This election, some of you may be winning or losing in prediction markets. Here I want to recommend the NYT needle, a tool that political scientists and pollsters alike praise. It can predict the final election outcome based on how the vote count is going, where the votes are coming from, and whether they're mail-in or in-person. From my own observation, I'd say that once the needle starts swinging, the result won't change 95% of the time, and you can get a confirmed result about an hour before the official call. But I still want to remind everyone: there's risk in everything, so be careful and avoid losing everything!
So it can help you cut through the misleading information from current vote counts and prediction markets and make money with peace of mind.
I've sworn off participating in prediction markets myself, so I'm begging you all to use it (Amex even reimburses it) — don't let the non-family members take all the money. Right now there are still plenty of House districts that are neck-and-neck in prediction markets, but here there's no suspense left, so take a closer look. And you can check it again in '28.

You mentioned that once the needle starts moving, 95% of the outcome is already locked in and you can call it an hour early—where does that number come from? Is it from backtesting past elections or personal observation? Also, if there's a vote-counting pause or a county reports late, could the needle give a false signal? The post doesn't seem to go into detail on that.