Ranking Women's College Hockey Teams: Ties and Home Ice

• Activity Number: 154
• Type: Invited
• Date/Time: Monday, July 30, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Sports
• Author(s): Michael Rutter*+
• Companies: Penn State Erie, The Behrend College
• Keywords: Sports ; Bayesian ; Rankings

Abstract

Methods for ranking NCAA college hockey teams play an important role in the sport as the ratings percentage index and pair-wise rankings are used to determine at-large births in the post-season tournaments. This poster continues the research presented in the September 2011 issue of Chance Magazine in which an alternative method of ranking women's college hockey teams was presented that includes modelling the effects of ties and home ice advantage. The model uses a Bayesian approach to calculate the probability of the three possible results (home win, home loss, or tie) based on a normal distribution describing the difference in quality of the teams playing each game. Of particular interest is the estimation of the effect of home ice advantage, both on a division-wide level as well as for each individual team. Three additional years worth of data have been added to the model, and a method to express the uncertainty in the rankings will be presented. A comparison of ranking methodologies for the most recently completed season will also be included.