Enhancing Predictive Accuracy in Forecasting Football World Cup 2022 Outcomes

This study evaluates key predictive metrics for the 2022 FIFA World Cup outcomes, comparing the predictive accuracy of the simple linear regression (SLR) and multiple linear regression (MLR) models. The analysis examines model fit and accuracy using statistical metrics and addresses key assumptions such as homoscedasticity, autocorrelation, and multicollinearity. Significant variables include the Soccer Power Index (SPI), offensive and defensive strengths, and simulated goal differences. To improve model assumptions, log-transformation is employed for the dependent variable. The findings demonstrate that MLR models outperform SLR in predictive accuracy, contributing to advances in sports forecasting methodologies. Recommendations are also provided for future sports analytics applications.