[cferneyh]

By using two goaltenders on the same team you've eliminated the variables that make inter-team comparisons meaningless.

You haven't proven that those variables don't invalidate GAA as a meaningful stat when comparing goalies on different teams.

If those goaltenders were on different teams and you wanted to dismiss the differences as "better scoring opportunities", or whatever, be my guest. But with parity being what it is these days, the difference of the top 10 or 15 teams defensive skill level is nowhere near as great as you are implying. Granted, having Detroit's D in front of you versus LA's is most certainly going to make a difference - I won't argue that - but it is laughable to imply that SJ or NJ have significantly different defensive abilities.

Good goalies tend to put up good numbers, bad goalies tend to put up bad numbers. It isn't really all that complicated. If you still want to dispute that fact - and I'm sure you do - take a look at the teams that allowed the most goals this season (Atlanta, Tampa, LA, Tor, EDM, Car, Ott, NYI) and tell me which of the starters on those teams should have been worthy of Vezina consideration, had their "weak D" not skewed their goaltending statistics.