Friday, August 3, 2012
Why the RealClearPolitics Average Is Not Reliable
For those of you unaware, Real Clear Politics is a website that tracks polling data and aggregates news. It's claim to fame is a feature, known as the RCP Average, that taken the mean of all public pollingf over the past 2-3 weeks. Unfortunately, the RCP average is very easy for corrupt polling outlets to manipulate.
As Rush has pointed out for years, a lot of polling is designed to shape public opinion, not reflect it. This longstanding problem has gotten much worse in the 2012 cycle. As long as 1-2 crap polls come out per month, it throws the RCP average off.
For example, the current RCP average has Obama up by 2.7. That average, however, includes this garbage poll from NBC News/Wall St. Journal and this craptacular specimen from Pew; Breitbart News demolishes the first poll here and the second one here. Once those two polls are removed, Obama's lead plummets to 1.2%.
This means that, out of the past 9 public polls, 2 crap polls are responsible for 55% of Obama's lead in the RCP Average; this phenomenon has been going on for years.
So, what should you do?!?
Personally, I stick with Rasmussen. They were the most accurate pollster in 2004, 2006, 2008, and (most importantly) 2010. Scott Rasmussen was the first pollster to recognize how the Tea Party has altered the American Electorate for the foreseeable future. Wisconsin is the first place we've seen these results. As Scott Rasmussen explained at RightOnline, we should expect similar results as long as the interrelated issues of unemployment and debt drive the conversation.
This is not a personal attack on RCP. RCP isn't corrupt. RCP's methodology, however, leaves them vulnerable to the corruption of others.