Self-referential poll

[BTCurious]

Please vote what you think is correct, before reading the rest of the posts.

[1713439004]

1713439004

[1713439004]

1713439004

[1713439004]

1713439004

[teukon]

Please vote what you think is correct, before reading the rest of the posts.

On this it would be useful if the first few posts here didn't give any opinions on specific poll options.  I wonder if I can resist

...
...
...

Whew, made it.

[BTCurious]

Here, I'll help resisting.

There have been votes.

[teukon]

There have been votes.

Hmm... Does "There have been votes" give users information on the poll?  Presumably if there had not been any votes that could imply that people are thinking very carefully about the poll.

Perhaps it's only fair to give cryptic information such as whether or not any of the options is currently correct (not mentioning specific options of course).

[BTCurious]

I don't think "there have been votes" gives info, since I try to vote for the option which I think will be correct in the long-term. If you just want your vote to be correct right now, then of course it gives info. The opposite of the info "there have been no votes", which would imply you should vote 100% if you want it correct right now.

[Explodicle]

Here's a fun article:
http://en.wikipedia.org/wiki/Superrationality

I assumed most of you would employ similar strategies to my own.

[BTCurious]

Here's a fun article:
http://en.wikipedia.org/wiki/Superrationality

I assumed most of you would employ similar strategies to my own.

Thanks, that was a good read!
I wouldn't assume everyone here is superrational though

[dree12]

I wouldn't assume everyone here is superrational though

That's what I did and I got it wrong .

[teukon]

Here's a fun article:
http://en.wikipedia.org/wiki/Superrationality

I assumed most of you would employ similar strategies to my own.

Thanks, that was a good read!
I wouldn't assume everyone here is superrational though

Damn, there goes my stategy.

[naypalm]

And no one picked 1% 

[payb.tc]

i didn't think about it for too long but i figured since there are 6 options, then the average would be 16% so 2-20% would be the one to choose.

however over time, more people would probably also realise this and vote on lower numbers, bringing them higher in comparison to options 4, 5, 6.

so i bumped it up a notch and chose 21-35%  ...it's currently on 16.7% but i expect that one might grow to fit correctly over time.

[BTCurious]

I've seen/done more of these polls over time, and usually the correct answer is indeed around the 25% range. I wanted to make it a bit more interesting, that's why I didn't make the ranges the same size; the high percentage is obviously much higher.
I find it funny there's always people choosing 100%. It's not like everyone is going to do that…

[Phinnaeus Gage]

Upon reading the supplied Wiki link, then reading via other links, I think I came across a great example of semantic loopholes: htttp://www.cs.virginia.edu/~evans/cs655/readings/purity.html

[I.Goldstein]

The most logical option will be the second to last one in most cases.

[BTCurious]

The most logical option will be the second to last one in most cases.

Why would that be the most logical? I can think of 2 interpretations of "most logical", and they both give a different option than the second to last one.

[I.Goldstein]

As the amount of voters increase, the more likely the chance a large majority of people choose one poll option. People's perceptions rarely differ and tend to correlate to one degree. The larger the sample, the more potent the accuracy. There can only be one, if you will. This will inevitably eliminate any possible correctness of most poll options and leave only 51%-99% as the only possibly correct option.

[BTCurious]

As the amount of voters increase, the more likely the chance a large majority of people choose one poll option.

That's not true. You cannot see the results before you voted, and you cannot change your vote afterwards. So it would be: As the amount of voters increases, the statistical deviation will get smaller, and it will get closer to its ultimate distribution.

[I.Goldstein]

As the amount of voters increase, the more likely the chance a large majority of people choose one poll option.

That's not true. You cannot see the results before you voted, and you cannot change your vote afterwards. So it would be: As the amount of voters increases, the statistical deviation will get smaller, and it will get closer to its ultimate distribution.

It is true. You have only misunderstood me. What I mean is the majority of people will stick to one common answer. People tend not to be snowflakes when it comes to this stuff.

In any case, the poll will either be totally incorrect or correct in terms of one answer. Distribution being even is not likely.

[Phinnaeus Gage]

As the amount of voters increase, the more likely the chance a large majority of people choose one poll option.

That's not true. You cannot see the results before you voted, and you cannot change your vote afterwards. So it would be: As the amount of voters increases, the statistical deviation will get smaller, and it will get closer to its ultimate distribution.

...based on rational voters, otherwise all bets are off.

[I.Goldstein]

As the amount of voters increase, the more likely the chance a large majority of people choose one poll option.

That's not true. You cannot see the results before you voted, and you cannot change your vote afterwards. So it would be: As the amount of voters increases, the statistical deviation will get smaller, and it will get closer to its ultimate distribution.

...based on rational voters, otherwise all bets are off.

http://jamin.org/blog/wp-content/uploads/2007/11/rational-emotial-service1.jpg

Mr. Gage gets it. I salute you.

