Here�s the url or the DirGames-L list. It�s an excellent resource.

http://nuttybar.drama.uga.edu/mailman/listinfo/dirgames-l

Mikael Wir�n wrote:

> Steven Sacks and Agustin Maria Rodrigues wrote:
> > DirGames-L would be a more appropriate forum to post this, but
> > you should know that AI is not easy to program, and nobody is
> > going to be able to give you a canned script.
>
> Thanks, you learn someting new everyday. Been using this list for the last 4
> years, i guess you get a little blind to the outside world. Didn�t know
> there was anything like dirgames-L. Where do you sign up?
> Yeah, belive me i know too much about the problems of fuzzy logics and AI.
> Thats why i phrased my question as i did. I�m not looking for premade code,
> i�m looking for peoles thoughts of where you want to draw the line between
> the computer making the best desicion and making a mistake.
> > There are many different forms of AI.  Some are more advanced
> > than others, some are much harder to program than others.
> > The problem with simple AI is that it is fairly easy to figure
> > out how to make the computer play a certain way every time,
> > therefore you can take advantage and win every time.
> Since it�s a turnbased tile game. It won�t be too hard programing the
> functionality of the (pseudo AI). It will have three different responses,
> move, attack, surrender. The action will be based on 1. Status (shield % and
> hull %) 2. estimated damage to the opponent (shield % , hull %) 3. range a.
> to the opponent b. to the edge.
> > Randomizing responses will end up with the computer making
> > stupid decisions, like retreating when it has 90% health and
> > the player has 10% health.
> Exactly, Thats why i�m asking for your experiences in the field. I�ve only
> made logic AI, learning to do estimates based on previous experiences in
> LISP, never something like this. And definitely not in director =)
> > One approach is to make the computer as infallible as possible,
> > and then tone it down from there.
> Yeah, this is the approach i�m taking. The user sees the whole area and can
> decide from that on what action to make, so the computer will have the same
> means. But if i leave it at that it will make THE best move possible in
> every situation. That won�t mean it is ubeatable, since hitting will be
> somewhat based on chanse and stats. But it won�t be fun to play. So i need
> to put in some random actions based on something logical to make the
> computer hesitate, run away, take a longer route or something like that.
> > A good start is to do a search at google or hotbot on programming AI.
> Done that, been there, seen that, written a paper on that =) But it doesn�t
> make me any smarter on where to draw the line between smart and stupid =)
>
> Thanx for your responses
> /Micke
>
> [To remove yourself from this list, or to change to digest mode, go to
> http://www.penworks.com/LUJ/lingo-l.cgi  To post messages to the list,
> email [EMAIL PROTECTED]  (Problems, email [EMAIL PROTECTED])
> Lingo-L is for learning and helping with programming Lingo.  Thanks!]


[To remove yourself from this list, or to change to digest mode, go to
http://www.penworks.com/LUJ/lingo-l.cgi  To post messages to the list,
email [EMAIL PROTECTED]  (Problems, email [EMAIL PROTECTED])
Lingo-L is for learning and helping with programming Lingo.  Thanks!]

Reply via email to