It'd be interesting to see the overall variance on quality vs. rewatachability. That could give some clue on the objectivebess of users' answers to quality. Perhaps quality is just one of those things that is more universally objective than it is subjective?