398 replies to this topic

[MustrumRidcully]

Heffay, on 03 November 2013 - 08:10 AM, said:

Only people who care about 3PV will vote/post on a 3PV topic. People who don't care won't bother even reading the threads.

That is why self selected polls are so useless. You're only polling the people who care about it in the first place.

But people care either positively or negatively about it. If someone actually wanted 3PV, and see this type of poll, wouldn't he reply?
You see it many other polls on the forums. Poll outcomes are basically never that one-sided.

And of course, votes of people that don't care either way are not relevant to figure out if a feature is needed. But if 90 % of those that care at all about 3PV vs 1PV don't want it all, is there any indication that it's a topic worthwhile to pursue?

[Heffay]

MustrumRidcully, on 03 November 2013 - 08:37 AM, said:

But people care either positively or negatively about it. If someone actually wanted 3PV, and see this type of poll, wouldn't he reply?
You see it many other polls on the forums. Poll outcomes are basically never that one-sided.

And of course, votes of people that don't care either way are not relevant to figure out if a feature is needed. But if 90 % of those that care at all about 3PV vs 1PV don't want it all, is there any indication that it's a topic worthwhile to pursue?

People who don't care about it at all (90% of the population) just don't care. They fit under the "it doesn't matter" bucket. And yes, they matter if you are trying to determine the will of the community. You can't throw out 90% of the population just because they have no opinion on a topic. If you want to see how big of a deal 3PV is, "don't care" is just as valid as "good" or "bad".

[MustrumRidcully]

Heffay, on 03 November 2013 - 09:01 AM, said:

People who don't care about it at all (90% of the population) just don't care. They fit under the "it doesn't matter" bucket. And yes, they matter if you are trying to determine the will of the community. You can't throw out 90% of the population just because they have no opinion on a topic. If you want to see how big of a deal 3PV is, "don't care" is just as valid as "good" or "bad".

Don't care means you better figure out what they care about and try to make that happen for them, because you're otherwise wasting your time.

I suppose no one really cares if the PGI office building is painted in white, black or pink, but that doesn't mean the developers should go off repainting instead of developing UI 2.0 or CW or something like that.

[Cimarb]

How do I unfollow a post? My Eridani Light Horse BeatStick ® is wearing out....

[Heffay]

MustrumRidcully, on 03 November 2013 - 09:05 AM, said:

Don't care means you better figure out what they care about and try to make that happen for them, because you're otherwise wasting your time.

OK, that is 100% legitimate. Of course, what they care about is not 3PV. We need UI 2.0, CW, etc.

[Windies]

[Farpenoodle]

Ur all dum

[Ghogiel]

Heffay, on 03 November 2013 - 08:10 AM, said:

Only people who care about 3PV will vote/post on a 3PV topic. People who don't care won't bother even reading the threads.

That is why self selected polls are so useless. You're only polling the people who care about it in the first place.

Nearly all polls suffer from selection bias. You think people responding to random phone, email, post or anything polls are not self selecting themselves. All it does is introducemargin of error, they can still be statistically representative.

Heffay, on 03 November 2013 - 09:01 AM, said:

People who don't care about it at all (90% of the population) just don't care.

bolded emphasis.

OK no. Just because they didn't vote doesn't mean they don't care lol. It means they didn't vote.

Edited by Ghogiel, 03 November 2013 - 10:51 AM.

[Heffay]

Ghogiel, on 03 November 2013 - 10:46 AM, said:

Nearly all polls suffer from selection bias. You think people responding to random phone, email, post or anything polls are not self selecting themselves. All it does is introducemargin of error, they can still be statistically representative.

No, real pollsters go out of their way to avoid selection bias. This "poll" didn't even attempt to get a representative sample. It was the absolute worst case scenario for a poll.

You see this in online polls all the time. There is always a blurb that says "the results of this poll are for entertainment purposes only, and should not be taken as an accurate representation of people's opinions." Or something similar to that. When you have a poll that ignores 90% of the people who say "I don't care enough to even vote", the margin of error is *huge*. Basically: 90%.

The simple fact remains (and is the basis for why Wikipedia refuses to acknowledge the issue): Self selected polls are useless for determining what people think.

Let me put it another way: How many people saw that poll and didn't even bother to vote? That entire population is discounted in the results of that poll. If you don't have a random sample (or a way to correct for a non random sample, which you *don't* have), your results are meaningless.

