Questions like "will I do x before y?" Are inherently un-marketable, as they can be resolved however the user wants without any external oversight. Unlike regulated markets, where you need to provide a lot of information to the Stock Exchange and SEC, these Questions are useless as they can't be verified whatsoever.
Example:
"Will I take a shot on date x?"
How would I trust you that you provide correct verification? Nothing obliges you to. 🤷
Additionally:
"Prediction markets, including ours, have proven to be incredibly accurate ar estimating the correct odds" (From Manifolds Website)
This can only happen if the subject(s) that the prediction concerns don't know of the prediction. Even knowing that there is a prediction would influence the result.
Because it's play-money, no company would
Probably change it's behavior just because of this, which means Manifold should work in most cases - except the ones mentioned above.
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ15 | |
| 2 | Ṁ15 | |
| 3 | Ṁ10 | |
| 4 | Ṁ8 | |
| 5 | Ṁ8 |
People are also trading
I think personal markets are not relevant to a lot of users, and empirically they don't ever seem to make it to the top of trending often. I can see why you'd find them annoying. A goal of Manifold is to provide socially valuable forecasts of important things lots of people care about, and yes, personal markets are somewhat antithetical to that.
But another goal is to make betting on questions approachable and fun, to make forecasting popular. By having a lax content policy and making question creation cheap and simple, we allow people to experiment in a low-stakes way with questions that they matter to them. I think anyone seriously chasing unique trader rewards, and anyone who just believes in prediction markets and wants to push the medium forward has to graduate from making fun questions to bet with friends, to making fun questions to bet with entire communities.
But not every creator wants to tryhard and that's ok. If people find value from small, silly questions then good for them. I think betting with friends can make people more rational in their personal lives. That's valuable even if the silly creators never convert into serious creators, even if a market that seems silly never turns out to be insightful.
The real problem is discovery. aka you should only see what interests you. The problem is not that the content exists, but that people who find it annoying see it. It's a similar problem to, like, people who want to bet on sports seeing LK-99 stuff they don't care about.
I think you might want to distinguish the following questions
Can the outcome be manipulated by the author?
Can the outcome be verified?
Is the market pointless if none of the above are true?
Should Manifold ban pointless markets?
Imo the answers are
Depends on the market, some things are hard to achieve even when you want to
Rarely but sometimes, and participants can refuse to bet on markets without verification promised
Probably not, given the fact that people bet on them
Pointless to whom?
@samb I think that has to do with what the platforms goals are. If the goal is to be the most accurate predictive market - maybe they should stop these kinds of questions… see https://manifold.markets/ChristophSchlapfer/will-manifold-ban-selfrelated-quest#itEFsiW731vKcWHmBu2X
@samb Because it’s not a prediction of the outcome if it can be manipulated, and users don‘t stand to gain anything from manipulated outcomes. It‘s not betting if the cards are fixed.
@JonathanBoswell That is a self fulfilling prophecy. If I predict I‘m going to do something, and the market predict’s I’m not - I got two options. Option 1: I can purchase a „yes“ position at good odds and do it - thus making Mana.
Option 2: I can not do it, and give all the people who predicted it a win 🤷 (Mana)
(The reverse also applies of course)
A predictive Market that you can influence and take gains (Even if it‘s just fake internet pixels!) is worth nothing 🤷
@JonathanBoswell very true. An example of a market of this type is my market on whether I will read 50+ books this year. Had it not been for this market, I'd very likely not have the extra push to read more. I also created another bounty market to get recommendations for books and ended up liking some of them to the point of actually reading. Also the market is fairly long term given it runs over the entire year.
@firstuserhere But unless and until you prove you actually read the books, it‘s just a pipedream. Why would I risk any of my mana for an unverifiable and thus arbitrary market?
@ChristophSchlapfer I've been logging books on Goodreads for 8 years now. I update there once I read a book. Many books I read are logged in apple books as read which afaik happens only when i actually flipped through the pages and it logs how long I read it. Sure, it can be manipulated, but then that takes away from the power of how it can influence me if I'm just gonna manipulate it. If i manipulate it, I'm giving myself less incentive to develop an actually good habit.
