🏅 Top traders

I made a mistake in the way I resolved this - I apologise. The question wasn't formaulted super well either which I hope to be better at in the future.

The way the description is set up does imply that if at any point more than 2% of Britons describe themselves as vegans then this resolves as yes. The way I interpreted the biannual tracker by YouGov (https://yougov.co.uk/topics/society/trackers/dietery-choices-of-brits-eg-vegeterian-flexitarian-meat-eater-etc) made me think that these people were polled probably before January but that seems irrelevant according to the description.

@gaurav When I resolve a market with this much of a clarity problem, I choose N/A.

@gaurav it's normal for questions to not count events before the start of the market because it's a prediction market, not free mana for using a search engine - e.g. literally this question is just "has Google ever laid off 1000 workers before this date", but clearly they mean "from market creation" as otherwise the question should have resolved immediately https://manifold.markets/Soli/will-google-layoff-more-than-1000-e

@Ramble @Panfilo for what it's worth, I asked the moderators in Discord whether there was a precedent/rule on whether "Will X happen in 2024?" created midway through the year counts or does not count events in 2024 from before the question was asked, and I think most agreed there wasn't.

But for this particular market, I think the important point remains that the presence of a January 2023 poll (showing 3%) had not been established until after the market was resolved. In that case, re-resolution made sense, and it happened that this market creator's meaning was that events before market creation were in scope.

Perhaps this is similar to "Will X happen by 2025?" which usually means "before 2025" but sometimes means "before the end of 2025." I rarely see such markets NA'ed in either of these cases, or on the basis that there was new information shared after the first resolution.

@Jacy there definitely is, this market was resolved negatively because the event proceeded the question being asked in spite of the creator explicitly stating that they would accept events prior to the market's creation, because that was considered so anomalous a detail to add so late in the market. https://manifold.markets/lyfeaquatic/will-a-manned-aircraft-be-shot-down.

Sorry I took some time off from this app will try and resolve issues with this question soon

@gaurav This needs clarification: according to YouGov's biannual tracker page, January 1st had 3% and July 11th had 2%. But this market was created well after January 1st 2023, and that poll may have been conducted in 2022, so should we assume it doesn't resolve Yes based on that info? https://yougov.co.uk/topics/society/trackers/dietery-choices-of-brits-eg-vegeterian-flexitarian-meat-eater-etc

But confusingly, the link you shared seems to show a poll run in December 2022, which might be the poll referred to on the overall timeline graph, as 2% vegan, meaning it would be below the line anyway. Now if we go by polls we know to be conducted in 2023, that would be the 2% on July 11th, and likely the 2% that was posted in January of 2024 (but held in 2023, if we care about that).

So there appear to be no polls conducted in 2023 by YouGov that show over 2% of Britons as vegan, BUT there is the slight possibility that the inconsistent data from the January 2023 survey, conducted in December 2022, was 3%. But if you had meant that to count, I imagine the market wouldn't make sense as it was made in August 2023, after the second 2023 poll was published.

@Panfilo yep, market resolves no.

@Panfilo yeah I agree I wasn't clear enough with this. I'd agree that the poll was probably done in 2022 and as such this resolved as no.

@gaurav @Panfilo this isn't accurate. YouGov ran one poll Dec 13-14 (the subject of the blog post linked in the description as a a one-off poll about Veganuary). That was not part of their biannual tracker, which was run on Jan 2 and July 12 in 2023, where they have been tracking this question since July 2019. These are definitely three distinct polls (Dec 2022, Jan 2023, and July 2023). YouGov is a reputable polling firm and would not mistakenly refer to a poll as being run on two different dates, especially when the sample size and results are different. So for the question:

Will more than 2% of Britons describe themselves as 'Vegan' or 'Plant-based', according to YouGov in 2023?

The answer is yes because the Jan 2, 2023 poll found 3%. More than 2% of Britons described themselves as 'Vegan' or 'Plant-based', according to YouGov in 2023. You can see that by going to the tracker and clicking "Download full data set." Here is a screenshot of what that looks like in Excel.

Sorry for not posting this earlier today. I've been traveling.

@Jacy I'm fine with losing the profit, but assuming the surveys are separate, that means this should resolve N/A because the "previous survey" in the description wasn't actually, and because the market was created long aftwards rendering it post-hoc.

@Panfilo I'm not seeing the case for NA. The "previous survey" (Dec 2022) was a previous survey at the time of market creation (Aug 2023), but more importantly, there was nothing saying the 2023 survey had to come after market creation. Some people specify that in descriptions, e.g., "Surveys published before this market was created do not count," but in the absence of that, it's just "in 2023," and this one was certainly in 2023.

@Jacy it's a "prediction market". All markets refer to events after question creation unless explicitly stated otherwise (or if described as a poll/clearly not a time-bound question).

@Jacy I think the future tense makes it pretty clear what the intention was.

@gaurav how does this resolve if some YouGov polls in 2023 show more than 2% and other YouGov polls in 2023 show 2% or less?