Author: Yngvi Bjornsson
Date: 09:56:23 07/13/00
Go up one level in this thread
On July 13, 2000 at 11:47:36, Masciulli Gianluigi wrote: >On July 13, 2000 at 11:41:21, Martin Schubert wrote: > >>I think it's better too improve a programs understanding of closed positions >>instead of avoiding closed positions. > >Yes would be better but is much more difficult to do! >I think the only doable way is to make the game "wilder" and more open. > >just my 0.02 euro. > >gianluigi Yesterday I attended a lecture by GM Joel Benjamin where he talked about his role in the development of Deep Blue. One of the fundamental design decisions (chess knowledge wise) was to make the program understand closed positions as opposed to have the program avoid them. Because this decision was made at an early state, necessary changes could be made to the chip design. The benefits of this were apparent in the famous second match game where the program willingly locked up the center with d5, but still played the position admiringly well (putting the rooks on files what could potentially open, played a f4 pawn lever, etc). On the other hand, implementing all this knowledge in software will compromise the search speed, whereas this is not (as much) of an issue in hardware. -Yngvi
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.