Statistical Analysis of Twenty20 International Cricket

Predicting the outcomes of sports tournaments such as cricket is a subject of significant interest. This thesis focuses on the statistical analysis of Twenty20 International cricket matches. In the thesis, we compare model-derived rankings with ICC T20I cricket rankings. The approach presented in this thesis is derived from the Bradley-Terry model for paired comparison data. The modeling strategy incorporates factors like the home advantage and elements that take into account the team's previous match performances, indirectly inferred through the T20I cricket dataset. Model fitting is accomplished through the method of maximum likelihood estimation. To validate the methodology, we demonstrated an application using the T20I cricket dataset from the past 9 months.