Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: What constitutes human-like play?

Author: Peter Berger

Date: 06:06:17 01/17/03

Go up one level in this thread


On January 17, 2003 at 08:53:01, emerson tan wrote:

>How can you tell if a program plays like human? Rebel and Hiarcs were claimed to
>be one of the most human-like programs but I dont know how to determine. I
>estimate my elo rating to be only around 1500-1600. Maybe it would be  better to
>ask the Grandmasters.
>
>Which among the programs were praised by Grandmasters as Human-like? Gulko
>played Fritz, Junior, Shredder and Hiarcs if I remember it right, which of these
>was he impressed the most?
>
>To me, human-like play is able to formulate a plan, play for a theme, so far I
>havent seen that in programs. The programs still play within its "field of
>vision"

I think it works better the other way around. What is a typical "computer move"?
A move extremely unlikely to be made by a (strong) human, often looking very
ugly or pointless. A good example is Bf8 from the second game Kramnik-Fritz -
there are many other examples.

A program that makes such moves very rarely could be thought of as more
human-like.

Of course the ideas what is a pointless computermove will be very different
depending who the judge is and his/her understanding of the game.

Many standard plans can be found in computergames, too. Alhough they don't
really plan their play can look logical and instructive. For most programs it is
quite easy to find extremely strange and illogical moves in other games though.

IMHO most of the human-like descriptions when it is about current chessprograms
are marketing hype. I don't see where Rebel or Hiarcs play more human-like than
Fritz or Tiger. Just good marketing I think :)

Peter



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.