I was wondering what AI techniques are generally/usually used for S/T RPG games like Final Fantasy Tactics, Fire Emblem, Tactics Ogre, XCOM...?

There's always a standard set of AI techniques which is used in almost every game:

1. Pathfinding (what is reachable, how to get quickly to goal area etc.)
2. Decision making (what is the best to do now ? )
3. Action execution (how to perform a certain action?)

Common implementation for these areas are
1. A* based on either waypoints or navigation meshes
2. From simple scripting over to behavior trees.
3. Motion controlling, animation synchronisation, sound synchronisation, steering behavior.

For a round based, grid based game I would recommend to look into
1. A* + waypoints (=grid cells).
2. Behavior trees
3. animation synchronisation/scripting

good luck :D

Is there a particular effect/decision-making pattern/type of AI behavior you're trying to achieve? Various games use a pretty wide array of different techniques depending on the game design, available resources, and so on. It isn't like the AI devs just open the Giant Reference Book to "tactical RPG" and copy out the code ;-)

There's always a standard set of AI techniques which is used in almost every game:

I know but what I was asking is if there is any particular "configuration" that pops up often with SRPG/TRPG's?

Is there a particular effect/decision-making pattern/type of AI behavior you're trying to achieve? Various games use a pretty wide array of different techniques depending on the game design, available resources, and so on. It isn't like the AI devs just open the Giant Reference Book to "tactical RPG" and copy out the code ;-)

No I was going to just roll my own AI but I want multiple AI opponents (edit- for different challenges) and was also curious how it was usually done, hence the question how its usually done. PC is the target platform so there are no real resource constraints.

My point is there's no such thing as "how it's usually done." Every game is going to be different in some substantial way.

My point is there's no such thing as "how it's usually done." Every game is going to be different in some substantial way.

Does anyone have any example implementations or descriptions of them then? Or a case study or post-mortem? Basically I'm interested in previous implementations.

There's a bit about Xcom in here.

http://www.gdcvault.com/play/1018058/AI-Postmortems-Assassin-s-Creed

Thank you kind sir... will take a look.

You might want to take a look into goal oriented action planners.

It may or may not be useful in your game, I find it more flexible and simpler to understand and implement than behaviour trees and finite state machines for game AI.

There are some good implementations out there under really permissive licenses.

Enjoy :)

Will do, thank you for the advice.