I'm surprised I was not quoted in this!
Depending on which projection you are looking at, Biden was closer~ to Trump in March maybe before the middle of the country was taking Covid serious. According to The Economist, they were projecting it as such
President—Forecasting the US 2020 elections
It's interesting to see them be so certain with predicting results. So many pollsters/Election experts are so, so, so, so scared to predict that the election will go one way with certainty because there is a small chance they could be wrong (like in 2016).
View attachment 72260
This Canadian election analysis site is pretty bullish about their predictions too in the same way as the above:
US Presidency
Real Clear Politics and the like don't weigh polling at all depending on the bias. So their polling averages are all out of wack. There are loads of polls that show Trump tied or up in the GE polling which is insane. These are all not to be trusted, especially when the show the data. They'll have like 30% of Dems voting for Trump and 75% of the Independents which will not happen at all.
Plus early voting data is a huge indicator on how this election can go. I saw today that Texas is almost at 100% of ALL votes in 2016, election day has not even happened yet. If turnout is massive, and Biden/Dem ballots have returned with 5-10% leads or much, much more in states like Texas, NC, Georgia, PA. It's already game over.