Author: Serge Desmarais
Date: 17:07:17 08/29/98
Go up one level in this thread
On August 29, 1998 at 10:58:13, Robert Henry Durrett wrote:
>Suppose you are playing chess against your computer and you have just reached
>the point where you are out of the computer chess program's tournament book.
>Suppose, too, that you have not handicapped the computer in any way, so that it
>will play it's strongest moves, limited only by the time limit settings.
>Finally, also suppose you have set the computer to play at your favorite time
>controls, such as standard. If you are playing against Fritz5, that means you
>are in "tournament mode."
>
>So, what will happen? The odds are, if you are not playing at the 2500 level or
>better, you will be mercilessly crushed! And, if the software has a voice
>feature enabled, as in Fritz5, the computer will "rub it in" with some very
>irritating snide remarks. Sound familiar?
>
>But you will not be intimidated! You analyze the game after it is over and find
>your first mistake in the game. You make the computer return to that move where
>you first erred and you take back your move, making sure that nobody notices.
>You then play the game from there. Then what happens? Chances are, you get
>crushed again!
>
>But you persevere. You keep on taking back your moves and playing from the
>position where you changed your move. After many defeats, you eventually get to
>the point where you have drawn the game or won it. Truly, the computer is
>completely helpless against this strategy. [True????]
>
>Now, you have an impressive line starting from the position where the computer
>first went out of book. If you memorize this line, will you then be ready to
>amaze and baffle your friends with a "show and tell"? Not yet!
>
>Before you invite your chess friends over, you check your line to make sure the
>computer will not vary, or deviate from the moves it played earlier. Well, if
>the amount of time you take for each move is less than before, chances are the
>computer will not have as much time available and may not play as well, and the
>computer may therefore make unexpected, but worse, moves than it had before.
>Clearly, if this happens, you are will not yet be ready for your show-and-tell.
>Too bad. Back to the drawing board.
>
>But, after more trials, you finally get to a line which can be repeated over and
>over. You will take almost no time to make your moves, so the game situation
>will be repeatable. You are now ready! You invite your friends over to show
>them your new chess program.
>
> You have deliberately set the computer to play from it's tournament book, so
>that the computer will play it's "best line." You thus know in advance exactly
>what moves will be played by the computer while it is in it's tournament book.
>That way, you have guaranteed for yourself that the game will reach the position
>you had studied privately in secret.
>
>Your friends are all sitting around your computer, ready for your demo. Just to
>introduce the software to them and to show them how easy it is to win against
>it, you quickly run through your game, seemingly playing "snap moves" almost at
>random, and lo and behold, you draw or win!!! And, . . . it looked so easy!
>
>Then you encourage one of your friends to try it. Naturally, your friend gets
>crushed! One by one, each of your friends try it and they ALL get crushed.
>They finally all go home, shaking their heads, amazed at how easily you drew or
>won against that software. Now, they are all "hooked" and will purchase copies
>of that software for themselves. [Sound like a good marketing strategy?]
>
>Sound wonderful? Ready to try it yourself?
>
>THIS BEGS THE QUESTION: How could the chess software designer defeat this
>strategy if he/she wanted to? How could more variety [or randomness] be added
>to the middlegame without significantly reducing the computer's middlegame
>strength?
First of all, that is what the "book learning" is for. So after, with all
your takebacks, you manage to win/draw, the program will NEVER EVER reach that
same position when it was out of book in your game. Other forms of learning also
prevent it since the applies once out of book ("position learning" and all.
That is why, in Crafty for example, you can have the computer only play ONE
defence with the same line over and over and as long as you draw or win, the
computer's play will always vary. It will always try something new from game to
game (the new try could be worse than the preceeding one, but will at least be
different!).
Serge Desmarais
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.