and how's that different from the projections they had for Hillary in 2016?
up to election night she had a 71% chance in most polls...
I wasn't paying much attention that year because it seemed a foregone conclusion that Hillary was winning. I don't know what this group said. I have watched since then and they've been very accurate. For several months leading up to 2020, they said Biden was winning while the media fretted and people I knew fretted. He won the states they'd been saying he'd win for a couple months.
What they do is better than most. Most polling, because it is easy I guess, just measures popular vote.
On the last midterm, my dad was telling me that there would be a big red wave in the House and they were also going to take the Senate. He said "the polls" said so, but 538 was telling me that the Democrats would get both Georgia Senate seats and that the red "wave" would be pretty small and tight.
Getting polling right is difficult, but Trump didn't get underestimated in 2020. Pollsters learned from 2016 and adjusted. You have to know how to weigh your data and how to draw proper conclusions. I know one of the things they learned in 2016 was that the way they were carrying out the poll was biased toward Hillary voters being more likely to be polled.
You're absolutely right: we shouldn't trust the polls. I guess we just need to mobilize as much of the population as humanly possible then vote blue all the way down the ticket.
You are saying one minute that Trump doesn't have a ghost of a chance and then another minute saying that you can't trust the polls, get out there and beat the streets.