On lies, damn lies and statistics

I made a promise to myself not to mention the election. I really don't want to thrash that dead horse and jump on the analysis bandwagon, or use another cliche in the next few lines. It seems so many were taken by surprise by the results, but really hasn't this happened before? Polls were taken, surveys composed and filled, people questioned, numbers crunched, and, well...

Can statistics be trusted? One fellow, an academic, it is said, got it right, and accurately predicted the outcome. To a former academic such as myself this is no surprise. The creation of those little figures and lists of data are the life's blood of so many academics. They can be used to pretty much prove anything, and I once joked that I could prove that the earth was flat. Some of my former colleagues have no sense of humour, but after a number of cans of disgusting Japanese canned coffee I was able to satisfy my critics. Japanese coffee is pretty toxic stuff, so I can't recall how it was done, and no argument is worth drinking large amounts of that wretched liquid again to relive the moment. However, I stand by my conviction that numbers, so innocent and pure, and so meaningless when standing alone, can be used to deceive, manipulate, and confuse when attached to information.

Statistics themselves are just numbers attached to some kind of meaning. They are no more deceptive or corrupt than the concept of "big", "small", "heavy", "light", or "here" and "there". When they are interpreted, or played with to illustrate a point, or to add validation to an argument, then their sweet, innocent nature undergoes a radical change. (A lot like myself after imbibing vast amounts of Japanese canned coffee).

These otherwise cute and informative sets of figures can become monsters, or, in some cases, just plain stupid.

While numbers and raw data can be quite pure, we humans are not. We all have agendas, and most humans have a tendency to want to surround ourselves with the like-minded, or change those around us to our way of thinking. When the poll results and election result predictions were coming out, who was questioning them? Far too much faith is given to a bunch of numbers said to represent some kind of reality. I could go and interview 100,000 people across Australia from different socioeconomic and age brackets, and discover that each and every one of them supported the (fictional) Yellow Party.

That I was questioning rank and file members at local chapters does not need to be mentioned. Of course, to make the figures sound a little more "realistic" I would include a couple of (fictional) Brown Party chapters. Viola! The Yellow Party must be the more popular group, or the better group, or the one providing the best leadership, or whatever it is that I am trying to prove.

After all, who is questioning all the statistics we are bombarded with constantly. Of course, the methods employed to prostitute reality are a little more sophisticated, but not much.

Of course with political elections, such things as mandatory voting, gerrymandering and preferences, not to mention how as a nation we like to support the underdog and hate being told what to do, all play a large role in what election outcomes will be.

I may one day run for office myself. Of course you can trust me, I know how to manipulate numbers to prove anything.