From my own perspective I don't necessarily see any immediate differences, but I'm newish to the field. What differences have you come across in regard to building for these two different scenarios (an rpg world and a story world), and should a builder consider what the world will be used for before beginning their building?
There is a little bit of subtlety I want to clear up in your question before answering. There are two ways to build a world independent of its use:
- Build a world then select a use for it based on its characteristics
- Build a world expecting that it can be use for any use
From your comments it sounds like you lean towards the first but I will try to discuss both.
The first idea is very similar to story creation when deciding what media it will be portrayed in. If your story has a lot of action sequences, you may lean towards a movie, a lot of dialogue may lean towards live play, a lot of psychological ideas may lend itself towards a novel. Anyone who works with stories knows that some stories are better suited for different media.
It is very similar for world building. For a world you have created you would look at different characteristics such as how well defined are the physics, how many minute details are included, and the events that go on in it. A small, very detailed world with well defined physics may lend itself toward video games or a tabletop RPG. If it lacks the physics definitions, it may be more suited to being paired with an existing tabletop RPG system. A loosely defined world with large events happening may work better for story-telling (novel, etc.). But there is no reason this has to be decided before you make the world. Build the world you want to build, and likely before you finish, you will understand what uses would best fit it.
The second bullet above, however, is the trickier of the two. If you want to build a world that someone else can pick up and say "I want to use this as a setting for a novel" and someone else says "I want to make a video game here" and yet another "I'm going to write the next script for a James Bond movie in this world", you are setting a very tall order. This sort of world almost requires a top down design process. You can't possibly design all the details for it, some of it has to fall to the individual authors. Therefore it must be designed to detail the parts of the world that would impose consistency between the different author's creation. Physics, political maps, geography would all be required. Detailed demographics would also be very important. It would need to be detailed enough to provide consistency without being so complex that the authors can't possible get it right.
This is also much more difficult to do if you plan on it being paired with a generic tabletop RPG system. An RPG system, by definition applies rules to the world. You would likely need to have the system in mind before designing the world to ensure that the rules the world makes is consistent with the rules of the RPG system. With a little work you could even make it compatible with multiple systems, assuming the systems themselves are consistent with each other. I don't think it would be possible to design a world that would work with any RPG system, however.
Note: since you mentioned DM in your comment, not Game Designer most of my answer will consider rpg world to be D and D like instead of Skyrimish.
The main difference is purpose. An rpg world is typically designed to allow multiple people to play characters in the world over a long period of time several times, such as D and D. Even a game like Skyrim gives a player multiple ways to build their character and explore the world. A story is written once, by you, under your complete control. That means an rpg world must be balanced and open ended. Everyone needs to have fun and should be able to reach the same level of inworld power. In a story you can have the prophecied one, someone with extreme or unique abilities. That doesn't work as well when multiple people are involved.
All this means that your first consideration when designing an rpg world is balance, you are creating a world for other people to tell stories in. Each person needs to have multiple choices available to them that allow them to have as much fun and impact in the story as anyone else. Since you hope the world will be used multiple times, you need a large world to give players plenty of area to explore. Mechanics need to be simple and well defined.
For an actual story, you only need to create as much of a world as your characters will interact with. Mechanics can be bent to fit the needs of the story. Since you control the story, you can have a much narrower focus. I think of it as, building a world for others to tell a story requires a telescope, a long range overview that gives plenty of room for exploration. A story is a microscope, a narrow focus on particular characters on part of your world.
As others have said, the differences are fairly subtle. In both cases you need a world that's well-enough developed to be plausible and interesting to the people consuming it (readers or players). But there's an important difference: RPGs have players.
Well duh, you're probably saying. Let me unpack that.
A work of fiction like a novel is controlled by one person (absent unusual situations like shared-world collaborations), who is usually the world-builder. You build the world for the purposes of your story, revealing its characteristics in the telling, and if you need to adjust something while you're writing, you do that.
An RPG has players who interact with your world. In the course of doing that they may bump into problems you don't want your world to have (such as inconsistencies) that you wouldn't have found on your own. You might decide to fix those issues rather than living with them. If you fix them, you have to decide whether to do so in-game ("you've just discovered a new magic manual and it says...") or in consultation with your players ("look guys, I didn't mean to make it impossible for you to create simple healing potions; forget that thing I said about needing unobtanium").
