@LarsDoucet You did the right thing, that is all that matters. Have a great day!

@LarsDoucet

. Better to stick with an imperfectly initially worded market than make it ambiguous,

I think ambiguity is okay if upfront. Like if you worded it "Will I judge that [...]" then I think it becomes mostly okay no matter how you resolve this. Still, I feel that you did the right thing also, given your circumstances.

@LarsDoucet Well done.

@JoyVoid I don't think that using subjective qualifiers are a very good solution because that just normalizes ambiguity. I really like the way @jack handled the challenge I made to his market about whether or not Musk and Twitter would settle. He temporarily closed the market while we discussed the question I raised and others could also comment before any change was instituted.

@BTE I really feel like there's two part of manifold, one where ambiguity is welcome and explicit, and one for more grounded-in-objectivity question. Might be worth to have more tools to make this axis more specific and clear.

@LarsDoucet Thanks for making the market anyway and for (in my opinion) handling it very fairly.

Better to stick with an imperfectly initially worded market than make it ambiguous, and just open a second one that's more specific.

I don't think that's quite what happened here, from my perspective. The original question already wasn't clear to readers on whether an attempt to back out counted or not - from the original question, I believe Eigel and I interpreted it as roughly "will he attempt to back out" while dreev interpreted it as "will he succeed in backing out". So Eigel asked about it, and you answered them with a reasonable attempt to clarify the ambiguity - at least, that's how it looked from my perspective. The second ambiguity about what would count as Musk being forced to close added additional confusion on top of that but even just the first already makes it quite tricky to resolve the ambiguity given that I think many people had already traded on different interpretations.

I don't think there's always a good solution here. If the ambiguity is some edge case that isn't a big deal, or there's a clear reason to think one interpretation is better than the other, or if there hasn't been that much trading impact from the ambiguity, I'd typically edit it and depending on the circumstance may offer to reimburse anyone negatively impacted. And in other situations I'd often close N/A and reopening the market with a clarified description.

*Eigil

@LarsDoucet NO sounds right! I was arguing that there were possible worlds in which N/A would've been fairest but we didn't end up in those worlds.

Correction: We are in one of those possible worlds where some people might feel misled if it resolves NO, but I'm not one of those people. So I'll leave it to others to argue for YES or N/A. Again, I'm personally good with NO but if anyone feels misled then N/A may be fairer.

@dreev I'll leave this up for a bit and let someone who insists on an "N/A" resolution make their best case to convince me, otherwise I'll resolve "NO".

@LarsDoucet Looks like it's gonna resolve no.

I currently think this market should resolve N/A because there are too many ambiguities that weren't resolved. He sent a formal termination letter, and then changed his mind about it and is maybe going to buy Twitter after all. To me, that seems like he tried to back out of the deal and then realized that he couldn't.

"If the terms are renegotiated and he winds up buying it, it sill counts as "NO" because he did buy it." -> indicates it should resolve NO

"I will say that if he formally reneges on the deal, and it's very clear he has every intent to back out, and the only thing forcing him to go through is a court holding a gun to his head, then sure, this resolves as yes." -> not 100% match because the court case didn't go to trial yet, but pretty clear fit imo, which indicates it should resolve YES. (Note that the vast majority of cases are settled out of court, doesn't mean the court case wasn't holding a gun to their heads to force them into a settlement/plea deal)

You see, I think I disagree there, respectfully. I didn't word it as precisely as I should have, but when I meant a court was holding a gun to his head, I really mean they are forcing him to go into it, which essentially cashes out to a court order or losing at trial.

If he settles out of court, that means he's doing it of his own volition, even if he's really really not happy about it, because he could still go to trial and he chose not to. The spirit of this market the whole time was about whether Elon Musk would choose to not go through with the deal. If there's a court order forcing him, then clearly he didn't choose.

Just because buying Twitter is a less worse choice for him than facing up to whatever has yet to come out of discovery that he wants to remain hidden, doesn't mean he isn't choosing to agree to the deal.

I'm not about to pull the trigger on a ruling just yet tho.

I'll let y'all deliberate and/or push back on this and see what everyone thinks.

@LarsDoucet I should add that I don't hold my view strongly and I think there are other reasonable points of view, but I still lean towards N/A at the moment. I might be convinced otherwise.

@LarsDoucet If it's renegotiated AGAINST Musk (maybe Twitter requires stronger clauses to drop litigation), I don't see why would it should resolve to YES.

If there's a penny cut of the price, the I think it should be solved to NO.

Bloomberg suggests the first option.

@MP Do you have a typo in there? Both statements seem to be in favor or resolving to NO: "I don't see why would it should resolve to YES" --> did you mean "I don't see why it WOULDN'T resolve to YES" ?

