Wednesday 22 May 2013

getting to bayes

There’s an instrument of disabuse for everyday assumptions and likelihoods that I had not heard of before called Bayesian Probability, after its proponent 18th century English poly-math and minister, Thomas Bayes. Intent on rescuing providence, rationally, from chance, Bayes championed a sort of inverted inspection of odds, imploring people to look to prior arrangements and question how the deck may be stacked and weighted in favour of certain outcomes. Although modern interpretations of Bayes’ thinking maybe over-reach his original context, the notion that probability—writ large and scientish, is based in part on belief is not something merely synonymous with gullibility and naivety and magical-thinking.

Rather, how we measure the likelihood of outcomes can hinge on personal experience, and maybe to a fault, since successes and failures (surprises and dis- appointments, too) are counted by past usefulness and go unnoticed and with indifference otherwise. Sometimes it’s an over-simplification to believe that the chance is 50-50 since we are better acquainted with either something working-out or not and not something in between. Something about the way we pose the question or prime the conditions may obscure our judgment. We are also, sadly, more accustomed to failure than success. This is a bit revelatory and makes me wonder what misguided influences might be tarnishing my choices—not that perfect and logical decisions seem all that savoury either as an alternative. I am remembered to something along these lines whenever I play the lottery but also know that though a long-shot, one only needs to be right once.