Sign in to follow this  
HexDump

Tips building IA on games like Yugi-oH, Might and magic, etc...

Recommended Posts

Hi all, I´m in the process of building the IA for a game that more or less is the same that the ones listed in the topic. I have little IA knoledge (well... I have done some platform enemy IA, tetris, etc... basic stuff), and I don´t see how to create good enemy players IA for a game like this. There are thousand of possibilites. I have thought about using a Neuronal nertwork but I should let it play lota time to get good results (or this is what I think). Could any one give any tip, insight, etc... on this? Thanks in advance, HexDump.

Share this post


Link to post
Share on other sites
Not to be a wise-guy or anything, but you are refering to AI (Artificial Intelligence), right?

Usually the rule of thumb is to make it look more advance than it really is, remeber - the illusion of intelligence is just as good as the real thing. :)

For instance, when the CPU has found another player:

if stronger than player, attack
if weaker than player, run away.
if equal strength, move closer to reinforcements.
etc...

The key is to add random elements to it, so that you really can't anticipate what the CPU will do. You will probably find that the simplest possible solution could provide just as much fun as a complex neuron network AI.

Hope it helps!

/Robert

Share this post


Link to post
Share on other sites
Hi Rasmadrak,

you´re right I´m refering to IA (AI in spanish :)), sometimes I misspelled.

Well, what you said is really true, but I see thousand of possiblities in this type of games, so... I´m a bit overloaded :D. This is why I asked for some help.

Anyway I have to think more about it,

Thanks in advance,
HexDump.

Share this post


Link to post
Share on other sites
I dont know how Yugi-oH works, but I did work on the Might and Magic series, if it's the PC game your talking about. For games with such complex simulations ( and they were very complex, with spells, counter spells, damage modifers, 2ndary spell effects, etc..). If i remeber correctly the way it worked was it actually ran the simulation engine and did a search of possible moves and ranked them. Essentially doing an A* search and at each step the state of the battlefield was tokenized and the simulation actually ran to search for the next best moves. Don't know how deep the search wen't, and I'm sure they wrote some huersitic for generating moves, so they can avoid totally incorrect ones.

This was ofcoruse expensive CPU wise, but with todays machine you can do a much deeper search and potentially have stronger AIs.

Good Luck!

-ddn

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this