Thursday, August 7, 2008

The Drunkard's Walk

I just finished reading a book by Leonard Mlodinow called The Drunkard's Walk. I picked it up because it reviewed well in the New York Times and seemed to be an interesting examination of probability (An interest of mine, even more so since my picture book on the subject was released). I was unprepared with how impressed I would be by the book.

I soon realized that the book was almost as much about psychology (and neurology) as it was about probability. It was not just about the mathematics, it was about why the human brain has such a hard time calculating probability objectively. It was about the human need to see non-existent patterns in randomness, to let our biases unconsciously affect our perceptions, and to ignore relevant mitigating events when assessing the odds.

It was a call to reexamine our beliefs in an objective light, something almost impossible for people to do.

It included experiments I am very familiar with from my recent readings on neurology: for example, the fact that animals can predict what a randomly blinking red and green light will blink next better than humans, because we are trying to solve the question of a pattern, even when one isn't there. It even included the Rosenhan experiment--an experiment I have seen cited, oddly, in the three of the last five books I've read. Before that, I had never heard of that experiment.

Random chance, or an indication the experiment is entering the zeitgeist? Probably a bit of both

The Rosenhan experiment was carried out to see whether eight different people, with no mental illness except for the (fictional) symptom of hearing the word "thud," would be diagnosed as schizophrenic by a psychiatrist. Seven were--the other was diagnosed manic-depressive. Subsequent claims to no longer hear the word "thud" didn't change the bias. The bias had already cemented the diagnosis.

How does this relate to probability? Our own biases, based on events caused by random chance (great success or failure), help form our opinions, even when our conscious mind knows that the events were luck related. We are always seeking to find reasons why people deserve their success or failure, trying to create patterns where there is only randomness.

Mlodinow urges that we judge by ability, not by results. A difficult proposition, but definitely a worthy goal.


Kasheri said...

We are always seeking to find reasons why people deserve their success or failure, trying to create patterns where there is only randomness.

I feel this is never more clearly demonstrated than when a person with habits considered healthy becomes ill. If a person who "never" eats "bad" foods and who "always" exercises gets something like, say, cancer, people just can't process the fact. People can not fathom the randomness of the event: that illness is seldom the result of what you deserve but simply occur. People will become angry if you suggest otherwise. They will even go so far as to attribute failure in the purity of the ill person rather than accept the idea that something like cancer is random. I have always found this sad and troubling because it places such limitations on some people's compassion.

Theater of Ideas said...

Yes, I think that it gives people the feeling of control, i.e. if I act in exactly the right way, I will never get sick and also get everything I want. And in order to believe that, people then say, if someone gets cancer, that it's somehow their own fault.