Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: What is the thinking game that gives programmers more money?

Author: Tord Romstad

Date: 00:22:32 06/10/02

Go up one level in this thread


On June 07, 2002 at 15:37:59, Russell Reagan wrote:

>Go programs have the potential to make more money I suppose, but currently I
>don't think anyone would buy a go program because currently go programs are a
>joke. An amateur novice level go player can beat current go programs. They are
>very weak.

This is somewhat exaggerated.  A novice would have to study at least a few
months before (s)he is able to beat the best programs, unless (s)he is very
talented.

>From the opening position of a game of chess, there are 20 options for white.
>Using alpha-beta and good pruning techniques, maybe you can get a good branching
>factor. For a go program, there are 361 options from the opening position. Using
>alpha-beta and whatever pruning techniques you'd like, you aren't going to get
>any kind of managable branching factor.

This is not the only problem.  Another (perhaps even more important) reason
that go programming is more difficult than chess programming is that it is
extremely difficult to design a good evaluation function for go.  In chess,
counting material and adding some positional bonuses is a good start.  In
go, there is no similar approach.

>What irritates me when discussing this topic with go players is that go players
>seem to think that there is something innately more difficult about their game.
>In other words, they seem to hold the attitude that their game takes more mental
>capacity, or one has to be more intelligent to play go or something, when the
>reality is that go is a much simpler game and the only thing making it hard is
>it's incredibly high branching factor when trying to analyze variations.

It is not the only thing making it hard.  Go positions are also much more
difficult from a strategical point of view.  I don't see anything in go which
is simpler than in chess, with the obvious exception of the rules of the game.

>I think it is very likely that even the best go players in the world play
>horrible when compared to a program that could solve go. I think in chess it is
>not so. I think that the best chess players in the world would not make complete
>fools of themselves when playing a perfect playing chess program.

This is absolutely correct.  Of course, the above paragraph summarizes
precisely what your go playing friends mean when they claim that go is
more difficult to play than chess.

Tord



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.