[losvedir]

As an Andrew Yang fan myself, I was shocked to see that PredictIt has him as the third most expensive (read likely) candidate after Bernie and Biden, at least when I checked. This is in stark contrast to his 0-1% polling rate in the most recent polls.

So, are they just very leading indicators? It got me thinking that it's a relatively cheap campaign strategy since people like to refer to those as some sort of "market based odds" thing. Is it legal and/or ethical for campaigns to buy themselves on the betting markets to drive up the appearance of demand?

We'll see come the Apr 1 FEC reporting deadline if it translates to campaign donations, which are on of the real signals of the seriousness of campaigns.

[SatvikBeri]

The major flaw in PredictIt is the low maximum cap and lack of interest paid. That makes it unprofitable to bet on small corrections even if you're right (e.g. 5% to 10%, or 5% to 1%). So PredictIt shouldn't be considered precise at small percentages.

[pixelperfect]

With PredictIt there's a max bet size, so the manipulation would have to be spread over many accounts. Seems like a very high risk/reward ratio considering how many people would have to be involved, and considering most people don't know or care much about PredictIt.

[dragonwriter]

> So, are they just very leading indicators?

AFAIK, there is no evidence that early standing in prediction markets is any better a predictor than early polling; prediction markets tend to converge to the ultimate result close to the election, but so do polling numbers.