What factors make it more likely that an organization uses violence?  This  paper builds a model that tries to answer this question and validate the model using out of  sample testing.  This paper draws from a new data base created through a collaboration  between the National Consortium for the Study of Terrorism and Responses to Terrorism  at the University of and the Minorities at Risk (MAR) Project focusing on ethno-political  organizations in the Middle East  to test factors that make it more or less likely that an  organization will choose to use violence. In this paper we argue that the choice of  violence is impacted the ideological motivations of the organization , the capability of the  organization, the behavior of the government and the place of the organization in “regular  politics.”  Analytically we use multiple approaches for data massaging, classification and forecasting to achieve high classification accuracies (measure in terms of precision, recall, and F-measure).