View Single Post
Old
08-08-2012, 05:54 PM
  #4
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,501
vCash: 50
Quote:
Originally Posted by Iain Fyffe View Post
I make this very point in the post. I think it's the most likely explanation for the negative correlation, however more work would certainly be needed to confirm that.
Sorry, I just looked at your chart and didn't have time to read through your blog at that time.

It's certainly a possible factor, but I would doubt it really explains negative correlations. Perhaps it slightly lessens the expected positive correlation, but there's not even actual evidence of that at this point.

You make a good point about the score effect, but wouldn't the leading team play defense in an effort to avoid the high quality scoring chances which may necessitate penalties? Otherwise, they would be A) giving up more shots (according to what you said), B) giving up the same % of high quality shots, and so a greater number of high quality scoring chances, C) decreasing their own shots and scoring chances. How would this help them protect a lead? I would guess they might decrease their own shots and scoring chances, in exchange for more possession and/or fewer high quality scoring chances for their opponents. This would make sense to me, but then it should not increase the high quality scoring chances for their opponents (which shouldn't increase penalties by the leading team). Also, the leading team, by giving up scoring chances of its own, would be seemingly avoiding penalties which might be committed to create their own offensive advantage, which might negate much or all of any effect that is there from them giving up more shots.

I think the ES GF/GA data vs. new PPs (or NOMS as you term it) would be the most helpful at this points. It's not my highest priority right now, but I may look at it in the future.

Czech Your Math is offline   Reply With Quote