Board Thread:General Discussion/@comment-27660480-20190224111749/@comment-35284508-20190225212856

I am the person who started the thread https://seekers-notes.fandom.com/wiki/Thread:126388 mentioned above. We jointly collected 80,000 observations with maximum gifting beforehand. Towards the end I collected another group  (6000 obs I think) without the gifting beforehand and could not find a significant difference. However, I never tried from an account that had never gifted. And I never checked talismans because it would have been too costly.

You are ambitious in the conditions you set out detailing exp lvl, gifting, talismans used and I admire that. But when you list results for a single day and try to compare to another single day you are ignoring the normal variability in the results. In the thread on statistics I started with data from 20 days as a baseline. As other people added results (for 80 player-days total), it just became more stable. But each person adding results had periods where their range was like your 52 to 29 spread.

I estimated that you cant tell what the result of your test was (within 10%) unless you have at least 5000 observations. That is 5 days of 10 charges per day on 100 friends/day. 10 days if you have 50 friends... etc. Better results if you can collect 20 days, Better still at 80 days. If you try to prove something with one days results you are like a guy who asked for his money back because he rolled his new die 6 times and didn't get a different result every time.