Voted

[koganbot]

Virtually no line, but the place was packed with voters. This was because Colorado has gone to paper ballots, which on the voting end makes things vastly more efficient. They give you a ballot, find a seat for you (it took 15 seconds for one to open up) at one of the many long tables, and you fill out the endless ballot (scads of referenda and judges to be voted on).

Over half of Colorado had already voted in advance, however; about 7,000 more Dems than Repubs, according to the radio (didn't mention independents), which isn't a big difference.

I didn't see any poll workers for candidates, though they may have been there. When I walked in, a fellow standing near me was being advised by an election official as to what to do, since his name wasn't on the list. The fellow thought that, since he's moved around a lot, he might be listed in some other precinct, and the election official was giving him advice on how to find out. Of course, the fellow could have been one of the people whose names had been purged. I hope he's persistent. Provisional ballot is a last resort, and I hope he tries other options first.

Comments

[koganbot]

Nate Silver on why you should ignore exit polls. Not sure how to tell which networks rely on them and which don't, however. Anyway, exit polls will overstate the Obama vote, says Silver; one hopes that the newscasters will bear this in mind.

[dubdobdee.livejournal.com]

i don't think there's much evidence that newsrooms understand statistics at all -- meaning they don't really know what they should ignore and what not (how size of sample and randomness of sample actually operate); even allowing for the fact that there's a pressure towards selecting for news-drama and surprise, as against accuracy -- the polling companies themselves are torn between need for accuracy (getting it stupidly wrong will reflect badly in days to come) versus need to be first with something striking ("zogby poll: astounding new york upset" -- which gets the name in early and loudly, and the fact of wrongness will be lost in the noise)

my hunch is that even the networks that have made a management decision to ignore exit polls will find it hard to resist stuff which is churning through internet back-channels like drudge

[koganbot]

Here's an excerpt from Mark Blumenthal on the subject:

2) The networks rarely "call" an election on exit poll results alone. The decision desk analysts require a very high degree of statistical confidence (at least 99.5%) before they will consider calling a winner (the ordinary "margin of error" on pre-election polls typically uses a 95% confidence level). They will also wait for actual results if the exit poll is very different from pre-election poll trends. So a single-digit margin on an exit poll is almost never sufficient to say that a particular candidate will win.

(And here's Blumenthal's essay on exit polls from 2006.)

[dubdobdee.livejournal.com]

"They will also wait for actual results" <-- top plan!

i have to say if there's one thing that causes UK observers to go "WTF!" in respect of US politics, it is the concept of the networks "calling it" -- obviously our TV political coverage does a bit of early prognostication also (there's a lot of time to fill with stupid vamping) and is often wildly wrong, but no one takes the slightest notice of this

"The decision desk analysts": everything depends on how high up the authority totempole these guys are (probably varies a little from newsroom to newsroom) --- jumping too soon is a disaster, but being badly second is a catastrophe --- and they self-evidently have NO sway on non-election days, in regard to how the networks deal with polls