Author: Andrew Slough
Date: 17:54:29 06/14/99
Go up one level in this thread
On June 14, 1999 at 19:45:09, Laurence Chen wrote: >On June 14, 1999 at 19:03:16, Francis Monkman wrote: > >> >>On June 14, 1999 at 18:56:06, Laurence Chen wrote: >> >>>What is the point of your post? I don't understand what message are you trying >>>to bring up. What is Nunn trying to say in his annotation? Is the move 16. Rf1 a >>>very good move or a very bad move? >> >>The point of posting this game is to remind (if necessary) that chess is about >>more than >>'strongest move versus strongest move", ie ping-pong. Chess is fundamentally >>about ideas, and although computers are very good at 'getting the idea' when >>it's there to be got, they are not so good at creating the conditions. I think >>you'll find that most/all programs are as amazed by this game, as John Nunn was. >>His point about move 16. Rf1, is that although it turns out to be best (leaving >>Black with just slight advantage), any player (and I would add human or >>computer) would have had to be very far-sighted in order to choose it. >> >>Francis >That's why opening books exist for both humans and computers. Why bother >re-invent the wheel if all the information is available? I would disagree with >you when you say that computers are not yet able to create conditions, it all >depends on the type of position. I find that Fritz 5 is very good in creating >conditions in open games, and I yet would like to see any human GM outplay Fritz >in an open game. Of course, the human GM would use anti-computer tactics, that >is, closed position with long range planning to outplay the computer. It is how >one uses the tool or weapon which wins the game, not the weapon itself. >Laurence Of course, the point is that the opening novelty (on whichever move it was) was not in Nunn's `book' and nor would it be in a computers book, as computer books are based on human games. Computers would be just as vulnerable as Nunn was, if not more so. I think the reference to not being able to `create the conditions' refers to alpha-beta based programs lack of ability in long term planning. In open games, there tend to be less concrete planning and more tactics and hence computers tend to get better results. I think your faith in Fritz's ability to outplay `any human GM' in an open position is misplaced. I always marvel at Bronstein's manhandling of supercomputers in open positions at previous Aegon events. I think that many GMs have shown that it's quite possible to outplay a computer without going for closed positions with long range plans (not as easy to achieve as it sounds anyway). Not least is GM Rhode's recent win over Rebel. Fritz still makes some positional mistakes because of the way it is written (quick/simple evaluation at terminal nodes). Andy
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.