Predicting the Most Valuable Player (MVP) awards in the NBA is a complex task that involves analysing player statistics and performance metrics. For the history of the tournament, MVP selection has been based on subjective opinions and votes from sports analysts, and votes from the players themselves. However, using mathematical concepts in machine learning techniques, it is now possible to make more objective and data-driven predictions. In this study, we use mathematical concepts from the field of machine learning: Random Forest [1] and SMOTE [2] (Synthetic Minority Over-sampling Technique), to predict MVP award shares.