Let's take a look at one of the polls posted earlier: http://mwomercs.com/...also-needs-1pv/

Less than 100 people voted for it. Over 2000 people looked at it. Since it only records registered users who look at it once (go ahead and test this out if you don't believe me), that means the VAST majority of people didn't even care enough about the topic to vote. They don't care about 3PV enough to even bother to click one simple button, which means *they don't care*.

Edited by Heffay, 03 November 2013 - 12:07 PM.

[Ghogiel]

Heffay, on 03 November 2013 - 11:59 AM, said:

[/size]

No, real pollsters go out of their way to avoid selection bias. This "poll" didn't even attempt to get a representative sample. It was the absolute worst case scenario for a poll.

You see this in online polls all the time. There is always a blurb that says "the results of this poll are for entertainment purposes only, and should not be taken as an accurate representation of people's opinions." Or something similar to that. When you have a poll that ignores 90% of the people who say "I don't care enough to even vote", the margin of error is *huge*. Basically: 90%.

The simple fact remains (and is the basis for why Wikipedia refuses to acknowledge the issue): Self selected polls are useless for determining what people think.

Let me put it another way: How many people saw that poll and didn't even bother to vote? That entire population is discounted in the results of that poll. If you don't have a random sample (or a way to correct for a non random sample, which you *don't* have), your results are meaningless.

Let's take a look at one of the polls posted earlier: http://mwomercs.com/...also-needs-1pv/

Less than 100 people voted for it. Over 2000 people looked at it. Since it only records registered users who look at it once (go ahead and test this out if you don't believe me), that means the VAST majority of people didn't even care enough about the topic to vote. They don't care about 3PV enough to even bother to click one simple button, which means *they don't care*.

No dude. All polls have selection bias. All it does is introduce margine for error.

And no, it doesn't mean they don't care about 3PV because they didn't vote in a poll. Inserting a reason why they didn't vote is asinine.
Maybe they don't care about polls in general. :shurg:
And actually that is what polls are for, using a small sample, ie those that did vote, to roughly gauge how a population thinks on an issue.

Democrazy is a hell of a drug.

[Heffay]

Ghogiel, on 03 November 2013 - 12:19 PM, said:

No dude. All polls have selection bias. All it does is introduce margine for error.

And no, it doesn't mean they don't care about 3PV because they didn't vote in a poll. Inserting a reason why they didn't vote is asinine.
Maybe they don't care about polls in general. :shurg:
And actually that is what polls are for, using a small sample, ie those that did vote, to roughly gauge how a population thinks on an issue.

I didn't say they don't have bias. I said they attempt to minimize it with random samples that accurately reflect the population they are trying to poll. And the 3PV polls have done absolutely nothing to do that. It's self-selected of people who care to vote. You want a more accurate result from the polls? Take the number of votes of people who are opposed to it as a percentage of the TOTAL number of people who saw the poll, not just those who voted. It's still a subset of the population, but includes the people who don't care. Something the original polls made no effort to do so.

And pretty much across the board, only 5% of that group of people even bothered to vote. So BEST case scenario, you're looking at 4% of the population who opposes 3PV, with a rather significant margin of error.

[Ghogiel]

Heffay, on 03 November 2013 - 12:28 PM, said:

I didn't say they don't have bias. I said they attempt to minimize it with random samples that accurately reflect the population they are trying to poll. And the 3PV polls have done absolutely nothing to do that. It's self-selected of people who care to vote. You want a more accurate result from the polls? Take the number of votes of people who are opposed to it as a percentage of the TOTAL number of people who saw the poll, not just those who voted. It's still a subset of the population, but includes the people who don't care. Something the original polls made no effort to do so.

And pretty much across the board, only 5% of that group of people even bothered to vote. So BEST case scenario, you're looking at 4% of the population who opposes 3PV, with a rather significant margin of error.

Um... lol?

You are completely fabricating random % and adding a ton of BS here.

[MustrumRidcully]

Heffay, on 03 November 2013 - 11:59 AM, said:

[/size]

No, real pollsters go out of their way to avoid selection bias. This "poll" didn't even attempt to get a representative sample. It was the absolute worst case scenario for a poll.

They try to, but do they succeed?

When someone calls me and asks to response to a survey (regardless of how public/officical/legit), I usually decline. There is a subset of people that do not want to be polled in such surveys.

So again, there is a self-selection bias.

If you're arguing:" Yeah, but they still get the people that care enough about the topic of a poll to respond"... Yup. Just like forum polls.

