Interesting read. Yesterday the first US election Reuters polls came were published since Biden stepped down. There was quite a gap between the polling and Polymarket. (on July 25th, 3pm CET) Who to trust? Personally I'm still inclined to trust the polls more. We don't know who is betting. Maybe the bet subconsciously represents the wish of the better? Or you could imagine political parties gaming Polymarket. To put in massive bets and hope it will psychologically change the outcome. So I'm arguing this way it could even be dangerous for democracy to rely too much on prediction markets for predictions. Reuters polls: 🔵 Harris 44% (+2) 🔴 Trump 42% https://x.com/ppollingnumbers/status/1815803619221950786 Polymarket: 🔴 Trump 62% 🔵 Harris 35% 🔵 Michelle Obama 2% https://polymarket.com/event/presidential-election-winner-2024?tid=1721825808906
We actually had a link of someone arguing for trusting the prediction market much more than a poll: https://news.kiwistand.com/stories?index=0x669906102282c163db19961f63cefe8622e09cd7af3df3e2c614529467941cd7aedbe418 Somehow I feel like there are good and bad prediction markets. Since some sports are a fact of chance, I think sports betting is a more nebulous activity than betting on elections. It is my understanding that exit polls, at least here in Germany, are interviews with voters right when they „exit“ after voting. And those are usually very accurate. On the other hand, if you ask everyone on a soccer team who‘s gonna win the match or whatever, that‘s not gonna lead to an accurate result unless some people have alterier motives. I‘m also not sure if the wisdom of the crowds, for example asking everyone in a stadium before a match, leads to a similar precision outcome than when „exit polling“ after an election. Misha, I guess the discrepancy between the numbers could be between Polymarket computing the likelihood of who‘s gonna be president vs Reuter‘s „who‘s more popular“ with voters. As Harris hasn‘t been officially nominated yet by the Democrats, maybe Polymarket is pricing in that risk, whereas Reuters tried to gauge popularity? Just my best guess… That said, I remember us also discussing this in the comments here last month or so, that betting is also a way of gambling and having more intense experiences (because you can decide your skin in the game).