Developing a New Metric for NBA Player Evaluation: Points Per Lost Possession
Gerberich, Derek C
Area of Honors:
Bachelor of Science
Andrew John Wiesner, Thesis Supervisor Murali Haran, Honors Advisor
NBA basketball moneyball Bill James Bill Simmons forecasting points per lost possession PPLP predictive models sports
The basic statistical box-score measures of the NBA have been recorded since the game’s inception in the late 1800’s. Is it not intriguing that we still evaluate player performance by the same measures that were used nearly a century ago? Surely there must be room for improvement. Where is the Bill James of the NBA? What is the NBA equivalent to the MLB’s On Base Percentage? Finally, how would one go about developing such a metric? Presumably, a new statistical insight into one of American’s most popular and lucrative ($4.3 billion) industries would be quite valuable. Can we improve upon a measure that has been accepted for more than 100 years? This paper aims to explore that possibility.