I said in the original thread in defence of the survey that response bias doesn’t invalidate the results, just limits the conclusions we can make, but this is a type of conclusion we absolutely can’t make due to the said bias.
90% of the respondents not very representing of the entire playerbase but who are likely invested in Mercy think this. That would be about right. It even says in the thread that many respondents are support/Mercy mains.
2,000 people is roughly the amount taken on exit polls to indicate the outcome of an election for countries with up to 60,000,000 and they are still extremely reliable.
The second point stands, the first point makes no sense if you actually understand statistics.
2k people randomly selected from the populous is MUCH more representative than 2k people gathered from the community that comes together to complain. It’s like grabbing 2k people out of a 60million from ONE city. They’re going to be biased.
Because seriously 10% of 10% of players is like 1% of the forum users lmao
Because the rough amount of players who use the forum on any forum in every game is usually around the 10% mark a forum poll is no where near an indication of what people want and what they dont.
The problem isn’t the number of respondents in itself though, as it seemingly was suggested. 2k would be a fair enough sample size were it fully unbiased. (Note: it’s not unbiased in this case.)
Just that every game forum ever has had forum involvement of less than a few % usually. Of that small representation overwatch has had mostly people who complain. Considering that it’s a tiny, select and vocal group, I’d say your average overwatch player is NOT well represented by those who seek out, post and in this case care about mercy enough to respond to the survey.
Shrug I doubt the survey says much, I didn’t read everything but I doubt any statistical programs like SPSS was used for example. All it says is that there are quite a number of people unhappy with her currently. We already knew this.
I find if you want truthful answers it’s usually best to hide the “actual” questions in a questionnaire with several other questions to lessen the risk of the survey itself only attracting a particular set of participants.
Besides, people who are disgruntled are FAR more likely to actively voice their concern, so results will be skewed even further.
This is called the vocal minority and that’s basically what the entire forum is, not even people commenting on Mercy. More often than not, people come onto the forums to complain, so you don’t get the perspective of those who enjoy with the game.
Since a huge part of the forum contain is basically guys whining about Mercy, nothing interesting here.
Make a survey taking a significant part of Overwatch players directly in game, people will finally understand that noone give a damn about Mercy.
Unfortunatly, this survey barely gives more than an insight on the topic.
2680 people’s opinion cannot be tought of relevant in a game of so many players.
But you could argue it represents still part of the playerbase. So
The Survey was advertized on the forum, a place that unfortunatly isn’t checked by everyone (if it was on the homepage of the game menu, it would be more representative). It was advertized on other platforms but not where most data could be collected.
As the author claimed,
Most people who answered are mercy main.
Tbh, making a survey like that doesn’t make sense in the first place. She just got nerfed so there is no point in asking if she is as fun to play as before because of course people will say no. It’s like asking the population if they liked the new increase in taxes. They might hate it but it might be for their own good in the future. (Nerfing mercy is maybe a positive aspect for the future of OW)
The survey gave us an insight, but it cannot be held as the sacred texts