IMO the entire election will come down to PA, MI, WI, IA, AZ, and NE-02 and ME-02. NC, GA, and FL will all give Trump or whoever the GOP candidate is a run for the money, but I predict that they will all go red. Note that if even one of them flip, however, the Democrats have a clear track to victory. Same for the Republicans if any of the three swing states in the Upper Midwest (not including Ohio) go red.
A big, big, big problem from 2016 is that the base learned so many wrong lessons. It took a sensationalist media, Russian propaganda, the electoral college, and the least supportive primary candidate since 1980 to bring Hillary down. Not saying she didn't make mistakes--she did, and contrary to popular opinion, she admitted it--but a fair analysis of her defeat requires not cherry-picking the facts.
This is subject to change, but right now the winning move for Democrats is a center-left woman candidate who can connect with the Upper Midwest, paired with a progressive running mate. This would do a few things: (1) Having a progressive VP candidate could be a sort of reverse Mike Pence effect, who was thrown in to appease white evangelicals (and it worked), but not that many other people chose who to vote for/against solely because of Pence; (2) A woman candidate would tap into the highly underrated force known as the woman vote that was stronger in 2018 than in 2016; (3) None of this is going to matter if the Democrats can't flip the Upper Midwest.
I have a couple names in mind but I don't want to say just yet who. Give me a few months and we'll see if they're still up to par.