Will all the NO traders in the market about whether "range" markets work agree that it was resolved fairly?
5
Ṁ100Ṁ786resolved Mar 14
Resolved
NO1H
6H
1D
1W
1M
ALL
I realized that in https://manifold.markets/dreev/is-there-or-will-there-be-a-success I should've first bought 1 mana of YES and commented with my reasoning for resolving to YES before doing so.
To make up for that, I'm resolving this market to YES only if all the NO traders there agree that YES was the correct resolution. Discuss!
Mar 14, 9:33am: Gurkenglas (who may have been the only NO trader in the other market) has spoken!
This question is managed and resolved by Manifold.
Market context
Get 
1,000 to start trading!
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ22 | |
| 2 | Ṁ1 | |
| 3 | Ṁ0 |
People are also trading
Sort by:
Looking on this with fresh eyes, the response is clear: A range market should allow you to pick at least two degrees of freedom in your distribution, such as mean and variance. That's not the case in PROB-resolved markets. To get the stock market's full probability distribution over future prices, you don't look at the stock's price, you look at derivatives.
Ah, yeah, I'm anxious to hear your conclusions there. Sorry again for not hashing this out in the original market before resolving it!
@Gurkenglas re: your last comment in the other market: The DPM issues that make it wrong -- do they also make normal non-range markets equally wrong? If they make range markets more wrong, then that would be an argument that range markets via resolve-to-PROB do not in fact work correctly. I agree that they work more restrictedly, like only allowing you to express triangle distributions, but I don't think that counts as not working correctly.
But this market is your call. I committed to resolve it based on whether you agree.
I panicked when I saw in Discord you had resolved this Yes, but on closer look I found I hadn't actually bet on your market. I was thinking of https://manifold.markets/Cyril/will-manifold-implement-range-marke
I don't think I'll trust prob resolution for ranges until I've seen a writeup using an interactive model to go through multiple examples, andor MM creates a market type specifically for ranges.
My thinking was that, first of all, I didn't know if such markets existed at all so being pointed to 3 examples was a big part of the way there. I also said the market had to work correctly, and that's pretty tricky to decide. But I think it was clear from my first comment (and the first NO shares bought) that I had in mind something like the possible bug that @Yev had seen. That turned out not to be related to resolving-to-PROB or to using that trick to do a range-style market. So I think those do work, just not as richly as Metaculus's version, as @James explained.