The problem in a nutshell is level is a poor predictor of actual roster strength. There are several ways of showing this, none of which appears satisfactory to my co-debater–since he can always say “but what about [next method]?” when I answer his previous objection. When I get to the end of his list of methods, he cycles back to the first one. Honestly it gets tiring.
People have been pointing out the flaw of assuming level=roster strength often to him, for a long time. See, for instance, this post from Kharee in October of last year:
Cycling through demonstrations gets old fast. I’m the latest person to decide that it’s not worth the effort.
Oh, and @Balusticballsac, not that it ought to matter, but I’m a man, not a woman.