MIND READING

There's one other intriguing discovery coming out of this work, which has even the scientists baffled: with approximately 85 percent accuracy, the subjects, separated by the distance from Los Angeles to Texas, can guess whether they're playing against a man or a woman. They appear to be picking up on subtle clues in the interactions that the scientists themselves haven't identified.

So here was my strategy. In total defiance of the social norms that should incline me toward cooperation and trust, I pursued the single-minded goal of amassing as many points as possible. Recognizing that the more I invested the more money there would be for both of us to split, on each round I sent all 10 dollars to my counterpart, who routinely returned $16 (of $30) to me--just enough over half to keep me going.

That is, until the ninth round, when, I calculated, the other subject could come out ahead by keeping the whole $30. So I got there first: I "invested" zero. I did the same on the last round and cleared a hypothetical $148 ($16 times eight rounds, plus $10 times two rounds) to her (or his) $112 ($14 times eight rounds). And I pulled off one more coup: I figured out, correctly as it happened, that I was playing against a woman. I reasoned that a man would have been just as competitive as I am, and guessed that I was going to betray him on the ninth round--so he would have kept all $30 to himself on the eighth round. At least, most of the ones I know would have, although maybe a sample consisting mostly of journalists isn't entirely representative. Out of such tiny insights, scientists are constructing a model for some of the most intricate and sophisticated decisions a fully evolved human being can face in the modern world. And maybe, in some small way, if Camerer and his colleagues are right, making the world a more trusting and cooperative--and peaceful--place.

WITH MARY CARMICHAEL

© 2004