The goal of prediction markets is to integrate information about an event and return a single probability that everyone can use to make decisions.
If one person has better information about a future event than everyone else does, ideally they'd be able to influence the market to have a new, more accurate probability. But as markets get bigger, the ability of a single individual to shift the probability gets smaller. Even if they sink in all their liquid cash, they may make a killing once the market resolves, but they won't necessarily have provided much useful information to others in advance.
Simplifying everything down to a single probability is also not always the most helpful result. It would be even better if everyone could see the information that's leading to a certain market probability; more granular information is often more useful to people when what they care about is anything other than the exact question posed by the market.
So prediction markets are more beneficial to society when traders share information with each other freely, and the prediction market functions as a Schelling point for new information about that event. But for each individual trader, they can profit more by keeping their knowledge secret. This presents a collective action problem: How can we get traders to share what they know with the rest of the market?
Right now Manifold's comment section is actually pretty good at this; it reliably contains good discussion about why the probability is what it is. There's a healthy spirit of altruism and cooperation over profit-seeking.
But I don't expect this to last as Manifold gets bigger. The more money is on the line, the more incentive people have to play their cards close to the vest. And the more people join, the harder it will be to maintain a friendly atmosphere. Profit-seeking traders are, after all, what make prediction markets function in the first place.
(I don't use Predictit, but I heard that it was common for people in their comments section to attempt to mislead other traders, in order to get them to bet poorly and lose money.)
Right now the only incentive to make good comments is tips. It seems to be working well so far, though realistically I expect most of those comments would have been made even without the lure of tips in return for good comments. (The current tip amounts are pretty trivial compared to the potential profits.)
But this is another collective action problem; it's better for me when other people make good comments, but individually I'd rather that informative commenters get rewarded by other people, since that way I don't have to spend any mana myself.
This market resolves to YES if Manifold impliments some other system to incentivise traders to post what they know, and if that system seems to be working. Otherwise it resolves to NO at the end of 2024.
Here are some butterfly ideas:
Market participants can optionally mark themselves as "whales" and/or "informants" on a specific market. If an informant has information that they believe the market has not yet priced in properly, they can write up a description of that information. That description is then shared with one or more whales. Any profits that the whales make on bets placed afterwards are shared with that informant. Then after a certain amount of time has elapsed, the writup is automatically posted publically for all other traders to view.
Some immediate problems that come to mind:
Informants could just make stuff up and hope it leads to a profit. So there'd need to be a downside to making false claims, maybe they also share the whale's losses.
An early informant shouldn't share in profits that come from other information down the line. So maybe they should only share profit on bets made within a certain time frame from when the whale read their writeup? Maybe there's a cutoff after a different informant makes a new writeup, or after the market probability changes by a certain amount?
No whales, no informants. Before making a comment, anyone can indicate that they think their comment will result in the market probability changing. If it does change, they get some reward proportional to the market liquidity and the distance it moved.
Problems:
Same as last time, you can say false stuff and maybe turn a profit. There would need to be a penalty if the market doesn't move.
You could make a comment right before information is scheduled to come out (e.g. election results) and have a 50% chance of turning a profit based on information that you didn't provide.
Anyone can get a very short-term loan on a market, with the loan amount proportional to the market liquidity. They can use this loan to place a bet that shifts the market to any desired probability, no matter how big the market is. Then they can make a comment. If the market stabilizes near the probability they shifted it to with their loan, they turn a profit. If not, they have to repay the loan and lose mana.
Problems:
The ease of account creation means people could use this to go deep into negative balances and take mana away from Manifold. But then again that's already possible in other ways and Manifold doesn't seem too concerned about it.
I haven't thought deeply about these ideas, they're just to get the discussion going. This market resolves based on whether any system to encourage information-sharing is implimented and seems at least partially successful; it doesn't have to be one of these 3.
1,000
1.00It would be really great if Manifold could include quality of life functionality for loans between users. The awkwardness of deciding terms adds on the difficulty of selecting worthy users to invest in, I hear. I've been lucky enough to receive a lot funding, but I've also been gregarious enough to ask several people.
Another idea: a quest that stipulates: make a comment that gets a like/bounty/bet reaction once per day
Posting my idea from discord: allow users to reward commentator's they agree with by investing in that users actions on that market and sharing in the profits (or losses). Instead of tipping, you would just add some mana to that user’s current position, (and possibly subsequent position changes). When they sell out, you would sell your corresponding position and split the profit your contribution made with them.
The notif could be something like: "@Charlie just added 100 extra mana to invest in this market"
On the topic of short term loans: perhaps a simpler way to do this would be using a standard leveraged trade? I go 100x long on a market and post my information. If the market moves against me I'll be liquidated but if i convince other traders I make outsize profit for the small amount I put in.
Cons:
it's riskier than simply buying uninformed trades from others as it's easier to get liquidated
Would need some additional human verification to prevent alts from generating excess risk
Pros:
It can reward those without sufficient liquidity to continue to soak up others' bad trades while they keep their info private
Requires less maintenance - The trader take the leveraged trade and releases the info rather than tend to a limit order or other strat to soak up other users' uninformed bets
We need margin for stonks anyways AFAIK
Do agree it would be nice to incentivise. If we accept we can't change human nature to value shared gain over personal gain, then we have to make it so sharing information maximises personal gain.
Very interested to hear if anyone else has thoughts on how to do this (not convinced by the current ideas haha).
Thanks for the discussion and ideas, this is a topic I'm very interested in. There's been some previous discussion on the topic which I want to link but I can't find it right now.
I think fundamentally it's an almost impossible problem to fully solve - I 100% agree with the statement that the more money is on the line, the more incentivized people are not to share their valuable private information.
An illustrative example: https://manifold.markets/Phi/the-prime-factors-of-n-are-both-2-m Multiple traders (including myself) factored the number, but we did not reveal this information for several days because the market incentivizes us not to! In this case, I was a whale on the market and I could have taken all the available liquidity right away. But it was much more profitable for me to trade slowly over time, getting more people to bet against me so I could take their mana. This is generally true, that liquidity is much smaller than volume. I think this demonstrates that the issue is even beyond "But as markets get bigger, the ability of a single individual to shift the probability gets smaller."
The main issue in my mind is that as markets get bigger, the profit incentive to keep your valuable private information to yourself grows. The issue you raised where a single trader has less ability to move the market with their trades is also true, but that seems like an easier problem to tackle - there are profit incentives pointing in the right direction there, and the ideas you suggested could help enable that. I think the much trickier problem is that the profit incentives overall point towards keeping your information private.
The example I gave also shows one way your proposed ideas don't fully solve the problem, although they may help somewhat. The potential profit from quickly moving the market to the correct probability is just much smaller in general than the potential profit from gradually moving it to allow other traders to keep betting against you without the private knowledge you have.
The potential profit from quickly moving the market to the correct probability is just much smaller in general than the potential profit from gradually moving it to allow other traders to keep betting against you without the private knowledge you have.
Yes and no. Bets give away information, so betting in a way that gives away less information can be beneficial. But if someone else figures out the information too, they get to take "your" profit. If you're concerned about that happening, it's better to do so all at once. (This just happened to me; I figured out a way to make a market resolve YES, so I placed a small bet towards YES in the hopes that I could bait people to bet against me. But then someone else placed a large bet towards YES, I suspect because they figured out the same thing, and I lost out on profit I could have made by being more bold.)
@IsaacKing Yes, there is a lot of trading strategy involved in which is better, and it certainly depends on how quickly you expect other people to catch on. If you suspect your information will not be easily known to others, that is the most valuable private information, and also the most profitable to keep private.
