Predict the Results of Professional Baseball Games ─Take the 21st Annual Games of Chinese Professional Baseball League for Example

• Main Authors: Chun-Te Lee, 李俊德
• Other Authors: Tien-Lung Tsai
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/86036512190040769764

碩士 === 國立屏東教育大學 === 應用數學系 === 100 === Baseball, known as the nation ballgame in Taiwan includes Little League, Senior League, and Big League. From 2009, CPBL(Chinese Professional Baseball League) is included as one item in the Sports Lottery of the nation. In which baseball fans can not only enjoy the exciting games but also bet on games at lottery stores to win the money. This research presents a case study of predicting the result of the baseball games on the 21st annual games of Chinese Professional Baseball League in 2010 based on the hitting and pitching of Home teams and Away teams as explanatory variables. When 50%, 70% and 90% of the game is played, the total number of hitting and pitching in each game is applied to the logistic regression model to construct a prediction function of the following 20 games of each team. The model can be constructed on the condition that the more the proportion of the games played, the more total number of the games played. When 70% of the game is played, the performance of the pitcher can help predict the result especially by using total number of first and third session of the game. Home team’s R(runs), SO(strikeouts), and WPCT(winning percentage) were selected most as explanatory variables; while Away teams’ are IP(innings pitched), RA(runs allowed), and R. R are the most important explanatory variables in both teams with prediction accuracy by 70%.