|
Activity Number:
|
481
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Sports
|
| Abstract - #307010 |
|
Title:
|
The Brave New Hockey World: a Statistical Assessment of NHL Rules Changes
|
|
Author(s):
|
Paramjit Gill*+
|
|
Companies:
|
The University of British Columbia
|
|
Address:
|
IK Barber School of Arts and Sciences, Kelowna, BC, V1V 1V7, Canada
|
|
Keywords:
|
ice hockey ; Bayesian model ; dyadic data ; NHL ; Poisson distribution ; sports statistics
|
|
Abstract:
|
The National Hockey League (NHL) last year approved the most radical set of rule changes in many years. The new rules have resulted in more penalties being rewarded and thus leading to more "power play" goals. Traditionally, the Poisson distribution has been used to model the number of goals in NHL. But with power play goals being so prominent, more appropriate distribution will be a mixture of two distributions accounting for non-power and power goals. We present a Bayesian dyadic model for the number of home-team and away-team goals in a game. The model incorporates home-ice advantages in the offensive and defensive abilities of the teams. The teams can be ranked by these abilities and the effect of power play scoring and penalty killing abilities on the final ranking can be studied statistically. Data from the last two seasons are used for modeling the effects of rule changes.
|
