I've got it into to my head that it would be a good idea and effective use of resources to give the the player a small number of actions but allow them to use those action in many differnet circumstances. Thus making the game more interesting. For example using the same action/ability in combat, to infiltrate an enemy base, or to help solve a crime. I call this lego gameplay since it based on legos(small variety of pieces large number of possible creations) I'm sure a few games have done this before but I can't think of any at the moment so if anyone knows of any feel free to mention them. I figure the key to get this to work is action nodes. Action nodes are entities that can be acted upon by one or more action to produce a result. These actions nodes can be combinded together to form an interesting series of actions. Ideally a system would be in place to generate problems for the player to solve based on their available actions and proficencies. So if the player is captured and thrown in prison the game will create one or more escape paths by setting action nodes based on the actions the player is proficient in. The way I see it all actions fall into one of the follow catagories Skill Tool Enviromental So lets say the player has a bag of rocks, rocks could be hurled at enemies, used to distract guards, break certain types of objects, knock small objects over. The the action never changes since each time you are just throwing a rock, but the differnt applications of throwing rocks make for more interesting gameplay. Does this sort of things appeal to people? Or would you prefer a large number of specilized actions instead of a small number of generalized ones?

One of the best examples that I can think of that does this kind of thing, is the Monkey Island series. Basically you have the ability to use, look, etc. Combine objects together. It doesn't have the same usability, but its the same sort of thing.

I personally like the idea of have a little bit of both. You have a generalized action that can work in many situations, but in the circumstances that you need something a little more specific you also have that.

--Ter'Lenth

The key is a good physical engine. If you don't have one, every action should be pre-programed, and your game won't be freeform anymore.

Quote:

Original post by someboddy The key is a good physical engine. If you don't have one, every action should be pre-programed, and your game won't be freeform anymore.

Why? A simple rule based system can in most cases be just as effective if not more so then a complicated physics system, and requires far less development resources.

Quote:

Original post by TechnoGoth Quote: Original post by someboddy The key is a good physical engine. If you don't have one, every action should be pre-programed, and your game won't be freeform anymore. Why? A simple rule based system can in most cases be just as effective if not more so then a complicated physics system, and requires far less development resources.

To an extent. The problem with most rule systems is that they are fine-tuned towards the game designers ideas about what the solution would be, rather than what the game player is thinking.

The example I see bandied a lot is with the staple of action/adventure/puzzle games, the pressure plate switch. Suppose the designer writes the rule that the player has to place rocks on plates to open a door. That's all well and good, but if the designer then gives the player the ability to drop weapons, they should also work on the plates. As should the player avatar if it is standing on a plate. Or an enemy guard if he is standing on the plate. And so on

The best solution is then not to code a rule for every object, but define a simple physics system, where everything has a weight, and if enough weight is placed on the pressure plate it is activated. The system doesn't need to be complicated, just sophisticated enough to deal with the game world in a logical, consistent way so the player doesn't scream and swear at the game designers boneheadedness [grin].

Oh, and I think you need to do something like this to implement your idea of the generalities of game actions over specific tools. If you want to use rocks to distract guards, you should code it that any noises will distract the guards. That way the player could throw rocks, daggers, hit pans together, or just jump up and down on a metal grate in order to get the guard's attention.

Those examples still don't warrant a good physical engine. Sure, in the pressure plate example you need very rustic physics, but it's still mostly rule based (the rule is that when pressure is applied to the plate something happens).

The rock distracting the guards is a better example of how you could easily use a rule based system approach. Every action has several consecuences (throwing has "attack", "make noise", "apply weight on landing"), and different actions can have partially the same consecuences (swinging your axe also has an "attack" effect), and you build rules around those consequences. That can give you a lot of flexibility.

shmoove