Abstract
The purpose of this thesis is to test the appropriateness of using win percentages as a basis for measuring competitive balance in a sports competition. Match results are simulated to generate end-of-season outcomes which are used to construct win percentages and corresponding ability ratings for the playing teams. Simple ordinary least squares estimation is used to analyse the relationship between competitive balance measures constructed using ability ratings and the corresponding balance measures using win percentages generated from the simulated data. Regression results show that although the overall relationship between the competitive balance measures of ability ratings and win percentages is quadratic, a linear approximation of this relationship over most observations performs sufficiently well. This would suggest that win percentages can accurately measure team ability and provide an appro priate substitute for ability ratings when analysing competitive balance. Therefore, the current use of win percentages in competitive balance analysis is appropriate and will produce results approximately equivalent to those based on a more involved analysis of actual team ability.