The problem with prediction markets is deciding the outcome. If you base it on an authority's decision, your market is a market on in the authority is captured by liars.
@cypher the problem I am trying to point to is that the resolution criteria becomes a mix of the question itself and political influence.
If you can have a reliably uninfluencable decider, then it's not a problem, but those are thin on the ground.
@WomanCorn @cypher You're not worried about the problem for people betting: what this harms is the attempt to use prediction markets to create consensus.
@ciphergoth @cypher yeah. I assume the bettors will take capture into account and can survive a few unfair losses.
But if we use prediction markets as truth oracles, capture takes us away from truth being a good bet.
@WomanCorn I like thinking about mechanisms which don't require an authority's resolution, or allow for easily-viewable observables which in theory everyone should have access to.
@WomanCorn Like, it seems likely to me if you just make a prediction market, and tell people to sell the stock when they think the event specified is less likely, and buy it when they think its more likely, this produces a consensus around what others are going to do, and still lets the stock be coupled to the resolution criteria without needing an authority to resolve anything themselves.
@WomanCorn Or you can have arbitration procedures for if a resolution firm doesn't resolve the market the way expected, like how we have arbitration procedures for medical malpractice.
@D0TheMath this is interesting. It seems to leave the possibility for cranks to keep the market alive well after it should naturally close. I don't know if that's a bug or a feature.
@WomanCorn This is not a problem, people buy shares in markets in which they trust that the resolution will be reliable.