I like using cumulative distributions because they make small changes in the data a little more obvious. e.g. if all the buckets are 10 but there's a section where they're 11, that difference will show up in a cumulative distribution as a bend in an otherwise straight line, which in my opinion is a much easier difference to see