Application of Game Theory and Markov Chains on English Premier League (EPL) Scorelines Analysis

Abstract

The aim of this paper is to obtain the optimal strategies of two competitive players using Game Theorem and to make future predictions of games using Markov Chains involving the EPL. All the teams that have participated since 2005/2006 EPL season to EPL 2019/2020 season were considered and the method of proportion of wins was used to select five best teams. Linear programming was employed to select the optimal strategies, while the predictions for seasons 2020/2021 to 2023/2024 are obtained by Markov chain method. The results obtained revealed that Man U is the optimal strategy for Player A, and that Player A has to choose Man U to maximize his profit, meanwhile, Chelsea is the optimal strategy for Player B and he has to choose Chelsea to minimize his loss. The findings of the results also revealed that for Man U or Chelsea to win their home games, it will depend on their current home winning against the team they are playing with.