@LarsDoucet Yes, I suspected I would mess with the double negatives.

If it's renegotiated AGAINST Musk (maybe Twitter requires stronger clauses to drop litigation), I think the market should settle at NO.

If there's a penny cut of the price, the I think it should be solved to YES.

Bloomberg suggests the first option.

@LarsDoucet if the spirit of the market is "about whether Elon Musk would choose to not go through with the deal", then it seems this market should resolve YES?
Elon clearly chose to back out of the deal, and it's looking like his back-outingness was unsuccessful.

@MathiasFoster I disagree, because from the very beginning at the top of the market was, "If the terms are renegotiated and he winds up buying it, it sill counts as "NO" because he did buy it."

IE, it's not about whether he wavered, or balked as a negotiating tactic. It's about whether he made a final decision. I introduced ambiguity by saying, "If a court literally forces him to, but otherwise he would have backed out it still resolves as if he backed out", but I think it's still clear that if we wake up one day in a world where Elon Musk owns twitter, and there was no court order or lost trial outcome forcing him to do it, then all his dancing around the issue up until that point ultimately cashed out to him not backing out of the deal.

Still going to keep the debate going, I want to be fair to everyone who put big bets down on this thing, gonna see if we can work out a consensus (and also need to leave some time for Elon the chaos monkey to make a few more moves).

@LarsDoucet I do think the possibility of this market resolving YES even if Musk (is forced to) buy twitter is a highly surprising to someone reading the title of this market, and that making the market resolution criteria ambiguous was not ideal. If he buys twitter, then he didn't back out, and the fact that he tried to back out seems immaterial.

You say the spirit of the market was about Elon's intent... and maybe that was your intent, but I'd guess that's not how a lot of people interpreted the market, despite the fine print at the top, and resolving on something as ambiguous Elon's intent, while there are much easier, non-ambiguous resolution criteria ("Will Elon buy twitter?", "Will Elon buy twitter at the terms of the original offer?", "Will a court order Elon to buy twitter?") - seems like a missed opportunity.

I think I would prefer a N/A resolution over resolving over a subjective interpretation of Elon's intent.

@Retsam19 Yeah I have some regrets. I think I'll hold tighter to original resolution criteria in the future.

That said I think there's still a good chance of this resolving unambiguously, or at least unambiguously enough. If he goes ahead and just buys it, and there was no court order forcing him to, that seems fairly clear cut to me personally?

@LarsDoucet

If he settles out of court, that means he's doing it of his own volition, even if he's really really not happy about it, because he could still go to trial and he chose not to. The spirit of this market the whole time was about whether Elon Musk would choose to not go through with the deal. If there's a court order forcing him, then clearly he didn't choose.

I don't really agree with this framing. If someone takes a plea deal in a criminal case and goes to jail, are they "making a choice" or are they "forced" to go to jail? That's not an exact analogy of course, but more similar is if someone takes a plea deal and pays a fine.

Another way to look at it: let's suppose hypothetically that Elon Musk and everyone else knew that the court would order specific performance. Then I think it wouldn't make sense to say that Elon made the choice of his own volition - no matter what he had done he would have ended up having to close the deal. And I don't think this hypothetical scenario is all that different than the actual scenario.

@jack I think the analogy doesn't quite hold because the typical defendant in a criminal case facing jailtime, typically supported by a public defender, is really pretty different from Elon Musk, the off-and-on-again richest man in the world, in a civil trial, not facing jail time, is not really comparable. He could go to trial if he wanted to.

Yeah, sure, it's not a very analogous situation. Nevertheless, my take on the situation is that M&A law forced Musk to close the deal. The law only very rarely is enforced through court order - most of the time people follow the law and just the anticipated threat of enforcement action is enough.

just the anticipated threat of enforcement action is enough.

Are you sure? The Biden administration is currently facing a string of defeats when it comes to court battles over M&A related regulations:

https://www.reuters.com/article/usa-antitrust-idINL1N30Z15M

So I'm not sure it's clear cut that anything was "forced" here or there was an inevitable outcome, even if we accept your framing.

Still open to being convinced, of course.

Ok, my claim about that might have been too strong. But my main concern isn't so much about the specifics of that clause, as about the broader question of whether Musk tried to back out of the deal.

Musk did formally send a termination letter. It might have been his intent to back out, or it might have been purely a negotiation tactic, but I think if the question had just been "Did Musk attempt to back out of the deal?" with no further clarifications, then the answer would be a clear yes. Since there are some clarifications that are somewhat contradictory, it seems pretty ambiguous to me.

Anyway, I already thought N/A resolution would make the most sense based on the previous conversation https://manifold.markets/LarsDoucet/will-elon-musk-back-out-of-the-deal#uucU4LWG7m1YvBPehUfi, now I think it also makes sense for a different reason. But I think I've said my piece, I want to hear what others think too.

