Will the Hubdub Community Correctly Predict the Outcome of this Question?

I'm sorry I'm not smart enough to wager on an unknown question. I don't really get this one.

nothing at the link - it just goes to a blank page. So we are wagering on hubdub's ability to get *any* random question correctly predicted, at least at the end?

It is irrelevant what question this supposedly refers to. As it states, it is yet to be created. I rather suspect this will be voided, but unless my logic fails me, an option must be at least 50% (actually, slightly above), in order to win, when there is only one other option. So, the answer must be yes....

I have a quibble with the definition of "correctly predict" here. Arguably, one can correctly predict a coin toss by saying it's a 50-50 chance. One can only verify the correctness of that prediction with a series of many coin tosses.

Still an interesting question.

oops, reading the question again, I see my previous statement may not be correct. I assumed the winning option refers to this question, but it seems it does apply to the yet to be created question, in which case, it all depends on how many options there are for the question yet to be created. If there are only 2 options, than what I say above is true, otherwise, it could be anything, down to 11% (since there is a max of 10 options...
Very clever, kruijs, but it seems totally game-able by the question creator....

But if it's a multiple outcome question, then the winning outcome may very well be less than 50 per cent. So as wild and crazy as this is (thanks again, kruijs!), it seems valid.

tuff,

it does not ask if any of the options has 50% - that is not what I meant by "winning option".

it asks if the option which is settled has 50% or more.

That's what was "Correctly Predict the Outcome" intended to mean.

BTW: it might take about 5 days until the question in question will be created

"Thou shalt not tolerate the discordian in your house, yard or bed, unless it be for the pain-free release from his misery by night." ;)

The head line question asks one thing, the background says it's not that but something that doesn't even exist yet.

We're up to market 40082: http://www.hubdub.com/m40082

Approx 220 created in the last day... so in a few days we should know the question we're predicting the outcome of.

Very clever kruijs!!!

Very clever indeed. Nice to see hubdub history thought provoke markets

I am starting to catch on here thanks to f-o-f. So this refers to a question that will randomly become the # 40500 market on Hubdub and may be created by anyone. This was not clear to me and probably other relative newbies on the site who may not be familiar with the history you refer to. I guess it is not as game-able as I originally thought. Still, it basically comes down to whether the particular question has 2 options or more than 2 options. The scientist in me wants to do the analysis to see the proportion of questions with >2 options over the past month to correctly estimate the probability, but I'm too lazy to actually do that....

Now that I think about it a bit more, even many of the multiple option questions have a large percentage of bets on the correct predictions. A quick look (unscientific) suggests the odds favor the yes option by a fairly wide margin. I'm in....
kruijs, this is bloody clever of you.

After one day there is 15K H$ activity and 123 bets on an outcome of a question yet unknown. Great!

these must all be made by Hubdubbers with at least 100K wealth, and thus on the Prophet status.

Others can't read the future and would just gamble, isn't it so?

tuff_sledding - you're still missing the point. The market in question could have any number of options, and it doesn't matter if ONE of them ends up at over 50%, it matters whether the CORRECT option ends up at over 50%. The one with the most votes is not the winner, the CORRECT one is the winner.

@ mattg: I get that. As I said, most of the time the group gets it right. You are missing the point of what I said. Many of the questions do not suspend until fairly close to the end and the group flocks to the eventual winner - go check it out.
@ kruijs: it is an educated guess, not gambling. Playing the percentages. Well, OK, it's gambling, but isn't that what this site is all about????

I meant, right now, without even knowing about the question in question - we can only guess whether or not we will predict the outcome of that question correctly ... isn't it?

BTW: I'd bet question #40500 will be in the sports category.

I don't think that the question is accurate. It only works for a binary, yes,no question. If the question has many possible answers then the Hubdub community might correctly predict the outcome but with a percentage of less than 50. The questions should read: "Yes, the winning option has the largest percentage. No, the value of the winning option is not the largest percentage".

.. but daddeoh, that condition exists to add some spice to this market :-)

well, if the purpose of this question was to make me sit here and think for, like, an hour following multiple links and reading all of the comments on this question until i finally get what the heck i am betting on - then mission accomplished.

I had no problem in thinking this out. either I am very smart, or the alternative.. which is more probable?
I also am not sure about the swine flu, with WHO making all the decisions. I thought WHO was WHAT is on second, I DON"T KNOW on third, etc.. (old abbott and costello routine)- it helps with these kinds of questions..

I think it depends on the question. I am waiting for it.
If the question will be something like this http://www.hubdub.com/m19310/Who_will_hold_most_of_the_US_debt_March_2009 than I have doubts about the collective wisdom of Hubdub Community.

