Alexi Giannoulias | Democrat for U.S. Senate

I expected my first state of the race to be on a different topic, but I just can't help myself when I see reporters cutting and pasting memos from Mark Kirk without any context.

Yesterday, the Kirk campaign released a memo claiming that he has a lead in the U.S. Senate race. 

Magellan Data and Mapping Strategies are pleased to present the topline results of a 772N autodial survey of likely 2010 general election voters in Illinois.  The interviews were conducted November 3rd, 2009.

But there's more to the story.

Three things jump out at me. 

1. It's a one-day survey, notoriously unreliable in the polling industry because you've biased your sample to not include people that happen to have regular items on their schedules on that particular day of the week (think PTA meetings). 
2. It was done on November 3rd, the day after the GOP base was ginned up by elections in NJ and VA.  Robo dialers pick up those who are more "inclined" to answer a poll in a given time period insted of measuering actual voter sentiment.  They're good for movement measuring, but not accuracy; and
3. Magellan Data and Mapping is not Mark Kirk's polling firm.  Kirk's real pollster is McGlaughlin and Associates.

See here, from his 3rd Quarter FEC report:

So, if you're releasing a one-day sample poll that is known to be unreliable, you juice it by running it on November 3rd and it's not done by your real pollster, how accurate can it be?  My guess is that he's trying to distract from the real numbers in this race, which look a lot more like the ones we released.