Quote Originally Posted by service_8572198 View Post
Forget all seeding, make the tree completely random, like the football cups here in UK (yes football not soccer :P)
Once the prelims are over, each player is entered randomly to the 512 tree, total luck of the draw
Ultimately 'the best' should win, but may give lower powers the chance to get further
I have a feeling that that is the way it is now. At least that is what I have concluded from the pairing that I have seen up to now.
I do agree with Matt. Since it is a single elimination tournament the standard seeding rules should be followed First with last, 2nd with next to last, etc.) The real issue is how to do the rankings. Any thing that is related to total power is not a good indication. It would have to be maximum ever attained power, like on the leader boards, or it could be manipulated.

One way to do it is by how far has the player has progress in prior CWFs. A player could get 1 point for each stage that he has reached in the last 20 (or some other number) CWFs. If there ate ties, then it is resolved by number of wins in the preliminaries and then maximum attained power. The reason I have the number of wins in the preliminaries as 2nd instead of first is that because there are not enough battles to make it statistically relevant to be used first.