Comparison of Machine Learning Models to Best Predict Game Attendance in Major League Baseball

To forecast Major League Baseball game attendance, this study employs six different regression models commonly used for machine learning.

To forecast Major League Baseball game attendance, this study employs six different regression models commonly used for machine learning. The models include Multiple Linear Regression, Decision Tree, Random Forest, Support Vector Regression, Multi-Layer Perceptron Regression, and Gradient Boosting. For the dataset, the 2022 and 2023 games of the Los Angeles Dodgers are used to reflect the recent trend of fan attendance and focus on a single random team so that different tendencies of each team’s fans do not become a confounding variable. Each regression model’s performance is evaluated by Mean Absolute Percentage Error and Root Mean Square Error. The performance evaluation suggests that Random Forest and Gradient Boosting predict attendance with the highest accuracy. Using these prediction models, each baseball team may facilitate staff management, event organization, and marketing before their games.

Explore More!

Research Paper