Long version (canonical): Will @Eliza provably know the answer to the million dollar MrBeast Salesforce whatever puzzle before it is publicly announced by MrBeast/Salesforce or their team?
I will not trade on this market.
Close date will be extended if necessary
edit 7.3.26:
If you as a trader or watcher believe that @Eliza did know the answer before its presumably forthcoming reveal, and that Eliza is declining to disclose this, you should expect to have until market close to adversarially prove it.
The close date will likely continue to be pushed out until the reveal occurs. Once the reveal occurs, you should not expect the close date to be pushed out again. Resolution may not be immediate, but it should be prompt.
1,000People are also trading
I’ll say it again, “provably know”, is a high bar. For example, “has acknowledged seeing the answer that turns out to be right”, doesn’t cut it because you need to show that Eliza believes the source isn’t lying. So even a screenshot of a responded-to DM from one of the three Manifolders on the winning team is not, on its own, sufficient here.
@JimHays I suspect you're just tacking on an additional clause to the resolution criterion that did not exist. It isn't "Eliza is convinced that X is the answer", but "Eliza knows that X is the answer". That's conguent to @marvingardens adding resolution clarification about adversiarial proof.
It's a clever wording play, I'll give you that!
@JimHays Same as every other "X knows a thing" resolution. When the answer is revealed, if Eliza knows the answer, and it can be proven that he did, it resolves YES.
But I'm not the creator, we can ask marvin if he'll resolve the question based on your additional clause
@Talesfromtim Do you have examples of other “x knows a thing” markets to point to as examples?
I don’t disagree with “When the answer is revealed, if Eliza knows the answer, and it can be proven that he did, it resolves YES.”
My point is that seeing the answer does not mean knowing the answer.
Suppose you send Eliza this message with no other context:
“The right answer is either:
R62 L39 R05 73 6 0 309 3 12 12 or
R62 L39 R05 6 0 309 3 12 12 73”
Then Eliza acknowledges receiving it, and it turns out one of those was right, I don’t think Eliza knew the answer, because of the uncertainty.
And likewise, if you send only one of those, and it turns out to be right, I still don’t think that demonstrates Eliza knew the answer, because there’s still significant uncertainty about whether you were telling the truth.
@Talesfromtim I don't have anything to add, but anyone can through ... --> liquidity in the upper-right
@Talesfromtim Also according to research by the subject of this market himself, adding liquidity to a market trading at the extremes is very inefficient; most of the added liquidity will just be lost.
@marvingardens In the case of a regular binary market, it's pretty good. It's only an issue in the multiple choice and other variants where they chose not to implement their own plan.
See Maniswap for more details about why it's okay in this market type.
Manifold-the-platform also makes the choice to not allow you more fine-grained limit orders below 5% so it's still hard to trade something in that range.
Trying to get a YES resolution by sending Eliza the answer is too contrary to the title and spirit of this market to be reasonable, in my view. Even if you tried to prove they knew it, they didn't solve it as the title requires.
@xjp I think I agree, and yet the description and explanation of the market say otherwise.
I don't think markets should try to adhere to the spirit of things, because that leads to a lot of ambiguity and confusion. And the creators comments below seem to clarify decisively that this would be ruled as YES.
@Talesfromtim I think if things get resolved on bizarre technicalities nobody would expect that's what leads to ambiguity and confusion.
@xjp Either the creator intended this to be allowed, or they didn't. If they intended it, then it can't be a bizzare techicality.
From https://manifold.markets/marvingardens/will-eliza-solve-mrbeasts-milliondo#lqqgfkoqq5 (27 days ago) I assume @marvingardens thinks "Someone tells Eliza the answer" is valid for this to resolve YES.
@Talesfromtim It's a bizarre technicality if it doesn't align with how a reasonable observer would interpret the question. With "solve" in the title, the description text has an unwritten assumption of Eliza knowing the answer after solving it. If OP wanted to make a market about Eliza knowing the answer before the public, then the title should have been "Will Eliza know MrBeast's million-dollar puzzle answer before it's revealed?"
I think it's good to have a decent amount of discretion for market creators to disambiguate their title questions in the market descriptions. But that's not what this case is, because there's no ambiguity. Eliza didn't/won't solve it.
In other words, it's okay for markets to have resolutions that align with some reasonable disambiguation of their titles according to the description text. But I don't think it's okay for markets to have resolutions which blatantly contradict their titles.
@xjp I wrote the "long version (canonical)", and I called it "canonical". My goal in doing so was to operationalize a maximally measurable version of the potentially loose and ambiguous idea in the title. I think that's been reasonably borne out, despite your statement to the contrary. In a prediction market context, loose and ambiguous also means adversarial.
If we just had the title, we could fight for a thousand years on what the resolution is. That's why there's the canonical criteria.
This market does not have a "resolution which contradicts its title"; it has not been resolved. What it has is resolution criteria, which have always been present, front and center, intended to minimize ambiguous situations. You are doing some complaining now, but there is much more complaining you could do in the alternate market you are proposing.
Please remember also that I created this market to cover the possible futures one month ago, not today. You can create a market that more narrowly targets only the futures that are still possible today if you like.
