Will Manifold's fixed-payout markets still be close-able in a month?

Basic

Ṁ215

resolved Jun 27

Resolved

YES

ALL

So way back in ye olden days, all our markets were Dynamic Parimutuel, meaning that the payouts could change after you placed a bet. We also instituted the ability to close a market, so that the payouts would be locked After some reflection, it turned out DPM is a bad model for options markets: https://kevin.zielnicki.com/2022/02/17/manifold/ Now, Yes/No markets use Maniswap (https://mni.fo/maniswap) and fix your payout at the time of purchase. I think we can do the same for Numeric markets (and Free Response too, fwiw): https://manifold.markets/Austin/will-manifold-change-numeric-market So the question is, should we continue to support the "closing" (aka disallowing of trading) on any market? I'm no longer sure this is needed or net beneficial. What does make sense is "an end date for the criteria of this market". E.g, if we're predicting whether Manifold will have raised a series A by end of 2022, then a natural end date is Dec 31 2022. But there's no reason the market shouldn't continue trading afterwards Reasons not to close fixed-payout markets: 1. Allows traders to cash out early without needing to wait for the resolution. No more "waiting for closed markets" 2. There's still room for ambiguity even after the close date has passed, as the resolver tries to figure out what the resolution ought to be. 3. After-close trading can inform the creator about how to resolve the market! Reasons to close fixed-payout markets: 1. Reduces the "premium on vigilance"/demandingness/attention cost of prediction markets, which Misha has talked a lot about 2. (from @SG): help out the liquidity providers and to prevent traders profiting from last minute news-trading (which is mainly bad because of how it affects the leaderboard). But those aren't that important, esp. if we implement liquidity withdrawals.

This question is managed and resolved by Manifold.

Get

1,000

and

3.00

9 Comments

7 Holders

17 Trades

Sort by:

A weak factor in favor of closing: I'm very interested in knowing when a market is /expected/ to resolve. I think it's important to keep some sort of "end date" associated with each market. if only so that I can sort markets easily to find the short- or long-term ones. Getting rid of closing is good (see my vote), but maybe roll out auto-resolve at the same time?

In the presence of liquidity withdrawal, closing lets the creator send a signal that all liquidity should be withdrawn. Of course withdrawing in response should be optional, and in particular people should be able to add liquidity afterwards, but it is useful for not all liquidity providers to have to be vigilant.

predictedYES

@Gurkenglas if the creator is sending that signal because this answer is now public knowledge could they instead resolve the market?

So they'd still resolve, just not close? Hmm. I'm neutral.

predictedNO

@MattP that's correct! Trading would just always be open up until the market gets resolved. What is "close" today will mean something like "time used for resolution criteria"

predictedYES

@Austin or... and hear me out.... what if you take this opportunity to replace the "close date" with an "automatically resolves on X date"? With the option for the market creator to pick for it to auto-resolve to PROB, N/A, or even YES or NO. The "time used for resolution criteria" could then just be in the question title if relevant, which should be the case anyway IMO. (aka "will X event happen by Y date?")

predictedYES

Seems that this would greatly help with the "abandoned market" problem - and hey, if you're removing the close date anyway, you've got to put *something* in that UI slot right? ;)

(and to be clear, the general assumption would be that market creators will resolve markets before the auto-resolve date - auto-resolve would just be a backstop on every market, which might encourage more betting since people won't be afraid their mana will be locked up forever... though of course getting rid of close dates also fixes that latter issue)

I like the idea of NO but I worry that I'm a power user and it'd be negative for new users, so buying YES.

Related questions

Related questions