Comment by Hammershaft

The source is a company that works with Polymarket and sells Polymarket data (as well as Kalshi and other gambling platforms).

> 12 hour ahead prediction

As the comments on that article rightfully point out, restricting the data analysed to 12 hours before the resolution feels like cherry picking.

Not really.

1. As pointed out in the comments, you can inflate your prediction success rate by predicting things that are 100% to happen. There are plenty of 99.9% bets on Polymarket with degens betting on some 0.01% lottery event trying to strike a jackpot against the odds, and those will inflate the perceived accuracy.

2. All you're really doing is paying insiders to leak information a few hours ahead of time. Insofar as Polymarket is unusually accurate on things that aren't ~100% to happen, it's likely because in the 12-hour-window that post measures is when all of the insiders place their last-minute bets telling you what will happen. This is extremely bad for society. It's wealth redistribution from stupid people to unethical people[1], and it could completely compromise national security when eg. an insider tells you 12 hours ahead of time that the US is about to launch an invasion of Venezuela. There is no societal benefit to this.

[1] Even if you have no sympathy for idiots who bet their life savings on markets without having insider information, gambling addiction has extremely detrimental effects on society and directly results in increased crime rates, divorce rates, etc. as people lose all of their money and do bad things in desperation, so it is a problem that becomes everyone else's problem.