I'll take a model that iterates often, with higher likelihood of variance in ideology, over one that optimizes for moderation at the expense of gridlock.