Celestia Forums

Since a lot of people lately seem to be thinking that factual addon announcements are getting overshadowed by fictional ones in the Celestia Addons forum, I thought that it may be a good idea to make a new forum, especially for announcing and talking about fictional addons. I am putting this thread here because I'm thinking about doing this myself and I would like to see what others think of the situation.

Michael Kilderry

You left out "don't care"

Michael Kilderry wrote:I don't quite think so, I have brung out the colours in the three images
...

Should There Be A Seperate Forum For Fictional Addons?

...and as a native English speaker you are teaching us poor foreigners really strange things

Bye Fridger

Please, nobody's perfect. When I make so many posts quickly I can't always get it right. I think 2 mistakes noticed out of over 200 posts is pretty good.

If you don't care, then you probably shouldn't submit a vote.

Michael Kilderry

Thanks for the votes, I might ask what Chris thinks about this situation, unless he replies here first.

Michael Kilderry

Michael Kilderry wrote:Thanks for the votes, I might ask what Chris thinks about this situation, unless he replies here first.

Michael,

You may wish to wait a bit... I mean, you only have 8 votes here. Hardly
enough to make decisions which will affect thousands of users, eh?

Take care, Bob

I completely agree with Bob and would have written the same if he was not ahead of me

Your vote is worth NOTHING from a statistical point of view.

If you do n votes altogether, the remaining uncertainty goes like sqrt(n). For 8 votes, sqrt {8} = 2.8

YES votes : = 6+- 2.8
NO votes : = 2+- 2.8

Hence 6 -2.8 = 3.2
2+2.8 = 4.8,

Obviously, there is TOTAL overlap among YES and NO opinions within the large uncertainties of +-2.8.

As you may easily convince yourself that this changes a lot once you have many more votes! That's the essence of polling...

Bye Fridger

8 people have voted. 2 voted no, 6 voted yes. There's no "uncertainty" there in the number of votes at all, and neither is there any uncertainty in where they were cast.

Fact is, one could argue that all online polls in all forums are statistically irrelevant. After all, usually no more than about 10% of the total forum membership even posts at all let alone posts frequently (and many more may just lurk and not comment at all), and of that you'll be lucky if even half of those respond to polls. That's pretty standard across all online discussion boards. Frankly, there are probably not even 100 people who post on these boards on a frequent basis, as can be seen from the memberlist.

But that said, there are 2187 members of this community. If we assume that 100 post regularly enough to be considered "active contributers to the board" then that still leaves only 8 who have cast a vote on this poll - the rest, apparently don't care either way or aren't interested.

Assuming Chris is even open to splitting up the forums as suggested - I think we should still just go with what the majority vote for. If those who don't vote either don't mind or don't care what the outcome is, then they won't mind if the decision is made for the boards to be split. If they DID care, they'd vote to influence the outcome.

Evil Dr Ganymede wrote:8 people have voted. 2 voted no, 6 voted yes. There's no "uncertainty" there in the number of votes at all, and neither is there any uncertainty in where they were cast.
...

You don't want to teach me statistics do you?
You would be right if all users would vote. Then you just have to count the votes and the largest number wins.

In polls like this one, however, only a small subsample of all possible voters actually submit their votes, whence the statistical uncertainty arises.

If 80 people would have voted, for example, the residual uncertainty would be sqrt(80)=8.9. Suppose 60 voted YES and 20 voted NO. Then this is a CLEARCUT result beyond doubt.

The YES votes are: 60 +- 8.9
the NO votes are: 20 +- 8.9

So in the "worst" case the YES guys have 60-8.9=51 votes, while the NO guys only make it at best to 20 + 8.9 = 28.9.

Hence with a factor 10 more voters, a 60:20 result means a CLEARCUT YES!

t00fri wrote:

Evil Dr Ganymede wrote:8 people have voted. 2 voted no, 6 voted yes. There's no "uncertainty" there in the number of votes at all, and neither is there any uncertainty in where they were cast.
...

You don't want to teach me statistics do you?
You would be right if all users would vote. Then you just have to count the votes and the largest number wins.

In polls like this one, however, only a small subsample of all possible voters actually submit their votes, whence the statistical uncertainty arises.

But why is this even relevant? Who cares if it's a subsample or not?

A question is asked, and it gets a limited number of responses. From those responses, it is known exactly where each person votes (and at the moment it's 7 yes, and 2 no). Unless they vote, everyone else can be assumed to not care about the outcome.

The statistical jiggerypokery is simply irrelevant. 7 people so far have voted yes, and 2 haven't. There's no uncertainty in those numbers, and nobody is even claiming that the votes cast represent all the users anyway.

I didn't say that I was going to make an opinion on the current votes the way they are, all I said was that I was going to ask Chris for his opinion. I think the Evil Dr is right when he says that the small subsample of votes should be taken seriously, even if not many people voted, if you don't vote, it's your loss if you were on the opposite side to who had the most of those votes.

Michael Kilderry