From the CNN partial returns for Wisconsin...
As I said, I think the whole vote-counting thing with projections is weird. See here as Kerry is projected winner even as Edwards currently has more votes.
Update: Obviously, projections is a technical business having to do with statistical sampling and analyzing which sets of returns have already come in and which ones haven't. I just thought it was amusing to see the numbers in that shape.
Posted by Curt at February 17, 2004 08:05 PM
It's not that hard to explain. Each district has a particular coeficient, similar to a stock's beta value, derived from polling over several months. So you can calculate that if candidate A did 44% in district A, the he is likely to get 41% in district B. Pure math, and it usually works, if the race isn't razor-close, as it was in Florida.
Posted by: M. Aurelius at February 18, 2004 07:14 AMNot that I don't love a good conspiracy theory now and then, but I'd say the odds are this is just a fluke mistake by whatever intern they had updating the site.
I work in the web business, and although there's PLENTY of technology out there that could update the results automatically, most places are too lazy to implement it. Why spend $XX,XXX on a decent server application when you can just make your unpaid or min. wage interns do it.
Posted by: Brian Douglas at February 18, 2004 03:09 PMOh, I'm sure that it's just because they knew the general voting behavior of each county, and some had reported results while others hadn't. It's funny, and it really does seem like there is a lot of opportunity for monkey business even when it's not being taken advantage of.
Posted by: tunesmith at February 18, 2004 03:21 PMThis would be a whole lot more interesting if Kerry had actually finished behind Edwards; as it is it seems to confirm that the news organization knew what it was doing when it projected Kerry as the winner.
Posted by: Pat Curley at February 18, 2004 05:26 PM