At time of writing in 2024, China's Yunnan province is extremely famous in the tea world and produces a lot of tea, with puerh tea being particularly well-known. This market resolves to YES if the quantity produced in 2030 is equal to or greater than what was produced in 2024, and to NO if the quantity has dropped by more than 50%. If the amount produced is between 50% and 100% of 2024 production, the market resolves to NA.
My underlying question for this market is: is there going to be no (climate, political, etc.) catastrophe that makes Yunnan unsuitable for tea production? If the resolution criteria are unclear, I'll use this underlying question as a guide to refining them.
Related questions
If the amount produced is between 50% and 100% of 2024 production, I'll resolve to 50% or NA, depending on how reliable the stats I can get are.
Assuming the stats are decent, you could resolve this market to 2 * ((2030 production / 2024 production) - 50%).
That's 0% (NO) if production falls by more than 50%, 100% if it grows, and scales linearly in between. Seems more intuitive than an arbitrary 50% or N/A resolution. What do you think?
@BrunoParga I considered it, but I'm a bit concerned this will lead to nitpicking about which stats to go by; just evaluating total number is already pretty difficult, and the question of whether there's a sharp decline is more interesting to me than how much exactly it declined. But if there's enough interest I could do that.
@jesyspa having clear and objective criteria such that anyone could resolve the market makes it a much better market.
What if you pass away? Become unable to resolve questions? (I obviously hope these things don't happen.) What if you leave Manifold?
You can choose a source and stick with it. E.g. the pdf linked from here https://firsdtea.com/2023-china-tea-report-production/ allows for a decent eyeballing of the production, and you could just say that data from that organization, or their successor, is what counts. There doesn't seem to be an actually good English-language data source, though.
having clear and objective criteria such that anyone could resolve the market makes it a much better market.
I agree, but I think a linear scale makes this harder, not easier. Instead of having to decide between two or three possibilities or calling it NA if it's something in the middle, you need to pick a specific percentage.
I do think that on second thought, resolving to 50 in case it falls in the middle is the worst of all options. I'm leaning in favour of something like:
NO: if drops by at least 1/3
YES: if no indication that it dropped by more than 1/6
NA: if the data suggests either is possible, or if data suggests the drop fell between those two.
WDYT?
With regards to picking a data source: since I'm interested in catastrophic drops more than in mild decreases, I suspect that in the case of NO, there should be news articles on the matter.
I agree, but I think a linear scale makes this harder, not easier. Instead of having to decide between two or three possibilities or calling it NA if it's something in the middle, you need to pick a specific percentage.
You don't need to "pick" anything. There is the 2024 number from the source you choose, there is the 2030 number, and the percentage flows inevitably from these two numbers. It doesn't get more objective than this.
WDYT?
With regards to picking a data source: since I'm interested in catastrophic drops more than in mild decreases, I suspect that in the case of NO, there should be news articles on the matter.
I think that these are probably three separate markets. One is what currently exists: YES if production grows, NO if it drops by more than 50%, unclear/undefined if it drops by less. (Unclear/undefined is bad, and I've suggested how to fix that in the way that is most consistent with what's already defined for 0 and 100%.)
The market of 1/3 and 1/6 is a considerably different market; I think it is bad because of the significant chance of N/A, and it would be really bad if you changed this one to be that one, but if you want to create a separate one, go ahead.
The third one seems to be what you are really interested in, so maybe you should N/A this one and ask just what you mean to ask: "will there be articles stating that there was a catastrophic drop in Yunnan tea production by 2030?"
I think that is a worse question because 1 - you'd still need to choose sources, but in a broader way (two options would be "major media" like the NYT or the Economist, but they just might not pick up on what is ultimately not a huge concern, or to choose specialized media as the resolution source) and 2 - the question would still depend on your subjectivity of whether any particular phrases in particular articles are enough to resolve YES/NO. Like I said, anything that depends on your subjectivity is not as good for the market as something objective. At least, this option should greatly reduce the scope of N/A; a market that N/As is a failure that should be avoided.
At least, this option should greatly reduce the scope of N/A; a market that N/As is a failure that should be avoided.
I think this is the principal place where we disagree. I think that a market which resolves to NO if something clearly happens, to YES if it clearly doesn't happen, and to NA if the situation is difficult to assess is perfectly fine. I agree that an objective market is better, hence I'm looking for criteria that someone else could assess, and I think "will production drop by at least this much?" works for that.
Thank you for your feedback, I will leave the bounds as they are and clarify that that in case neither applies, the market resolves to NA.
@Tasty_Y Darn, you're right, I mixed up YES/NO in the title vs description. :( I'll go by the title, I think, since people are more likely to buy based on that, but can compensate folks if they went by the description (though given how close it still is to 50%, I hope the harm is minimal).