Slam Dunk or Air Ball? Predicting NBA Scoring from College Performance (2023)
Undergraduates: Abrar Ahmed, Edward Gray, Emma Dutka, Marilyn Castillo-Ibarra
Faculty Advisor: Jonathan Jensen
Department: Exercise & Sport Science
Every year, NBA teams draft players they believe will positively impact their team. However, these prospects are not always successful at the next level. Other research in this area found that college statistics are the best predictor for longevity in the NBA(Kannan et al., 2018) and we aim to examine how college statistics can be used to predict how much a player will score in the NBA to help teams make decisions when drafting. We collected college and NBA data from all NBA players from 1979-2018 and analyzed it through a hierarchal model which returned a linear equation that calculates how many points a draft prospect would score in the NBA. Results found that a high college field goal percentage and/or points per game increase NBA scoring, while a high weight and/or games played in the NCAA decrease NBA scoring. This research can be utilized by NBA teams to determine which draft prospects they pursue.
Link to Poster