I’m going off memory here, so correct me. I believe Nate Silver at 538 was one of the few pollsters that gave Trump a 15-20% chance on election day in 2016. He was criticized for being way too generous to Trump with that estimation.
I thought that they began predicting/projecting out the election odds by August and yet I do not see that yet on the sight. I see popularity and national polling averages, but no odds/percentages on winning the electoral college.
Has he said he is not doing that this year or does he wait for the conventions?
Discussed somewhat in the thread about Biden’s VP pick. Off the top of my head, I remember Silver starting earlier in previous years as well. He did not used to wait for the conventions, as I distinctly remember post convention bounce maps. These included analysis before hand of how big a bump was expected vs. what actually happened, the Republican vs. Democratic convention maps, etc. Of course I also remember him giving Trump a closer to 30% chance rather than the 15-20% you recall, so it could be I’m not remembering things correctly . Memory is funny that way.
I have not followed 538 so far on this election, but I am guessing about a 75% chance of Biden winning at this point. A real serious chance for Trump, but Biden has a real small lead.
The Economist has it at 90% for Biden. I think that’s a little optimistic but I like where the odds sit.
One of the things I have to point out to my fellow lefties all the time is that Trump had a 1 in 3 chance of winning and 1 in 3 things happen ALL THE TIME. Roll a die. Did it come up 1 or 2? Bang! 1 in 3 chance paid off.
I think 538’s (and others) analysis is both right and wrong. In a free and fair election, the odds do not favor Trump. But this election isn’t going to just come down to the votes cast for each candidate, but instead the number of votes counted for each candidate. The Republicans are going to do their best to corrupt the election by decreasing those numbers. They’re going to shut down polls (due to COVID and/or security reasons), perhaps have federal agents (and local police where they have control) “protecting” the polls, disrupting and delaying the delivery of mail-in votes, and calling the election at the earliest possible moment it is legal to do so (and maybe even earlier and hope to drag it through the courts), if Trump is leading at that moment, and dragging the election out as long as possible if he is not. Sadly, I think they’re going to succeed because the stakes are very high. The Republicans cannot afford to have Trump go to prison, and they cannot afford to have the amount of corruption and criminal activity by other members of the administration, e.g., Barr investigated and punished. While, of course, they can never lose the 40% that are brainwashed into their cult, it has the potential to be catastrophic among undecided voters.
The election will NOT be fairly contested. No chance of that at all. Republican cheating will be very substantial.
What’s difficult to say, of course, is how much that will affect the outcome. I would guess it will be roughly equivalent at a 3% swing in popular vote, but that’s a very rough guess indeed.
. Memory is funny that way.