A prediction of Korean baseball score using Conway-Maxwell Poisson distribution
Jae Hyuk Cho1 · In Chul Jung2 · Ilsu Choi3
13Department of Mathematics and Statistics, Chonnam national University
2Baseball Planning and Research, Kia Tigers
2Baseball Planning and Research, Kia Tigers
Correspondence to: 1 Master Candidate, Department of Mathematics and Statistics, Chonnam National University, Gwangju 61186, Korea.
2 Pro, Baseball Planning and Research, Kia Tigers, Gwangju 61255, Korea.
3 Professor, Department of Mathematics and Statistics, Chonnam National University, Gwangju 61186, Korea. E-mail: ichoi@jnu.ac.kr
2 Pro, Baseball Planning and Research, Kia Tigers, Gwangju 61255, Korea.
3 Professor, Department of Mathematics and Statistics, Chonnam National University, Gwangju 61186, Korea. E-mail: ichoi@jnu.ac.kr
Received May 28, 2024; Revised July 4, 2024; Accepted July 11, 2024.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
- Abstract
- In this paper, we analyzed the score prediction distribution and winning percentage of baseball games considering the previous game’s scores and runs, home and away. KIA’s home scores have an average of 4.2 and variance of 9.1, and away scores have an average of 4.4 and variance of 10.6, which shows that they are more overdispersed than the Poisson distribution. Therefore, the Conway-Maxwell Poisson distribution, which can apply overdispersion, was used. In addition, the posterior prediction distribution was derived using the bivariate gamma distribution, which is a conjugate prior distribution. For scoring and conceding runs, the model was simplified using a method of modifying the parameters of the prior distribution. Based on the games from April 1, 2023 to July 27, 2023 for all other teams, focusing on KIA, we predicted the posterior prediction distribution and win rate for the next game. In particular, in the game between KIA and LG, it was predicted to be 2:4 at home and 4:3 on away. Additionally, when KIA is at home, the probability of winning is 23%, the probability of losing is 63%, and the probability of not being able to decide the winner is 14%, and when KIA is away, it is 55%, 30%, and 15%, respectively.
- Keywords : Conway-Maxwell Poisson distribution, Korea Baseball Organization, posterior prediction distribution
Full Text(PDF) Free
