Action buttons: how will we do the next round of voting?

Greek letters that look a bit like Playstation layout (ΔΠΩχ Φψ)

  • Perfect (>99% of perfection)

    Votes: 11 16.9%
  • Pretty good (67%-99% of perfection)

    Votes: 12 18.5%
  • Average (33%-66% of perfection)

    Votes: 11 16.9%
  • Pretty bad (1%-32% of perfection)

    Votes: 12 18.5%
  • Abysmal (<1% of perfection)

    Votes: 19 29.2%

  • Total voters
    65

Granitehead

Advanced Member
Joined
Oct 16, 2009
Messages
3,011
Be thankful I wanted to something brainless for a moment anyway. :p

Vote away.
 
Last edited by a moderator:
Thanks for the addition voting threads.

This is my voting thread. There are many like it, but this one is mine.

This voting thread is my best friend. It is my life. I must master it as I must master my life.

This voting thread, without me, is useless. Without my voting thread, I am useless.
 
I don't get what percentage's have to do with any of this?

Edit: I also find it funny that Greek letters scored pretty well on the other poll, and it's pretty much rated Abysmal on this one :)
 
Last edited by a moderator:
Edit: I also find it funny that Greek letters scored pretty well on the other poll, and it's pretty much rated Abysmal on this one :)
See the number of votes in both polls and you may have your answer :p

EDIT: Also it seems to be tied for the winner of "most votes in the perfect" category? Seems to be a polarizing choice.

EDIT2: Oh, and the previous poll didn't allow opposing some choices AND preferring others. You could vote everything except those you oppose or only those you prefer. This one allows both making the results much more polarized.
 
Last edited by a moderator:
Thanks for the addition voting threads.

This is my voting thread. There are many like it, but this one is mine.

This voting thread is my best friend. It is my life. I must master it as I must master my life.

This voting thread, without me, is useless. Without my voting thread, I am useless.
full metal jacket :) nice!
 
given how bottom heavy most of these graphs are this mode of voting seems best suited to weeding out those combinations liked least, (Roman numerals scoring especially low I notice) but not what was liked best.    Perhaps it would have been better to include all options on this poll, and then use this to narrow the field...

- Neelix
 
T4b, you waited two days to declare a result from the first poll? Is 72 votes really enough to determine the voice of the community? I'm quite sure there are many people that don't read the forum every day.
Yes, I think we should give the first poll some more time. The name poll has 187 voters, the SoC poll has 314 voters, the case color poll has 127 voters, I think we should at least wait until there are more than 100 voters.

Also I'm not sure if this way of voting is very good -- it's interesting to see how polarized the opinions are, but it also means that the preference of people with a polarized opinion will have more weight: compare someone who thinks one proposal is perfect and all others are abysmal, versus someone who think all of the proposals have their merits and disadvantages (not perfect, not crap) but still prefers one of them over the others, so e.g. he votes "pretty good" for that one and "average" for the rest). If we're going to vote like this, the polarized vote will have a bigger overall impact than the "moderate" vote. That in turn leads to people voting more polarized than their true opinion.
 
Last edited by a moderator:
True. If one could do a poll like that, all polls wouldn't be binary.
 
Last edited by a moderator:
I don't think the opinions are too polarized - maybe the words I put there are too polarized for the liking of some people, but I would just interpret them as

Acceptable?

Yes!

Yes.

*shrug*

No.

No!

Strategic voting has to be held in mind while analysing, but doesn't damage the results. I think it's okay to do it: If you really want your opinion to dominate, then you choose Yes!. If you aren't quite sure but think so, then you choose Yes.

Maybe I just should have actually labelled them like that.

Just after starting the poll, the results were worse, of course. For a vote with more choices you also need a bigger sampling size.

Votes are generally binary because that's easier to interpret and because generally you need to decide for one thing or the other. But if you want an accurate picture of the opinions of people more varied choices give more information.

So all in all I think it is the right kind of poll for this kind of task.

Now was this too early? I don't know, but the other thread already seemed to move on from the poll to discussing how irrelevant the results were and how each ones preferred choice was actually clearly the right one based on the poll and calculating in it's mistakes. :p

Edit: Also considering that the options currently ranked lower in the other poll got like no votes at all - so they seem clearly unpopular. On the other hand there were options added later. If there are significant changes in the first poll I'll admit I was too early, otherwise I'll keep it a "Not sure, doesn't need to be".
 
Last edited by a moderator:
I'd rather think a poll where you rank the options from best to worst is actually better. That makes it impossible to vote one perfect and rest abysmal, since the votes are practically normalized for each voter. How strongly each voter feels the options relate to one another is and should not be relevant when tallying a popular vote since people have different scales of emotion; some polarize immediately and vote only for the extremes, some are mostly reserved and vote the middle options, reserving both extremes for special cases. By normalizing these votes we get something actually comparable irrespective of how strongly each voter reacts to the poll.
 
