Healer Changes Straw Poll from Reddit

[Warkupo]

The important takeaway is that with 900+ participants, the poll is more than enough to have statistical validity. Of course a bit of background would help determine the standard deviation but it clearly shows a majority of players aren't happy with the healer changes as proposed.

SE should take that feedback to heart.

It's not really representative though; The vast majority of the playerbase does not frequent the forums or reddit. Furthermore the people who are upset at the healer changes are a lot more likely to seek council from people who share their opinions, meaning that any poll you put out is going to have a natural bias to it to begin with. Very rarely is a poll unbiased, and you should always consider bias factors when analyzing data.

[Kurando]

snip

Pretty much this. Unless it's an official poll were every single player gets involved then you really should take polls you find through 3rd party mediums (Reddit, Discord etc) with a grain of salt as it does not show a true representation, with bias playing a big role. Content/happy players have less reason to complain and even then it would get drowned out by those on the other side of the spectrum so SE will be cautious when trying to establish which group is actually the "majority".

[Deceptus]

It's not really representative though; The vast majority of the playerbase does not frequent the forums or reddit. Furthermore the people who are upset at the healer changes are a lot more likely to seek council from people who share their opinions, meaning that any poll you put out is going to have a natural bias to it to begin with. Very rarely is a poll unbiased, and you should always consider bias factors when analyzing data.

Pretty much this. Unless it's an official poll were every single player gets involved then you really should take polls you find through 3rd party mediums (Reddit, Discord etc) with a grain of salt as it does not show a true representation, with bias playing a big role. Content/happy players have less reason to complain and even then it would get drowned out by those on the other side of the spectrum so SE will be cautious when trying to establish which group is actually the "majority".

That's not how polls work. You try to avoid selection bias by trying to get as many people as you can from a wide variety of sources. You take the data that people give you. You can't just say, I don't agree with the result, lets discredit those who participated.

That's why I said with the number hitting 900+ respondents (now 1000+), it was statistically valid. At 1026, it would have 95% confidence interval at +-3 % deviation which any pollster would LOVE to have.