Good Results from Bad Data 

If you're data is flawed... then the results from that data can't be accurate. It's a simple rule that most people who claim to follow the scientific method know well. It's a rule that commentators and pundits have ignored the past couple of days. If you listen to the major cable news networks, there are two election stories.

1. Why were the exit polls off by so much?
   
2. What can we learn from the exit poll data about why people voted the way they did?
   

Hello! If the data was proven wrong in predicting who won, what on Earth makes you think you can gleam useful information about why people voted the way they did? The data is flawed. That means that ALL results from that data are tainted - not just the predictions on who would have won.

Vodkapundit is doing some gloating about a pre-election prediction he made regarding leaked exit poll data. It's a well deserved gloat. I commented there about it, and it deserves repeating here I think. One of the theories floating around the blogosphere, and the punditry on the news networks, is that the Kerry campaign possibly flooded polling locations, where they found out exit pollers were, with shills who would gladly answer the long questionaire and answer Pro Kerry in order to affect exit poll data. This does not mean they voted illegally. I'm not claiming that at all. All they would have to do is walk into the polling location, not vote, and then walk out as if they had. Since exit pollsters aren't allowed within the polling locations by law in most states, it's pretty easy to pull off this scam.

Does the DNC not remember what happened in Florida during the 2000 election? Let's recap. Before the pan handle polls closed (because they're in a different time zone), many networks called the State for Bush. Many people have said that one of the reasons Florida was so close in 2000 was because voters went home in the pan handle thinking the state was already won. The pan handle counties generally vote Republican.

So why would Kerry want to leak data making it look like battle ground states were won? Wouldn't they be scaring off their own voters who didn't want to stand in insanely long lines for nothing? Could that have pulled in more Bush voters who didn't want to see 4 years of Kerry? Could that have been the stupidest decision of the election? Just a thought on a wild ass theory.