From kragen@dnaco.net Fri Aug 21 16:28:26 1998 Date: Fri, 21 Aug 1998 16:28:25 -0400 (EDT) From: Kragen To: leonf@magnolia.net Subject: Re: SOCIETY'S EFFORTS TO CORRECT THE PARADOX Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Keywords: X-UID: 1330 Status: O X-Status: In http://www.magnolia.net/~leonf/sd/xvp-6.html, you say: Not only do we have trouble estimating the true probabilities, but we, in fact, tend to be biased in our estimates. Humans, if given a choice, tend to take a sure thing over a probable expectation even if the expected value is higher for the probable event. See this essay on "Decision Making" for a discussion. (btw, the link to "Decision Making" takes me to www.psych.nwu.edu, which redirects me to pangaea.pratt.edu, which doesn't answer at the moment.) This may, in fact, be rational under some circumstances. For instance, I'd much rather have an assured monthly income of $2,000 than a one-in-a-thousand chance each month of getting $2,000,000 (assuming I have no other income.) I think this is because my utility function is decidedly nonlinear -- the utility of $2,000 a month is much, much more, perhaps 1,000 times more, than the utility of $0 a month. ($0 a month would probably kill me in less than a year -- again, assuming I had no other income, and I'd probably finagle some somehow. Assume for the purpose of the argument that I agreed in good faith to forego any other income in return for the chance at the $2 million, and would keep to that agreement even unto death. On the other hand, I can live comfortably for several decades, perhaps even a century, on $2,000 a month.) On the other hand, the utility of $2,000,000 is perhaps 10 times the utility of $2,000. So my expected utility from the monthly one-in-a-thousand $2,000,000 is maybe a hundredth my expected utility from the assured monthly $2,000, even though my expected monthly income is exactly the same. (Not that I'm arguing that people are perfectly rational -- just that what appears to be an irrational choice when analyzed on the basis of expected payoffs may, in fact, be a rational choice when the nonlinear relation between payoffs and actual utility is taken into account, and expected utility is analyzed instead of expected payoffs.) Kragen -- Kragen Sitaker We are forming cells within a global brain and we are excited that we might start to think collectively. What becomes of us still hangs crucially on how we think individually. -- Tim Berners-Lee, inventor of the Web