# Statistically Speaking

My college education had a number of prerequisites for graduation. Since I’d signed up to be an engineer, classes like “Fundamentals of Engineering” seemed like wise additions. Other university requirements, like “Walking,” were less obvious choices. Mostly, though, I had to deal with class after agonizing class of Math. I endured hours of instruction in Geometry, Algebra, Linear Equations, Discrete Analysis, and at least three classes with Calculus in the title. One branch of math that holds a special place in my heard, though, is Statistics.

What a great branch of mathematics statistics is. It is unmistakably the most bogus field of study ever, yet it is used by experts around the globe to shape every facet of our lives. Your paycheck, your insurance payments, your health care premiums, what TV shows stay on the air, and what flavors come in a pack of Starburst all depend upon statistical data. Considering its massive role in our lives, you would expect Statistics to make sense. You would be wrong. Here is an example given to us on our first day in class by the professor. Statistically, what is the most likely outcome of a flipped coin? The answer? Landing on edge. That is the average outcome of a fair coin toss. You think that’s bad? Statistically speaking, the average number of legs per human is slightly less than two. I could try to explain that it has to do with the integer nature of legs and the fact that there is a critical shortage of three legged people, but the main point is that we are dealing with a realm of mathematics that has a tenuous link to reality at best. And that’s when things are done correctly!

Once you start playing fast and loose with the language, you can start producing convincing statistics to prove practically anything. Let’s say you want to prove that cigarette smoking is good for you. Well, did you know that, while approximately 25% of Americans smoke, less than one 1% of people in hospitals are smoking? Clearly people who chose not to smoke are much more likely to be hospitalized, and thus smoking is healthy! All you have to do is ignore the minor detail that smoking is prohibited inside of hospitals. The words are weasely, sure, but the math is sound. If you are willing to tweak the initial sample after the fact, you can get even more remarkable statistics. Sure, 4 out of 5 dentists agree that a medium softness toothbrush is best, but if you leave off that last dentist then you’ve got a 4 out of 4 consensus, haven’t you?

Weasel words are cheating, though. Let’s get back to the real math. How about this statement: About half of the people in the United States are of below average intelligence. Sounds like we are a nation of idiots. I’m not saying that we aren’t, but about half of us are above average intelligence, too. That’s because “average” is an imaginary line drawn through the center of the total population. In a normally distributed sample set, half are above average, half are below. That’s just the way math works most of the time. Sure, if you sample 10 people who weigh 100 lbs and 1 guy who weighs 1000, then the average is around 182 lbs. That means most of the people are below average. Generally speaking, though, there aren’t too many human manatees throwing off the average, so usually saying that half of anything is below average is like saying blue things are blue. But it is statistics, so we take it to heart.

I haven’t used my statistics learnin’ that often in my professional life, and not at all in my personal life. The one thing I have used is potentially the most idiotic unit of measure ever conceived, the erlang. Let’s say you have a crap load of people answering phones, and you want to know how many craploads of callers they will be able to handle. (crapload is an industry term) You need calculate some erlangs. I won’t bore you with the details… well… with any more details, but I’ll tell you this. The calculation includes sums and exponents and factorials, and the units you end up with in this case are minutes per minute. If that makes sense to you, you are either a mathematical genius or you have serious mental problems. The fact that the two can be confused says about all that needs to be said about statistics.

Tags: ,