The Poisson Regression and Quasi-Poisson Regression Analysis on FIFA World Cup Games

• Main Authors: Kang, Yue Teng, Che Him, Norziha
• Format: Conference or Workshop Item
• Language: English
• Published: 2023
• Subjects: QA Mathematics
• Online Access: http://eprints.uthm.edu.my/12485/1/P17880_012cee6c36971a5dc5ed3405996a084d.pdf; http://eprints.uthm.edu.my/12485/; https://smpu.uthm.edu.my/uploads/P17880_012cee6c36971a5dc5ed3405996a084d.pdf

The FIFA World Cup's surging popularity has attracted a diverse fan base, including passionate enthusiasts in Malaysia. This widespread interest has motivated researchers to delve into the details of the tournament, using it as a crucial platform for predictions, team performance evaluations, and an exploration of international football competition at its top. This study aims to predict number of goals in World Cup matches, employing Poisson and quasi-Poisson regression models. The optimal model is determined through a comprehensive assessment, considering AIC, BIC, standard error, and p-values. Utilizing a dataset from Kaggle, originally sourced from the official FIFA website, the findings consistently associate variables such as goal inside the penalty area, goal outside the penalty area, left inside channel, right channel, attempted defensive line breaks, completed defensive line breaks, yellow cards, passes, and own goals with a raised probability of number of goals. Significantly, the quasi-Poisson model exhibits a superior fit, as evidenced by its lower AIC value of -50.0853 and a reduced deviance value of 38.7935. Consequently, the quasi-Poisson regression model emerges as a more suitable choice than the Poisson regression model, particularly for addressing the overdispersion inherent in the data.