I have read commentary to the effect that Biden is in better shape now than Clinton was at this point in 2016 because he has a bigger lead in the polls than Clinton did. I believe that observation is flawed because it seems to rely on conglomerations of polls (like those tracked by RealClearPolitics and FiveThirtyEight), the quality and composition of which varies from day to day, and which has probably changed a lot since 2016.
I follow White House Watch at Rasmussen Reports. It’s superior to most polls because it’s a tracking poll that samples the same group of likely voters throughout a campaign. I suspect (but don’t know for sure) that there’s a lot of overlap between the 2016 sample and the 2020 sample.
The graph below compares Clinton’s and Biden’s lead or deficits against Trump, given the number of days left before election day. As that day draws nigh, Biden is actually doing worse than Clinton.
Notably, the final Rasmussen poll in 2016 hit the popular vote gap right on the head: Clinton “won” the mythical nationwide popular vote by 2 percentage points. And it did her no good because her popular-vote “victory” was the result of lopsided outcomes in deep-Blue States (e.g., California), where extra popular votes didn’t translate into extra electoral votes. I expect the same kind of result this year, though some of the States that went narrowly for Trump in 2016 may flip in 2020.