Conversation

I try hard not to be dogmatic about these things: what works works, and sometimes that changes. Going into the election, I think prediction markets > 538/polls. Post-election, I think 538 > prediction markets. I think they're good at different things.
Quote Tweet
The election was like a month ago and I'm still seeing people beat themselves off over the efficacy of prediction markets over polls. Gun to my head I can say with absolute certainty prediction markets were embarrassingly awful. 1/2
Show this thread
2
19
Prediction markets can easily be off by 10% b/c of cost of capital. 538 doesn't have that problem! But also, if 538 is off by 80% there's no 'trade' you can do to correct it.
1
7