A doubly self-exciting Poisson model for describing scoring levels in NBA basketball

TL;DR

This paper introduces a doubly self-exciting Poisson model to analyze scoring patterns in NBA basketball, capturing both within-game and within-season temporal autocorrelation.

Contribution

It proposes a novel doubly self-exciting Poisson model with a Bayesian estimation approach for detailed scoring analysis in basketball.

Findings

Model effectively captures scoring autocorrelation

Applied successfully to 2018-2019 NBA data

Provides insights into scoring dynamics

Abstract

In this paper, Poisson time series models are considered to describe the number of field goals made by a basketball team or player at both the game (within-season) and the minute (within-game) level. To deal with the existence of temporal autocorrelation in the data, the model is endowed with a doubly self-exciting structure, following the INGARCH(1,1) specification. To estimate the model at the within-game level, a divide-and-conquer procedure, under a Bayesian framework, is carried out. The model is tested with a selection of NBA teams and players from the 2018-2019 season.