Author: Gordon Rattray
Date: 10:27:28 09/20/01
I've been wondering what level of chess could be achieved by using a combined effort of computing resources? i.e. for a given amount of hardware (separate machines) and various chess software, and some sort of time limit, is there a best strategy for making use of this? For example, imagine a grandmaster challenged this discussion group to a correspondence game. How could we jointly use our range of software and hardware to best effect? I'm making the assumption that we'd use little or no chess knowledge of our own. Instead we'd reply on our computers and knowledge of computer chess - however, I do admit that in practice it would be impossible not to be influenced by human chess knowledge to some extent. I guess some aspects of this were seen during, e.g., the "Kasparov vs The World" game. However, I'm thinking of more of an emphasis on computer involvement and instead of voting, the group would work more closely together. For example, the analysis could be distributed in such as way as to try to minimise duplicate effort. Maybe people could be given sub-positions to analyse with results being posted back?! etc. The issues arising would include: which programs to rely upon for a given type of position; how to resolve cases where evaluations differ significantly; how to make best use of all resources; best use of time; etc. I'm not suggesting that this would be easy or fun to do in practice - it may be tedious and difficult to coordinate. I'm just curious and considering theoretically as to how good this combined effort would be. How much better would it be than, e.g., one person using one computer with one fixed program, analysing the root position for the time period? Could "CCC combined" beat the best correspondence players? ;-) Thoughts? Gordon
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.