June 4, 2023
Manifold Markets is a play-money prediction market site where anyone can make a market. After making an account several months ago, I’ve been using Manifold seriously for the last several weeks and I’ve come to the conclusion that it sucks. This is mainly for two reasons.
After you place a bet on Manifold, you get 2% of your initial investment back every day in the form of an interest-free loan. So if you bet 100 mana on Monday, then on Tuesday you’ll get 2 mana back, and on Wednesday you’ll get a little bit less, and so on. In a little over a month, you’ll recover half of your initial bet, and in 8 months you’ll recover over 99%. If you reinvested this money in the same market every day for a year, you could turn your initial 100 mana investment into over 800 mana. Your leverage is theoretically unlimited, and the money is completely free. You only need to pay back the loans when you sell shares or the market resolves. And all users are considered creditworthy.
Of course, you don’t actually need to pay back the loans. If you screw up and end up with a negative balance, there are no real consequences. You can just quit Manifold, make a new account, whatever.
Ironically, this policy is a major problem for markets that deal with events in the far future. Take for example the market “Will AI wipe out humanity before the year 2100?” Manifold poweruser Mira wrote a good explanation of the problem. The market is practically the same as “This market will resolve negatively in the year 2100,” so the Yes shares ought to be worthless regardless of how you feel about the issue. Nevertheless, some people are buying Yes shares, and we can divide them into two groups:
At the time of writing, the second largest Yes position is held by Eliezer Yudkowsky, an infamous AI doomer. This market is also the one in which Big Yud is most invested: 15,100 mana, of which 10,606 is borrowed. I am certain that Big Yud has no plan to cash out early or pay back his loans. He’s only buying Yes because it makes him feel good.
The loan system was intended to encourage people to bet on markets with a long time until expiration. But instead it has turned these markets into purely speculative investments with no connection to reality.
Manifold allows you to convert your mana into charitable donations. This is pretty much the only thing mana is good for, other than the satisfaction of having a high score. But let’s do some math here. Manifold makes its revenue and user activity public. At the time of writing, they have roughly 800 “engaged” users, where an engaged user is someone who has used the site at least two days per week for each of the last three weeks. This number has been quite stable for the last few months. They’re also making about $2,000 per month by selling mana. (There was a recent spike due to the Whales vs Minnows market, but that was just one user being a huge idiot, and it was mostly unwound.) Manifold currently gives out up to 25 mana each day to encourage people to use the site. This means if there were a little over 250 people logging in every day to collect their 25 mana, then that alone would be enough to wipe out the $2,000 they make from mana sales. With 800 engaged users, are there 250 who are active enough to do that? Maybe. Of course, those users would have to convert their mana into charitable donations. If they never do that, then there’s no problem. I’m guessing Manifold is counting on their powerusers to just hold their mana forever, and if they ever decided to turn on the firehose then it’d be a big problem. That might happen if the powerusers lose confidence in Manifold’s ability to continue operating.
How long can Manifold continue? I’m not sure, but I have serious doubts about their ability to make a profit in the long run. So I’m going to continue using Manifold, but I’m going to focus on converting my mana into charitable donations while that’s possible.