Home Ice Advantage?

by Hornecker on Jan 30, 2012 11:21 AM CST

Over the all-star break I was bored and missing hockey so I decided to look into the home team winning percentage in a little more detail, since every team having a 55% home ice advantage struck me as not entirely accurate as some teams are better home teams, some are better road teams, and some just flat out can't win games.

What I did was compile all of the games from the 2006-07 season to the current season, and here are the results:

	HW%	HPTS%	RW%	RPTS%
2006-07	0.550	0.607	0.450	0.507
2007-08	0.537	0.591	0.463	0.520
2008-09	0.559	0.618	0.441	0.497
2009-10	0.561	0.622	0.439	0.500
2010-11	0.519	0.583	0.481	0.538
2011-12	0.562	0.620	0.438	0.492

And the average for all of that time is:

	HW%	HPTS%	RW%	RPTS%
League	0.548	0.607	0.452	0.509

What struck me as the most interesting here is how widely the "home ice advantage" varied from year to year. In full seasons it ranged from 51% to 56%. So then I decided to look at individual teams winning percentage. I ran a weighted average using the following equation:

HW%= ((wins from ‘06-‘07 to '10-'11)/(number of games) + 2*(wins in '11-'12)/(number of games))/3

That gave me the following for each team:

	HW%	HPTS%	RW%	RPTS%
ANA	0.537	0.568	0.311	0.420
BOS	0.602	0.650	0.627	0.648
BUF	0.484	0.568	0.419	0.440
CAR	0.506	0.557	0.284	0.387
CBJ	0.385	0.449	0.254	0.317
CGY	0.591	0.641	0.386	0.453
CHI	0.627	0.688	0.474	0.528
COL	0.533	0.544	0.460	0.504
DAL	0.562	0.608	0.472	0.493
DET	0.793	0.830	0.508	0.529
EDM	0.496	0.556	0.271	0.314
FLA	0.473	0.606	0.425	0.497
LAK	0.481	0.551	0.459	0.569
MIN	0.591	0.657	0.415	0.478
MTL	0.429	0.544	0.436	0.478
NJD	0.554	0.596	0.544	0.570
NSH	0.610	0.668	0.551	0.580
NYI	0.440	0.513	0.353	0.420
NYR	0.633	0.682	0.588	0.635
OTT	0.569	0.602	0.459	0.536
PHI	0.513	0.583	0.599	0.644
PHX	0.465	0.539	0.443	0.508
PIT	0.612	0.657	0.535	0.580
SJS	0.613	0.662	0.560	0.636
STL	0.666	0.734	0.391	0.464
TBL	0.575	0.612	0.326	0.387
TOR	0.507	0.586	0.459	0.494
VAN	0.614	0.675	0.586	0.614
WPG	0.556	0.602	0.325	0.399
WSH	0.682	0.717	0.393	0.444

*Winnipeg was calculated using Atlanta's record

This gave a better picture of home ice advantage, ranging from a 38.5% winning percentage (Columbus) to a 79.3% winning percentage (Detroit). From this the next step would be to calculate the winning percentage for each combination of teams. I did that using the following equation:

HW% = ((League Average)+(Home Teams Winning %)+1-(Road Teams Winning %))/3

RW% = (1-(League Average)+1-(Home Teams Win

ning %)+(Road Teams Winning %))/3

So for Detroit's next 10 games the odds are:

A	H	HW%	APTS%	HPTS%
DET	CGY	0.544	0.519	0.624
DET	VAN	0.551	0.519	0.641
DET	EDM	0.512	0.519	0.581
DET	PHX	0.502	0.519	0.573
EDM	DET	0.690	0.412	0.718
ANA	DET	0.677	0.464	0.718
PHI	DET	0.581	0.576	0.718
DAL	DET	0.623	0.501	0.718
NSH	DET	0.596	0.544	0.718
SJS	DET	0.593	0.573	0.718

This seems to pass the eyeball test for me, but any thoughts or input?

Tweet Comment 2 comments | 1 recs |

Do you like this story?

Comments

Great writeup

My only statement on it is that I wish all games were worth the same number of points, so it would be easier to use points% as a better indicator of winning % (since right now, two teams can play a home-and-home where they each win in OT and leave their points% above .500)

by J.J. from Kansas on Jan 30, 2012 2:31 PM CST reply actions

I agree. Right now if I run those numbers out through the end of the season the wings win 7 more away games, 11 more home games but come up with 40 more points to finish second overall (behind the Rangers with 109 pts) and 1st in the Central. From these numbers the team that benefits the most from the charity point through the remainder of the season is the sharks, with a predicted 5 more charity points.

by Hornecker on Jan 30, 2012 2:54 PM CST up reply actions