What is the appropriate resolution for a question where the answer is clear, but only after revealed by the question asker? For example,

Is there such English word that if we hash it with md5 10,000,000 times, we get afc407fd6a4e9717a10410c1bd4b7ffa?



One line of thinking is that it's invalid not because "it isn't fair" (as you mention in another answer), but because it's impossible for a reporter at dispute round N to know whether the answer was known at market end time.

In other words, the market is improperly formulated.

If it had been the intent of a market maker to use Augur as a "bounty market", they would have been better off asking "Will [answer to NP-hard question] be known by [date X]?", specify a [blockchain Y] address as the resolution source, and the procedure by which the answer will be delivered to said address in the additional details.

The "blockchain" part is required for unambiguous timestamping, and explicit procedure description to ease verification/reporting.

EDIT: Slightly OT, but "oracle systems" in general (and Augur in particular) are well-suited to resolve temporal events, and not so well for event-less statements about knowledge.

"Is there such a word..." creates a certain ambiguity for a market trader when evaluating a market maker's intent; whereas "will a word be revealed..." (with specific instructions) less so.

Finally, the two existing markets that have (probably) sparked your question specify "general knowledge" as resolution source.

Ask anyone in the street, "Is there an English word that has md5 hash equal to CD880B726E0A0DBD4237F10D15DA46F4?", and they'll answer "I don't know". In fact, even those who know what an "md5 hash" is likely won't be able to answer.

In other words, the answer to that question is not in the realm of "general knowledge". :)

EDIT: It has been brought up elsewhere that "general knowledge" these days means "googlit". This can be argued as a valid argument, provided the reporter doing so knows to check the timestamp... Which I wouldn't count on.

P.S. Finally-finally: the markets lack additional details, which hinders verification even by a well-equipped reporter.

Character encoding, endianness, byte padding? Definition of "English"? (British? American? Outback? Pidgin? All of these? Which dictionary? African or European?..)

  • ok... so what if its just md5 once? then its trivial and easily answerable. Wheres the line? twice? certainly before 100000000000000000, but somewhere above 1? seems arbitrary. The entire point of the hard-fork-as-a-threat is that when it comes down to it, when someone asks a question like the preimage of md5("password") where the md5 has a million google hits, the network is going to hard fork into a "yes" if the disputes go on far enough, but "invalid" at 10,000,000 if its not revealed and nobody-knows-what if it is revealed. – hedgedandlevered Oct 11 '18 at 22:21
  • Yes, it's arbitrary. Considering the fact that the reporting (and, by extension, forking) mechanism is essentially a plutarchy, I'd argue that there's no amount of reasoning that can give a "hard, final" answer on "where do we draw the line". Logic does not draw the line; REP budgets do. – Noel Maersk Oct 12 '18 at 20:08
  • Its not a plutarchy if the hard fork threshold is reached. And thus, the threat of the hard fork is the ultimate assurance it will resolve appropriately ("logically"). The point of this question is to ask what is appropriate, not what is likely to happen. – hedgedandlevered Oct 16 '18 at 18:46
  • "It's not a plutarchy if someone can foot the bill." ;) As to the rest - I understand what you mean; my point is that the question itself is inappropriate. (I've edited the answer above many times, and almost wrote the sentence regarding temporal events here.) – Noel Maersk Oct 16 '18 at 21:07
  • No, its not a plutarchy when everyone in the entire system votes, which is what happens with the Augur fork. I'm referring to the mechanism in Augur's code, not hard forking ethereum. Nobody is footing the bill. Wealth is transferred from those who didn't report correctly. I agree its inappropriate, but nevertheless it must resolve at some point. – hedgedandlevered Oct 17 '18 at 12:58


This question isn't fair, so it's invalid.


Whatever the actual answer is

However, this shouldn't matter. Suppose the answer is yes. The asker intends to dispute the market indefinitely with the real answer (yes), then reveals it later.

They can be confident that if they do this enough where a fork is necessary, revealing the word will ensure everyone moves their augur to the correct universe.

Anyone who bets on this is at the mercy of the question asker, so anyone who chooses to enter a market or bet on the validity of the question will ultimately be giving the question asker a freeroll. Everyone is better off just ignoring it.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.