Black Swans

I read an interesting interview with a statistician in New scientist. The basic premise was that humans are good at predicting straight linear type probabilities – like a die roll or coin toss but hopeless at non-linear predictions. He didn’t call them non-linear events but dubbed them as black swans. I guess he had addressed lay people before and seen eyes glaze over.

He sited several examples of how hopeless we are at predicting these events. The examples included economists hopelessly predicting markets (oil price and interest rates to be specific), the outbreak of war and the importance of the computer/internet. He also talked about when some one predicts one of these events. His explanation is that there are all sorts of people making all sorts of outrageous predictions – only the ones who fluked the right answer gets remembered: the thousands of misses are ignored. I would have to agree with that point.

Initially I was impressed with his logic and argument. I then reflected on my education and experience with non-linear control systems and though duh.

He did make one very good ponit and that was that when some one gives predictions of interest rates or the price of oil or whatever they should also give a confidence level on the number they predict. Now that would frighten the crap out of everyone! Imagine a government setting monetary policy when the basis of policy has an extremely low likely hood of being reality.

So anyway this got me thinking about some of my predictions of gloom and doom about the effects of climate change and peak oil on my circumstance. May be I am hopelessly wrong; I am trying to predict non-linear outcomes. Random factors and other unknowns may make me one of the erroneous predictors. There is the famous example of the late 1800’s where it was predicted that our cities would be neck deep in horse shit by the 1950s. Mind you I have no source for this prediction and suspect it may have been made up by some one who was trying to protect their right to pollute.

I certainly hope my predictions are hopelessly wrong, that some unexpected technology or event changes our current course. Mind you booms, busts and cyclic hardship are a long tradition of our species (and life on earth in general) so we are in for a hiding. I only hope the when is outside my timeline.