They're better predictors of what people think will happen.

That can correlate with what does happen, but they're not the same thing. As much as Kalshi et al's marketing would like to suggest otherwise. They're not oracles or crystal balls.