Exit polls don’t make sense until after the election is over and they can reweight w.r.t. actual results. Until then we have no idea.

💯 also, those places tend to be very blue regardless, so not like we could really tell if it’s students or professors showing up anyway.

Mostly, think it’s not worth pondering at this point. We can just wait and get returns now

Forgot to add, MU was on campus in Nov 2020

Just as anecdote, my wife is the LEO here in Columbia MO, and the college students have been coming out of the woodwork for the past two weeks.

Which is to say, I think you are right to caution, but I think student turnout IS probably up. But we’ll never be able to suss that out of the data

We want high turnout in urban areas, the GOP wants high turnout in rural/exurban areas. You really need both numbers to get a sense of what’s happening, as well as what happened previously. So it’s not always straightforward.

That being said, I think turnout right now looks good for Dems.

Early exits are like every other kind of early data—garbage.

A corollary here is that early voting is likely to favor urban environments, because…well, because there are more voters there, so more infrastructure is available.

It was a boring election. I put my kids to bed and went to sleep, that's about all I remember

All these polling "models" are the essence of scientism--they took a bunch of shitty data and did...things, and the output ~looks~ scientific--odds! Quantified uncertainty! But it has absolutely no rigor, nor even a pretense at basic mathematical justification. It's all just bullshitting.

Another great example is the obsession with polling averages, but never really taking the time to ask hard questions like "Why are all the averages different even though they have basically the same data?" or "How do you reconcile a poll with a 3pt MoE and a poll with a 4pt MoE?"