In today's contest, problem E is rated 2900. Problem F is rated 2800. But, both in the contest and after, more people had E solved over F. This is good evidence to show problem E is easier than F — but it is still rated lower. Are there other major factors in determining problem rating other than solve count?
On another note, many people gave up after the unrated announcement, and long queues prevented some AC's. Thus, problems D and E especially became pretty overrated. I think ratings of overrated problems due to these issues should be brought down because when practicing, it gives a false sense of the real rating of the problem.