## Saturday, September 27, 2008

### Lies, Damn Lies, and Statistics

I've always been a bit of a numbers man. (no, not those numbers) Using simple algebra I used to prove to my math students that 1 equals 2.

Using facts and statistics you can get away with a lot. And there is good reason why there is an old saying that says "Liars figure and figures lie". Mark Twain said it differently, “There are three kinds of lies: lies, damned lies, and statistics."

Much like people, if you torture numbers, they will confess to anything.

Here's a simple example:

"Let’s look at a typical scenario of statistical lying. Say that you live in an area where the housing market has taken a beating to the point that no one is building homes and no one is buying any existing housing, and to make matters worse, asking prices are down some 50% off of their peak.

Just when you thought things were totally down for the count, some fool falls off of the turnip truck and decides to build a new house. Since there had been no other building permits issued thus far in 2008, the headlines would appear as; Building Permits Rise by a 100%; as a home building craze returns to Mayberry."

But sometimes the numbers lie in much more sublime way. Here's a good example:

What reduces emissions more?

A. Someone swapping their old SUV (which gets 12 miles per gallon) for a hybrid version (18 mpg) or

B. someone upgrading their 25 mpg compact to a new 46 mpg Prius?
(ignore for a minute manufacturing issues or driving habits and assume the miles driven are the same).

The surprising answer (for those who don't work it out) is A.

It's easy enough to see why this is the case. If the driving distance is 100 miles, then for case A the saving in fuel used (and hence emissions) is 100/12-100/18 = 2.8 gallons, while for B, you have 100/25-100/46 = 1.8 gallons.

The confusion arises because people like to think linearly about numbers, not inversely, and so tend to assume that a similar change in mpg has a similar impact on fuel usage. This is not however the case - improvements in efficiency at the low end of the scale are much more useful at reducing emissions.

This is actually a very general point - when trying to raise efficiency it is always sensible to start with the least efficient processes."

Of course, you could also make the point that the SUV driver should buy the Prius instead.

Numbers are remarkably subject to shaping. If you came in second in a two man race you were also last. The difference I suppose is your mood and your subjective view of things.

Last night's debate numbers are apparently showing that Obama did better than McCain. I didn't see it that way as I watched it. But, today, after seeing that McCain never once looked at Obama, that he withheld that gentlemanly courtesy in some kind of weird psyche-out ploy, I now see why so many CNN viewers saw it differently.

Obama looked competent and unruffled.

Obama's numbers are looking better and better.

Let's hope they are, but

Let's do every thing we can to make sure,

that the Liars, and the Damn Liars,

don't have an opportunity

to make him a statistic.

Labels: