I have an entity in mind. You can ask a question in top level comment. The question must be formatted in a way that I can answer it Yes/No. The comment must start with a prompt "HERE IS MY Nth QUESTION: " followed by a question, which contains at max 10 words. "Nth" has to be replaced with the number of questions the person has already asked +1.
I will answer in such way, that at the end 66 out of 100 my answers were truthfull. This is made by using a shuffled list of "be truthfull" / "lie". Here is MD5 hash of the string with the list and the secret entity: 59c92e996388c4bf29275736f07faff8 .
I will not answer the questions about the word/letters.
Only Yes Holders are allowed to ask questions. The amount of questions allowed for a person is his_yes_shares/10.
The description may be updated to add rules (for example: against sabotage).
Resolves Yes, if the entity was mentioned in the correctly formatted question.
Resolves No after 1d since the 100th question was answered or when close date is reached (whichever happens first).
Any discussions and offtopic should be in child comments or will be hidden.
This is the first in the series of markets, where we will find out, how many lies out of 100 questions still allow Yes side to win. If you win, the truthrate drops. If you lose, the truthrate rises.
1,000
3.00@KyleWan so, the need to buy questions stops people?
I hoped there would be a new balance. If yes doesn't buy, then no has möre chances the answer is not found, they will buy no and make yes shares cheaper.
So any new bet for No reduces their chances. Negative feedback loop...
HERE IS MY 1th QUESTION:
Is it biologic ?
@KyleWan With the YES shares, we have 6 questions (and only YES holders can ask questions)
@dionisos Random generator might give you 80 truthfull answers in 2 consequtive markets. We would have to repeat market more times to get results for different values. Also random generator makes it possible for less than 50% truthfull answers, which makes it fundamentally impossible in my opinion to solve.
With iterative approach the amount of lies will be a series which converges to the answer I am trying to find. With randomness I would have to organise much more markets to get a valid smple for statistical conclusions.
@KongoLandwalker Ok, so I think the correct way to do it, would be to generate the series of lie/not lie randomly before, fix it so there is exactly 66 out of 100 truthful answers, and then use this list.
You can even share a hash of the actual list if you want (with salt).
It is a way to not be biased by the questions and the context you are in. (I ask for it, because otherwise I fear it will not be random enough)
Also, a hash of the correct answer could be nice :-)