Will Manifold add Metaculus style probability-distribution answers before 2033?
12
97
250
2033
70%
chance

Resolves yes even for an experiment, or if it's later removed. Implementation doesn't need to be exactly the same, just the core concept.

Get Ṁ200 play money
Sort by:

I currently use multibinary markets for something similar. Example: /Mira/if-mira-trains-a-transformer-model

It kind of looks like one of those normal distributions:

I made many similar markets, and I think "linear interpolation between Fibonacci sequence starting with a custom base" or "linear interpolation between evenly-spaced multiples" could be a really good design. (depending on the quantity being predicted)

@Mira cc: @JamesGrugett in case this gives you ideas for implementation. I think you could give these buckets a custom chart type and the frontend for casual readers of markets would already be really close.

Later, you may also want a custom betting interface and AMM that allows atomic bets on multiple legs, much like how stock brokerages let you buy/sell an entire stock + option leg position in one go.

Logarithmic interpolation is really annoying to bet on without a custom interface, so I don't recommend starting with it.

predicts YES

@Mira Interesting. We have thought of that before, but we were never that enthused by it. Maybe it works pretty well with a few buckets, but is only a small improvement for a lot of buckets.

I think what we didn't like was that it was kind of hard to explain to traders.

bought Ṁ30 of YES

oH?

bought Ṁ10 YES from 58% to 60%

like this message if you thought the image posted was an updated UI ☠

More related questions