Thursday, December 21, 2006

More OPPGATT and other madness

You may recall in a previous post I invented a new statistic, called Opponents' Points Per Game At The Time, or OPPGATT. This is an attempt to measure how good a team is when the target team (in this case, the Sharks) plays them. Specifically I was using it to try and evaluate the difficulty of Nabokov's starts vs. Toskala's, and I came to the conclusion that Nabokov has faced much tougher opposition. Purely by the luck of the draw, since they have played in alternate games every time this season.

Aren't there other factors we can use to evaluate how tough an opponent is? What I'm trying to avoid is specific game-type decisions and team status, stuff like injuries, scratches, and starting goalies. Then we get into the minutia of individual match ups, scheduling issues, and other stuff. What I'm looking for now is more macro-level type stuff.

Isn't home vs. away such an indicator? As of today, this season, teams are on average getting 1.17 points per game at home, and 1.02 points per game on the road. By my count, there are only 5 teams in the NHL (the Rangers, Penguins, Sabres, Senators, Bruins, and Lightning) that have better records on the road than at home. Some teams, like the Wild, are dominating at home (13-3-1) but are just godawful on the road (4-11-1).

The thesis is this, and not that controversial - road games are harder to win than home games. But the real question is, how much harder? One more useful number: the overall average of points awarded per game is 1.099. That means an away game is a 0.07 point handicap, and a home game is a 0.07 point advantage.

We can apply this modifier to my OPPGATT number directly, since they are also in the units of points/game. All of Nabby's toughest games were at home, so they weren't quite as tough using just OPPGATT. However, two of Toskala's toughest three games were away, so those games were a hair tougher than we thought.

Overall, it closes the gap a bit in terms of toughest games, but the outcome is the same, mostly because the discrepancy was so large to begin with.

I might try to use these number to evaluate the quality of starts in a future post. Shouldn't losses against bad teams hurt a goalie's (or team's) respectability (for lack of a better term) more than a loss to good opposition? More to come...

No comments: