Quote Originally Posted by VelKallor View Post
My point was specifically on using it as a trend in terms of player retention. Taste and opinion are not a factor.
It is a factor because it decides if looking at steam charts is valid or not. If the people are largely the same and like/dislike for the same reasons then looking at steam numbers is valid. The cutoff point for statistics being good enough to not fall pray to statistical errors is at roughly 800 people for sample sizes. Way above that and you dont need to fear statistical errors skewing your results. What sample size wont account for is systematic error. If you want to invalidate using steam charts as sample you need to show that stormpeaks only refering to steam charts is commiting a systematic error. You would do that by showing that console and standalone players have significantly different interests/tastes.