RPG Generator Idea

Platform: PC (Windows)

Genre: Role Playing Game
Theme: Any



Core design goals

- Creates a game separate from what even the developer chose to create.

- Has a nearly random generation, but slightly modifiable user input system to create a setting where the player can feel comfortable.

- Allows every players experience to be unique.

This idea purely theoretical, but at the current time I would like to explore this idea for a little while. Why not right?

My goal is to create a universe for each player to feel at home and have the best experience possible, not necessarily the easiest or most boring. I want to create, essentially, a RPG world generator.

The best part about this goal is that this CAN be achieved. All it has to do is use a bunch of "survey" answers, to generate a world in which a player can feel unique, and where the quest line fits what they want to do in that instance.

Question I am looking to ask here is:

Who else is believes that this idea could go somewhere in the right hands?
Who has some good ideas for some the basic survey questions?
What exactly goes into YOUR favorite RPG experience?

Here is what I am currently working with/trying to fill out in terms of this "Survey":

- Name

- Enemy/Rival Name
- Alignment
- Favorite Stat (Strength, Intelligence, Agility, etc.)

- Favorite Color
- Fantasy cultures

- Favorite School Subject
- Time Period
- Element Selector (maybe each thing on the periodic table has a unique use?)
- Epic Destiny (maybe a fill in the blank that searches for key words?)

Notes:
- This might be posted in the wrong section, I'm not sure where to put it, but input from a community sounds pretty cool.
- I would like to include additional questions that make the story line a little more "random", that help define how you level up, what your abilities are like, the currency you use, and the how items are generated.
- Maybe this could be a community project for fun? (I Don't know quite how this community works, viewed 3-5 posts where people seemed to be helpful)

Thanks in advance for the advice, and I'm looking forward to going further into depth on how this community work and how I can become a more common member in it.

I think sometimes the fun (or at least half the fun) is with not knowing. The way the world and characters are presented to me and the order it happens in has a immense impact on my love for some games. Now I'm not knocking your idea but I think that if the survey is too direct it will take away some of the magic that happens when one truly enjoys a game for the first time.

Also will all games generated share the same battle system?

Who else is believes that this idea could go somewhere in the right hands?
Who has some good ideas for some the basic survey questions?
What exactly goes into YOUR favorite RPG experience?

The last one I sort of answered above, it was the experience... learning the world, meeting characters...

Yes the idea could go somewhere in the right hands but I would also like to mention what you want to accomplish is alot harder to implement than you think. At least assuming you want to make the game more engaging than simple linear "point a to point b" quests.

I'd need to think abit as to what survey questions I'd come up with.

I think sometimes the fun (or at least half the fun) is with not knowing. The way the world and characters are presented to me and the order it happens in has a immense impact on my love for some games. Now I'm not knocking your idea but I think that if the survey is too direct it will take away some of the magic that happens when one truly enjoys a game for the first time.

Some good points

The goal would to be to generate something different every time so even the person developing the game could have a new experience.

I assume at some point this includes the battle system too.

In terms of graphics, yes this would be difficult to implement.

At face value of how you determine stats would be entirely based off of how many choices you give them, and if you give them 20 choices for 30 questions the possibilities equal 20^30

At some point the amount of work to generate, for example DNA in the human cell (which uses four nucleotides to generate all genes in the human body.) using nucleotides out ways the amount of work it would take to implement literally every single human being in existance, or that has existed.

"a nearly random generation"

Good World Generators have to be mostly NON Random as their sub-pieces that are created have to interrelate in logical ways and those interelations usually are irregular patterns of logic/fit. Things have to fit together or you just get a disjunct mess (unless you have an utterly simple world).

Simple example is terrain generation which has pretty standard (understood) patterns of how terrain has to be put together.

When the major features are "seeded" (thats the random part) and then the more detailed building follows those hinted patterns and expands hierarchically into the final result. The original seeds (and internmediate results) themselves can conflict and require backtracking to patch and resolve into a cohesive result. (Ie- a waterfall next to a desert is rather hard-to-fit combination)

Terrain is simple compared to ecologies or societies which have magnitudes more complex 'proper' interrelations.

When specific features with significant limitations are allowed to be specified the generation becomes driven by those key factors and frequently ALOT of the processing is trying to get those details to fit together (with each other and the more generic less constrained world details)

Out of curiosity what methods (do they have a name) will you be using to implement the generator?

Good World Generators have to be mostly NON Random as their sub-pieces that are created have to interrelate in logical ways and those interelations usually are irregular patterns of logic/fit. Things have to fit together or you just get a disjunct mess (unless you have an utterly simple world).

Simple example is terrain generation which has pretty standard (understood) patterns of how terrain has to be put together.
When the major features are "seeded" (thats the random part) and then the more detailed building follows those hinted patterns and expands hierarchically into the final result. The original seeds (and internmediate results) themselves can conflict and require backtracking to patch and resolve into a cohesive result. (Ie- a waterfall next to a desert is rather hard-to-fit combination)

I think what I meant to say about this stuff was more along the lines of what your saying, but rather a way more complex version, and of course it would be difficult to put together, but I assume it would be possible to put it together with enough work.

Out of curiosity what methods (do they have a name) will you be using to implement the generator?

I have no clue what they would be called, once again this is what if it was possible, how would it work.

The best part about this goal is that this CAN be achieved.

Maybe I read out of context but the impression I got from the line above is you had an idea of how you were going to go about implementing it. There have been discussions on dynamic story generation before but your survey idea is pretty new. Although the first thing I thought of was I read the survey part was IIRC Ulitma, and the questions they asked at starting a new game.

Maybe I read out of context but the impression I got from the line above is you had an idea of how you were going to go about implementing it. There have been discussions on dynamic story generation before but your survey idea is pretty new. Although the first thing I thought of was I read the survey part was IIRC Ulitma, and the questions they asked at starting a new game.

When I said that I meant in terms of generating the data, opposed to the graphics. I meant how the story line goes through. By what I'm saying, each choice has a numerical value that helps determine something. For example, creating a way to level up based on these questions would be:

- Favorite Number 1-1000 (Determines how much XP it would take to reach the next level, and in turn determines how much XP each monster at level 1 would award)
- Favorite Stat (What each level would give the most "points" to, saying that each time you level up, each stat gets a certain amount stronger)
- Favorite Time of day ((Yes, it could possibly even be this random.) The hour could then, for example, be used to divide the XP to get to the next level to determine the average weapon strength)
- Length of the player's selected name (An additional numerical value that could determine how much each stat goes up with each level?)

This is more or less the type of idea I had about implementing it, though I have nothing solid, I figure with the right random questions you could collect enough data to generate a layout of a RPG world that player would enjoy, and then take that data, and have it actually build it with what the player choose.

Lets say for a quest A you need item A. But item A was randomly placed in castle B. The guards block you from entering castle B until you complete quest B. You because this is random can't get quest B until you finish quest A. If this scenario or one like it was generated randomly the player would be stuck. There are many things you need to look out for when generating quests, have you thought through what you need to look out for or even how to represent quests so that you can check for situations like the above?

survey part was IIRC Ulitma

This + dwarf fortress

Yes, I think that it is possible, because the survey is just a hidden configuration (in ultima you configure your character). The best and most deep world generation (including history generation and conflicts etc.) is used by dwarf fortress.

I would check out dwarf fortress, I think it is mod-able and you can quickly test-out your survey idea. Dwarf fortress is although a good game to compare your goals. Does DF deliver what you have in mind ? If not, then you will have a problem, if more than that, you could be lucky. DF is by far the most complex fantasy world simulation I know and implementing such a beast is one of the most challenging tasks in game development.