Interesting article that supports

A theory I've been chewing on for several weeks now. Most current poll models have biased outcomes due to widely varied participation rates from poll to poll due to hot trends (debates, news and so on) dramatically changed demographics over the last decade and land line vs. cellphone vs. PC only/no TV homes. Before you slam me this theory is supported by Nate Silver in the included link. There is also a YouGov poll detailed that attempts to adjust for issues like this and has received very interesting and supportive results for the President. It shows little to no change of support for the President since before the first debate. It also shows strong and steady support in the swing states. It's a good read...YouGov part is toward bottom of article.

http://www.huffingtonpost.com/2012/10/18/obama-debate-poll_n_1979490.html?utm_hp_ref=politics