@ChristophSchlapfer I agree that self fulfilling profit is antithetical to allowing one’s self to be influenced by a predictive market. Perhaps the better question is “Will Manifold disallow a the question creator from betting on their question?”
@ChristophSchlapfer another example, i and others often create markets about "will manifold introduce the feature <X>" where X can be a poll or removing non predictive markets or stop creators from betting on their own markets etc. Using those markets, the devs can gauge usersbase interest and these markets have actually incentivised some decisions made. Are they predictive?
@JonathanBoswell that question already exists. I can see the usefulness of multiple choice or „Should I do x / y“ questions, but framing your future actions (that you have active control over) into a prediction is not worth more than doing some statistics in excel that extrapolate how much reading you did and plot a line. 🤷
@ChristophSchlapfer I would urge you to try such an experiment and see if it is worth more or less. Personally I've found it more useful because I get competitive when there is something to lose and it motivates me more than what I'd do in isolation
@firstuserhere But that is not the same thing, except if you‘re a developer for Manifold. And asking „Will X do Y“ when X can see the question is the same as „Will I do Y“. If the Question is put like this: „Should Manifold implement Feature X until 2023“ - that can be used to gauge interest and help make decisions, but that‘s not a prediction, but a Poll.
@firstuserhere But that‘s the point, I can‘t loose, because I can always buy options towards what I want, no matter If I did it or not. I can even be truthful and still manipulate the market 🤷 If by loose you mean not achieving the specified Goal, I still can‘t see the effect other than setting the goal itself🤷
@ChristophSchlapfer yeah that's true. That's a risk. I hope people won't manipulate such markets but obviously they will. See Levi's "will i go on a date with a cute nerdy guy" market. Keeping aside the matter of what manifold will do or not do, which are administrative things, i personally believe markets can serve as incentives. See this market. I don't know if manifold has ever done it, but I'd be happy if they notice this market, see how high it is, and decide to implement this. Is it really predictive when they can see and bet on it with insider info? But it sure can be useful.
@firstuserhere The premise of Manifold is „Prediction Markets, including ours, have proven to be incredibly accurate at estimating the correct odds“. If the market you linked predicts the devs are going to do it this year with 73%, is it a still a prediction, or is it a poll? 73% of capital that voted on the question wants to see it happen. That is not a prediction.
@ChristophSchlapfer i disagree. Devs have not bet in that market. People could put the chance at 15% if they want. Market could indicate the event to be unlikely but it's not doing that. I don't think it's a poll, because people are predicting based on their knowledge of what manifold is like, what changes Devs are willing to make, etc.
@firstuserhere But answer me this: what motivation does a company have to do things? Especially for apps, it‘s about DAUs, MAUs, engangement. If the devs see that Market at 73%, they don‘t have to bet in the market to change the outcome. Next Question: Are the Devs interested in making more Mana available to users for free? Probably not. Why? Because if you get it for free, why buy Mana from them?
Now it comes down to predicting - Will the devs prioritize Engagement Metrics or Mana Sales?
But that is not the question the Market poses. That‘s a second order prediction, which doesn’t touch the answer relating to what‘s the more beneficial outcome for them: yes or no on the halloween theme.
@firstuserhere I have another point: prediction markets work, because it‘s assumed that the market price reflects all buyers opinion and knowledge. As with ChatGPT, that can be confidently wrong if the buyers have false information. In the best case, one or more subject matter experts weigh in. Generally, I would assume, the more information that there is (in the open) about a subject the better predictions can work, because more people can read and digest more information.
How does that work on a personal scale? Are there many subject matter experts on your past behavior? Is there a lot of information regarding your current life situation that help buyers to decide the market?
@ChristophSchlapfer there are subject matter experts on my past behavior - my friends, family, and co-workers.
@ChristophSchlapfer yes, or at least they used to do it more, but now a lot of them don't use manifold much