Not That I’m Opposed to Obama Winning the Virginia Primary…
Posted on February 12, 2008 Posted by John Scalzi 27 Comments
… But is it too much to ask that when CNN calls Virginia for the man, they count more than 0% of the ballots? And actually have the man in the lead?
Jeeze, John, get with the program. The election itself is over and done with, reporting actual, verifiable results is just one of those pesky nuisances we have to put up with… like a root canal.
This sounds like one of those physics problems involving spherical cows on frictionless planes doing statistics with imaginary numbers.
(Their math has improved: with 14% reporting, Obama leads 61% to 38%.)
I’m always amused at the consistent difference between the BBC and NYT headlines when it comes to reporting on these primaries. The former will have “Soandso is projected to win Wonkadelphiland”, the second “Soandso wins Wonkadelphiland [and is set to win the whole field!]”.
Latest I’ve seen is with 51% reporting, Obama has a 30-point lead over Clinton. Ouch. That’s what we referred to back in the day as a “drubbing.”
That graphic shows the percentage of white voters who voted for the candidates according to exit polls. So it was mislabeled or something. He was projected the winner instantly due to their exit polling which apparently showed a “significant” margin between them, so they figured they’d go with it I guess. But as for those specific numbers, the CNN talking heads say it shows that he’s made progress on the share of the overall white vote compared to Clinton.
It’s news so it is more important to be first to report it than it is to be the first to report it accurately. Don’t worry they will get it right on the next hours reporting cycle. Maybe.
#3 and #4 have it pretty much right: Sen. Clinton is on the downside of an electoral landslide in VA tonight, and the exit polls, on which CNN quite correctly refused to pronounce until the polls closed, made that very clear. Calling it for Obama was perfectly appropriate – but rather than putting a big checkmark next to a (lower) percentage of ~0% of the vote, they should have just put in a sentence about a predicted victory.
I’d like to _think_ there’s merely a glitch in whatever spreadsheet they’re using, that they actually have some basis for their call.
It’s alchemy, John. You can’t hope to contain it. Or understand it.
Poof, lead is gold.
Good catch, John. It’s one of those things that makes living in America so infuriating . . . even if it’s an illusion, I don’t like the press dictating what we’ll do in the (near) future.
9:35, and CNN’s done it again… called MD for Obama right after the polls closed, even though the first official numbers that went up showed Clinton with more votes in the first returns ( less than 1%, mind you)
MD 0% * Obama 33% Clinton 50%
I’m presuming that in both MD and VA, the exit poll trends have been pretty definite…
Yeah, I would suppose that as well.
jim: It is obvious to the casual observer that spherical cows aren’t frictionless. (grin)
Dr. Phil
The raw numbers I saw that went up first with both MD and VA (>500) were so small that I suspect that they were based on one single precinct reporting. Kind of like in NH when they report the results from Dixville Notch. They really should wait until at least 1% of the vote before reporting percentages.
That is an an amalgamation of actual results and extensive exit polls. That’s why they always say “we project that…” It’s usually right, and since the polls are closed by the time they report it, what does it matter?
Oh. I thought we just look at the check mark and the name next to it. You mean we have to compare the numbers for ourselves?!
As long as the numbers are not coming from Sylvia Browne, I can live with such brief cognitive dissonance from the new networks.
newS. newS.
And no, I am not a poli-cow…
NBC’s calling of D.C. for Obama with 0% reporting was even more puzzling. There were no exit polls done in D.C. What’s up with that?
No excuses for independents in Ohio – go vote. http://www.akron.com/akron-ohio-community-news.asp?aID=1514
“As for nonaffiliated voters (who are often called independents), Williams said they can vote in either party’s primary. They must sign a form stating which party they will vote with in the primary.”
Remember that old story about it doesn’t matter who votes, it matters who counts the vote?
Seen this time and time again, and it’s something to remember when poll-watching so you can sound like a veteran….
What happens is that the first numbers reported (those “0% precincts reporting” counts) by the county election commissions are the advance/absentee ballots, some of them filed weeks before the election date. Advance/absentee ballots are largely filed by the elderly. But the media outlets project the winner based on day-of-vote exit polling, which covers the entire range of actual voter demographics. Not that exit polling is always all that accurate…but a really wide trend is utterly dependable. Like the 2 to 1 results.
So you get an early “flash count” with 0% of precincts reporting that is often wildly at variance with how the poll results will come in.
Oh, and the election commissions don’t really have a lot of leeway about holding ANY results back to let the numbers accumulate. They report them as they have them. If they didn’t, they’d be crawling with hungry lawyers and conspiracy freaks. It’s public information, and the media outlets report the numbers as soon as they’re registered. They usually have people right there AT the election commission to grab the numbers as they’re registered.
What they have first is the advance ballots. They count them ahead of time to get them out of the way, and post them the moment they get confirmation that all polls have closed.
One might infer from the sarcasm that you think: a) election predictions are ever based on actual collected data, and b) the accuracy of election predictions might improve with the use of such actual collected data. I have my doubts.
Maybe they should handle it like food labeling, and note anything under 2% as “trace amounts.”
I’m just saying how the reporting you see actually works. Exit polling has its own problems and skews, they’re tough to adjust for, and getting tougher. If the race is actually anywhere near close on the exit polling the media people will hold off on a call until the actual numbers confirm the exit polling. If it isn’t close, they’ll “project” the winner as soon as the polls close.
The first “actual vote” numbers you see reported when they show 0% precincts reporting are the advance vote numbers. The “call” for one candidate or another in a non-tight race is based on the media’s exit polling. That simple, but when they’re presented together and disagree they can definitely trigger a WTF? response.