Author: Graham Laight
Date: 02:32:47 03/17/00
Go up one level in this thread
One way to calculate the quality of two equal chess players is the proportion of draws they score. Generally, I think I'm correct in saying that, the better the players, the more likely that their game will end in a draw. If I had time, I'd get a chess database, and try to correlate likelihood of a draw with playing strength. But if this assertion is correct, then when you get the point where the players cannot beat each other, the game is solved. Again, if I had the time, I would try to create a graph of the proportion of draws in computer games over time. It might tell us nothing - but it might show us where we are in relation to "solving" chess. If anyone would have a stab at doing this in an impartial way - even if it was only a "quick view" rather than a "rigorous study", I think that many of us would love to know the result! Maybe the SSDF database could be a good starting point? -g On March 17, 2000 at 00:56:19, Vincent Vega wrote: >On March 16, 2000 at 23:48:43, Hans Havermann wrote: > >>On March 16, 2000 at 10:41:56, blass uri wrote: >> >>>My program calculated an upper bound(ignoring things like side to move, >>>50 moves rule) and found 3.7010630121207222927827147741452119115968e46 >> >>>Retko v.tomic found a smaller number and I do not remmeber the number >>>but it was not less that 1e46. >> >>The estimate 64!/(32!*8!^2*2!^6) ~ 10^43 is given by Shannon in his >>seminal paper "Programming a Computer for Playing Chess", Phil. Mag. >>41 (1950) 256-275 (also in D. Levy's "Computer Chess Compendium"). > >Whatever the exact number is, it means that chess could be solved in 50-60 years >if Moore's Law holds. That's a big if, but it could happen even sooner if >advances like practical quantum computers occur.
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.