Generalized findings derived from historical case studies should not be taken as prescriptions for upcoming or ongoing operations. For example, while we found that insurgencies last about ten years, we do not suggest that any specific insurgency will last ten years. Predicting specific outcomes from general assumptions is a common logical or, by some scientific definitions, an ecological fallacy. Quantitative findings alone cannot—and should not—shape COIN campaign planning. Further, these findings are correlative and not necessarily causative. In other words, addressing only one factor, such as the availability of sanctuary or external support, will not necessarily create a tipping point that will mark the beginning of the end of the insurgency.
From "How Insurgencies End" (http://www.rand.org/content/dam/rand/pu ... _MG965.pdf), courtesy of the RAND Corporation.