[Ghogiel]

MustrumRidcully, on 03 November 2013 - 01:00 PM, said:

They try to, but do they succeed?

When someone calls me and asks to response to a survey (regardless of how public/officical/legit), I usually decline. There is a subset of people that do not want to be polled in such surveys.

So again, there is a self-selection bias.

If you're arguing:" Yeah, but they still get the people that care enough about the topic of a poll to respond"... Yup. Just like forum polls.

Phone polls are the classic example of selection bias. It excludes everyone with unlisted numbers or people who simply don't have phones. Not that there is anything wrong with that, all that means is there is a small margin of error.

[Heffay]

Ghogiel, on 03 November 2013 - 12:46 PM, said:

Um... lol?

You are completely fabricating random % and adding a ton of BS here.

No. Look at the number of people who viewed the polls. Then count up the number of people who voted.

How do you explain that 95% of the people who saw the poll didn't even bother to vote? Do you think the distribution of those people even remotely matches what people are claiming the poll shows?

If you don't care enough to vote, you're in the "I don't care" bucket. And that is 95% of the population.

MustrumRidcully, on 03 November 2013 - 01:00 PM, said:

If you're arguing:" Yeah, but they still get the people that care enough about the topic of a poll to respond"... Yup. Just like forum polls.

Phone polls take into account hang ups and "don't care" answers. The forum poll is only people who bothered to vote. If 95% of the people just said "I don't care enough to answer this question", that would be reflected in the poll, under the "I don't care" portion. But they don't just throw those numbers out.

MustrumRidcully, on 03 November 2013 - 01:00 PM, said:

They try to, but do they succeed?

That is why they have a margin of error. And that is why they post that margin of error. For the 3PV polls, there is no posted margin of error, because the people who quote those polls are throwing out 95% of the data that is collected that they don't like the answer to.

Edited by Heffay, 03 November 2013 - 02:20 PM.

[MadcatX]

Although I can't give any percentages, I can say that, like any other gaming forum, a sizable percentage of people come to forums with the intent to complain (be it a valid complaint or not). But that's just a result of at least the western civilization: If we're content or not concerned about something, we generally won't comment unless required to keep the status quo. But when we're upset about something, then we're more then happy to vent our frustrations in a public venue.

Edited by MadcatX, 03 November 2013 - 02:23 PM.

[Sandpit]

Heffay, on 03 November 2013 - 09:01 AM, said:

People who don't care about it at all (90% of the population) just don't care. They fit under the "it doesn't matter" bucket. And yes, they matter if you are trying to determine the will of the community. You can't throw out 90% of the population just because they have no opinion on a topic. If you want to see how big of a deal 3PV is, "don't care" is just as valid as "good" or "bad".

Ok let me point somethign out here....

3000 people care and don't want it
100 people care and do want it
90% of the population doesn't care one way or the other.

So you're saying that you should do somethign that ****** off the 3000 people and appeases the 100 since the 90% don't care. According to your argument that's exactly what happened.

[MadcatX]

Sandpit, on 03 November 2013 - 02:42 PM, said:

So you're saying that you should do somethign that ****** off the 3000 people and appeases the 100 since the 90% don't care. According to your argument that's exactly what happened.

Then his argument would be valid in this case, because that is exactly what happened.

Edited by MadcatX, 03 November 2013 - 02:58 PM.

[Literally]

Someone mail Heffay a tuition bill for the schooling he's getting here. No free rides!

[Heffay]

Sandpit, on 03 November 2013 - 02:42 PM, said:

Ok let me point somethign out here....

3000 people care and don't want it
100 people care and do want it
90% of the population doesn't care one way or the other.

So you're saying that you should do somethign that ****** off the 3000 people and appeases the 100 since the 90% don't care. According to your argument that's exactly what happened.

Well, it's more like 95% of the population, but let's not quibble over a few percentages. And yes, it's much like when Blizzard decides to do something like nerf Hunters. Yes, a portion of the player base gets REALLY upset, but sometimes it has to be done in the best interests of the game.

Splitting the queue is bad for the game. 3PV is good for the game. Sorry, but that is basically what you're left with.

Literally, on 03 November 2013 - 03:01 PM, said:

Someone mail Heffay a tuition bill for the schooling he's getting here. No free rides!

When your "evidence" is thoroughly discredited and you have nothing left to stand on, all you have left are ad hominem attacks. And that doesn't win any debates.