This thesis builds on the research and ideas of the school of thought that believes strategy is the most important factor in predicting war outcomes. One shortcoming of that school is the inability to explain why strong actors would implement a strategy that does not provide the highest probability of victory. This project uses a game theoretic model to illustrate how a seemingly non-optimal strategy may be rational for initial phases of the conflict. However, this rationale does not apply beyond initial stages of conflict. To explain non-optimal strategy selection in prolonged conflicts, this project analyzes strategy drivers—factors that influence strategy selection and implementation. Probability of victory is only one of the factors found to influence strategy implementation. Other than probability of victory, this study finds that the institutional predisposition of a military is the most important because it is the most consistent and the most controllable by the military. With this conceptual basis, the project analyzes U.S. involvement in Afghanistan since 2001. It also takes a cursory look at U.S. operations in Iraq since 2003, the Soviet occupation of Afghanistan in the 1980s, and the U.S. involvement in Vietnam. The model and case studies illustrate a U.S. military institutional predisposition with an excessive disposition towards direct attack. As such, this thesis recommends taking action to provide the U.S. military with a more neutral institutional predisposition.