Digital engagement, competitive balance, and stadium attendance in Indian Women's Football: An explainable machine learning and PLS-SEM framework

Abstract

Attendance underpins the commercial sustainability and visibility of women's football, yet research on attendance demand in emerging markets remains scarce. This study proposes an integrated predictive and explanatory model for Indian women's football, combining conventional inferential statistics with seven machine learning algorithms, explainable artificial intelligence, and composite PLS SEM. The analytical dataset comprised 1,000 match level observations spanning economic, venue, team quality, competitive balance, match context, digital communication, digital engagement, fan base, and environmental indicators. Prediction was evaluated through an 80:20 holdout split with baseline and tuned models, alongside nested 5x5 cross validation. Model performance was assessed using R squared, RMSE, and MAE, while SHAP, LIME, and partial dependence and individual conditional expectation analyses supported interpretation. Theory guided structural relationships were tested through a Mode B composite PLS SEM model with 5,000 bootstrap resamples. Random Forest achieved the strongest tuned holdout performance (R squared = 0.155, RMSE = 4,946, MAE = 4,128), while SGD Regressor produced the lowest mean nested cross validation RMSE (4,855). Match importance and weekend scheduling emerged as the strongest predictive signals. The structural model explained 18.0 percent of the variance in attendance and showed positive predictive relevance (Q squared = 0.167). Match Context was the only construct with a strong, statistically significant direct structural effect on attendance (beta = 0.418, p < 0.001, f squared = 0.211), while digital communication and engagement showed weaker and structurally uncertain effects. These findings indicate that, within the joint predictive and structural framework, match context is the most actionable lever for scheduling, marketing, and fan development decisions in Indian women's football.