Files
Abstract
The National Hockey League became the rst North American professional sports league to lose an entire season to labor disputes when the league canceled the 2004-2005 season. The NHL emerged from the lockout with a new collective bargaining agreement and a new set of rules. This study uses the logit framework to model the impact of team performance factors on the probability of winning to determine differences in play between the pre-lockout 2003- 2004 season and the post-lockout 2005-2006 season under the new rules. In particular, the results of the logit estimation reveal that special teams play is significantly more important to determining game outcome in the "New NHL." Stochastic frontier modeling is also used to further di erentiate between the two seasons in terms of NHL goal scoring efficiency by team.