Scholarlink

In recent years, data analytics has transformed college football by providing teams with insights into performance metrics that influence game outcomes. This study investigates the role of offensive and defensive metrics in predicting team success in college football, focusing on offensive efficiency, yards per play, and turnover margin. The problem addressed is understanding which statistical indicators are most strongly correlated with team wins and how these metrics can be used to classify play styles and predict score margins. Using a dataset covering college football seasons from 2013 to 2023, this analysis employs statistical methods, machine learning, and clustering to evaluate the impact of various performance metrics. The methodology includes descriptive analysis to identify top-performing teams, correlation analysis to uncover relationships between metrics and win rates, Ordinary Least Squares (OLS) regression to model key predictors of team success, K-Means clustering to categorize teams by play style, and Random Forest regression to predict score margins based on offensive and defensive factors. The results reveal that offensive efficiency and yards per play are the most significant predictors of wins, with a noticeable upward trend in these metrics across the decade. High-performing teams, such as Oklahoma, Alabama, and Ohio State, consistently rank at the top in offensive efficiency, suggesting the critical role of offensive power in achieving game success. Clustering analysis identifies three distinct play styles, separating high-offensive-efficiency teams from those with more conservative strategies. The predictive model for score margins, based on Random Forest regression, achieves a root mean square error (RMSE) of 6.82, with offensive metrics contributing the most to score predictions. These findings imply that college football teams can enhance their performance by focusing on offensive efficiency, optimizing yardage per play, and maintaining strong turnover margins. This research provides a framework for data-driven strategy development in college football, allowing teams to benchmark against top performers and adjust their tactics based on empirical insights.

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

Statistical Analysis and Machine Learning in College Football: Insights from Offensive Efficiency, Yards Per Play, and Clustering of Play Styles

  • 章节

  • 发表日期：

    2025年

查看全文
  • Said A. Salloum,
  • Rose A. Aljanada,
  • Khadija Alhumaid,
  • Khaled Mohammad Alomari,
  • Aseel M. Alfaisal,
  • Abdalla Elnekiti,
  • Raghad Alfaisal

摘要

In recent years, data analytics has transformed college football by providing teams with insights into performance metrics that influence game outcomes. This study investigates the role of offensive and defensive metrics in predicting team success in college football, focusing on offensive efficiency, yards per play, and turnover margin. The problem addressed is understanding which statistical indicators are most strongly correlated with team wins and how these metrics can be used to classify play styles and predict score margins. Using a dataset covering college football seasons from 2013 to 2023, this analysis employs statistical methods, machine learning, and clustering to evaluate the impact of various performance metrics. The methodology includes descriptive analysis to identify top-performing teams, correlation analysis to uncover relationships between metrics and win rates, Ordinary Least Squares (OLS) regression to model key predictors of team success, K-Means clustering to categorize teams by play style, and Random Forest regression to predict score margins based on offensive and defensive factors. The results reveal that offensive efficiency and yards per play are the most significant predictors of wins, with a noticeable upward trend in these metrics across the decade. High-performing teams, such as Oklahoma, Alabama, and Ohio State, consistently rank at the top in offensive efficiency, suggesting the critical role of offensive power in achieving game success. Clustering analysis identifies three distinct play styles, separating high-offensive-efficiency teams from those with more conservative strategies. The predictive model for score margins, based on Random Forest regression, achieves a root mean square error (RMSE) of 6.82, with offensive metrics contributing the most to score predictions. These findings imply that college football teams can enhance their performance by focusing on offensive efficiency, optimizing yardage per play, and maintaining strong turnover margins. This research provides a framework for data-driven strategy development in college football, allowing teams to benchmark against top performers and adjust their tactics based on empirical insights.