A Performance Metrics–Based Model for Predicting Match Outcomes in the 2023 ICC Cricket World Cup

The purpose of the study. To develop and validate a logistic regression model for predicting match outcomes in the 2023 ICC Cricket World Cup using selected in-game performance indicators and to determine the relative contribution of each variable. Materials and methods. Data from 47 matches of the 2023 ICC Men’s Cricket World Cup were analysed (one match decided by the Duckworth-Lewis-Stern method was excluded). Independent variables included Toss outcome, Opening Partnership Score, Runs and Wickets Lost in Powerplay, Total Number of Fours and Sixes, and Total Wickets Lost in an Inning. Binary logistic regression was applied to predict match outcomes (win/loss). Model goodness-of-fit was evaluated using the Hosmer-Lemeshow test. Results. Wickets Lost in an Inning was the strongest predictor (OR = 2.324, p < 0.05); each additional wicket lost increased the odds of losing by 132.4%. Each additional four reduced the odds of losing by 13.7% (OR = 0.863). Total sixes and other variables showed weaker or non-significant effects. Toss outcome and Opening Partnership Score were not statistically significant predictors. The final model demonstrated good fit (Hosmer-Lemeshow test, p > 0.05) and acceptable predictive accuracy. Conclusions. Preserving wickets throughout the innings and maximising boundary scoring (especially fours) are the most critical factors influencing match outcomes in the 2023 ICC Cricket World Cup. The developed logistic regression model offers a reliable tool for performance analysis and strategic decision-making in limited-overs cricket.