Hi guys! I''m designing an RPG (isn''t everybody) and I need some feedback on how to go about breaking down the design doc. Maybe you guys could give me ideas about all of the sections that you should put into the doc and how to break stuff down, what detail level to get into, etc.. I have read a few books on the subject, and have been working pretty hard on things, but I just wanted to get some other people''s input on the the entire game design issue. At this point I have sections on, Basic concept, Story, Game interface including the various screens, Game World, Main Characters, Towns, Services (i.e. special people in towns like blacksmith etc.), Tribes (like classes or races), Economic system and bartering, Character Stats, Character Skills, CHaracter Abilities (the rpg will uses a much different appraoch to building your character, involving practising skills and abilities, that is to say, the more you use a certain skill the better it gets), Weapons, Weapon Modifications, Armour, Monsters, Combat System, Scripted Events and Dialogues And that''s it so far. Can anyone tell me something I am missing out on completely. Or does anyone have any advice for feshing out any of these categories, which is something I have already started to do, i.e. making tables and lists and all of that. The main difference, I feel, in my game design is that I actually have the programming knowledge to make it into a game ;oP, with the help of a few of my friends of course. Anyways, if anyone has any advice for me I would be more than happy to listen to it. Thanks again guys Kevin

You need to have a more concrete view of the large amounts of data to be handled. I''d suggest you create some spreadsheets to clarify the data structures for all the categories of information that are used. Also load some sample data that you can manipulate and organize. You can then size and calculate how you''re going to implement the classes.

ZoomBoy
Developing a iso-tile 2D RPG with skills, weapons, and adventure. See my old Hex-Tile RPG GAME, character editor, diary, 3D Art resources at Check out my web-site

Your categories look good, I personaly think you can''t have too many. Here is my opinion on a good design:

What *I* think is necessary is good organization, go through each topic and decide what it ''belongs to'' everything should fit somewhere, if it doesnt it''s a bad category.

Obviously the global topic would be Game
Game has: Story, Game interface
Story has: Basic concept, Main Characters
Game interface has: Menu Screens, Game World
and so on...

In my experience creating a tree like organization helps out a lot. Also some topics fit more than one spot, like Main Characters might be also be categorized as belongin to the Game World.

Eventually you''ll see a good balance forming, and it should be easier to see what more you need, and what isn''t necessary for your own design.


Just because the church was wrong doesn''t mean Galileo wasn''t a heretic.
It just means he was a heretic who was right.

Thank you so much!
Atavist, you personally! That tree-like sort system cleared up tons of stuff on our project! I used 3 overlaping trees, so it came out quite a complex scheme, but it''s so clear which part goes where and which ones are cross-refferenced...! The major problem though was drawing the tree on my PC, cuz I don''t have a specific software for that... Do you know of any? I feel that my project''s trees will be remade several times till we get the perfect combination, so it''s gonna be pain in the @$$ using Illustrator all the time.

One more time: thanks!


Boby Dimitrov
boby@azholding.com

Thanks for all of the tips guys! I like the idea of organizing the doc into sections with each relating to one another. Right now I just sort of have them all in order one after the other, not in any real order. It makes sense to do it in an organized way though.
As for the data structures and all creating some test data, should I wait to do that until most of the creative ideas have been thought out such as the story and characters and game maps, or should I begin on that right away. I have been doing a mix, making up some numbers and equations as well as working on the creative and idea side only.
I have heard that you should save all of the Design Spec stuff for later, all of the numbers and data structures what not until your creative design is complete.
What do you guys think?

Thanks again,

Kevin

Hmmm.
1. I''ve always created my docs with pen and paper. I think the tactile effect helps me think better

2. As for data structures, if you have programmers who are able to take an object oriented aproach. The data should be able to be aranged exactly like your idea tree. My cardinal rule of data is to hide everything from a certain chunk of data that it doesn''t care about. If it doesn''t use it, it shouldn''t know about it.

Just because the church was wrong doesn''t mean Galileo wasn''t a heretic.
It just means he was a heretic who was right.

Well I am the main programmer and I can take an object oriented approach if I want to, well I can try at least. I must admit my C skills are better than my C++ skills.
As for the arguement about local data, I agree with you 100%, there is no point letting any data be known to any functions that don''t need to know. It can only cause probelms in the end.
I''m still wondering about how you are supposed to design all of this stuff. I don''t work in a very systematic way, I just seem to add on ideas whenever I get them to whatever section. I think it just seems a little overwhelming at first.

Kevin

Well programming style is a huge topic.
I like classes, really like classes

Game has: Story, Game interface
Story has: Basic concept, Main Characters
Game interface has: Menu Screens, Game World

class Game{
Interface mInterface;
Story mStory;
};

class Interface{
MenuScreen mIntroduction;
MenuScreen mStatistics;
WorldScreen mWorldMap;
WorldScreen mTownMap;
};

and so on... yes it is overwhelming, yes it is more complicated than this simple example, but the concept is sound. After your able to break things down into smaller units simply work on one unit at a time. Program the unit so it doesn''t care about most other units, so it only works with stuff it needs to. And then start the next unit. Sometimes you''ll find you need to go back and fix one unit, but it shouldn''t require a complete rewrite to extend its capabilities.

Just because the church was wrong doesn''t mean Galileo wasn''t a heretic.
It just means he was a heretic who was right.

The reason why I suggest spreadsheets of test data is so that the programmer can sort through the minimum information for each class. It''s easier to spot what is needed and you can always add another column. The designer can set up the 1st spreadsheet with data and hand it off to the programmer. They can test their assumptions about what''s needed.
If you have a list of weapons, then you might think about how they are used. A player would have one list but an NPC would have another list and a 3rd might be a weapon shop which uses the data diferently(retail/wholesale price, Buys low, sells high).

ZoomBoy
Developing a iso-tile 2D RPG with skills, weapons, and adventure. See my old Hex-Tile RPG GAME, character editor, diary, 3D Art resources at Check out my web-site

I agree with the use of spreadsheets for data, it is pointless to try and make large tables in a word processor or something else. After all that is what spreadsheets are used for. Maybe you had another point in there that I''m not quite picking up on.
That is a great idea about the classes as well. It makes sense that if you break your design down properly then a class system should follow from it easily.
Unfortunately I think that using all of these classes and breaking everything down will only cause your game to run slower. Maybe that''s not the case, but I''m sure that a basic C++ is slower than it''s C counterpart.
But this isn''t a forum for programming discussion, so I will shut up now, and thank you guys for your design advice.

Kevin