Consider this hypothetical scenario: Say there is a prediction market "Will Alice's portfolio value be at least M$10,000 at the end of January?" Say Bob sees Alice currently has M$5,000, so he buys YES on the prediction market and transfers Alice M$5,000 to get the market to resolve YES.
This is based on couple of recent controversies, but I've abstracted and changed the scenario somewhat to focus more on the question in principle and less on the specific people and profits/losses involved.
Is this acceptable? Scroll the the bottom of the comments where I have posted a poll - you can vote either YES or NO by liking the appropriate comment. The poll closes at market close (Jan 14). Resolves to the percentage of votes (YES) / (YES + NO) when I check the vote tally shortly after the poll closes. By voting, you agree to vote honestly.
See fine print of the poll mechanism here.
You are also encouraged to discuss your thoughts in the comments.
(If it matters, assume the market in the hypothetical scenario is run by a reliable third party who is guaranteed to resolve correctly.)
Related:
1,000
3.00I think the case for YES has been made fairly well in other comments threads https://manifold.markets/jack/how-acceptable-is-it-to-buy-mana-to#kIc3VCNbKyGVjqOmLoc3 or e.g. at https://manifold.markets/IsaacKing/will-i-lose-my-trustworthy-ish-badg#lSbUcFfdhuJAFuFPym7Z.
I'm interested in hearing people's arguments for NO.
I'm particularly interested in why there's a big gap between this one and /jack/how-acceptable-is-it-to-buy-mana-to . What difference do people see? (Or perhaps is it an artifact of polling error, i.e. different people voting in the polls?)
A direct mana transfer or "transfer market" are obvious, but you can also give them money by waiting for them to buy YES on a near ~100% market, buying NO, and hoping they continue buying more YES. No need to even communicate with them.
Or finding any news that could resolve a market and giving it to them for free.
Or finding a set of related markets and taking arbitragable positions, so nobody casually glancing through the web UI finds an obvious case.
Or noticing them betting on 50% markets and then taking secret action to make those markets resolve profitably for them. So no communication or collusion is necessary, they wouldn't even be aware of you.
@Mira Agreed. I'm curious, does doing it in such ways change how acceptable you think the behavior is?
@jack I think market creators shouldn't try to profit from insider knowledge of their own markets(popping everyone's limit orders before resolving would also be naughty). And if somebody causes them to profit just because they made the market, it should be discounted.( but then there's an asymmetry because causing them to lose money wouldn't be considered bad...)
But there's 4 parties: The market creator, the target of the market, the trader attempting to profit, and the trader being profited off of. If the target is not the market creator, and if the market creator doesn't bet in their own market, then at least the judging is fair so I'm not sure there's a problem.
If the judging is fair, one can limit risk by not putting too much into one market and refining the market rules as more and more cases come up. Or limiting the maximum position size of the market so there's not much profit motive in manipulating it, but still some in predicting it.
So I tend to say it's okay for a trader to do it, as long as the market creator is impartial.
@Mira In my hypothetical example, I left the market creator unspecified because I don't think it should matter. We can equally well assume the market is run by a smart contract that automatically fetches the data from the Manifold API and resolves completely objectively and correctly.
@jack In the smart contract case, anything would be fair game because the judge would be impartial. So I would say it's fine.
It would just be a stupid contract to bet on... one side trying to loan the target money at lower and lower interest rates(including negative), while the other side tries to borrow money at higher and higher interest.
But things like position sizing or capping the profits or taking a weighted average over time(so traders have some time to reduce exposure) might be good things to include in the contract.
@Mira Ok, so your position is that it's fair game to manipulate the metric (portfolio value) that the market is resolving on. I think the people voting NO think that's fundamentally against the spirit of the market, or something along those lines - they want the market to measure Alice's "real" portfolio value, not be manipulated.
@jack I agree with Mira that the market creator/resolver matters. At least for perceptions. But also because the act of creating a manipulable market is something that could be judged.
@MartinRandall Fair point. In this hypothetical, please assume the creator is an impartial third party who resolves correctly.