My friends must be very unscientific!

What’s up with the polls? One week Kerry is ahead by eight points. A week later, Bush is ahead by a similar margin.
I hope my readers will bear with me in this train of thought, as I am not going to spend the next hour tracking down poll after poll to prove my point, which is that the polls suggest a highly erratic, fickle voting public — a substantial portion of which changes its mind on a daily or weekly basis.
I must be running with the wrong crowd. I have friends who are for Kerry, and friends who are for Bush. I don’t talk to all of them them daily about politics, and I don’t need to. I know who is for Bush, and who is for Kerry. Not one is undecided, and not one is likely to change his or her mind.
Am I alone in this assessment? Where are all of these fickle, volatile people? (Via the justifiably skeptical InstaPundit.)
I know a lot of people. How come I don’t know one single fickle voter?
How do the pollsters manage to find so many of them?
Surely they’re not using junk science!