I am not sure how you got that I can't see your data, but I already reviewed the math behind it. And everything looks good to me. Also, importing from other timers coincides with the statistics. Furthermore, if there are some unusual solves, they must be for sure at the beginning of your data...