For everybody asking about Arizona, here is Nate Bronze's explanation:
Arizona. We may need to do a longer post on Arizona later. What’s
left to count is mostly mail votes that were returned late in the process — on Monday or Tuesday. There’s some ambiguity about how many ballots this actually is; Edison Research
seems to think around 450,000. If so, Trump would need to win those votes by 21 points to overtake Biden’s current 93,000-vote lead. They’re disrupted fairly evenly throughout the state.
Wait — outstanding mail votes? Shouldn’t those be good for Biden, as in other states? Well, not necessarily, because Republicans have a fairly strong mail voting program in Arizona and — this is the key part — the mail ballots that were returned later in the process (the ones yet to be continued) were significantly redder than the ones that came in earlier on, as Democrats sent their votes in early. For instance, the party registration of the
votes that came in Monday and Tuesday were: 23 percent Democratic, 44 percent Republican, and 33 percent independent or other parties. That is to say, a 21-point GOP edge, which would put Trump on track to tie things up.
But … here’s the bad news for Trump. Party registration may be a misleading indicator in Arizona. It has a lot of ancestral Republicans who have now turned into swing voters. Biden also had a
big lead among independents in polls. And earlier batches of mail ballots were considerably stronger for Biden than party registration alone would suggest. So probably these ballots are going to come in more for Trump than for Biden, but not as strongly as he needs.
There’s also the fact that two other news organizations, the Associated Press and Fox News, have called the state for Biden. I’d assume they’ve put more work into looking into this than I have, so that shifts my priors a bit, but you never know and you will get incorrect calls occasionally. Overall, I’d say this is
Likely Biden but I don’t think the state should have been called yet.