[BTCurious]

It is true. You have only misunderstood me. What I mean is the majority of people will stick to one common answer. People tend not to be snowflakes when it comes to this stuff.

Actually, the results are usually pretty spread. People are irrational, or try to be rational but get it wrong. Wait and see, you'll see there's not "one common answer".

[I.Goldstein]

It is true. You have only misunderstood me. What I mean is the majority of people will stick to one common answer. People tend not to be snowflakes when it comes to this stuff.

Actually, the results are usually pretty spread. People are irrational, or try to be rational but get it wrong. Wait and see, you'll see there's not "one common answer".

Science is always the best answer.

However, irrational is not always sporadic nor is rationality always relevant in the first place. We're limiting our perspective here.

http://en.wikipedia.org/wiki/Bouba/kiki_effect

Common psychology may be key to predicting the outcome.

[I.Goldstein]

Also, 51%-99% is considered a majority. It is the "common answer" to which I am referring to.

A rational person would consider this the most likely choice they should pick.

Anyways, I know nothing.

[BTCurious]

The question that I can't refrain from any longer:
If you've got it all figured out, then why is your choice wrong?

[I.Goldstein]

The question that I can't refrain from any longer:
If you've got it all figured out, then why is your choice wrong?

I don't have it all figured out. My choice is only the most certain.

[dree12]

I've seen/done more of these polls over time, and usually the correct answer is indeed around the 25% range. I wanted to make it a bit more interesting, that's why I didn't make the ranges the same size; the high percentage is obviously much higher.
I find it funny there's always people choosing 100%. It's not like everyone is going to do that…

V cvpxrq 100% onfrq ba na nffhzcgvba gung V nz engvbany naq rirelbar ryfr vf engvbany. Gurersber, fvapr jr ner nyy engvbany, jr jbhyq boivbhfyl nyy cvpx 100%.

Gur bcgvzny fgengrtl vs rirelbar ryfr orunirq veerthyneyl jbhyq or gur enatr vapyhqvat 16.67%. Gur ernfba sbe guvf vf gung V pna nffhzr V nz vafvtavsvpnag.

Gur bayl jnl gb fhpprrq ng guvf vf gb cvpx bhg gur pbeerpg engvb bs engvbany crbcyr gb veerthyne crbcyr. Vs unys gur crbcyr jrer engvbany, naq unys jrer veerthyne, bar pbhyq znxr n puneg nf sbyybjf:

Nalguvat <50%: Ab, orpnhfr gur engvbany crbcyr jvyy nyy pubbfr gur fnzr nafjre.
51-99%: 8.3% (veerthyne) + 50% (engvbany) = 58.3%, va gur enatr.
100%: Ab, orpnhfr gur veerthyne crbcyr jvyy abg ibgr sbe guvf.

Guvf cbyy unf fb sne erirnyrq 10.3% bs crbcyr cvpx 0-1%. Guvf bcgvba qbrf abg znxr frafr rira vs rirelbar jrer veerthyne, naq znxrf yvggyr frafr vs rirelbar jrer engvbany. Gurfr 10.3% pna or nffhzrq gb or veerthyne, cynpvat n zvavzhz ba gur ahzore bs veerthynef ng 62%.

Xabjvat bayl n znkvzhz bs 38% bs crbcyr pna or engvbany, V oryvrir gung gur zbfg bcgvzvfgvp bcgvba cbffvoyr vf 36-50%. Gur zbfg crffvzvfgvp bcgvba jbhyq or 2-20%.

(notice: the previous response is not written in German. It is rotated 13 letters of the alphabet to spoiler the discussion.)

[I.Goldstein]

Theorem 1

Proposition: That this poll will having options that match a following electorate.

Corollary: If an option states a majority, then it will require such.

Theorem 2

Proposition: A rational actor will choose the most probable answer and choose the majority.

Corollary: Non-rational or non-determined actors will choose from any of the above.

Sub-corollary: If there are many determined and/or rational actors, there will be consistent results; one result will likely match. If the majority is sporadic, then the poll will have no consistency; it is unlikely that one result will match.

[BTCurious]

Theorem 1

Axiom: That this poll will having options that match a following electorate.

Corollary: If an option states a majority, then it will require such.

Theorem 2

Axiom: A rational actor will choose the most probable answer and choose the majority.

Corollary: Non-rational or non-determined actors will choose from any of the above.

Sub-corollary: If there are many determined and/or rational actors, there will be consistent results; one result will likely match. If the majority is sporadic, then the poll will have no consistency; it is unlikely that one result will match.

