Male Viera as a playable character! show your support

[Cerat]

Thanks for the explanations. So I'm asking because I'm a bit confused but, since people can't dislike posts, how would 1577 likes be an accurate sample size?
Even though I linked the poll in my first post, I'm not counting that since people could take the survey multiple times.

Forum posters are only a very small fraction of the community, which the active playerbase is somewhere over 580,000 potentially, maybe even smaller as that figure includes potential alt characters. When taking a sample you don’t need a LARGE percentage of the potential respondents, generally a few hundred-a thousand is sampled in most major statistics depending on what they’re testing. Obviously a larger response pool would be more accurate but that’s not feasible with a population this size in most cases. So they used the data that is available and able to be worked with to determine a sample size that would be large enough to show demand , in this case 384 likes would do so. Think of it this way, in any poll you read about, the average sample size is anywhere from 500-1000 and that’s usually accounting for TENS of MILLIONS of people, far more than the community. Statistics allow you to take a smaller sample size and compare it to a larger pool and that’s where the margin of error, deviation and confidence level come into play to say that while it can’t be 100% accurate with this size this is what it’s believed to be with X confidence and could be + or - X%.

[Senn]

...

When you're talking about sample sizes, aren't those normally done by picking people at random, and not people who all agree on something? The 1577 likes don't take into account the people who don't want male viera. This isn't a random sample size, so would using this still be accurate?
I'm not trying to play devil's advocate or anything, I just want to understand.

When I was calculating the margin of error for myself I used a poll from a much earlier forum thread. It actually gave me a lower margin of error so I'm not sure that is what they are using, but suspect that they are. I'm on phone and can't be near a computer for a few hours, but if you want the poll then later I can link you it.

Is the poll you are referring to the one I linked to in the OP? If not, go ahead and link it.

[Cerat]

When you're talking about sample sizes, aren't those normally done by picking people at random, and not people who all agree on something? The 1577 likes don't take into account the people who don't want male viera. This isn't a random sample size, so would using this still be accurate?
I'm not trying to play devil's advocate or anything, I just want to understand.

It is picking at random since you are not controlling who is liking the thread OP. In this case Virus figured the number of individuals that would need to like (aka show support) to show that there is an overall demand since you cannot tell if the non-likers are in agreement or disagreement or ambivalent. That’s why the number is only 384 needing to like in order to show demand instead of simply saying the 1577 itself shows demand. Essentially on the OP there only needed to be 384 likes, with a + or - 5% margin of error, in order to statistically show there is a demand within the overall community. 1577 like have been given, almost 5x’s the needed number.

That all being said, if I’m mistaking anything with this, and I am only human, Virus or anyone else better versed in statistics can feel free to correct me. this is my explanation for the data provided and I’m fairly sure that’s how it was intended based on my own knowledge on how to read it ^_^

[Senn]

...

Gotcha, thanks for explaining.

[VirusOnline]

snip

snip

Yes all of your responses are correct in interpreting ^^

Deviation is a bit harder to grasp at this time of night, but consider it a necessary factor into considering how many people we must poll in order to relate it to the entire active playerbase.

Unfortunately we can't see dislikes as said. We must go upon responses who are against mViera, other threads that exist that are also against mViera and its number of likes as well which are all vastly low compared to this thread and support in general. What the statistics are virtually saying is that we can take this thread's unique likes + the sample size theory + the poll and we can extrapolate that if SE were to stick their hand into a bowl and grab 330-360 responses from this forum or 384 responses from the active playerbase, one could be 95% confident that at least 50% of those 384 responses would be in support of mViera, give or take ~5% (margin of error). One could call this considerable demand.

For the poll, I made the margin of error greater in the poll only because I'm unsure if duplicates are possible. If we see the worst case scenario and observe that even then in such a worst case margin of error that there is still high demand, well then one can imagine the best case scenario. But the poll I do believe is viable information. It is able to be spread outside of the forums and thus has higher visibility to those whose agreeance/disagreeance varies further helps the random poll idea that can't exactly be had on the forums.

Edit : This isn't even considering the added support for fHroth. The data could be combined but that's treading into the arena of probability & mutually exclusiveness/inclusiveness statistics.