You need a larger sample size. There are millions of players playing this game. You sampled 200(in the endzone for the people who do not care for ranked that much no less). Your sample size is 0.0002% of the playerbase, and that is if it is just 1 million. Now if we went for a more realistic number of 100 million(there are 6+ billion people on earth, and this is one of the most popular games) your sample size is 0.000002% of the playerbase. Nowhere close to being even 1%...
It has been ~3 weeks since the thing started, and even though I might of played 25 games since then, other people have work/school/lives that they have to tend to(and some might only play 1 ranked game a night, with the rest being normals for all you know). And your plat 2 friend seems shady, since he couldn't even be bothered to show his data.
I also question why you chose Gold, which people would end with since they get the skin reward in, and not Bronze, Silver, Plat, Diamond, Challenger, or Master? Any of those would be more accurate than Gold as people still have motivation to try to keep trying to rank up. And this data is just from one server, of which there are ~10-15.
Oh, and as Eedat pointed out, the badges are not as accurate as, say, going into the profiles and checking yourself. For all you know, ~10-30 of those people have played within 10 games of the badge requirement, but I guess you don't believe in giving multilayered statistics.
Another thing, someone else brought up the point that(apparently) many people smurf. Now this absolutely could not mess with your findings, could it? The fact that some of the people could be playing ranked, but on a different account, which would have the ranked games played not be recorded on their currently unused profile?
This post is way too statistically unsound for it to be of any use. And since you didn't post sources for the data you used, and I do not know you or of you, I cannot trust that you did not fudge the numbers in your favor.
