Data analytics has begun to take over the world of sports. Coaches are making educated gametime decisions based on these advancements. However, leagues, such as the National Football League, still do not allow coaches to access in time analytics. Therefore, this raises the question: how do NFL teams make risky decisions? While there are studies that apply this question to kickoff strategies as well as the injury risk for players, studies regarding actual play calling have yet to be published. Furthermore, the concept of risk differs from sport to sport. Because it rests so heavily on uncertainty, play calling in football is actually transformed into a subset of game theory.

Even if the majority of the decisions are based upon game theory, there are still a myriad of factors that could affect the ultimate choice, even if they are incognizant. These can be broken into factors of the actual game, season, team, or for the coach who oversees all of the play calling. These factors can then be run in a logistical regression against an example of a risky decision in football. The one chosen in this study was going for it on 4

th down. After establishing what factors are significant, a final predictive model can be created that in turn displays what plays the largest role in delicate gametime situations. Finally, this model can hopefully provide a framework for companies outside of sports, such as ones in finance, to evaluate different investments and projects.