 81 percent) of the battles — a statistically significant difference (but recall earlier disclaimers) if this were a random sample. A puzzling aspect of the raw data concerns those situations where deception was not employed and plans were consistent with preconceptions. It would be expected that this would have a low incidence of surprise, yet all eight cases (beware of small sample sizes) resulted in surprise--weak support for what some observers have termed, "the inevitability of surprise." The next principle suggests some reasons that help explain the prevalence of surprise.