Each year, the NBA gives out awards for players that played well throughout the entire season. These awards include the All NBA-Team, All Defensive-Team, All-Stars, and the coveted MVP (most valuable player). Besides just winning, the player awards also have ramifications on salaries as well. An award results in higher maximum salaries for players (bonuses) and can make a difference between several million dollars added to their contracts! Because of this, these awards change how teams operate to conform to the league’s salary cap limitations, which are very strict. The problem with the award results is that many selections aren’t agreed with, even among the expert committee of 100 members who are in charge of voting.
In order to create a non-bias solution that can solve these unexpected surprises that affect teams’ salary caps, a machine-learning algorithm was implemented for the Athlete Projection System to predict the probability that a player will win a consequential award for a particular season. The best features were selected and run through numerous data science algorithms/models to predict the best players for each award. This predictive product model will be used by front offices of any professional sports teams (not just the NBA) to project a player’s future salary, determine the best plan for the teams’ future salary cap allocations, and operate accordingly with player transactions. Sports agents can also use this predictive model to inform their clients of their expected maximum earning power.
Built by: Justin Shin, Juan David Palomares, Nicholas Kim, Fuhua Liu, Ken Opamuratawongse, William Chen