Abstract
The National Basketball Association (NBA) is the premier men’s professional basketball league in the world. Thirty teams play during the regular season for the sixteen playoff spots and the opportunity to compete for the Larry O’ Brien NBA Finals Championship Trophy. In this paper, we introduce a logistic regression model that can be used to predict which teams will make the playoffs at any point in the NBA season. In addition, we discuss potential applications of this ranking scheme that can be used by an NBA front office, “arm- chair” GMs playing fantasy sports, and professional/non-professional gamblers. More importantly, we move beyond the rankings that are commonly seen on the popular websites by providing a measure of uncertainty associated with our predictions. Finally, we introduce an R package (ballr) that can be used to access data from basketball-reference.com