Heffay, on 01 November 2013 - 11:59 AM, said:
So are we all in agreement that the original 3PV polls are meaningless and no one will use them to say what the will of the community is?
Heffay, on 01 November 2013 - 12:30 PM, said:
So plan B is that I no longer point out how those surveys are flawed, and others continue to abuse statistics all in the name of mock outrage. Got it.

I’d like to point out that Heffay tried this line of argument in this thread (
http://mwomercs.com/...also-needs-1pv/) and demonstrated that he does not understand how statistics, population samples or polls work. He even claimed to work with statistics every day, which is obviously not true. Do not hesitate to remind him of this any time he starts arguing about polls in a thread.
The forum polls about 3PV were the largest recorded, most unified, community response to any issue ever. I’ve been here since the first month of open beta, and the only thing I can remember coming close was the poll on “should SRMs have damage upped to 2 from 1.5” (another overwhelming “yes”). The first poll, made when PGIGP first introduced the idea of 3PV back in the spring, had more than 3000 votes, <90% of which said that 3PV shouldn’t even be in the game at all. In the thread linked above, a current poll indicates that a full 30% of users who still browse the forums and reply to threads don’t play because of 3PV.
The idea that forum polls have no informational value is at best a misunderstanding and at worst a lie. It’s a misunderstanding that betrays a lack of understanding of how samples work and why its important for them to be random, or it’s a lie designed dismiss information that doesn’t line up with the liar’s world view, or rather what the liar wants your world view to be.
I’ll repost here my last argument against the “forum polls have no value” idea from the linked thread:
Heffay, on 01 November 2013 - 05:02 AM, said:
Instead let's just say that anyone who believes that a self selected sample is an accurate representation of the population as a whole clearly has no idea what they are talking about and I doubt that any amount of evidence will manage to break through your confirmation bias, so let's just agree to drop it?
Not going to drop it because you're dead wrong, and I'm not conceding that point to you. This lie/error is a key pillar of belief for the shills/knights of PGIGP that must be knocked down if the game is to improve.
First of all, I will admit that the data that we have (forum polling, player/website/magazine reviews, etc.) is not ideal. Ideal data would be a polling a sample of every individual who has ever played MWO, not just those currently playing MWO. While it would be possible for the devs to do the latter, the former is probably impossible.
Second, that does not mean that there isn't information that can be gleaned from forum polls. In fact is is the ONLY information that we have. Dismissing the forum polls and then coming up with some sort of explanation for why they're wrong isn't creating new information, its dismissing the only information you have which puts your argument on shakier ground than anyone using forum polls ... in fact it makes your arguments groundless.
Third, there is no a-priori reason to assume that the self-selected population of forum users isn't representative of the population of MWO players at large. The fact that someone uses a forum and posts on it is indicative of a higher level of motivation and dedication to the game. Level of player motivation and interest is not conceivably correlatable with specific opinions about that game.
Fourth, as I stated before, the existence of feedback forums, containing polls, for MWO and every other game out there now is an implicit assumption that forum users are at least a reasonable approximation of the game population at large. The existence of these forums rests on the assumption that there is useful information to be gathered from them. Of course that doesn't stop bad devs and their shills from dismissing this information when it doesn't fit their pre-conceived worldview.
Finally, self selected populations are used when necessary in other fields. The entirety of drug and biomedical research involving human populations rests on self selected populations (especially the healthy controls). You can't force people to participate and you can't randomly call people and ask them to participate (as you might do in a phone poll). You put up advertisements and the most interested and motivated (sound familiar?) individuals apply. I'm pretty sure the same approach is used in marketing focus groups. In each case there is no way to account for the """silent majority""" of individuals who were not motivated enough to participate ... and yet these studies are treated with validity.