The pollsters hand-pick 500, 1000, or 1200 people to continuously "sample", delete the data spikes that don't uphold the results they are paid to produce, then report their garbage as if it were divine revelation. It's still B.S. Luckily for the pollsters, the vast majority of sheeple have never endured a college statistics class.
Here's a poll taken from a sample of 8713 (and expandable) of a huge population, Facebook. With 500 million uniquely identified members, 150 million from the USA, allowing no multiple votes, allowing no deleting of data spikes, here's what you get.
Should Sarah Palin run for office in 2012?
Votes: 8713
61% Yes, Sarah Palin should run for office in 2012.
37% No, Sarah Palin should not run for office in 2012.
1% I don't know. Please see my answer below.
I am loving that poll. Did you notice that at least half of the first few comments were by people for whom English was not their first languge? LOL