Lib here. I live in very-blue MA where there’s one or maybe two “Trump 2020” signs in my town. From what’s around me, I can’t really see what’s going on in the nation (or in the electoral college, which is very much a different thing — Rs keep winning with a minority of the popular vote). So I track using FiveThirtyEight. The best predictor, IMO, is their election forecast by simulation where “We simulate the election 40,000 times to see who wins ” — and they they tally them up.

As of my writing this, it’s 28/72, so a bit lower for Trump than in 2016. But, as I said repeatedly in the run up to 2016, a low chance is not zero chance. At 28/72 that’s still very, very true.

As a side note, this doesn’t match up with how well Trump/Biden are doing on a national level as Trump has a huge Electoral College advantage (states where there are more EC votes per registered voter tend to be red). I’m afraid a lot of farmers are actually going to vote to be bankrupted. Hope I’m wrong.