Thanks for posting, it is very interesting to see the data from a real pricing experiment. I'm not sure what you're using for transaction processing, but I think that could affect the rankings due to larger percentage transaction costs on smaller transactions.