🏅 Top traders

This is not the first time this has happened with Tomek K btw.
https://manifold.markets/NicoDelon/will-itstomekk-still-be-trustworthy

Will @itsTomekK be trustworthy-ish through 2023?

57% chance. Tomek is regularly inactive on their markets, resolving them late or altogether missing crucial events that should trigger resolution. More often than not, their resolution criteria are vague. They create a lot of markets and trade actively but this has been at the expense of what one would expect of a trustworthy-ish creator. Resolves NO if they lost their badge by then (even if it is later reinstated), YES if their badge hasn't been removed before the end of the year. @itsTomekK is the user named Tomek K.

Who wants to guess the entire next tweet?

https://manifold.markets/Roma/what-will-be-the-text-of-the-next-t

What text will Trump tweet next? (Levenshtein distance)

TLDR: What option will be most similar to the next Trump's tweet? If Twitter/X user "@realDonaldTrump" posts again in 2023, I will: Take the post text, Remove all links, Make all characters lowercase (for example "A" becomes "a"), Remove all characters other than a-z. If there's no text remaining, I will discard the post and wait for the next one. If there's text remaining, I will calculate the Levenshtein distance between each answer and the text of the post after all transformations. I will resolve the market to the answer with the lowest Levenshtein distance. If there's a tie between multiple answers, I will resolve to all of them with equal proportion. If no eligible post is created before close, I will resolve to N/A. Important caveats: Only text written by the user "@realDonaldTrump" counts, for example a repost without any additional text is not eligible. Replies count. Any other types of posts count as long as there's text written by "@realDonaldTrump". Text in images/videos/sound doesn't count. Answers added after an eligible post is published will be ignored! Do not add an answer after it's known! I will never resolve to "Other" because there will always be some answer with the lowest distance even if that distance is very high. Other caveats, just in case: I will apply the same transformation/normalisation to the submitted answers before calculating the distance. The formula: levenshtein_distance(normalize(answer_option), normalize(post_text)). If two answers have 0 distance between each other acording to the formula above, the one added later will be ignored. Basically, don't add "CoVfefe" if "covfefe" already exists. If X/post/re-post/etc are renamed to something, the same renames apply to the resolution criteria (assume it uses the new names). If the account is hacked or Trump didn't write the post himself, that doesn't matter as long as the post is created by "@realDonaldTrump". If the post is deleted before this market is resolved, it becomes not eligible. If the post is edited, the first original version will be used. But if the first version has no usable text, the next one will be considered and so on. The text is taken from the post's page on Twitter/X for web (twitter.com). For example: https://twitter.com/realDonaldTrump/status/1347569870578266115 . Just in case other apps display text slightly diferently. Other users' mentions via "@" remain in the text even though they're technically links.

Multiple Answer Resolution for this one , or N/A

It's gone on too long, and it's clearly contentious, so I'm changing my stance to it should be resolved N/A. Also I'm muting this market, good luck y'all :D

@Mqrius I was an actual predictor on this market who did not try to snipe and you're insinuating that my trade should be cancelled for something I did not participate in. It's not my fault that people tried to snipe afterwards, you should have respected the fact that the market was closed and waited for the creator's decision (which, of course, should be resolving to a link or similar because there was no 'word'.)

@ShadowyZephyr Nah, I never bet on this market

@Mqrius oh, my bad, I was thinking of someone else

@ShadowyZephyr No worries. I'm mostly saying it should be N/A because this market has become a clusterfuck and N/A makes most people equally unhappy

Ok, so I'm reporting Tomek K. for fraudulent activity.

@itsTomekK CLOSE THE MARKET

@JarkorMadriz It's clearly a difficult scenario to resolve. allowing the market to settle into an equilibrium seems wise at this point

@cloudprism allowing the market to settle into an equilibrium, u mean market manipulation

@JarkorMadriz Manifold could add timestamps for when options were added such that they can exclude them after a particular moment in order to disqualify non-predictive bets

@cloudprism Or the current configuration could be preserved, because it incentivizes trading on ambiguity correction, timely resolution, and improved accuracy. This could be left up to market creators

@cloudprism Yes, that would be a solution. Although by default it is understood that it is the OP's job to close the market when it is due. I certainly won't bet on a Tomek K market again given this negligence.

@JarkorMadriz This is not negligence. Prompt resolution is important, but correct resolution comes first.

@cloudprism If it is not negligence, it is because it is deliberate. He has had several days to make the decision, this is simply ridiculous.

@JarkorMadriz The standards for the trustworthy badge are not high enough.

And people will make excuses for things like this on the discord. I totally agree with you. Although I would still bet on Tomek K markets because this isn't a repeated pattern, this is still a problem.

@cloudprism
1. It's been 4 days
2. You can CLOSE a market WITHOUT RESOLVING IT.

So... You're going to let people keep betting on this? Wtf

While we're all here... round 2?

Remarkable that the “picture” option did not start to go up with the publication of the tweet but with the open discussion about the resolution criteria.... xd

@Joshua

Okay I admit that got a laugh out of me 😅

We've now entered into the Keynesian Beauty equivalent of The World's Ugliest Dog Contest.

Wow, Donald have surprised us again - he probably saw this market.

Of course it won't be resolve to any word - since there is a link... and a picture, pasted separately (not parsed from the website).

