Pseudolympics

Contributed by
Aug 20, 2008
<?xml encoding="utf-8" ?>

I am not really a big Olympics fan. I used to be, but then they started letting pro athletes play, and the "pageantry" became ridiculous, and now China is hosting them and using them as a political tool in a way that strongly reminds me of

1938 1936. I thought it couldn't sink much lower.

But I was wrong, of course. Because, it turns out, astrology rules the Olympics! At least, it does according to a totally credulous article posted by Reuters, which parrots a bunch of silly claims by "statistician" Kenneth Mitchell -- that's in quotation marks, because his statistics are very, very suspect.

This kind of article steams me pretty well. It is written slightly tongue-in-cheek, but it still simply repeats without any rebuttal the Mitchell's statements, which are easily shown to be false. I'd go into details, but happily Josh Giersch already has, and quite well. From Mitchell's bad math to the cherry-picking of data, Josh slams him pretty hard.

There is one thing I want to mention, though. In the article, Mitchell says something totally bogus:

Explaining his eureka moment with all the zeal of a statistical crusader, he concluded: "Did you know that the distribution of Olympic swimming medallists against the tropical astrological zodiac signs can be almost exactly mapped by a polynomial function of the third degree?

"That's one to shut people up at a pub."

Actually, I'm sure it would, since most people don't know what a polynomial is, and thus can be bamboozled by such a falacious use of math. A polynomial is a function (like y = 8x3 - x2 + 3x + 6) that can be used to plot points on a graph. It can also be used backwards, so to speak, using the points from a series of measurements to derive the equation. So you can measure some series of events -- like the astrological signs of Olympic winners -- and then find a polynomial equation that best describes them.

The thing is, in general the higher the order of the polynomial (the biggest exponent of x), the easier it is to fit the points. A third degree polynomial (like the one above) can be fiddled with easily to fit a series like that very well, no matter what the points really are. In other words, whether or not astrology is correct (and it ain't), if you plot the birth signs of the athletes, you are very apt to find a third-degree polynomial that'll fit the points. You can swap around the dates, pick the losers instead of winners, or measure the athlete's hair length and still find some polynomial that describes them pretty well.

So what Mitchell said is meaningless; it only sounds cool. It's an empty claim, just like all of astrology. I'm not surprised he said it, but what really irks me is that the Reuters reporter swallowed it wholesale. I dream of a world when journalists actually do some research for their articles. I'll be dreaming a long, long time.

Too bad these are the summer games. Otherwise, I could make a joke about astro-luge-y.