Polls and Predictive Fallibility

Apparently, this article was completely wrong. The vast majority of polls predicted Obama's victory correctly... to some degree they even under-predicted the margin of his victory. It is time to let go of wishful thinking and embrace scientific research. When many polls find the same result, the meaning is that the margin of error is actually smaller. If you want to better understand this, look up the Central Limits Theorem. Each poll is a distinct sample. Several polls are a sample of samples, whose distribution is a sampling distribution... look it up... it's real.