Which resolution criteria is best for a popularity contest?
24
resolved Feb 20
Rank by number of YES buyers
Rank by probability
Rank by a separate poll
Other

This is for advice on resolving /singer/who-are-the-top-10-ai-researchers-a

For "rank by probability" I'm imaging something like what Adrian suggests here:

It’s not a perfect method, but “resolve to top 10 by percentage excluding last minute whale manipulation” is much better. Especially if you randomize the close time.

Comments with suggestions are appreciated and will be tipped.

Get
Ṁ1,000
and
S1.00
Sort by:

It depends on what your goal is, if it's to find out who Manifold thinks the top 10 AI researchers are, then I would honestly vote to just N/A the market and then create a poll. If you just want to create a fun market for the fun of it, then I would vote to just stick with your original resolution criteria.

My ideal resolution method would be a poll where people submit their ranked top ten, and being ranked #1 is worth more points than being ranked #10. But I don't think that's really feasible. Still, I think some sort of poll that's independent of the market is probably preferable, even if it's a little wonky to extract a top ten.

The idea with the "separate poll" is that users would bet on the poll outcome. Example pair:

/singer/have-dating-apps-caused-more-good-o

/singer/does-manifold-like-dating-apps

@singer the only problem with the poll option is it's laborious since Manifold doesn't have an option like "you can bet on 10 of this list" and only lets you select one option

You really should not rank by probability. All that does is incentivize whales to just bet up really low options right before market close.

@benshindel What do you think of this proposal?

https://manifold.markets/singer/who-are-the-top-10-ai-researchers-a#jyf9fet1ofr

It’s not a perfect method, but “resolve to top 10 by percentage excluding last minute whale manipulation” is much better. Especially if you randomize the close time.

@singer I mean, I could just set up a 10k limit order on Vladimir Putin at 90% if that's the resolution method lol

@benshindel If it's not done at the last minute, I think that in theory the prediction market is supposed to be able to adjust to that type of manipulation (and you'll lose a lot of mana).

@benshindel yes but there's literally no incentive for anyone to bet on their "favorite"?! I could just bet on these as if they're stocks... the resolution criteria is self-referential so it's meaningless

@benshindel "bet on the best people and the best people are determined by if you bet on them"

@benshindel "something something Schelling points" would be the answer, but I don't have much else to back it up with

@singer I mean, that's kind of fair, but if everyone sees "Daniel Shapiro" or some random AI researcher at 90%, they may assume that person deserves some spot in the top 10 based on info they're unaware of

Related questions