View Single Post
Old 02-22-2015, 11:36 AM   #459
Jay Random
Franchise Player
 
Join Date: Aug 2005
Exp:
Default

Quote:
Originally Posted by CorsiHockeyLeague View Post
4. I'm not going to bother arguing the usefulness of possession statistics as a predictive measure with anyone on here. At this point people who dismiss those metrics are the hockey equivalent of the anti-vaccination folks. Believe what you want, if you're happier living in ignorance, enjoy.
The best predictor of future success in hockey is goal differential. Puck possession proxies are not even close. There are only two reasons why people muck about with the proxies: (1) they are available at a stage of the season when the sample size is too small for goal differential to be useful, and (2) they are new and shiny and have been tagged by silly people as ‘advanced stats’.

I get the strong impression that some hockey stats people have this romanticized image of themselves as Bill James, fighting the good fight against the reactionary old boys, and think that Corsi and Fenwick are the equivalent of OBP and slugging percentage – wonderful new statistical tools that will totally replace traditional understanding of the game. Maybe one day, given actual data on puck possession and real-time tracking of every player on the ice, such tools will be developed. But they don't exist yet. The Bill James of hockey doesn't exist yet, and the ‘old boys’ are already hiring people to do the most advanced statistical analyses they can come up with.

I think nearly every person on this forum, including probably all those who have played hockey themselves, understands that Corsi and Fenwick don't capture the whole game or even the biggest part of the game. So basically, you've just called everyone on CP Jenny McCarthy. Would you like ketchup for that foot?
__________________
WARNING: The preceding message may not have been processed in a sarcasm-free facility.

Last edited by Jay Random; 02-22-2015 at 11:38 AM.
Jay Random is offline   Reply With Quote
The Following 7 Users Say Thank You to Jay Random For This Useful Post: