Oh, wow! Thanks for collecting these stats, this is quite interesting to think about.
While the results are not surprising, it's good to have some sort of hard statistics. Most of the arguments that have been made about things like trash talking and massive teams are based on personal experience and believability, but we don't really take specific numbers and records. This is, at last, a collection of factual data.
And it's good that you did it over multiple servers, so that there were fewer players in the same games. However, perhaps next time someone does it, they should take it at different points during the day, on different days of the week for even more detailed research, as this would show stats from both busy times like the weekend and quiet times like early in the morning or during school hours.
In addition, different regions? c:
It'd be really awesome if you got some people from different regions to recreate the experiment in different regions and see how the different regions are different.