Statistical logic of scoring interaction in NBA playoffs (2014-2019): The influence of time and space

Abstract

Basketball is a sport of deceptions, but also of numbers, and the knowledge of its internal logic has benefited for decades from the contributions of colleagues such as Sampaio and Ibáñez-Godoy. Accepting that NBA playoffs are the most demanding competition in our sport, we set two complementary goals for this study: to provide the a priory probabilities associated with field shots based on space and time, and to show the way we followed to calculate them in the R environment.
Our analyses allow us to provide the probability distributions associated with field shots based on clock-time and court position, as shown in Table1. In addition, by simple logit models we can see and messure the impact that distances and moments dring the match (period and minute) have on the probability of success.
The best basketball players in the world maximize the utility of their shots without being able to escape the space-time logic of the game: the balance between attack and defense comes to light in the form of probabilistic functions with high linearity, especially in terms of space, and points to interesting forward-looking questions. In this sense, our study provides general statistical basis for understanding this logic, allowing in the future to advance in better informed, bayesian study models.