And sometimes, as the players discuss what they're experiencing, they'll speculate about your world and you might hear something that you like even better than what you were doing, so you might want to incorporate it. This happened to me once in a game I was running and I found myself rewriting the explanation for an important artifact in real time because a player had said something that sounded way better than what I'd done and I wanted to run with it. This can be exciting (you're now part of the collaborative story-telling just like your players) but also nerve-wracking (you need to not mess up any of your other secret information, and you need to be good at thinking on your feet in multiple dimensions). I don't recommend doing this often, but in an RPG setting you have the option in a way that doesn't come up as much with fiction where, at best, you can address complaints in reviews in a later episode.
Bottom line, an RPG gives you the opportunity for more collaborative world-building even if you aren't explicitly inviting your players to the design table. Depending on how rich your world is and how story-focused your game is, your world-development and progressive-revelation processes could be very different from those you'd use in a story. Here are some links from the Worldbuilding blog that discuss these issues in more detail:
RPGs (these are all by me):
"Behind the scenes" fiction (various authors, related to stories posted on the blog):
There are a number of good answers, but I'm hoping to condense and shorten the ideas presented:
a RPG-world includes "empty spaces" for the player to take actions and define themselves or their world.
a Story-world includes "empty spaces" which are open to interpretation and guessing, but knowing what happens during these spaces does not affect what will happen for the remainder of the known story.
Both cases do not necessarily require previous knowledge of which you are going for, before creating the world. However, it may require more world-building in some areas, and less in other areas, depending on what you eventually decide to do with it.
So, you could create your world, leaving more specific parts blank for whether you go for a story or rpg.
Or, you could create your world, and flesh out everything for both the story and the rpg, knowing that some of that information may get thrown away later.
If you want to do both story and RPGs in the same "universe", you could look at which pieces of your world are more relevant to a story. (Probably pieces of the world's history and major events). As well as look at the events and places that are not as defined - and allow your RPG game to take place where the players would have more influence.
An example might be the Lord of the Rings universe. The story of Frodo and the Ring is defined, but many RPG games focus on the doings on what some other person (the player) is doing somewhere else, or even at, some of the major events of the story.
The difference lies in the relative scope and detail levels; the world of a novel has less scope but more detail than the world of a RPG. Novels follow set lines created by the author so they only ever use a small part of the world they are set in but they use those areas very heavily so they have to have really full and lush details. RPGs have a planned outline but Player Characters (PCs) will go off and do unpredictable things in unpredictable places so the world has to be mapped completely because you don't know what you may need but PCs also don't tend to spend too long in one place so they can be kind of shallow in their details except where the GM needs them for the overarching campaign.
I think the differences are fairly subtle if you want to do a good job with either. The main difference in my mind is how apparent the different aspects are to your target audience.
In either case you are creating the settings for a story. You need to figure out the nature of the environment (climate, geography, resources, etc.), the form of the society(s) (population density, main pass times, cultural rules, technology level, etc.), and at least an outline of how things got to be the way they are (historical events, migration paths, wars, etc.). A key point is making sure you have internal consistency.
For example, having a major city on top of a mountain or in the middle of a desert wasteland doesn't make a lot of sense on the surface. What brought people there in the first place? Why do they stay? What resources support them? How have the above influenced their culture? Crafting your world to provide reasonable and reinforcing answers to those questions makes a big difference in how immersed people become in the stories.
The difference between building for a story and a game is in how you present things. In a book you have all of this framework that you hang your story on, but none of it may be apparent to the reader if it doesn't advance the plot. It should however be apparent to you that the assumptions and actions of the characters fit with living in this setting and their individual backgrounds. Which brings up the point of character development, in a book type environment, it is all on you to figure out the traits, background, and motivations for each character the reader sees. In some ways each character that comes to the foreground is a world building exercise in itself. One major advantage of the book type setting is that it supports an iterative approach, you can go back and adjust and revise things until you have a cohesive whole.
In a game setting the perspective changes. The audience is not passive and focused by the scope of vision you have presented to them (or at least not in most good games). Instead you have to split the load and let the players help shape the plot as the story progresses. You as the world creator need to give them enough background to know how their characters fit into the setting, while allowing them to discover the rest as they move through the world. This is a difficult line to walk. Giving the players too much kills alot of the mystery and discovery that can keep a game interesting. Not enough information makes the game feel disjointed and contrived. And it is almost guaranteed that at some point your players will take off and run in some direction you were not expecting. The more complete your view and understanding of the world is, the easier it is to figure out what the outcomes of such actions should be.