I meant the "early raw exit polls".
That night, I first thought it was a DNC attempted coup, possibly finding out which counties the pollsters are working in and perhaps sending heavy groups of blue voters early before lunch to counties where the pollsters were. As the exit polls information got out, it was supposed to depress turnout of Bush voters. If the networks followed their 2000 template, announcing winners earlier than expected, it could further discourage red-voter turnout in states where polls have not closed yet. But they didn't do that this year.
Now, I'm looking at that county by county map and question how the pollsters can weigh each county properly, or even predict the final turnout in each of those counties. If the exit pollsters over-weighted the heavy urban centers, they can easily produce the wrong results --- Virginia is the most obvious error, close in the exit poll but a blowout for Bush, but if someone overweighted Fairfax, Arlington, Alexandria, Richmond City, they can sort of get the virginia close race in the exit poll theory, and they'd be wrong --- which they were.
I doubt the pollsters can cover all the counties, or even project the turnouts correctly. Which is probably why early exit polls always get it wrong every time.
Virginia : us.cnn.com
Fairfax : us.cnn.com |
