3

How would you develop a game that could end up with complex rules but you need to work on the rules (add, tweak, balance) a lot? I have looked at rule-based languages but I haven't found quite useful enough information regarding this.

UI etc. will be developed later, first I'd need to iteratively develop the rules and formulas and test them out between iterations. The game in question would be a tactical two-player game where players select "troops" and a large part of the game is choosing the correct troop setup. So the rules could be something like

If attacker's skill A is greater than defenders skill B and defender does not have extra skill Z then ...

That's obviously a very simple rule, I expect there to be dozens, if not hundreds of rules, with paths (if A then if B....).

For testing I would write a test framework that can run the rule sets through with large number of iterations and logging that allows me to see how the latest changes affected balance. It would also be useful to be able to define acceptable values and a possibility to see easily the changes. What tools are there for this?

The language of choice is either Python or Java (depending on whether I want to target Android or not - probably I will).

Improve this question

3 Answers 3

Reset to default
2

As with larsmans I'm going to suggest that you use Prolog for your rules development. It's just the best language in semi-common use for doing exactly the kind of thing you want to do. I will, however, instead recommand tuProlog as looking like the ideal environment for your needs given that you want to write the game in Java. TuProlog is intended as an embeddable Prolog environment (can be embedded in Java or in .NET) with very finely-tuned library inclusion so you don't have to carry around a whole, bloated Prolog world with you when you're using just small parts of it.

Here's the blurb from the web site:

tuProlog is a light-weight Prolog system for distributed applications and infrastructures, intentionally designed around a minimal core (containing only the most essential properties of a Prolog engine), to be later configured by (statically and dynamically) loading/unloading libraries of predicates. tuProlog also natively supports multi-paradigm programming, providing a clean, seamless integration model between Prolog and mainstream object-oriented languages -- namely Java, for tuProlog Java version, and any .NET-based language (C#, F#..), for tuProlog .NET version. It is also easily deployable, just requiring the presence of a Java/CLR virtual machine and an invocation upon a single self-contained archive file. Interoperability is further developed along the two main lines of Internet standard patterns and coordination models.

Improve this answer
Sign up to request clarification or add additional context in comments.

1 Comment

Image
I will look into this first because after a very quick look at all the suggestions it looks like this holds the most promise. Thanks!
1

It's an interesting question, though a bit broad and generic. I'll try to answer it according to my interpretation of what you're asking ;)

Your main concern seems to be about being able to express the rules in a concise, readable way that would allow you to always keep the definition clear and change it rapidly. Since you say the rules are probably going to get complex, I believe your best bet would be to write a DSL for them.

Try to writing a grammar that would be sufficient for describing your rules and then see how to plug it into your game. ANTLR could be very helpful there, especially because it supports both Java and Python.

Improve this answer

1 Comment

Image
I need to have a look at ANTRL at some point as well. Thanks!
0

try The A.I. of F.E.A.R.. there FSM (finite state machine) was combined with A* pathfinding. instead of finding a path in a terrain the engine found chains of goals to implement generation of "intelligent" behaviour of agents on the fly. maybe there is something inspiring in there for you.

Improve this answer

1 Comment

Image
I don't see how this is relevant to the question (unless you'd want to implement an AI that exploits those rules for its own gain). I have to add that the paper is quite interesting, 'though ;-)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.