idea: Manifold can inspire itself from resolution algorithms e.g. from UMA, Augur, Chainlink, etc.

https://medium.com/uma-project/polymarket-integrates-umas-optimistic-oracle-7fa89cae493e

Tl;dr — Polymarket has integrated UMA’s optimistic oracle. UMA’s OO provides automated and decentralized resolution for the information markets displayed on Polymarket’s interface.

@o yeah my idea https://manifold.markets/jack/poll-should-it-be-possible-to-dispu was also inspired by optimistic oracles. I still think it's a good idea, and there also seems to be agreement that it would be a good option to have, but Manifold team seems strongly opposed to mandating it for all markets.

Made a related market

Which direction on this will irritate @Predictor the most? I have a bone to pick.

@AndrewHartman No

Nope. (I lost money on that market too, but if Manifold starts stepping in and overriding user resolution, it loses something distinctive)

@SneakySly Not reversing this would actually just be a genuinely dumb mistake.

Gross.

@SneakySly He's partly on the all-time leaderboard because inflation in the Manifold economy means that big scores in November 2022 count for a lot more than big scores in May 2022.

Would you like to have the constant fear that at any moment a market creator will have a bad day and start resolving markets you've put your time and energy into predicting on incorrectly?

@Predictor I don't think I'd be very afraid of that. I worry a lot more about losing Mana in correctly resolved markets... and it's exactly that worry which makes prediction markets fun, in a sense. Will my predictions be correct? Who knows...

@Sjlver I appreciate that.

Let's put this to the test!

Definitely not. Authors should have the final say in resolution, and Manifold should always maintain a strong hands-off approach to market resolution. This was easy mode.

@journcy Do you think it's better if the community can override, rather than Manifold?

I can kind of understand the sentiment, but I have a hard time squaring it with the reputation features. Is it really hands-off if Manifold slaps a big yellow caution icon on your author profile? Sure, the logic for this is in the hands of the community, but it's already the case that someone other than the author is deciding whether the resolution was correct, and I think the difference between that and overriding the resolution seems to me more a difference of how far you go on a spectrum, not a binary difference.

@jack It seems like a big difference to me. With warning icons, if you do something controversial on Manifold, your reputation is damaged. But anyone doing something controversial already expects to end up with a poor reputation, whether its de facto or de jure, this just makes it formal and legible. Any community that wants to make meme markets can continue participate in yellow icon markets at their own risk.

But when resolutions are getting actively rolled back, unsanctioned activities on the platform aren't just being discouraged, they're being made totally nonviable.

@jack I think it's better, but still the wrong move (as we're discussing on the other market).

Like Michael says, I think there's a big difference between Manifold implementing features that apply equally to everyone and just make legible simple information ("people on the platform objected to something I did") that I as a user can take or leave, and Manifold making ~editorial decisions about whether a given resolution is correct.

I don't think the Manifold devs can do this well, because it sounds really hard, and I don't think we should expect them to. Obviously much like spam is the constant practical exception to censorship concerns, if bad actors maliciously resolve enough markets I could see revisiting the issue, but I don't think we're anywhere near there yet.

To preserve the free wheeling atmosphere of Manifold, I propose the opposite. Allow market creators to set a list of users that can override their resolutions, by consensus if more than one specified. This addresses the view that Manifold decisions may not be bias-free and builds a trust network. This could also be used as a fallback for markets that no one resolves, which will become important for long term markets where users may leave, get ill, or die.

I'm slightly leaning towards "no". Although it made some people mad, there is some humor and irony in that market's resolution. Plus, the behavior was clearly allowed by Manifold's guidelines at the time.

@Sjlver Guidelines? The entire purpose of the site is to make correct predictions. Resolving a market correctly if you can is the guideline. If there is doubt, resolve as N/A and move on.

@Sjlver Actually, you're right! There is some humor in this, and I appreciate it now.

@Predictor Hihi... you did actually make me lose M50 on the EU member states ;-) Not too bad... thanks for the ProPublica donation!

@Sjlver You're welcome, at least someone appreciates it. ProPublica inspires me, as public trust is what I aspire to achieve too!

My proposal for improperly resolved markets:

For now ( ~5 improperly resolved markets a month):
-By default, Manifold admins reserves the right to override improper resolutions. We'll have a policy of fixing clear-cut, high profile improper resolutions.
-Creators can opt-out of this while creating a market, in the advanced section. The market page will show that the creator opted out.

-On Markets where the resolution is controversial but ambiguous, the admins will make a special effort to consult with the community in the comment thread and/or creating a market on the subject, and then decide based on that feedback.

In the longer term (30-500 improperly resolved markets a month)
- Implement a robust "disputes" system, like the one @jack suggests.

@FRCassarino Yes, this sounds very reasonable to me. The "for now" proposal you made is perfectly aligned with what I proposed too. The main point of discussion I wanted to raise was not how exactly to operate the mechanism, but whether authors should have total control over resolution or whether it should be disputable at all.

See discussion in https://manifold.markets/jack/poll-should-it-be-possible-to-dispu