Here's my take - we can't apply interpretation to people's predictions because there's no clear line between what's "obvious" and what's not. The policy at least since I've been running this is that what you post is what counts. I think we should (and will) cap each character at 100% and 0% because it's impossible for a character to go outside that bound, but doing anything beyond that requires us to make a judgment of your intentions.

Also, I warned him about his mistake soon after he made it, and he had almost a week before the match where he could have checked his predictions on the website, so there was plenty of chance for him to fix it.
---
R.I.P. Robert Jordan, 1948-2007