According to rules, it seems to me it should be resolved 50/50 link/picture, but would like to hear you out, what would be a correct resolution.

@itsTomekK I have no stake, and Roma convinced me that that's the correct resolution as well.

You should close this without resolving it right now, IMO.

So people don't bet on this conversation.

Closing is probably a good idea. It's also a bit problematic that someone added the "DONALDJTRUMP[dot]COM" option, but I think that one should be ignored because it's already covered by the "link" option.

My argument is that the link was "first" and the picture was the "second word". So I think it's 100% link. Maybe could see an argument for 10% picture because of ambiguity.

But yeah you can close now and then re-open or resolve after discussion.

Per the resolution criteria:

"If the repost, picture or a video are published together with a text post, this market will resolve to its first word."

This reads to me as a picture not mattering, because there was also text. And that text was a link.

There's now quite a bit of mana now depending on whether me or @SirCryptomind see your next comment first, unless you close the market first lol

@Joshua My admittedly self-interested counterargument would be that the question description doesn’t seem to conceive of a link as text. The ‘--it will be a link—’ option is listed along with the image one as one of the options in case the tweet ‘does not contain any text’.

@itsTomekK The first thing that appears in the tweet is the link, and THEN the image, so the first thing Trump tweeted was a link.

Also: «If the repost, picture or a video are published together with a text post, this market will resolve to its first word.»

@itsTomekK you're keeping us in a lot of suspense here >.>

@Tsiolkovsky an interesting observation... on further inspection the description lists 4 attributes:

If his post/tweet does not contain any text, other possible resolutions are:
--it will be a picture--
--it will be a video--
--it will be a link--
--it will be a repost--

then further down the description states the rule:

If the repost, picture or a video are published together with a text post, this market will resolve to its first word.

which includes 3 of the 4 attributes above, a repost, a picture, and a video... it does not mention the 4th attribute, which is a link, so one can argue that we cannot include the current outcome under that clauses' resolution conditions.

@Blomfilter
1) "This market will be solved to the first word of @realDonaldTrump's first post" -> Here it is clear that the solution is ordered, i.e. it respects the order in which the items appear. Although not all possible combinations between the different elements are specified, it is clear that this is the way in which the OP wants to avoid ambiguity when several of the elements appear together.

2) "If the repost, picture or a video are published together with a text post, this market will resolve to its first word." This DOES apply since the picture was published together with a text post. Text post which consists of a link.

Arguments 1 and 2 are independent of each other

Yeah, at no point does the resolution criteria ever suggest the potential of resolving to multiple options. I think this market should always resolve to 100% whatever is determined to be the most correct answer.

If someone had submitted "It will be a link and a picture" or "it will be a link to a video" or something else like that I think it would have been ruled an invalid answer, because the OP already added the non-word answers as their own distinct answers that cover the non-word cases.

Likewise, if his first tweet was "I am back" it would have resolved to 100% "I", not 50% "I" and 50% "am" or to someone submitting "I am". There's really no reason to complicate this, IMO.

I have no stake here.

I am at about even between (Resolve to link) and (Resolve to link and picture) with slight preference for the former. The whole "other possible resolutions" clause suggests that the author does not think links or pictures count as text. The use of the singular pronoun "it" implies that the possibility of more than one non-text item was not considered. But "it" could also be interpreted as "the first thing in the first post".

Still, though, I feel that this was not a possibility foreseen by the criteria, so it's ambiguous.

@BoltonBailey Yeah, there's a case to be made for N/A because of the ambiguity.

@BoltonBailey link and picture was not added until after it happened.

@Mqrius You bought a bunch of donaldjtrump.com to snipe for 50% and now that everyone agreed it shouldn’t count as a word, now you’re saying N/A (which punishes actual predictors like me)?

@itsTomekK In fact, to be honest, most people have voted after the event, and the link option went up to 95% because it is obviously the right choice xD Then a lot of people started betting on the image option to try to profit from the ambiguity, but even with those, currently 77% of investors agree that the right choice is the link.

@Tsiolkovsky There might have been a tiny bit of rent-seeking on my part here 🤏

@Tsiolkovsky Ambiguity should be traded on to reduce ambiguity in the future

@Roma I don't think there is any real ambiguity, as I have already explained, I think it is clear from the statement that the order in which the items were published was relevant, and the first thing Trump published was a link. It is followed by a photo as it could have been followed by a sentence, it is irrelevant.

@itsTomekK This market is all about what comes first. What comes next is entirely irrelevant. The markets first reaction was also to bet that option fully up, because that is the right answer.

@gigab0nus If the first tweet does not include words, but does include a link, then that makes sense. But as it stands, there IS a first word in the tweet.

@Tsiolkovsky This is just how twitter shows tweets with images. Images always go to the bottom.

@cloudprism I even interpret the „if there is no text“ as „if the first thing is not text“ but that is up to discussion or clarification from @itsTomekK .

@Roma If Trump had posted a text followed by an image, the response would have likewise been the first word of the text, not the image. That twitter's mechanics when displaying images in tweets prejudices the image option is irrelevant xD that was information we all had.

@Tsiolkovsky Good point that we all had the information!

On the other hand, description doesn't say that order matters for anything other than words. And if the tweet contains any words, then the first word is DONALD.