Archive for the 'Statistics' Category

Dec 03 2009

Hey! They didn’t ask me…

So apparently some researchers wanted to study the effects that pornography has on the male psyche. Naturally, in order to do so, they wanted to compare their study group of naughty-video-watchers with a “control group” of pure, never-saw-the-stuff young males.

The problem? The researchers discovered that their control group was what statisticians call a “null set.” They couldn’t find any young men who had never viewed porn.

Now, I can’t claim any particular moral purity; I am, after all, bent on achieving Complete Global Domination. But a high-speed connection to my Bunker of Doom was way too expensive.


No responses yet

Dec 09 2008

Rock On!

Once again, I feel like a mosquito at a nudist colony — I have so many options, I just don’t know where to begin!  Auto maker bailouts!  Corruptocratic politicians in Illinois!  International intrigue!

So of course, I’m going to talk a little bit about floating rocks.

I guess it’s a sign of either our sophistication or utter ignorance that we seem to be indifferent to the fact that the universe is filled with huge flippin’ rocks, many of them mountain-sized, just randomly floating around above our heads, with the occasional one just taking it into its rock-hard head to plunge to Earth in a roaring fireball, hopefully only taking out an abandoned mud-hut in Siberia, but every once in a while triggering a mass extinction event.

For some reason, I’ve witnessed three visible meteor strikes in my lifetime (which might explain my super powers), and even though these were insignificant little measured-in-pounds meteor runts, they’re still damn impressive to a hairless ape looking up at the display, mouth agape, from the ground.  Events like this make you start thinking about how very wrong it would be if you were taken out, utterly at random, by one of those suckers (Well, maybe for some of us — “Hey!  You hear about Mind Scalpel?” “What?  What???”  “Killed by a meteor strike!” . . . “Really?  Somehow that makes a strange sort of sense….”).

Anyway, good ol’ Ron Reagan famously speculated that the only way humanity would unite for the common good was if we were faced with a common enemy, like invading space aliens.  His critics had a field day twisting this, of course, but fundamentally he was right on — there’s nothing we’re quite so up for as a good rabble-rousing focused on demonized “others.”  Especially when we don’t have to scare up the torches and pitchforks ourselves — we’ve already funded the military with our tax dollars, after all.

But we only seem to get all het up when we’re able to put some face on the enemy.  If we were told an actual alien intelligence had malevolently set a bunch of mountains floating around, with one of them scheduled to hit Earth cataclysmically every few millenia, we’d come boiling up out of our gravity well in fleets of nuclear-powered space warships like army ants looking for honey.

But since they’re just randomly floating around for no apparent reason, well, hell, someone else can deal with it.  Maybe we’ll make a really crappy movie out of the potential for a world-ending meteor hitting us, but the production of that really crappy movie and the money thrown away on it will far exceed the money we’ll spend dealing with this potential problem.

“What’s the big deal?” you ask?  “The odds of a huge asteroid destroying the human race are minimal, like a fraction of a percent.”

First, I’d point out that they’re asteroids when they’re just floating around in space, minding their own airless business — they only become meteors when they enter our atmosphere intent on whacking us.  Second, quick lesson in risk management!  One has to consider both the probability of a particular adverse event happening, and the magnitude of the event if it happens.  Otherwise, you’re just dealing with one-half of the equation.

So someone hears that the frequency of a meteor strike with an explosive force of 10 megatons (roughly equivalent to 70 Hiroshima-sized nuclear bombs) is about once every 1000 years, and figures “hey, no big deal” — because he’s just focusing on frequency.  Let’s say there’s only about a 1 in 20 chance of that strike encompassing a major population center (let’s say Los Angeles).  Even better, right?  Oversimplifying the math, we can call it a 1 in 20,000 chance in any particular year that such an event will occur.

But…Los Angeles has about 10 million people in it.  Let’s assume they’d all be vaporized by such a strike.

So every year there’s a 1 in 20,000 possibility that 10 million people will be vaporized by Big Pappy Meteor.  Working the averages, that means that on average, we’ll lose about 500 people per year to such meteor strikes.

And this does not factor in the secondary effects of such an enormous event — disruptions in the economy, food supplies, distribution, shipping and communications networks.  Such an event could potentially end a nation the size of the United States.  If the credit default swap market doesn’t finish the job first, that is.

And by the way, meteors with an explosive force of 1 megaton (a mere 7 times a Hiroshima-killer) enter our atmosphere every 100 years.  Let’s say there’s only a 1 in 50 chance that one of these will hit an L.A.-sized target (smaller blast radius).  That still works out to 2000 people per year, on average, killed by meteors (yeah, yeah, all you probability and statistics people — I know the exact number’s a bit different — lighten up!  When you’re dealing with nukes you don’t need pinpoint accuracy!).

And if we’re talking about a world-ender, well, even if something like that only happens every million years, there are maybe 6.75 billion people on Earth, so that translates into about 6,750 people annually being killed by such an event (on average, not factoring in population growth).

So maybe a little prevention’s worth it.

Then again, I do believe in the wisdom of crowds.  Maybe there’s a hidden reason people might be willing to risk a meteor strike without any means of defense.

I know!  They figure the possibility that a medium-sized meteor will take out Washington, D.C., however slim, makes it a gamble worth taking….

No responses yet

Jan 27 2008

Public Service Marketing = Demonize the White Guy

(Non-humorous screed alert!)

Alright, this one was so damn subtle it almost slipped by me. But not subtle enough!

Currently, there is in at least one East Coast market a radio ad campaign for “Wireless Amber Alerts.” One ad features someone who is obviously supposed to be a police dispatcher making a radio call for an Amber Alert — the abducted child is said to be something like a “six-year-old African American girl.”

The suspect?

A “Caucasian male, 30-35 years old, blue shirt, black jeans, 180 pounds…”.

I’m in favor of the Amber Alert system, and I think the addition of the Wireless Amber Alert text messaging system is a great idea — in fact, I went to the web site to sign up for it.

Then I heard the above radio ad again, a few times.

And I thought, wtf?

Why, exactly, was this particular racial combination (African-American victim; Caucasian suspect/offender) chosen to be the example for the Wireless Amber Alert ad? How often, based on actual statistical evidence, is that particular racial combination of the suspect/offender and victim present?
Continue Reading »

One response so far