Let me collectively subtweet the predictors and their scrum over forecasting uncertainty. I only have the ability to understand such things intuitively and not in any sense of putting formulas together - though about a year ago I did devote a couple days to trying to ...

2/ understand Bayes. But I would say this election has a uniquely high level of uncertainty. Uniquely high and very high. There's significant uncertainty to how the election will even happen. The president is actually trying to sabotage the postal service to foreclose ...

3/ one channel of voting. The economy is a disaster but because it's disastrousness is mainly caused by one on-going exogenous shock there is at least the possibility it can rebound much faster than wld normally be plausible. The President can probably win an electoral ...

4/ college victory while losing the popular vote by as much as 4 percentage points. Perhaps most importantly, and not unrelated to all of these factors the President is likely to commit a number of criminal acts over the next three months, supported by the attorney general ...

5/ to try to influence or subvert the election (possibly with foreign actors). So for all these reasons, I'd say there's quite a bit of uncertainty. One curiosity I have is how the forecasters are even grappling with this kind of uncertainty some of which is extra-political ...

6/ and extra-electoral. For the moment I will simply say that there's a lot of uncertainty and a lot that is outside of conventional politics. Having said all this the President is very unpopular and - critically - consistently unpopular. Even if he were to bounce ...

7/ back to his relative highs a reelection victory would still be pretty tough to pull off. This is separate but important point to remember given all the excuse making about COVID. President Trump has ALWAYS been poorly positioned for reelection. ALWAYS. He's just ...

8/ gotten more unpopular - mainly because his malevolence and incompetence finally came into play on a critical, historical challenge that directly affects almost everyone in the country. He's been a bad President - a judgment a clear majority of the country agrees with.

9/ He's failed on a historic level with the pandemic and gotten 10s of thousands more killed than needed to be the case. He's overseeing an economic catastrophe which he'd likely get blamed for even if it weren't his fault but he's significantly at fault.

10/ If the election were today it's overwhelmingly likely that he'd lose. And there's not a lot of reasons to think he's odds will improve over 90 days. So I think it's quite likely he loses. But uncertainty? Yeah, a ton of it. Mostly but not entirely because he's trying ...

11/ to sabotage the election, his a whole party of accomplices, and may succeed in doing so. I have no idea how to quantify this kind of uncertainty.