Well, i studied statistic a little on college ...
And i know there is quite important "IF" in creating a sample size.
I mean, concidering it contains as many variant people as posisble it would be enough.
And the less variant your sample is, the more skewed results you get.
Say, there is quite good chance that reddit will be probably used mainly by Americans (that is not racist, its just not so favorite in central europe ... i dont say we dont use it, but we have other sites ... simmilar as Twitter / Facebook), usualy between 20-50years, mostly computer games players ...
But your sample would need equal representation of people who play computer games, people who play DnD, people who played BG 1&2, people who did not, variety of ages, maybe even genders, and preferably from as variant corners of the world as possible.
And that is something reddit just can never asure.
Last edited by RagnarokCzD; 02/06/21 12:54 PM.
