Author: Aaron Tay
Date: 01:49:40 03/22/02
Go up one level in this thread
On March 21, 2002 at 15:23:38, Dann Corbit wrote: >On March 21, 2002 at 15:13:12, Steve wrote: > >>Thanks to all of you for your responses. Is it possible to have programs >>recognize "fortresses" by instructing them to decide, when the evaluation has >>remained static throughout a certain number of ply, that the position must be a >>draw? It seems rather pointless to add this though in cases where you already drifted into a drawn position. I suppose for some programs, you can see some of the scores dropping, as the 50 move rule starts to come closer, those it's already useless in cases when it's drawn.. >Hossa has done some work on detecting walls. I think that fortress detection is >a good idea too. The problem is that I don't know of a single algorithm to >detect such a thing. A lot of times, we simply recognize a fortress. But the >question is "How did our brains arrive at the 'stalemate' conclusion?" [D]8/5K2/8/8/5N2/4p3/3qN2k/8 b - - Easier to solve i think, the human sees that the queen alone cannot mate. And the king cannot move. Any attempt to sac the queen draws.. [D]8/8/1q6/1kp5/2p5/2P5/KP6/4R3 w KQkq - Hmm this one is harder. I think the human sees the queen alone cannot win. Yet bringing the king forward doesn't "help"...
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.