The fact is that in order to design tight resolution criteria, you will often create classes of cases that will be contrary to what some percentage of readers would have assumed from the title. That is totally unavoidable and in my view a quintillion times preferable to every market relying on handwaving about the "spirit" of the market.
Tight resolution criteria. The title is just a title. If every aspect and every case can't fit in the title, then the title is just an approximation.
@xjp I completely agree with everything you said. And yet, I currently believe the Manifold community standard is "If the creator clarifies the market in the description/comments, that's valid".
I don't think it's okay for markets to change from the title this significantly. And yet, I believe it's what the standard is, unless those community rules and expectations get changed.
Forcing titles to better reflect the resolution criterion, would very much reduce a lot of "What was intended / Was this misleading" comments. So I'd support that change.
Personally, I am quite pleased that my criteria, written a month ago, account for this unexpected (to me) case where the answer to the puzzle isn't immediately announced when the puzzle is solved and over. We could have had a lot of people complaining that Eliza solving the puzzle after it's already solved isn't fair. But I covered it.
@Talesfromtim There's a huge flaw in this idea: An enormous proportion of markets /do not have/ resolution criteria beyond the title, and rely on vibes-based resolution of all but the most expected scenarios. If I hadn't written my operationalization, you wouldn't be able to tender this complaint; but it would be a far worse scenario.
Consider this perspective: The canonical question is wherever the resolution criteria are. The title is a title, like the title of a book. It tells you what it's about, it isn't the contents. It has a character limit. It's communicating a topic. The contents are specified inside. The mechanism of the market is inside.
All but the simplest markets will have some set of cases where the resolution will be unexpected for many readers. This includes, by the way, most markets which are thought by their creator to be too simple to have any ambiguity.
The difference here is that I have added extra words so that you can know what to expect in a larger set of scenarios. That moves the center of gravity from the title of the market to the description. That is a good thing, not a bad thing.
Believe this: When markets don't specify, you have far more potential points of disagreement with the creator. They just haven't told you about them ahead of time, so you can't worry about them ahead of time. That's worse.
@marvingardens I agree that all titles are ambiguous and that the description should generally take precedence.
My disagreement is that resolution criteria should narrow the semantic space of the title, not break far outside of it.
"Will BTC hit $100k in 2026?" -> "Resolves YES if daily closing price on Coinbase >$100k USD." -> Valid.
"Will the winner of Mr. Beast's $1M Super Bowl puzzle be a Manifolder?" -> "Winner includes anyone on a winning team who is promised a share of the prize money." -> Valid.
"Will Eliza shake hands with Donald Trump?" -> "Resolves YES if Eliza sees a picture of Trump shaking hands with anyone." -> Not valid.
It's not like the title and contents of a book, unless the title is "Introduction to Organic Chemistry" and the contents are pictures of brutalist architecture.
Btw, I did read your description but I interpreted the part about "provably knowing", when combined with the title, to contain an implication of after having solved it themselves. I didn't read your comments which made it more clear what you had in mind.
@xjp Not a bizarre technicality no one would expect. In the comments 17 days ago I literally said, “there could potentially be a big lag between the two” (announcement the puzzle is solved and announcement of the solution), having this exact scenario of somebody sending it Eliza in mind
Please construct for me an operationalization of what it means to "solve" a puzzle. Note for this project that Eliza had already solved parts of the puzzle; parts of the total solve were being drip-leaked to the audience, and the intended solve path appeared to be to incorporate those leaks; Eliza had already indicated more interest in solving the puzzle than in submitting the answer; the winning team ended up being a groups of dozens, each of whose contribution is unclear; Eliza posted parts of the solve publicly, becoming for a time the only place on the public internet where those parts were posted, meaning his work was part of the continuum of official and unoffical shared informationn everyone was working with; it ended up looking as if the solvers didn't quite know exactly when the puzzle was done, and needed to brute-submit a number of possible "final answers";
I can imagine an endless number of likely squibbles over what it means to solve a puzzle, and what I did was write a tighter question that lopped many of those off.
I agree that all titles are ambiguous and that the description should generally take precedence.
My disagreement is that resolution criteria should narrow the semantic space of the title, not break far outside of it.
First: If all titles have ambiguous edges -- and I would say they are mostly ambiguous edges -- more-elaborated resolution criteria will sometimes break inside of what your first expectations might have been, and sometimes break outside. If your position is that it's better for them to break inside than outside -- good for there to be more NO resolutions than you would have expected, and bad for there to be more YES resolutions than you expected -- I would say: why on Earth?
But second: You are still starting from the title and becoming frustrated at the criteria in the contents, when I genuinely just think that for any market with tighter criteria, it makes far more sense to start from those and see the title as a topic.
The contents are not brutalist architecture. Please remember how unlikely it would have been a month ago to know that the eventual solvers would be on Manifold, and there would be a huge delay between the puzzle being announced as solved and the solution being revealed. We are in a tiny corner case! A corner case that I am thrilled to have been prepared for. The fact that that preparation for this corner case goes against the vibes you would have otherwise expected had you only read the title, a scenario which does not exist and frankly would not really be a concern if it did, does not diminish my thrill of proper preparation.
@Eliza there is another market that says there is a 95% chance certain literal text was in the final answer so as long as I don't read anything containing that text I'm safe.