View Single Post
Old
11-23-2012, 02:12 AM
  #71
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
Quote:
Originally Posted by seventieslord View Post


thanks. looks like you ended up coming to the same conclusion I banged my head against the desk for hours to come to.

question: after playing with the spreadsheet, do you see any way in which Espo could have improved Orr's numbers? I had to give up that way of thinking before I could make the numbers come out properly.
Actually, I think there are multiple solutions... but that we need to find solutions where Orr w/ Espo > Orr w/o Espo. So my first solution seems invalid based on that constraint.

I may have found a more singular solution based on these additional constraints:

Orr on/off = 2.19/1.10 = 1.977 = (O+E)/(E-O)
Espo on/off = 1.64/1.41 = 1.149 = (O+E)/(O-E)

About as close as I can get (values were 1.972 and 1.149... as we want closest match for Espo):

E+O 2.36
O-E 2.06
E-O 1.20
-(E,O) 1.04

The question is, how accurate is the assumption that Orr & Espo's ES ice time was not correlated (IOW did they play together more at ES than would be expected by chance)?


Last edited by Czech Your Math: 11-23-2012 at 03:15 AM.
Czech Your Math is offline   Reply With Quote