"If no news - send polls."

I was listening to the radio the other day, and wsa surprised to hear that "the American People" thought this or that thing. The newscaster positively crowed that a *thousand* people had been
polled. Wow, a whole thousand? And from this, you know what *I* think? Why, it's magic Mr. Wizard!

No it's not, Johnny. It's our old friend, statistics. Statistics in this context is a con-game, one that
claims that a random sample of anything (of sufficient - but not complete - size) is descriptive of the
whole thing. The con-game here is that the sampling done is not random, and is not of
sufficient size. The results of the poll are useless. So why tell us?

The reason is, of course, a lack of actual news. Now, don't get me wrong - there's tons of
actual news out there -- people are being killed, and being born, and achieving all kinds
of interesting things. But finding all that stuff out is expensive -- that is, it costs actual
money to send actual people to find actual news. So instead, many news organizations
instead rely on that great artifact, the "press release." Press releases are just a story,
pre-spun the way whoever releases it wants. Press releases are often investigated, sort
of, but the story is usually passed on to us as-is. Press-releases are generally self-serving,
and dull. So what's a news director to do? Poll! It's cheap, it generates numbers, and you
can "analyze" it endlessly. What a godsend!

Oh yeah, numbers. In the news game, "numbers" equal "truth". If you can present it as
a number, it must be true, and therefore, "hard news". Ever wonder why the stock market
numbers get read to you every night, or every hour, or every minute? Easy - it's "truth"
(= "numbers") - and best of all, it changes every gosh-darn second! Wow, you could
found a whole news-network on that. No wait, they already have. And analysis! It
went down, it went up, it stayed the same - it might go up, or down, or not. Sheesh.

But back to polling. Nowadays, the people doing the polling have finally realized that
the scientific-ness of polling really doesn't matter. Have you noticed that CNN (and other)
"news organizations" (it's gotta be news, it says so right here on the label) have started
"internet polling" about issues? Sure, they say things like "this is not a scientific result, and
blah blah blah blah blah." (I'm not sure what it is they say, they talk *real* fast through
that part.) But they still announce the numbers, and something about what they "mean."
(Actually, a lot of something about what they "mean.") Any statistician would tell you
that the numbers mean nothing. They aren't random, they aren't 'fair' (you can
typically 'vote' multiple times), and so on. But so what?

So what should you do? Well, whenever you get polled -- lie! (Most of you already do,
according to my own non-scientific-but-nevertheless-valuable-and-you-should-believe-it
poll.) Tell them what you think they don't want to hear (this works really well on phone polls
conducted by marketing research and political polls.) Don't be extravagant -- if they
ask you where you live, don't say things like "the planet Zarkon", but give another
street address. If they ask you what you do, don't say "I'm president of Pixar", but
puff it up a little. What the heck? It's not like you're testifying in court or anything -- some
idiot thinks he can take *your* opinion and make it fit what he wants people to think.
If we all pull together, maybe these idiots will leave us alone. And if they leave us
alone, it's only a small step to getting those telemarketers to stop calling! What
a wonderful world *that* would be, eh?