Tuesday, September 14, 2010

Uncertain Odds on my Ambiguous Ignorance

So . . . as I've been beating you over the head with so far in this blog, we are a predicting, gambling, prognosticating species. We are compelled to try to determine what will happen in the future (and we are bad at it). Today I want to write about different types of future randomness, those of uncertainty, ambiguity, and ignorance (or to use dated political analogies, Paul Martin, John Kerry, and George W. Bush).

Let's start with uncertainty. An uncertain situation is one in which we know the potential outcomes and their probabilities, but we don't know what will happen. Think of a card or dice game; using math you can determine the odds of every possible outcome. If you roll two dice, there is a one in 36 chance of getting a total of two. Uncertaintly is extremely useful in this regard (but not in much else - more on that later).

Ambiguity is just as it sounds - you are ambiguous about the probabilities. You have some knowledge of the outcomes and probabilities, but not complete knowledge. One example would be a sporting event. Before a hockey game (assuming you know a little about hockey) you have some idea about who is more likely to win, but you could not put specific odds on it, no matter how sophisticated your math. Bookies try to, but these are based on past performance, thorough knowledge, and intuition. It would be impossible to say with certainty that one team has a 57% chance of winning, partially because it would be impossible to assess whether that was correct unless the same game was played the same way multiple times.

Ignorance describes a situation in which you don't know the outcomes or the probabilities. Imagine trying to guess the outcome of a hockey game if you had never heard of hockey before. You might expect that there would be a winner, but not even that is certain (ties, overtime losses, etc).

Our natural tendency as humans is to select situations where there is uncertainty as opposed to ignorance. We don't like to feel dumb. Ideally, we'd like certainty, but no one has certainty in their life (and trying to find it leads to other social problems, like being certain your religion is correct or certain that Mike Myers would have a long and illustrious film career).

The problem with all of this is that we treat most situations as though they only contained uncertainty, whereas in most real-world situations ignorance rules (and I'm just talking about future randomness here, not even what is, or isn't, in people's heads, though that statement could still be correct). According to Nassim Taleb (author of The Black Swan, what, you haven't read it yet? Go now!), the financial meltdown was because there were a lot people mistaking ignorance for uncertainty. Part of ignorance is blindness to possibilities; many of the financial models used ignored low-probability, high impact situations.

In other words, we treat situations in our day-to-day lives as though they were a card game. If I go to a job interview, I would look at it as we either get the job or not. But I may perform so poorly that my reputuation is sullied and word gets around the industry that I am not to be touched. I may perform so well that I get a better job than the one I applied for. Both are of very low probability, but possible. It is impossible to think of all the possible outcomes (go on, try - I bet you didn't think of the one in which the interviewer is a carnivorous alien who eats you - ok, maybe that one isn't possible). Yet we behave as though we know them.

Next time you think you have a handle on future outcomes, think of Lindsay Lohan. No one knows what she's going to do next. Anything is within the realm of possiblity. Now apply that to your own life. Scared yet?

No comments:

Post a Comment