Content
@
0 reply
0 recast
0 reaction
Nounish Prof ⌐◧-◧🎩
@nounishprof
One problem—more than half the voters are women. I’d estimate that less than 20% of polymarket users are women. If it’s not accurate, that’s likely the reason. I’d be stunned if this is a landslide for either candidate. It’s likely coming down to one or two states that could go either way. (GA and NC in particular) Polls — especially phone polls—likely skew older and closer in gender split. Prediction markets skew younger and male, and include a lot of players outside the US. And yes, polls can be manipulated just like markets. My guess is the polls are closer in accuracy and this election is a coin flip atm. I do love seeing @polymarket in the national discourse though!
18 replies
0 recast
88 reactions
tldr (tim reilly)
@tldr
Interesting point – hadn't thought of this. The counterpoint would be that *if* markets are rational (this is their core hypothesis), then men who want to make money should realize that regardless of their individual biases, there is significant arbitrage opportunity by buying Kamala back up.
2 replies
0 recast
7 reactions
Nounish Prof ⌐◧-◧🎩
@nounishprof
Good point — and it’ll be interesting to watch this evolve over the next couple weeks. My guess is that this divide will narrow on polymarket if markets are rational. What appears to be happening atm is a mix betting on the odds as well as trying to influence the outcome. Will be a really interesting post mortem.
1 reply
0 recast
3 reactions
tldr (tim reilly)
@tldr
1000% Im trying to ask myself in real time: “what would it take for me to think the prediction markets were wrong?” And “what would it take for me to think the polls were wrong?” Everyone will start spinning their stories backwards when they see the results so arming myself with what im testing for now
1 reply
0 recast
3 reactions
Nounish Prof ⌐◧-◧🎩
@nounishprof
There’s always talk after about the impact of polls on behavior. It’ll be interesting to see the takes on the impact of prediction markets for the first time.
0 reply
0 recast
1 reaction
