SVMAYOL wrote:
> hi sir,
> 
> thanks for your time. i am looking for a solution of an agent that will be my 
> opponent in the game basketball..im following the P.E.A.S. (performance 
> meaasure/environment/actuator/sensor)...and i dont know where to start...and 
> one thing, i dont have enough material for this..not even a book (Artificial 
> Intelligence: a Modern approach by stuart russell) our country stopped 
> publishing this book because it is not being used by the majority of school..
> 
> regards,
> sv

That problem is still too vast.  You need to take small, bite-sized 
pieces you can work on one at a time.  For example:

1)  When to move.
2)  Where to move.
3)  When to pass.
4)  When to shoot.

And each of those can vary by player and probably each one has subsets 
of rules to apply.  For #1 above, you could have two opponents - one who 
is very aggressive and one who likes to hang back.  Obviously, the goal 
is getting the ball into the basket (#4).  So, the computer could 
decide, "if I move here with this player (#2), this player will likely 
move here, which could open up a #3 opportunity with a probability of #4 
occurring with a 60% chance of success of making the shot and a 10% 
chance of getting the rebound."

A simple, first-pass implementation could just simply move players 
around trying to get in front of or away from those blocking them from 
getting the ball (i.e. creating a scenario to get the ball).

-- 
Thomas Hruska
CubicleSoft President
Ph: 517-803-4197

*NEW* MyTaskFocus 1.1
Get on task.  Stay on task.

http://www.CubicleSoft.com/MyTaskFocus/

Reply via email to