@jack This I think is the most persuasive of your arguments. A formal termination letter is a big, pretty clear move, and then it's down to whether the title or the text body, or the clarifications therein move the needle. (I think there's a pretty clear reading of the text & clarifications, but there's a case to be made it's ambiguous). I'll let some more people weigh in before I decide.

@LarsDoucet (There's also the salient fact that I have a lot of fake money riding on "NO" that I think may weigh just as heavily as any other criteria in resolving this). I think despite the case I've been consistently arguing for above, I feel myself tipping towards N/A and just learning some lessons on how to craft a good market in the future.

But please continue to debate the point while we keep following the news.

@LarsDoucet I'm with @jack but my biggest argument for N/A is that I think the thing being predicted wasn't just clarified but changed. It started as "Will Musk manage to back out?" and became "Will Musk's actions constitute a genuine attempt to back out?".

If I'm wrong and the second version was the right interpretation all along then I'd probably argue that an intermediate resolution would be fair, like saying it's 60% true that he attempted to back out or something. How to settle on a number like 60% I don't know. Unless he doesn't actually end up buying Twitter, then 0% aka NO is the only answer and it should resolve to that.

In short, let's see what happens, see how people think it should resolve according to how they interpreted the question, and if there are strong cases for different resolutions then resolve N/A. Otherwise resolve to the consensus.

PS: To clarify, it's not just strong cases for different resolutions that would justify an N/A resolution. If there's an ambiguous question and there's an equally strong case for resolving YES and for resolving NO, I think resolving to 50% is best. The answer to whether X would happen is that it 50% happened. That's a fine resolution capturing the inherent ambiguity.

But if there are strong cases for different interpretations of what the question actually asked AND reality plays out such that those different interpretations matter, then the only fair resolution is N/A.

Like if we ask "does 10 + 10 = 100?" and if, hypothetically, there's some reasonable reason that some people thought the numbers were in binary then resolve N/A. If we ask "does 10 + 1 = 11?" then the binary/decimal ambiguity doesn't matter and it can resolve YES. If we ask "is 100 a big number?" then it may be infuriatingly hard to pick a fair answer but the ambiguity is on the table from the start and it's ok to resolve it as fairly as we can. Unless the binary vs decimal thing matters materially in deciding how true it is that "100" is "big".

PPS: I think I'm overcomplicating it. It's just that one should resolve N/A if there's no other way to resolve without credible claims to being misled. If Musk buys Twitter but this resolves YES because he tried to back out, I would claim I was misled because I bought NO before the addendum was added. If it resolves NO then I have no complaint, but others might.

@NathanpmYoung Okay

@ian This would still be easier if there was a way to formally give a suggestion that lars could just accept.

@NathanpmYoung @LarsDoucet Would it be reasonable to argue for an N/A resolution, on the grounds that some people had bet based on the original description before the addendum was added? Originally the description seemed clear that unsuccessfully backing out would be a NO resolution. (See further thoughts in my comment below, "addendum changes everything".)

But no big deal if I'm the only one that interpreted things that way!

@dreev I dunno, there is quite a lot of money here. I hate resolving big markets ambiguously.

@NathanpmYoung Agreed, but I guess I'm claiming it's the least-bad option if an ambiguity in the question led to different bettors betting based on different interpretations. And normally you could say "caveat emptor" and expect people to pay attention to possible ambiguities but it feels like in this case the question fundamentally changed when the addendum got added.

Like originally I thought you were clarifying that if Musk ended up buying Twitter it then "it still counts NO because he did buy it". But if I'm the only one then probably an N/A resolution is more anti-climacticness than it's worth.

For anyone who doesn't know, N/A means everyone gets their money back. Also I don't think you should end it now regardless. We still want to see how the probability on this evolves! Just at the end of the year or whenever, decide then what feels most fair based on others' input.

Speaking of ambiguous resolutions, even without opting for N/A, some potential ambiguous outcomes (Musk somehow sorta sabotages Twitter and the price takes a huge dive?) may make it feel most fair to resolve to some intermediate probability rather than NO or YES.

Point being, we might as well see how the Musk/Twitter saga plays out before deciding what's fairest.

@NathanpmYoung @NathanpmYoung I never would have bought this market if I hadn't seen the comment from Nathan asking to change the question to say "successfully or not". I believe there are many people who bought NO under the exact opposite discussion, but those purchases were made under very different circumstances. I suspect the NO outweigh my YES considerably, but I don't see how you can avoid upsetting either me or them. I can't see how you can resolve it correctly by choosing YES or NO, so N/A is the only outcome that shouldn't affect the market makers reputation, so I feel obligated to flag this market as incorrectly resolved if the resolution is YES or NO.

