Predicting the success of American football draft prospects has been a topic of interest for years in the field of machine learning and data science. To be able to predict the success of a prospect with such accuracy could be very beneficial to multiple parties. In terms of American football, coaches could utilize a model such as this one in order to correctly predict the success of one of these prospects and be able to recruit a valuable player for their team which is insanely important, especially since there is a huge money investment in prospects. Secondly, players can utilize a model such as this one to analyze which drills are more important when participating in the National Football League Combine. This paper proposes several different types of machine learning algorithms which could accurately predict exactly how successful a player will be once they reach the NFL. The algorithms utilize features and data from various drills at the NFL Combine and implements models such as Random Forest Classifier, Naive Bayes Classifier, K-nearest Neighbors, Decision Tree, Support Vector Machines, Gradient Boosting Regression, Random Forest Regression, Linear Regression, and Decision Tree Regression. The feature
importance analysis determined the 3 Cone drill to be the most important out of the practices in the combine. The model developed using Random Forest could predict NFL prospect success with an accuracy of 0.81. Overall this paper presents a model to easily predict the career success of NFL draft prospects.