Abstract
Sabermetrician Bill James (1980) advanced a formula predicting a baseball team’s winning percentage based on runs scored and runs allowed called the Pythagorean method. Cochran and Blackstock (2009) revised this method to estimate a hockey team’s winning percentage using goals scored and goals allowed over the course of a season. While it is valuable to estimate the end of season winning percentage, it is equally important to know whether the Pythagorean method can be used to predict winning percentages at critical moments during the season. Thus, the purpose was to compare the estimated winning percentage at various points in the season using the Pythagorean method to the actual winning percentage at the end of the season. Using archival data from every NHL regular season game from the 2005-06 through 2010-11 seasons (N = 7,365), the results indicated the Pythagorean method estimated both home and away winning percentages with a high degree of precision. For home winning percentage, the actual end of season value was 54.95% while the estimated values were 54.96% at pre-Christmas, 55.01% at the All-Star break, 54.52% at the trade deadline, and 54.84% at the end of the season. As for visiting winning percentage, the actual end of season value was 45.09% while the estimated values were 45.04% at pre-Christmas, 45.09% at the All-Star break, 45.57% at the trade deadline, and 45.26% at the end of the season. The results are discussed in relation to player personnel decisions.