Can the Home-Ice Advantage be Measured Statistically?
View Single Post
08-27-2012, 12:01 AM
Join Date: Feb 2004
I take it, as a personal rule of thumb using a Poisson model (see Robin Lock's CHODR... which is one flavor... he negates OT and ENGs) that home ice is worth about +/-5%... that is 5% to this more on offense... 5% less on defense. If you figure about 3 goals a game each way this means a goal in your favor roughly every 3.5 games.
Home ice is relatively weak in hockey. Feel free to investigate it more. Also, when trying to do "team by team" comparisons please note, like a lot of these investigations, is that statistical power is weak and your mileage may vary. You would think crowds, home cooking, and last change and the rest would affect this more but the plain reality is that it doesn't. Go figure.
I suppose once you get in the 200x85 its mostly between the ears... but hockey being a very much "last mile, last step" type sport. Eh.
Its been awhile since I've looked on this on NHL data... but the results are usually within the same ballpark.
Last edited by Patman: 08-27-2012 at
View Public Profile
Visit Patman's homepage!
Find More Posts by Patman