I disagree with this axioma: "A rational actor will choose the most probable answer and choose the majority."
You give no reasoning why it is rational to choose the majority. Actually, you're trying to prove that that is the rational thing, which makes this whole thing a circular reasoning.

dree12 does a much better job at explaining things, although there are still a number of things that don't quite fit. For example: you can see from the example above this post, that people who try to apply ration, can still get it wrong. Actually you yourself showed that, with your choice.
And on a sidenote, when you calculate the 62%, it should actually rather be 52%, since the difference between 8.3% and 10.3% is 2%, which is what is relevant here.

[I.Goldstein]

I picked 100% based on an assumption that I am rational and everyone else is rational. Therefore, since we are all rational, we would obviously all pick 100%.

This rationality is based on nothing. My rationality is based on probability. If the poll is going to have any stablility in the end, the majority will have to go with the 51%-99% answer. This is true objectivity. dree12's is not truly objective but rather rational in the Randian sense ergo "Classical music is the only rational genre."

If the poll can achieve no consistent result with not a single trace of people trying to pick out common behaviors, then there is nothing to predict thus making this whole conversation void.

[BTCurious]

If the poll is going to have any stablility in the end, the majority will have to go with the 51%-99% answer. This is true objectivity.

Why? What makes you say that is stable? What if the 60% of the people thought 21-35 was correct, and the rest thought 51-99 was correct? There's nothing unstable about that, it would result in a poll with 60% of the votes on 21-35, and 40% of the votes on 51-99. That is not unstable! Sure, there's no right answer, but no one said there was.

[I.Goldstein]

You give no reasoning why it is rational to choose the majority. Actually, you're trying to prove that that is the rational thing, which makes this whole thing a circular reasoning.

It's the rational or rather the determined response because it requires a majority to fulfill. Thus it makes it the most likely one to be correct. If you're wrong, the poll will be a mess in the end thus showing there is no consistent human behaviour in this test.

[BTCurious]

You give no reasoning why it is rational to choose the majority. Actually, you're trying to prove that that is the rational thing, which makes this whole thing a circular reasoning.

It's the rational or rather the determined response because it requires a majority to fulfill. Thus it makes it the most likely one to be correct. If you're wrong, the poll will be a mess in the end thus showing there is no consistent human behaviour in this test.

If your axioma is: "Humans have consistent behaviour." Then that would be an okay reasoning. Also, your axioma would be horribly wrong.

[I.Goldstein]

If the poll is going to have any stablility in the end, the majority will have to go with the 51%-99% answer. This is true objectivity.

Why? What makes you say that is stable? What if the 60% of the people thought 21-35 was correct, and the rest thought 51-99 was correct? There's nothing unstable about that, it would result in a poll with 60% of the votes on 21-35, and 40% of the votes on 51-99. That is not unstable! Sure, there's no right answer, but no one said there was.

This would be peculiar and possibly a coincidence. If people always chose 21-35, it would have to be correlated with a common behavior which I doubt exists. It's very unlikely this will happen and be a consistent result over various sample trials.

I am not saying that couldn't happen in one poll but if it happened once it would be very little. It would have to be repeatable to have any meaning in a scientific sense.

[I.Goldstein]

You give no reasoning why it is rational to choose the majority. Actually, you're trying to prove that that is the rational thing, which makes this whole thing a circular reasoning.

It's the rational or rather the determined response because it requires a majority to fulfill. Thus it makes it the most likely one to be correct. If you're wrong, the poll will be a mess in the end thus showing there is no consistent human behaviour in this test.

If your axioma is: "Humans have consistent behaviour." Then that would be an okay reasoning. Also, your axioma would be horribly wrong.

Patterns can be found especially in the visual and logic realm. My axioma would not be horribly wrong especially in the study of human behaviour.

Example:

http://en.wikipedia.org/wiki/Bouba/kiki_effect

[dree12]

You give no reasoning why it is rational to choose the majority. Actually, you're trying to prove that that is the rational thing, which makes this whole thing a circular reasoning.

It's the rational or rather the determined response because it requires a majority to fulfill. Thus it makes it the most likely one to be correct. If you're wrong, the poll will be a mess in the end thus showing there is no consistent human behaviour in this test.

If your axioma is: "Humans have consistent behaviour." Then that would be an okay reasoning. Also, your axioma would be horribly wrong.

Patterns can be found especially in the visual and logic realm. My axioma would not be horribly wrong especially in the study of human behaviour.

Ergo:

http://en.wikipedia.org/wiki/Bouba/kiki_effect

This is an interesting answer. However, I would like to point out that your logic, when applied further, gives an answer of 100% (which is, in effect, 99.5%+).

[BTCurious]

The realm of human logic is far, far away from the realm of human basic impulse processing and classification.

What if the 60% of the people thought 21-35 was correct, and the rest thought 51-99 was correct?

This would be peculiar and possibly a coincidence. If people always chose 21-35, it would have to be correlated with a common behavior which I doubt exists. It's very unlikely this will happen and be a consistent result over various sample trials.

