SC: Adam Geller/National Research (R): About what you'd expect

[Alcon]

This turned out to be a pretty good poll. Nailed Kasich's strength in Beaufort as well, which I never saw anywhere else.

Nice find.  Isn't it nice to be able to confirm a pollster is real using crosstabs for once?

[Alcon]

So this is the year where random pollsters we never heard of do better than the gold standards?

This is the year when we realize that what we frivolously call "JUNK POLLS" and what we frivolously call "gold standard polls" are all actually just polls and are overrated individually, while also underrated collectively. Hopefully.

The 'pollster horserace' stuff can get pretty ridiculous.  One of my favorite things in 2008/2012 was seeing pollsters boosted for having smaller errors than you'd expect given the MoEs of their sample size.  It's one thing to see people not want to spend their time on stats, but it was pretty amazing to see people actually calculate average error, and then ignore the fact they were basically lauding pollsters for dumb luck and/or fraudulent herding.