Statistics and political polling

I am currently taking a Statistics course and seeing the real world uses everyday. The election polls that we here about over, and over, and over again really seem to use those formulas as much as possible. Ralph K.M. Haurwitz discusses this in his article “the lowdown on higher ed” http://www.statesman.com/blogs/content/shared-gen/blogs/austin/highereducation/entries/2008/10/28/odds_overwhelmingly_favor_obam.html. He references a UT professor who I’m sure is much smarter than I and has much more data than I could ever hope to look at. He utilizes a software that predicts future events used in investing. The prediction in the data on the web page of the company says there is a 88% chance that Senator Obama will win the election. The professor says the software is incorrect and that actually there is a 99% chance Mr. Obama will win. Basically Professor Love states that because of a bunch of algorithms that rely on random sampling the results are much higher that Mr. Obama will win.
I am more of a common sense kind of guy that looks at people a little more simply. On one hand I can see that a certain percentage of Obama supporters are younger voters that can’t really be polled because they only have a cell phone and no home phones which are used for polling calls. That could make up a huge percentage of votes that can’t really be counted. Along with the normal undecided voters that are unaccounted I can see an X factor that represents many votes for Obama. Lets look at the other side though, Senator McCain has an issue with his own party in that many Republicans don’t have a great feeling about him, his central voting record, and some of his policy stances don’t follow party lines. Those represent many voters that will eventually vote for McCain just to keep Obama out. Also, there are some of the undecided voters that will swing to the McCain side as well. I look at polls a lot like I look at weather forecasting. Many times these things are right, but in the case of polls I think that many, ok most have been wrong. Kerry was going to win in a landslide, McCain had Bush in 2000 by huge numbers, even this year Hillary Clinton was way ahead of Senator Obama through much of the primary. Don’t even get me started on the media and the fact that the only reporting they’ve done in the past two weeks is to report on each other’s polls and what that means. NBC’ references an ABC poll, CNN reports on a CBS poll, and ABC leads their news with a CNN poll that is there “breaking news”!!
Statistics should be focused on real data, that is numbers, populations, sales, dividends, and other real data that doesn’t rely on a human choice which can be overstated, understated, and even lied about. As for me, I think I will move to Fiji and see if I can fire up a coup.