Author: Michael Neish
Date: 19:18:00 02/02/00
Go up one level in this thread
On February 02, 2000 at 06:27:18, Jürgen Hartmann wrote: >I feel pity for the poor programs winning tournaments. They always get told how >meaningless this particular event was. Poor Stefan Meyer-Kahlen who only won >three completely negligible World Championship titles is my favourite example. I think the previous poster meant was that the number of games and the result were not sufficient to draw any conclusions about who is stronger. Of course all these tournaments are interesting to witness for everyone, and we are all glad that they are played. And they are good for computer Chess, no doubt about it. But when you start to say things like "Computer X seems stronger than Y" based on X's scoring several points more than expected, then you are walking down a dangerous path. And by "dangerous" I mean that you will be puzzled and confused later on when Computer X which did so well in this tournament, stumbles against a supposedly weaker opponent later on. Think about it. All programs are improving. I've read in some cases that new versions are supposedly 10 points stronger or 20 (if you're lucky) than the previous version. It takes a lot of games to measure this difference. You will not notice it in a single tournament. We are speaking different languages here. The tournaments are interesting from a Computer Chess point of view, or whatever personal reasons we may all have for being interested in them. But you cannot, really cannot say that any computer is stronger than any other based on the Cadaques results. After several more Cadaqueses, maybe, but by then there'll be new versions out and we're all back to the beginning again. If anyone wants to disagree with what I'm saying and take the results at face value, then of course it's fine. I'm just sounding a note of caution. I hate to be a wet blanket, but I prefer caution. Cheers, Mike.
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.