This is my first, of hopefully many, article involving more in depth statistical analysis. I’d like to hear everyone’s thoughts on it and what sort of things they would find interesting to look at. I am working on devising a statistical method to accurately predict which team will win an upcoming game as well as a Power Rankings formula, both of which I hope to unveil in a few weeks time. As part of those projects I have tried to identify what teams win games. Do defensive minded teams win, or do offensive minded teams win. To do this I analyzed all the game results of the 2003-04 season and came up with the following results.

 Home Team Record 589-470-171 .548 win% Team with better offence 634-414-166 .591 Team with better defense 637-414-170 .591 Team with better offence and defense 444-224-101 .643 Home team has better offence 351-180- 76 .641 Home team has better defense 349-178- 84 .640 Road team has better offence 283-234- 90 .540 Road team has better defense 288-236- 86 .543

A team was said to have better offence if they, over the course of the full season, had more goals for than their opponent. A team was said to have better defense if they had fewer goals against than their opponent.

What surprised me is how little it matters whether you have the better offence or better defense. So long as you are better than your opponent at one of them, you have a good chance at winning. Of course, being better both in scoring goals and stopping them is the best. Also, playing at home is a significant advantage too.

General hockey wisdom is that teams frequently play differently at home than on the road. At home they prefer to play an offense oriented game to put on a show for their fans while on the road they play a more disciplined defensive style game. The fact that home teams score about 10% more goals than road teams would seem to back this up, but it could also be due to the extra fatigue/stress caused by traveling and living out of a suitcase. Whatever the reason, home teams score more goals.

In the above table we used season long (82 game) goals for and goals against data, but what if we use home and road (41 game) goals for and goals against data. i.e. when determining which team has the best offence we look at the home teams goals for in games played at home and the road teams goals for in games played on the road. When we do this the results look quite different.

 Team with better offence 656-381-166 0.614 Team with better defense 539-514-170 0.51 Team with better offence and defense 284-146- 61 0.641 Home team has better offence 446-253-110 0.619 Home team has better defense 91- 20- 7 0.801 Road team has better offence 210-128- 56 0.604 Road team has better defense 448-494-163 0.479

Now those are some dramatically different results. The shocking thing is, if you can play good defence at home or good offence on the road, you improve your chances of winning dramatically. Also, it appears to be a big mistake to play defensive on the road.

The five teams with the fewest goals against at home Dallas, San Jose, Detroit, Colorado, and Ottawa. The five teams with the most goals on the road are Colorado, Ottawa, Toronto, Tampa Bay, and Vancouver, all very good teams with very good records. All of those teams finished with 100+ points except Dallas which ended up with 97. Also, Detroit and Dallas had the two best home record and San Jose finished fourth and Ottawa fifth.

Conversely, the five teams with the fewest goals against on the road are New Jersey, Boston, Calgary, Vancouver and Minnesota. The five teams with the most goals for at home are Detroit, Ottawa, NY Islanders, Tampa Bay and Buffalo. Two of those teams didn’t make the playoffs and two others had fewer than 95 points.

Moral of the story: Playing offensive at home and defensive on the road isn’t necessarily the best strategy.