Pattern Recognition

A few rambling thoughts on the mind of a gambler and pattern recognition.

Back in the mid-’60s, I attempted to write a heuristic checkers- playing program on an IBM 7040. (A 7040 was a bit more expensive than a New York Upper East Side townhouse and a bit slower than the chip that controls my dish washer.) I gave a lecture on heuristic programming at the University of Pennsylvania when I was 17, but really had no idea what I was talking about. One thing that I did learn — using pattern recognition with a limited set of examples can result in very poor conclusions. Humans are famous for making poor decisions based upon limited or selective pattern recognition. People have a predisposition to believe that which they would like to be true. Look at the traffic jams generated when someone spots the image of the Virgin Mother in his oak tree.

The human mind also has a protection mechanism that can interfere with decision-making based upon pattern recognition. The mind is capable of remembering both good and bad instances. However, the mind normally can remember only the actual physical sensation of pleasure. It cannot remember the actual feeling of pain nearly as well. People who have the additional capacity to remember actual pain often drive themselves mad. I think that this is why, after a Poker game, there are no losers. Ask each of the players post-game. Of five players, two will say they won and three will say they broke about even. But during the game, they will all complain about the cards and that they cannot draw a good hand. Where did the money come from?

Human pattern recognition is also influenced by a person’s outlook on life, optimistic or pessimistic. I get about the same number of comments from users of my software that my random number generator is biased toward negative counts as I get that my random number generator is biased toward positive counts. Or that the dealer busts more often than a real dealer, or less often than a real dealer. This is selective memory resulting in recognition of patterns that do not exist.

 © 2009 Norman Wattenberger