@LarsDoucet Meant to tag you above.

@NathanpmYoung @LarsDoucet Changing the language after taking many bets is a major risk and it shouldn't be without consequence if ANYONE is upset and can reasonably argue their case. Your change in language as I read it was an opportunity to profit off the people who purchased prior to that language being adopted. think you should resolve this N/A because that preserves the market makers reputation. So this is advice more than an argument in my own favor. Any honorable bettor should flag YES or NO as incorrect resolution.

@BTE Not a bad argument. I wish there were some better internal mechanisms for dealing with this, but I think it's good advice going forward to avoid ambiguous clarifications.

@LarsDoucet I think there should be a committee of regular market makers who volunteer to collectively review market resolution criteria for any market that reaches a predetermined volume. It would be like peer editing or peer mediation. Over time that group can create and publish standards and best practices, fundamental quality assurance. This is actually what I do so I would love to be a part of it. What do you guys think of this idea, maybe worth a discord thread? @jack @JamesGrugett @Austin @SG

@BTE I'd love for the ability of market creators to flag their own market as publically resolving. I like that Manifold allows for some markets to be solely resolving by their authors, but that should not be the only option in my opinion

@JoyVoid I like your suggestion to allow market markers to opt-in for public resolution by committee. Perhaps even the bettors should be able to ask for clarification from a group or arbiter prior to resolution? Precision in resolution criteria is not much different from liquidity in terms of increasing market trustworthiness.

@BTE I agree, and I like that then an ambiguous/subjective market is just like a public resolving market where the comittee is comprised of just the author.

But by far the failure mode of manifold is not so much ambiguity that can be solved by clarification, but situations that no one had seen in advance (like with Petrov Day, no one would have asked "What if the site had a bug in its implementation", though I suppose people could have asked "What if it gets nuked but the admins decide to undo the nuke")

@JoyVoid I just read the retrospective for Petrov Day so I understood the context of your comment. My take is the exact opposite. There absolutely should not have been any errors in the code because "what if the site has a bug in its implementation" is solved during quality assurance when done competently. A simple test for whether or not a new user could take the site down would have been enough. "What if it gets nuked but the admins decide to undo the nuke" is an obvious question in advance of an event like Petrov Day and it is disappointing to find out that they chose to effectively call a 'mulligan' because they were embarrassed. But props to the Less Wrong team for full transparency into their decision making and owning their mistakes.

@BTE I mean, it may have been an obvious question, but I do not see it asked in any of the Petrov Day manifold markets. So the questions of "Why wasn't it asked", "how can we ensure obvious questions are asked", and "what do we do when an obvious question was forgotten to be asked" are still very relevant in my opinion.

Though, at this point, I am unsure why prefer manifold over plain metaculus, to me the subjectivity is a strong point in favor of manifold, though it seems people differ in their preferences over this

@JoyVoid I like them both a lot, and I agree the subjectivity is cool and not having to worry about your brier score is also cool. I have improved significantly on Metaculus over the last year, but I am maybe last place in profit no Manifold because I took the mega risk that a judge was going to explicitly order Musk to do something instead of just intimidating him into it.

I have more fun making markets on Manifold than betting, which is a way to indulge my instinct to make overly confident predictions and now I don't make that mistake as much on metaculus. Long story short, the CIA needs to adopt Manifold internally if they want their analysts to achieve .032 Brier Scores.

@BTE I heavily agree with "I have more fun making markets on Manifold than betting", which is a problem because then I don't know how to create engagement for them :D
I wonder what fixes there are to make betting fun in and on itself (as opposed to incentivizing it with eg. streaks)

@JoyVoid The more controversial the better. Engagement is actually where I think I excel in my market making, not necessarily coming up with good questions with clear resolution criteria. One thing I did, until recently, was provide extremely generous providing liquidity to my markets. I would add at least $400 as soon as I created it and would add more or subtract as dictated by the interest of bettors.

I really think the thing that could set Manifold apart as a "new kind of thing" and not just another prediction/forecasting app is the social and publishing component. I recently went back and read my historical comments and I decided I am much more proud of and entertained by the material I have posted here in 8 months than anything I have been able to do on any other social network in a long time. Maybe it is the nature of the type of people who use a site like this that they are more interesting, funnier and motivate me to engage, and they seem to respond the majority of the time.

Another thing to consider, Manifold could be a pipeline for trading talent identified at absurdly young ages (think high school). The Quantum Fund of the future is going to be run by Gen Z and it isn't going to be a subsidiary of Goldman or whatever. It should be seeded by SBF and traders recruited form challenges on this site.

I should apply for a job at Manifold considering how much time I spend thinking about it already...