On Fri, 21 Jan 2005, Nathan Hampton wrote:

> What percent of hockey games are won by the home team? 66% may seem
> reasonable but it depends upon the level of parity in the sport.
>
> Academic drivel: Hockey, with its dominant teams mixed with so many weaker
> programs making the jump from DIII to DI may have less parity, and so a
> smaller percentage of home wins, because the dominant teams win home and
> away -- particularly true for intraconference games. On the other hand, when
> schedules are made, no weak team is going to be hosting a strong team but
> rather the strong teams (with bigger arenas?) have a disproportionate
> ability to be the host (home) team in cross-conference matchups which may
> make hockey's winning percentage even higher than basketball.

So the sensible thing to do is to use a rating system which takes into
consideration the strengths of the two teams as well as who's at home.
There is a natural extension of Bradley-Terry (aka KRACH) which takes
home field/ice into account (see section 2.2 of
http://www.arxiv.org/abs/math.ST/0412232 for the mathematical
details).  This Bradley-Terry-Davidson-Beaver rating system is
realized for college hockey as KASA (KRACH Adjusted for Site
Advantage) <http://www.uscho.com/rankings/?data=kasa>.

                                          John Whelan, Cornell '91
                                                 [log in to unmask]
                                     http://www.amurgsval.org/joe/

Enjoy the latest Hockey Geek tools at slack.net/hockey