Example of human reasoning: "I'm unique, there are not a lot of people like me. So if I pick 21-35%, there will probably not be a majority of people voting this."
People reasoning like this is not unthinkable.

According to you, because humans have consistent behaviour, there would always be one clear winner in elections.

[I.Goldstein]

The realm of human logic is far, far away from the realm of human basic impulse processing and classification.

What if the 60% of the people thought 21-35 was correct, and the rest thought 51-99 was correct?

This would be peculiar and possibly a coincidence. If people always chose 21-35, it would have to be correlated with a common behavior which I doubt exists. It's very unlikely this will happen and be a consistent result over various sample trials.

Example of human reasoning: "I'm unique, there are not a lot of people like me. So if I pick 21-35%, there will probably not be a majority of people voting this."
People reasoning like this is not unthinkable.

According to you, because humans have consistent behaviour, there would always be one clear winner in elections.

http://en.wikipedia.org/wiki/Heights_of_Presidents_of_the_United_States_and_presidential_candidates

Presidents are usually the tallest of society during their respective period. If you look at top candidates in recent elections, most of them stand over 6 ft.

[I.Goldstein]

My only point is we all respond to visual stimuli similarly. It's very primal.

[BTCurious]

How the hell do you see an online forum poll as a "visual stimulus"?…

As a counter-example: Here are the first election results I found when there's more than 2 things to choose from.
http://www.sfu.ca/~aheard/elections/results.html
There are no landslides, no consistent behaviour between humans, and no majorities > 40%.

[I.Goldstein]

The aforementioned poll has no relevance. It's based on intangible traits. All I am referring to is numbers, shapes and sounds. How we perceive these visual things has a lot to do with genetics.

Numbers are visual. They are processed a lot like shapes.

[I.Goldstein]

Also, no rational person is going to assume everybody is going to pick 100%. I don't understand the reasoning behind that at all, dree12.

[dree12]

Also, no rational person is going to assume everybody is going to pick 100%. I don't understand the reasoning behind that at all, dree12.

100% does not mean 100%. It means >99.5%, which is your argument applied in a stronger fashion.

[Phinnaeus Gage]

I've seen/done more of these polls over time, and usually the correct answer is indeed around the 25% range. I wanted to make it a bit more interesting, that's why I didn't make the ranges the same size; the high percentage is obviously much higher.
I find it funny there's always people choosing 100%. It's not like everyone is going to do that…

V cvpxrq 100% onfrq ba na nffhzcgvba gung V nz engvbany naq rirelbar ryfr vf engvbany. Gurersber, fvapr jr ner nyy engvbany, jr jbhyq boivbhfyl nyy cvpx 100%.

Gur bcgvzny fgengrtl vs rirelbar ryfr orunirq veerthyneyl jbhyq or gur enatr vapyhqvat 16.67%. Gur ernfba sbe guvf vf gung V pna nffhzr V nz vafvtavsvpnag.

Gur bayl jnl gb fhpprrq ng guvf vf gb cvpx bhg gur pbeerpg engvb bs engvbany crbcyr gb veerthyne crbcyr. Vs unys gur crbcyr jrer engvbany, naq unys jrer veerthyne, bar pbhyq znxr n puneg nf sbyybjf:

Nalguvat <50%: Ab, orpnhfr gur engvbany crbcyr jvyy nyy pubbfr gur fnzr nafjre.
51-99%: 8.3% (veerthyne) + 50% (engvbany) = 58.3%, va gur enatr.
100%: Ab, orpnhfr gur veerthyne crbcyr jvyy abg ibgr sbe guvf.

Guvf cbyy unf fb sne erirnyrq 10.3% bs crbcyr cvpx 0-1%. Guvf bcgvba qbrf abg znxr frafr rira vs rirelbar jrer veerthyne, naq znxrf yvggyr frafr vs rirelbar jrer engvbany. Gurfr 10.3% pna or nffhzrq gb or veerthyne, cynpvat n zvavzhz ba gur ahzore bs veerthynef ng 62%.

Xabjvat bayl n znkvzhz bs 38% bs crbcyr pna or engvbany, V oryvrir gung gur zbfg bcgvzvfgvp bcgvba cbffvoyr vf 36-50%. Gur zbfg crffvzvfgvp bcgvba jbhyq or 2-20%.

(notice: the previous response is not written in German. It is rotated 13 letters of the alphabet to spoiler the discussion.)

Is there an APP for this? For a second, I thought I was reading a new Wordpress theme.

BTW, I voted 0-1% believing it was a trick poll and most would have chosen that option as well. To date, that would make me a ten percenter. Not bad for my first self-referential poll.

[BTCurious]

I declare the 2-20% people right.
Of course, that still might change.