|
Activity Number:
|
122
|
|
Type:
|
Luncheons
|
|
Date/Time:
|
Monday, August 9, 2004 : 12:30 PM to 1:50 PM
|
|
Sponsor:
|
Section on Statistics in Sports
|
| Abstract - #302250 |
|
Title:
|
Simeon Poisson Meets the Fastest Game on Ice
|
|
Author(s):
|
Robin Lock*+
|
|
Companies:
|
St. Lawrence University
|
|
Address:
|
Dept. of Mathematics, Computer Science, and Statistics, Canton, NY, 13617,
|
|
Keywords:
|
sports ratings ; ice hockey ; Poisson model
|
|
Abstract:
|
Recent controversy over computer models in the BCS system of ranking NCAA football teams has focused interest on statistical methods for assessing and predicting team performance. In recognition of meeting in Toronto (home of the Hockey Hall of Fame) we consider a multiplicative Poisson model for scoring in the game of ice hockey. The model is used to rate and rank teams while accounting for varying schedule strengths, predict future game outcomes, and calculate probabilities of game events. We assume that the expected scoring rate in a game depends on a team's offensive ability, the defensive ability of its opponent, and an adjustment for home ice advantage. Relevant parameters are estimated from previous game scores. We examine the performance of this model using college hockey data on men's and women's teams playing at the NCAA Division I level and compare results to other ranking systems such as the ratings percentage index (RPI), pairwise comparisons, and media polls.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2004 program |
