Dayren

exhoost fume

Absolutely (or at least in terms of what people spend more time on): I meant that in the context of sample size in player-made surveys (which the post covered), not as a general statement. I don't know though if a significant amount of people on the forum would be as disingenuous as you suggest: to lie about what they enjoy just for the sake of complaining (seems silly but I can't rule it out, don't visit here often). What I really think the discrepancy you describe primarily comes down to is previously mentioned here and in the doc as well: the source of the respondents. It's possible -

That's a funny thing about statistics: obviously, you can't ever poll a group of people in its entirety BUT you don't need to. Even a seemingly small fraction of the whole can give you reasonably accurate insights, even when you're using random sampling (which is really the only option here since, well, we don't really have the information needed to do quoata sampling). In practice: the sample size in this survey would be good enough to produce data with a 99% confidence level (academics normally use 95%) with the margin of error being 3% for a group consisting of a MILLION people - and, while