AI ...Some things are hard to achieve

Balvamar

I set myself an AI project to do over the Easter break, and had partial success, but ended up frustrated and feeling dumb
I didn't want to just copy someone else's ideas, as I know there are some topics on the forums about this. After trying several times to achieve my goal I have given up for now. I got on the net and read about AI in games, and found it is indeed a very difficult task, depending on what you want to do. It made me feel a little better knowing I had set myself a goal that was beyond me at the moment, as I am only a few months into learning game maker.

I was wondering if someone could set me a little homework task

If I wanted to learn how to develop AI for a game, what is a small problem I could tackle and have a better chance of success?

Not a whole game or anything, just a little prototype to think about. Any help is appreciated. Cheers

tatiang

It would be better if you gave an example of the type of AI you want to create. Are we talking about choosing the next move in checkers, attacking the nearest target in tower defense, or something else entirely?

Balvamar

I was attempting to have an actor move to a goal through a maze. I was using booleans to identify whether collisions were north south east or west.

tatiang

You might check this out then: http://forums.gamesalad.com/discussion/68703/intelligent-a-i-path-finding-with-gamesalad-has-anyone-ever-created-an-a-style-path-finder-setup/p1.

pHghost

And most importantly this: http://www.policyalmanac.org/games/aStarTutorial.htm

With many of these advanced features, you need to start from some established method, otherwise you end up burning your energy on reinventing the wheel, when it's there all along. Read up on A* pathfinding and once you understand the concept, then you can start figuring out how you want to make it work in GameSalad and/or make it more effective, which will be the original work you contribute.