Yemack
Registered User
- Oct 30, 2007
- 8,246
- 5
Okay, that's a little better, but you're still calculating the chance to lose against Montreal based on past history against Boston, for example. If you dismiss that as an impossible, meaningless correlation, then your chi2 goes down.
I guess all I'm saying is don't read too much into the numbers, but if you've already acknowledged that then no biggie, we can move on.
I see what you are saying. Yes you are right in a sense that it's not a 100% pure sample group however, I never intended to find independent underlying causes or anything. I just simply tested a factor vs a factor. whatever that factor maybe.
If I did pearson correlation or other descriptive stats, it would have been wrong to use that arbitrary group as a factor. But not in this case in my opinion.