energy nationally, the reason being that Rs lead Ds in the models of registered votes cast 52% to 39% with 7.1 mil votes in. Like Montana, a wider margin than early voting in 2016, high turnout is not so bad for Rs in Texas. Beto came pretty close, but Trump is more popular than Cruz, and Abbott still won the gubernatorial race that year by more than a million votes. I think coming within 4-5% would represent a good result for Biden, which is what the Siena/NYT, UT, and UH polls show. Could be wrong obviously, but that's the reasoning. Biden and Bloomberg have so much money it's probably not a bad strategy though, and helps for the future.

North Carolina is one Trump is worried about. He keeps holding rallies and is spending as much money as ever there. Early voting shows 3.1 mil votes cast with Ds leading Rs 50% to 44%, which backs that up. For a state to flip Ds need to be leading before election day if the idea Rs prefer to vote in person is accurate. That's the case in NC, AZ and the rust belt. FL too, but it doesn't look like the margin is big enough there.