Not true. PredictIt was predicting Trump for 3 weeks prior up until 27th where it took a dive. This is likely due to over-reacting to the Puerto Rican island garbage joke at MSG on the 27th. Not saying prediction markets will be perfectly accurate but they will certainly be better than pollsters.
I didn’t say it always predicted Harris the winner. I said that it was predicting her to win just before the election. She was also leading during the entire period between August 17-October 10, and likely somewhat earlier (I can only see the 90-day history on my phone).
The point here is that there is no “the prediction markets” one can speak of as a cohesive unit.
PredictIt was predicting the opposite outcome up to the day of the election.