To expand on this a little bit, one of the things that really confuses people about Kenpom and similar systems is that winning or losing does not matter. It's looking at efficiency margins (how much you are winning or losing by) and adjusting them for quality of opponent. It then creates baseline ratings where teams can be compared to each other, and puts out an expected score for games, which allows these ratings to be continually refined. Kenpom expected us to beat Coppin State by 13, but we got a big bump since we beat them by 38. It expected us to lose to ODU by 2, but we actually moved up a spot or two since we only lost by 1. We took big drops in the three Bahamas games since we were expected to win all three by 2-3 possessions but lost comfortably in all of them.
Of course teams that win more will tend to be higher, but once you understand that the rest of it falls into place. Fordham is still in the 170s on kenpom despite being 12-1, because they have mostly beaten the worst teams in D1, sometimes by small margins. If Mason were to play Kansas tomorrow and lose by 1, we'd probably get a big bump! Same as if we played IUPUI or someone terrible and won by one possession, we'd take a big drop. The schedule has been smack in the middle of D1 average, but the wins are mostly by solid margins and the losses aren't as catastrophic as they felt at the time. Note that this is relative to the question of how is this different than Paulsen - the losses were catastrophic for any chances of an at-large, and 108 on kenpom isn't exactly good.