Yeah, I know it's early and it's a small sample size and the same number of voters haven't voted (I tried to cover that with the "as of this post" comment).
Usually %'s are the great equalizer, but I'll admit that with #'s this small, it's not a great indicator.
I actually think that a shift of roughly 15-20% is about appropriate when assessing a 3-1 series compared to a 3-2 series. This kind of shit interests me though.
Bookmarks