The forum software makes such a poll almost impossible to make, that's why I didn't consider it.

I agree that the polarizing you describe happens that way, but I don't think it's a problem (as long as the people voting are aware of it and know that if they want their vote to count a lot they need to vote for the extremes). I think it's good if opinions of people who don't care as much as other people are counted less than the opinions of people who care a lot.

I think when letting people arrange them according to their liking of them then it should be possible for people to select varying distances from one option to the next and to place the whole arrangement wherever they want on a scale from hate to love.

If someone coded something like this it could be visualized by dots on a line which you can arrange. On that line there would be at least the markers "neutral/can't decide/don't care" (don't directly influence the outcome, but show that you don't think it's as clear of an answer as the people who vote for the extremes think), "absolutely in favour" and "absolutely against it" (maybe in better words than mine).
 
I mean people have different internal multipliers for their feelings. The "I'm kinda against that" is the same for one person as "THIS THING IS THE SPAWN OF A PURELY EVIL ENTITY AND MUST NOT BE ALLOWED TO EXIST EVEN IN PEOPLE'S THOUGHTS!" is for another. This is mitigated by the normalizing. It's not about getting more accurate data because of finer granularity, but compensating for natural offsets and multipliers people have to get comparable data.
 
I mean people have different internal multipliers for their feelings. The "I'm kinda against that" is the same for one person as "THIS THING IS THE SPAWN OF A PURELY EVIL ENTITY AND MUST NOT BE ALLOWED TO EXIST EVEN IN PEOPLE'S THOUGHTS!" is for another. This is mitigated by the normalizing. It's not about getting more accurate data because of finer granularity, but compensating for natural offsets and multipliers people have to get comparable data.
But at the same time you also make it impossible to vote accurately - there is no difference between "A is slightly better than B, but both are okay", "A is slightly better than B, but both are terrible" and "A is great, much much much better than B, which is terrible", say.
I think one is better off explaining very exactly when a choice should be made than just ignoring a lot of important subtleties in people's opinions.

When a choice should be made:

Yes! - when you are strongly in favour <including or> want your opinion to carry a lot of weight

Yes. - when you are in favour, xor when you are strongly in favour personally but don't think it must essentially be the right choice for everyone and would also be okay with another choice winning if that is more popular

Neutral - when you aren't sure whether a solution is good or bad, you think it's irrelevant to you, you don't care

No. - when you are against the proposition, xor when you are strongly against it personally but don't think it must essentially be the wrong choice for everyone and would also be okay with this choice winning if it is more popular

No! - when you are strongly against the proposition <including or> want your opinion to carry a lot of weight

(or something along those lines - English isn't my native language, obviously. Just try to take the gist and guess what I probably meant if something is strange (-: )
 
Last edited by a moderator:
This poll don't mean anything. Yes!/Yes./Neutral/No./No! is senseless, who care about subtleties of feelings of each person?

T4b haven't even explain why we want to change the labeling, so people who didn't read the precedent poll can't understand.

We should make a unique choice poll with rounds. So that each person would choose THE layout they want and we could do a ranking.
 
Last edited by a moderator:
But first we discuss the right way to do polls, I think this is interesting.

I might even throw together some Javascript for my ideas... maybe.
 
Last edited by a moderator:
I mean people have different internal multipliers for their feelings. The "I'm kinda against that" is the same for one person as "THIS THING IS THE SPAWN OF A PURELY EVIL ENTITY AND MUST NOT BE ALLOWED TO EXIST EVEN IN PEOPLE'S THOUGHTS!" is for another. This is mitigated by the normalizing. It's not about getting more accurate data because of finer granularity, but compensating for natural offsets and multipliers people have to get comparable data.
But at the same time you also make it impossible to vote accurately - there is no difference between "A is slightly better than B, but both are okay", "A is slightly better than B, but both are terrible" and "A is great, much much much better than B, which is terrible", say.

I think one is better off explaining very exactly when a choice should be made than just ignoring a lot of important subtleties in people's opinions.


When a choice should be made:


Yes! - when you are strongly in favour <including or> want your opinion to carry a lot of weight


Yes. - when you are in favour, xor when you are strongly in favour personally but don't think it must essentially be the right choice for everyone and would also be okay with another choice winning if that is more popular


Neutral - when you aren't sure whether a solution is good or bad, you think it's irrelevant to you, you don't care


No. - when you are against the proposition, xor when you are strongly against it personally but don't think it must essentially be the wrong choice for everyone and would also be okay with this choice winning if it is more popular


No! - when you are strongly against the proposition <including or> want your opinion to carry a lot of weight


(or something along those lines - English isn't my native language, obviously. Just try to take the gist and guess what I probably meant if something is strange (-: )
My point is those subtleties are not comparable because of individual differences. IMO global comparability is more importantant than fine individual granularity.
 
Back
Top