Update: Basically I am extremely interested in the results tonight, and I want to know where things are headed – but I am not willing to hang on every detail. And counting may take a long time. So I have cooked up a way to track results to get an idea of what’s going on. Here’s my Geek’s Guide in PDF form, which includes a polling-error graphical calculator. Also download this Google Sheet.

I have two basic views. First, I want to use polling to identify which races are most in question. As long as polling is no more than, say, 5 points off, then we can do that. Second, this year I care a lot about downticket races that can affect the conduct of democracy in 2024. So I collected that polling (see previous post).

Tonight I will be tracking key races where I think it will be possible to estimate the final margin fairly early. That includes the New Hampshire races – and maybe, thanks to Jon Ralston, Nevada races. I’ll be plotting that in the Polling Error Calculator that is in the PDF. As more races come in, it will become clear whether there is a pattern of systematic error in the polls.

Once I have that error, it may be possible to assess who will control the Senate – and how many crazies will be in control of vote-counting in 2024.

I’ll update this later, trying to get this up right quick…have at it in comments!