Sign in to follow this  
  • entries
    74
  • comments
    187
  • views
    34005

Intelligence in Gaming?

Sign in to follow this  
NickGeorgia

83 views

I come from a background in Electrical Engineering (EE). When I say that to someone I meet, he/she often says something along the lines of... "Great!, you can help me fix my VCR!" or "My computer is malfunctioning, can you get it working again." Sure, I like to pop open a toaster and reconnect wires just like anyone would, but what these people don't realize is that EEs deal with more than just circuits and wires--they also deal with information.

My speciality happens to be in intelligent systems which deals with problems associated with information and controls. Often techniques such as clustering, neural networks, data mining, data fusion, A*, fuzzy logic, fuzzy-neural networks, etc. are potiential solutions to problems such as how to optimally control a mobile robot or UAV, image compression, health monitoring, search engines, etc. The list goes on and on... A favorite saying is: "If you beat on data long enough, it will confess to anything."

You may ask yourself, what does intelligent systems have to do with gaming? Some techinical types of problems involve optimally rendering images to the screen so that they are pleasing to the average eye, or possibily removing insignificant polygons from a model to improve performance, etc. Consider other types of intelligence injected into a game such as when sprites automatically hide behind cover or flee from a battle, a tic-tac-toe game which learns from it's mistakes, the computer's automatic moves in a Mortal Combat game, Sim City, Madden NFL, etc. These are examples of adding intelligence to the system. The computer makes decisions based upon the state of the game and the chosen algorithms. How we implement these algorithms will determine the intelligence level of our game. Sure, a game can look great and sound great, but if it isn't challenging or... interesting, how long will it hold the player? We would assume that the algorithms we select would not make the game so challenging so that the average user could not be successful, but how do we do this?

Ask yourself this question: how many times have I written an AI routine where there were a lot of if-then-else statements and it still behaved like trash? Is there a better way?

In my journal, I intend to every so often bring up a practical, simple example of some intelligent systems ideas that may or may not be useful to a game. We shall explore some of the possibilities informally together and do some experiments. Hopefully, in the end someone will get something out of it. Hey, it might be interesting, right? And don't worry about the complexity, as someone used to say to me, "hey, it ain't partial differiential equations...."
Sign in to follow this  


5 Comments


Recommended Comments

Umm.. can I request ideas which pertain solely to the 2D RTS genre, especially those written using SDL by someone whose avatar depicts at least 2 sharp blades?

Blah, I need to get units running around before I even think about A*/ANN/Toasters. Nevermind, I guess.

Back to worksies.

Share this comment


Link to comment
An AI-oriented journal? Sweet! I never got to welcome you to journal land. Welcome! Supperhappyratingsrape! One more for the programmer team. And good luck! I'm writing a 2D sidescroller right now, so I'll be reading this purely for the sake of retaining the information. Write good stuff though.

-IV

Share this comment


Link to comment

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