When Pelosi and the DEMs pressured Biden to withdraw, he was falling behind in the polls. One assumes that DEMs were scared by the trends they were seeing in the polling data. So are we talking about polls that were using the same methodology that is now thought to be over-correcting for the errors of 2016? Was Biden in better shape than it appeared? Your assertion seems plausible if so. Either way, I have far less faith in the polls than I did and hats off to Selzer for actually keeping the science in data analysis. What other posters like Nate silver are doing with the data seems awfully capricious.