Newsgroups: comp.ai.games
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!torn!watserv2.uwaterloo.ca!undergrad.math.uwaterloo.ca!descartes.uwaterloo.ca!tbannist
From: tbannist@descartes.uwaterloo.ca (Tyler Bannister)
Subject: Re: AI routines for RPG game?
Message-ID: <D5r357.F17@undergrad.math.uwaterloo.ca>
Sender: news@undergrad.math.uwaterloo.ca (news spool owner)
Nntp-Posting-Host: descartes.uwaterloo.ca
Organization: University of Waterloo
References: <D5E5oo.GDw@emr1.emr.ca> <3k29na$8g0@sphinx.Gsu.EDU> <SMISHRA.95Mar14120549@kiwi.acns.nwu.edu> <3k4r05$nj0@news.ccit.arizona.edu>
Date: Mon, 20 Mar 1995 17:55:55 GMT
Lines: 23

In article <3k4r05$nj0@news.ccit.arizona.edu>, hendrix@bigdog.engr.arizona.edu (Bryce Hendrix) writes:
> Sunil Mishra (smishra@kiwi.acns.nwu.edu) wrote:
> : In article <3k29na$8g0@sphinx.Gsu.EDU> gs01rmb@panther.Gsu.EDU (Rolf M. Bunchner) writes:
> 
> : That's probably what they currently use. The problem is that you have to
> : look at a long term strategy, which would involve trying to figure in what
> : your opponent will do.
> 
> : Gets quite hairy. Try having a look at some of the classical planning
> : literature, I don't have any references because it's not what I am really
> : interested in.
> 
> Computers can almost always see deeper into strategy than humans can. All 
> the computer has to do in this case is look 3 or so moves deep, the human 
> opponent will almost never look deeper, so the computer then can narrow 
> its options to 9^4 moves assuming 9 directions of motion allowed per move.
> 

	However in RPG games nothing is assured, everything you do has a chance of failure.  Plus there are many, many options.  The opponent could cast spells, could run away, could charge to the attack.  I'm sure that making a good A.I. for an RPG game is VERY difficult...

-- 
Tyler Bannister
tbannist@undergrad.math.uwaterloo.ca
