The unexpected hanging paradox regards a prisoner who is told that he will be executed on one of five days, but that he will not know which day the execution is on until the executioner knocks on his cell door.
This market is one of five markets that will resolve on August 21 to August 25. They can all be found in the Unexpected Hanging Paradox group. I will resolve one of the markets to YES on the day mentioned in its title, and the other four will be resolved to NO. However, the day will be a surprise: No one will know what day it's going to happen until I resolve one of the markets YES.
I will not bet on any of these five markets, since I will know what day they are going to resolve.
Sha256 hash: 3f75dc7c733335e5af1dcf7b90f384a16e6943e8e44bd95b7c4544185a7c04b3
Related questions
🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ557 | |
| 2 | Ṁ387 | |
| 3 | Ṁ279 | |
| 4 | Ṁ140 | |
| 5 | Ṁ117 |
Notice this has gotten lower, not higher, just as I predicted yesterday.
I guess I'll be the one to say it. The execution can't happen on Friday because otherwise on Friday morning before the executioner knocks on the cell door, we could deduce that the execution will happen on Friday, and therefore it wouldn't be a surprise, contradicting the premise. Prove me wrong.
@MatthewBarnett People are thinking 1 second after midnight Thursday (i.e. beginning of Friday). This is different from the normal story which requires the execution to happen at a certain time of day. If done in this way arguably no one would know it was going to be Friday, because they wouldn't be certain whether internet latency was responsible for Thursday not resolving yet.
@DavidBolin The real reason I didn't specify a specific time is because I don't know exactly when I will be free, so I don't want to say that it will resolve at a specific time and then end up missing that time.
@MatthewBarnett then it can't happen on Thursday because if Friday if impossible, on Thursday the prisioner would know that he's going to be hanged that day. And so on, therefore Friday is just as valid as the rest because no day is valid
@NiciusB So the other days are not valid because at some point you will know for sure which day it is? You should just bet heavily on that day then.
@rpominov yeah I have limit orders on every day betting towards 20%. I would get Friday closer to 20% but I'm too heavily invested in YES and although the expected value is higher I don't want to take the 1/5 chance
@NiciusB Are you aware that @JosephNoonan has put 100M YES on the "will it be a surprise" market? They do want to make it a surprise. And remember that this experiment is not quite the real paradox because it depends on market probabilities being 50%. I'm confident that, if we get to Friday and the Thursday market hasn't resolved yet, the Friday market will shoot up to close to 100%. (Risk free profits in <1 day!) I'm also confident that Joseph/Plasma believes this too. We're dealing with a human here, and that human is trying to make it a surprise. They won't have gone for something so easily foiled.
I'm less confident in how the markets will react on Thursday morning when the Wednesday market hasn't resolved. Will Thursday and Friday zoom up to 50/50 ish? Or will it be more like 90/10 because of the aforementioned Friday logic? Either way, it seems pretty likely that the Thursday market will make it above 50%. So I think Thursday is out but I'm less sure.
I'm even less confident what will happen on the other three days. I think the strongest argument is for Monday because it would take some courage for bettors to push Monday all the way up to 50% when there's still a week to go. Or maybe there's an argument for Tuesday, because Joseph/Plasma thought the markets might catch onto the strategy of bidding the "next" day up each day, meaning no day would be a surprise. Maybe they've gone for Tuesday hoping that we'd lose confidence in that strategy after it failed on Monday.
The fact that Joseph/Plasma has so little skin in the game makes me reluctant to put all my mana on the line betting this down much lower, but I really do think it can't be Friday, and is very unlikely to be Thursday.
@Fion I agree, if I were him I would resolve on Thurdsday to make it still exciting but maximize traders, and also people will get mad if it resolves to Friday even though their logic doesn’t work in my opinion
@Fion I put more mana on YES in the "Will it be a surprise?" market, but this one doesn't seem to have responded yet.
@MatthewBarnett The only solution to this paradox is to assume a probability >0 that the creator is lying (i.e. It will not be a surprise). Then the prisoner can be hung on Friday, so it could technically be a surprise on Thursday, Wednesday and so on. @JosephNoonan here has specified that "surprise" means a probability below 50%. So with a probability of ~20% right now that he is lying, a resolution on thursday would also mean he is lying (because the market will be sky high at that day). The problem now is to estimate the probability that he is lying (he seems to be a very honourable person until now) and deduce the most likely resolution day from that. I guess he wants to make it as exciting as possible (so probably not on Monday) but also play safe to not hurt his reputation by making a decision that means he has previously lied.
@gigab0nus And yeah actually lying means these markets could be N/Ad because he has lied in the description.
@Fion I agree with all of this. The argument that if it cannot be Friday it cannot be any day is false; you do not need a logical explanation to know this, the same way you don't need to logically refute the argument that in a limited repeat prisoner's dilemma they are automatically forced to defect every single time even if it runs for 1000 rounds.
@NiciusB I think they believe it will be resolved 1 second before or after midnight on Thu, meaning that people will basically not know whether it is Thu or Fri until it happens.