View Single Post
Old 11-26-2017, 09:33 PM   #1171
kevman
#1 Goaltender
 
Join Date: Apr 2004
Exp:
Default

Quote:
Originally Posted by edslunch View Post
By comparison, this is what strong correlation looks like (Goal differential vs. win % since 1946):



http://hockeyanalytics.com/2008/01/t...key-analytics/
That is a strong correlation. But it isn't a very good predictor. It's no surprise that the team that scores more goals wins the game. That's kind of the point. It's also no surprise that the team that has won more games scored more goals than they let in.

If you look at the current standings you see that they already follow the goal differential closely. So if you were to use it as a predictor you would say that Chicago and San Jose should be doing better and Pittsburgh should be doing worse. Other than that almost every team is exactly where they should be.

Is there a connection between winning big and where you're expected to finish in the standings? Is Vegas with 31 points a far better team than Nashville with 31 points because they win bigger and lose closer? Is Pittsburgh worse than their record because when they lose, they lose bad?

To quote Itse, would making brackets from goal differential do a good job at predicting the playoffs?
kevman is offline   Reply With Quote
The Following User Says Thank You to kevman For This Useful Post: