10/10 is Super duper restrictive, 0/10 is not restrictive at all
I will create a poll asking manifold users to vote on one of these options, and will resolve this market based on the option with the most votes. Note that ties will mean the resolution will split among the winners
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ215 | |
| 2 | Ṁ112 | |
| 3 | Ṁ18 | |
| 4 | Ṁ4 | |
| 5 | Ṁ0 |
People are also trading
@Bayesian In retrospect I feel like this methodology is a bit susceptible to noise, like a minority on one number could win if people on the other end of the scale happen to split their votes more across adjacent numbers. Maybe in the future it would make sense to take the average response and round it? Or a median.
@0xseraphim From my usage this doesn't really seem to happen that often, in fact I've been using it for some research coding + side projects and never actually noticed any requests getting re-routed or denied (that said, this could've happened as sessions were going without me noticing).
@Jasonb It's possible they're adjusting claude code routing server-side and they'll make it less aggressive. I did get a good result using fable 5 in the web interface earlier. But tokens be expensive so...