haha, "Better Bettors" came to mind, hmm..

That's a ridiculously cheesy pun.
....I like it. Suits me well. Better Bettors it is!

That reminds me: do you have any ideas on how Spread Betting will work with this? I know that the concept of spread betting itself draws from betting in sports-- is there a suitable sports-based analogue for this kind of 4-way competition? I had an idea, but it's a little more complicated than straight spread betting. Basically, you set over/under numbers for each character, and bettors have the chance to wager on any or all of them. For example, the first set of lines could be Knuckles (28), Rikku (24), Vaan (20), Yoshi (28). A bettor can bet Knuckles > 28, or Knuckles > 28 AND Yoshi < 28, etc. Each character they bet on nets them 0.5x the amount bet originally if they get every match right (for a maximum of 2x their original bet), but if they err on even one of the characters they bet on, they lose their bet. How does that sound?

As for the scoring system for the Oracle challenge, I'm more for the side of rewarding people who get the winners right in some way (or rather, penalizing those who get them wrong). For example, assume the following situation for the first match:

Knuckles 28%
Yoshi 27%
Rikku 23%
Vaan 22%

Now suppose that one player guesses Knuckles 30%, Yoshi 30%, Rikku 20%, Vaan 20%. Under a straight percentage scoring scheme (each character worth 25 maximum), this player gets 90 points. Suppose a second player guesses Yoshi 30%, Rikku 25%, Knuckles 24%, Vaan 21%. This player gets the same 90 points despite getting the winners wrong. If there's a penalty system for missed winners, I think that since there are more characters in play, it'll be harder to hedge predictions to stay on the safe side, while still providing incentive to make bold winner predictions. (Heck, some players have won the Challenge outright by doing just that. >.>) Also, it's consistent with the philosophy of the original scoring system.
...I wrote this up without noticing that cyko had put up a much more thorough example, so, uh, I'll also say that his example works excellently and I agree with it. But I'm not going to be dismayed or anything if you choose not to implement a penalty for incorrect winners.