20 Sep 12, 04:09AM
The design of this experiment is raw, with limited variables but it does give us a broad sense of popularity. I'm sure we can improve upon this experiment to make it more accurate, so don't overanalyze it. Take it less specifically and use it to get the general idea.