Hmm with the help of these comments I see... take question m39823 - Tigers ended up at 43%, but won, that would be a No then (if m39823 was the 'random question'). So the key variables include - how many options, how long before the 'decision' time was the 'suspend' time, and to a limited extent how the initial outcomes were (will be!) set vs. how good the predictors were. By and large - 'in running' questions should reach near to 100%.

So you could vary this question and require a threshold of say 60%, or 80%, for a ''correct' prediction...

...oddly, the prediction of the question creator when first setting the odds could be better than the 'community' prediction at the end, and yet the result could still be Yes, and vice versa...

I think I'll wait till I see the 'random question'...

a wise decision, bookie ;-)

question #40300 has just been created ...

Or this http://www.hubdub.com/m39120/Will_Iran_see_an_Earthquake_by_the_end_of_April.
6% is surely less than 50%, and it is a yes/no question.
The outcome of this heavily depends on the question that will be chosen so 50/50 odds until the question will be named seems more than reasonable.

How can it be a random question? Once the question is chosen, it is a specific question, and then is no longer random.

For example, this is certainly not a random question. That, however, might possibly be?

Great comments here. had to think awhile as well.
I recall long ago here when this market was first cleverly created. Thanks for bringing this back.

Great comments here. had to think awhile as well.
I recall long ago here when this market was first cleverly created. Thanks for bringing this back.

This is no game of chance. The question 40500 will be gamed. Its suspension time may be before or after the suspension of this. Category Editors may have a say since they can reopen a question, suspend, alter or void.
I am waiting for the question only to see in what direction can it be gamed easier.

OK, now I think I got it. There is no way of knowing. Even a dichotomous question (only 2 possible answers) could easily settle with <50%. I was still stuck on the concept that the question is the one above. So, anything goes. Kind of cool and mysterious. Sure generated a lot of confusion (mostly mine) and a lot of good discussion. Congrats again to kruijs for bringing it back.
Good luck to all.

Of course this could still be voided - say if question 40500 refers back to the outcome of this question... :-)

I think 'game related' questions should perhaps be set in a separate category... sort of meta-HubDub...

Discovered the HubdubDerivatives. Great, Kruijs, more of this !

http://www.hubdub.com/e/Market/How_well_can_the_Hubdub_community_predict_the_behaviour_of_the_world_markets_next_week_3327/view

@bookie
yes, correct. you un-covered my intention ;-)
(no, not really)

@youbet
exactly. I thought of re-creating that one. but maybe infernalmachine can do it himself as he is still an active member. and maybe destry could re-animate this one:
http://www.hubdub.com/m7663/Wildcard__Odd_or_Even
:-)

question #40400 has just been created ...

after thinking about this it depends on what the question is that we'll be betting on is as to how easy this will be to game. if very little is bet on the question it can be gamed fairly easy. even nigel admits or recognises in details that it might need voided if its being gamed. for you new users gaming a question is a BIG NO NO! i strongly suggest you don't try it.

question #40450 has just been created ...

dieseldog, that case is already covered in the background-text.
... or if the question is voided ...
read: in case of obvious gaming

here we go:
http://www.hubdub.com/m40500

I do not get this question

@coolcraft - I did the dumb blonde thingy. I bet yes and no. Now if I were only blond.

coolkraft:
short explanation:
This question settles "YES" if the referenced question (http://www.hubdub.com/m40500) is settled with the option which has a value higher than 50%.
Examples:
- this question settles YES
(a) if the referenced question is settled YES, and the YES option is 50% or higher
(b) if the referenced question is settled NO and the NO option is 50% or higher
- this question settles NO
(a) if the referenced question is settled YES but the YES option is lower than 50%
(b) if the referenced question is settled NO but the NO option is lower than 50%

The semantics of this question could now read:
Will the Hubdub community correctly predict whether or not they are collective geniuses who can correctly predict the movement of 12 international stock indexes.

Hope this helps ;-)

At present, the markets are tending towards the paradoxical prediction that the 'HdC' is/will be rubbish at predicting the indices - this may be logical, if you say that at suspension time each day there is plenty of uncertainty and at least 3 options still available (against a 60% certainty requirement). (I went the opposite way, on the theory that there is partial information available prior to settlement, enough to rule in or out some of the outcomes, and that markets tend on the whole to have some positive correlation.)

However, Kruijs seems to be (have been) betting on (already) 'losing' options on the indices question yesterday at the end, which shows that in principle these questions can be gamed (sorry Kruijs).

I'm taking my losses, and while interested in the final outcome in theory, this is not a fair test of how responsive the HdC is to changing news nor of whether the HdC has any 'oracular' power.

ups, did I?

Currently, this question settles "No"

which, as I know now, is a wrong assumption.

it will settle "Yes"

argh