Before 2026, will Manifold innovate its way to a market correctly pricing a 1-in-100,000 random chance of Yes as 0.001%?
32
6.6k
2026
1.8%
chance

At the start of 2026, I'll create another market that lasts 1 week. After that week I will go to random.org and draw a whole random number from 1 to 100,000, inclusively. If it's a 100,000, then I resolve that future question YES. Any other number and I will resolve NO.

Then after that, this market will resolve to YES if the 1-in-100,000 market was correctly priced at 0.001%. Any significant digits after the "1" will be ignored, for simplicity. But if that market doesn't converge on 0.001% within its 1-week duration, then this market will resolve NO.

The correct 0.001% pricing must be displayed in the Manifold UI as the market price. User-made derivatives or tricks (such as amplified markets) wouldn't count.

Get Ṁ600 play money
Sort by:

Expected value isn’t the only aspect of pricing. Would you say that real world lottery tickets are priced incorrectly even though they differ wildly from EV?

predicts NO

I think the easiest path to this might be amplified odds markets that have built-in Manifold support to display how the base probability and derived probability map to each other.

predicts NO

@jack market networks would be super cool. what if you could bet on whether relationships between markets will hold?

Doesn't the UI only show only decimal place?

@NeonNuke *1 decimal place

@NeonNuke Right, that's another thing that would need to be changed for this to resolve YES, and correspondingly, for the 2026 market to correctly show 0.001%.

Will you reveal the number? Or do we get only a binary YES/NO resolution?

@Mira I will reveal the number, and then resolve the market.

@Jotto999 I can't think of a good way to do this in the current system, even if Jeff Bezos pledged his entire net worth on subsidies or protective limit orders, and especially without derivative contracts.

There'd be too much money in shorting this contract and screwing with the probability. And it's just so cheap to screw with the probability close to 0% or 100%.

@Mira In theory, I guess sufficiently-big limit orders could do it. But you're right that the current system would require far too much mana.

Limit orders can't do it because they can only go down to 1%. But even if limit orders allowed going to 0.001% the market math is too lopsided to work in practice - every $1 invested on YES requires 100k on NO.

Inspired by an exchange with @jack.