Scholarlink

<p>A football match is modeled as a finite-horizon dynamic contest game with multiple battles. Each period consists of two stages. There is a battle for ball possession in the first stage. If one team gains possession and creates an attacking chance, there is another battle in the second stage, where the attacking team attempts to score a goal while the other team tries to defend its post. The game is won by the team that outscores the other team. We characterize the unique subgame perfect Nash equilibrium and provide illustrative examples to better understand the equilibrium behavior. We then use a dataset of 4338 football matches from four major European leagues to demonstrate the model’s predictive accuracy&#xa0;by generating probabilistic predictions for the respective match outcomes via model simulations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A game theory approach to football predictions

  • 期刊论文

  • 发表日期：

    2025年8月12日

查看全文
  • Kerim Keskin

摘要

A football match is modeled as a finite-horizon dynamic contest game with multiple battles. Each period consists of two stages. There is a battle for ball possession in the first stage. If one team gains possession and creates an attacking chance, there is another battle in the second stage, where the attacking team attempts to score a goal while the other team tries to defend its post. The game is won by the team that outscores the other team. We characterize the unique subgame perfect Nash equilibrium and provide illustrative examples to better understand the equilibrium behavior. We then use a dataset of 4338 football matches from four major European leagues to demonstrate the model’s predictive accuracy by generating probabilistic predictions for the respective match outcomes via model simulations.