It’s still too early, but worth updating my US election preview from 23 Aug
https://twitter.com/JohnRentoul/status/1297643119291387904
Biden has an average 7-point lead in national polls, which is significant, and more than the 1-pt lead Hillary had at this stage projects.fivethirtyeight.com/polls/presiden… en.wikipedia.org/wiki/Nationwid…
But what matters is Biden’s lead in the state needed to get him over the 270 electoral college votes to win
Currently, that state is Pennsylvania, where Biden’s lead is an average of 4.9 points projects.fivethirtyeight.com/polls/presiden…
If he wins Pennsylvania’s 20 electoral college votes (EVs), & every state where his average lead is greater, he would win 289 EVs
The @538politics average is adjusted for quality of polls; the @RealClearNews simple average for Pennsylvania is Biden by 4.3 pts realclearpolitics.com/epolls/2020/pr…
So Biden’s real polling lead in the state needed to win is 4.3-4.9 points; which is not much; & anyway it is still too early
Showing my working here (and it is 290 EVs for Biden, not 289 as above, because of Nebraska’s 2nd district – thx @willcooling) docs.google.com/document/d/1K4… 
So Biden could win by the narrowest poss margin, 270-268, if Trump wins Pennsylvania, meaning Biden needs Arizona where he has a 5.1-pt lead projects.fivethirtyeight.com/polls/presiden…
But that’s much too close for sensible forecasts
• • •
Missing some Tweet in this thread? You can try to
force a refresh
