Abstract
This paper re-examines the calculation of the relative standard deviation (RSD) measure of competitive balance in leagues in which draws are possible outcomes. Some key conclusions emerging from the exchange between Cain and Haddock (2006) and Fort (2007) are reversed. There is no difference, for any given points assignment scheme, between the RSD for absolute points compared to percentages of points. However, variations in the points assignment that change the ratio of points for a win compared to a draw do result in different RSD values, although the numerical differences are minor.