Since there is no possible situation in which OP can press the resolve button (as OP can only do so if the button has already been pressed) this market will remain unresolved forever. This means there are only two avenues for profit:

• Predicting what others will bet, betting that, then selling out

• Predicting the probability that OP will dishonestly resolve the market and betting based off of that

None of those options seem to be actually profitable without actual precognition, so I will not bet.

if we count percentages as partial no·s then a resolution of 50% is consistent but unprofitable

if we count them as distinct then uhh 1% res is close enough? i was gonna say something with d100·s but it's not working out

The question asks about the future, and there is a guideline "if you already know the answer, resolve". So he can resolve right after making a decision. There is no time-order problem.

The problem is in the fact that any resolution would contradict itself.

"This sentence is a lie".