So, folks, we are about 7 days away from the big dance. Even though the polls show Biden up by 7-8 points nationally, there is a good amount of evidence that Trump is very much in this fight. Trump does not act like he is down 8, nor does Biden act like he is up 8. Even Biden's campaign says that their internals admit that this is a close race. Could they be saying that to help with fundraising and/or mobilization? Possibly, but again, they do not need any more money (they won't be able to spend it all on time anyways, most likely), and campaigns usually release good internals to mobilize voters, not ones where they are down. So, this race could go one of two ways, either a close race or a Biden blowout, and most signs point to the former.
Dave Wasserman (nicknamed Dave Wasteman of the Election Mafia) clearly sees this race as being the latter. Here is his tweet surrounding the election:
You can clearly see he's coping. Let's pick this apart.
First of all, Biden is not up 52-43. 538 includes internals which can skew the aggregate. Historically, Dems do release more internals than Republicans, even in red wave years, so this is no surprise. Biden leads just 50.8-43.0 in RCP, and Biden has struggled to consistently poll above 51% (Obama's 2012 margin). It will be interesting to see if that may tighten if undecideds who approve of Trump come home to him in the last week, in regards to a more "stable" lead. Biden's lead has been more stable than Clinton's, yes, but it does not matter if the turnout models are garbage (as they have been in the past, and they have been criticized this cycle by experts such as Robert Barnes and Richard Baris).
Speaking of undecideds, yes, there are far less than there were last time. However, there is still 6-7 percent of the electorate that remains undecided. In the final two-way aggregate, there were 9-10% undecideds, and several of those were going third party (and there should be far less third party vote this time). Also, we can look back to 2012, when Obama overperformed the average by more than 3 points and there were far less undecideds, about half of what exists now (by this logic, Obama won them all and a decent part of Romney's share). Trump's approval is at 45-46% among Likely Voters, so him getting more than his approval rating is reasonable as most Presidents have in recent history. If he gets 47%, but Biden stays around 50%, this is anyone's game.
As for district-level polls, they are very volatile. A poll out of OK-05 came out today showing Trump up by more than 2016, similarly to the one out of NY-11 that showed him winning by a similar margin. 2018 had district level polls look all over the place. I don't like this argument either. We will see on Election Day. Biden may win, but I expect the Election Mafia to have many of their ridiculous points proven wrong regardless.