Where they’re pretty provably crappy is for *counting* voters. For whatever reasons, they do a bad job of figuring out what percentage of real voters were Obama voters or Hillary voters.

But figuring out, say, the psychology or demographics of Obama voters on election day, that’s easier. You just need to talk to a large enough sample of Obama voters, and you don’t care how many of them you have compared to Hillary voters. There can still be distortions (and I’m sure there are), but I think the “internals” are probably a lot more reliable and useful than the “who’s winning” data.

There is, alas, also an unpleasant possibility in this race drenched political cycle. And yes, it might be the so-called Bradley effect. Could some guilt tripped whites feel uncomfortable telling someone else, especially if they are a minority—that they did not vote for “Barry” Obama?

You know, reports like:
“It’s raining really hard now…”
“There’s a lot of police activity here at the scene…”
“The heat from this fire is intense…”

(How many years of journalism school do you need to do that?)

It does, however, give us insight into the empty-headedness of the blathering class. Without a feed from a Tele-Prompt-R, they couldn’t recite their birthdate.

But let’s face it, life in real-time just isn’t that interesting.

Scott