Posted by J Russell
I’ve definitely never written this before, and it’s likely of zero interest to anyone but myself. Still, that doesn’t usually stop me.
What I am suffering from when flying on an airplane, it seems, is an attack of subjective probability: the credence I give to the likelihood of any plane that I am on suffering some sort of catastrophe. Apparently, my credence should really take into account statistics on flying, which means that the whole business is more safe than houses. But, being awkward, or maybe simply more attuned to, and wary of, the hidden variables, I don’t. I take a very simple gambling perspective on flying: the plane will either get there, or it won’t. That’s a 50/50 chance.
It seems then, that in my fear of flying I start to take a Bayesian approach to probability, rather than a frequentist one; well at least until the hidden variables become too much for my brain to compute and I boil it all down to a binary sum game of survival: fly or crash, live or die, do or don’t. If everyone took this approach there’d probably be far fewer people in the air, but, amazingly there are still people willing to take far shorter NASA quantified odds on their survival.
Astronaut Chris Hadfield is a case in point. The first Canadian in space, he accepted odds of 9/1 surviving to go up in the space shuttle, and he did this more than once.