Has anyone attempted an Elo-style rating system for tournament play?
In games with races you could rate by the game to get the sample size up high enough. So A beats B 9-7 would be treated as player A winning 9 of 16 matches vs. B's rating. Is the variance in tournament play too great for...