Friday, October 16, 2009

Guessing

Risk is purely a probablistic concept. Probability requires a significant number of events. Playing poker once or for one night or for your life's saving is not a risky proposition. It is an uncertain event. The probablities associated with the game of poker are not informative if you're playing for your life's saving. The "right" play, the play with the highest expected value, doesn't mean anything if you lose. It doesn't mean anything if you win. All that matters is winning or losing.

My example is not a very good one, and I will work on a better one. But there are two ideas here. First of all, risk has no meaning unless you're dealing with a repeatable and replicated event. I would argue we haven't had enough service economy recessions to understand the risks associated with our economy. I would argue we don't know the risks involved with a public health care system in the U.S. Second, "risky" thinking is dangerous when one explicitly considers uncertainty. People who try and predict recessions will mislead people, because they are thinking about the problem in a fundamentally wrong way. People who discuss a public health care program in terms of outcomes and events will mislead people, because there is a fundamental error in there thinking. We don't know will happen, and we can't argue in probability terms.

I am not saying what I want to say, but my final point is that assuming and proclaiming that you know the distribution of outcomes when you really have no idea is (might be) dangerous and wrongheaded. We (some times) need to admit when we're guessing.

Poker isn't the best Risk, the idea that events can be broken up into chance

No comments: