Author: Dann Corbit
Date: 08:50:22 06/22/99
Go up one level in this thread
On June 22, 1999 at 11:09:51, Harald Faber wrote: [snip] >OK, so the final decisive question: Are chess programs allowed to participate, >and if so under which conditions, in "normal" human tournaments? To my way of thinking, even that condition is not strictly necessary. It should, however, be a real tournament of some kind with something significant at stake. Ed's contests with Rebel will qualify in my opinion. The reason I think that there has to be a condition like this is that people may play very differently when there is nothing at stake. As an example, you may have a Winboard program running on the internet. A GM plays 20 games against it and loses 12. But it turns out, she was really just testing out a new opening and looking for tactical holes which your computer located. So the experiment was not testing what you thought at all. When there is something at stake, and the person's reputation is on the line, they are much more likely to try their hardest to win. Once 30 games have been played by a single program on known hardware under these conditions, I think all reasonable doubt will have been removed for that program.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.