World building in role playing games is a strange beast, with about as many approaches as there are people who do it. I tend to do a lot of it, myself. I rarely get particularly into pre-written campaign settings, for a variety of reasons. Being really into history, religion, and mythology, I tend to find things in published settings that will strike me as silly or unnecessarily complicated. I also tend to dislike the idea that one of my players may have the same book at home and have spent as much time with it as I have. Given my natural love of tinkering, by the time I've made the adjustments to taste, fixed glaring inconsistencies, and then rewritten parts of it so the players don't already know all of the setting's secrets, I might as well have written my own setting anyway.
Once the decision to write a setting is made, the questions are "how?" and "for what purpose?" The latter will largely determine the former.
When I was younger, I took a lot of time in writing extremely long-winded backgrounds, deep and detailed settings, wrote cultural notes and all kinds of nonsense. I could start a campaign by handing out a fifty page primer to the setting. Pro tip: Don't do this.
The problem with creating that sheer volume of information is that your players are never going to read it all. Once in a rare cosmic alignment, your vision for what you wanted a setting to be will so line up with the tastes of a single player who just happens to also love reading setting material that they will, at some point during the weeks or months you spend in that campaign, actually finish your handouts, but this is such a rare occurrence as to be unworthy of mention.
If you absolutely must write a fifty page world bible on your campaign, then be prepared to handle that information the way a novelist does. A writer knows that you simply can't info-dump on your readers, whether we're talking about a prologue of exposition written in the beginning of the book, or pages scattered throughout. Any time you spend more than a paragraph or two outside of the flow of narration, you are in danger of losing your audience. Instead, you have to sprinkle tidbits throughout and show through the characters actions and reactions how thing go. This latter part can be tricky in a role-playing game, where you can't control everyone's actions, but you have an additional advantage. While readers can pick up on a hint and simply have to wait to find out more about it, a player who is curious about a piece of information can actually ask you to expand upon a subject. You should encourage the behavior.
If you're very into the world-bible style of world building, check out this article by Writer's Digest. It has a lot of interesting tips for this kind of project.
As for myself, what do I actually wind up doing these days?
Ultimately, every setting I create for gaming falls somewhere between two extremes.
On the one hand is traditional sand box, hex-crawl style preparation. I create a list of relevant elements and then populate an area with NPCs, factions, things that can happen, hooks to roll with, and so on.
On the other end, I've taken to embracing shared setting creation. If you're unfamiliar with the term, the idea is that rather than the GM having sole narrative discretion over the world, the players can invent elements and introduce them on their own. Depending on how far you want to go with this, it can mean anything from allowing the players to introduce details through their character's background and mannerisms, to sitting down and making the "lets make up a setting" the first session of the campaign.
In practice, what usually happens for me is some combination of the two. I will very often create a specific piece of a setting, often a single city or town, and through creating it establish the broadest strokes of the setting itself - general technological era, the kind of country / region it is in, etc. I'll usually expand on this, creating a very small sandbox at the beginning of the game, but inviting the players to create additional details about the world, its setting, its history and cultures, etc through their background and the player's mannerisms.
This approach has a bunch of advantages. The GM's investment up front is relatively minimal, but I can still plant mysteries and hooks as needed or desired. The players also find themselves more invested in the world as they establish facts about it. It ceases to be "my" world, and becomes "our" world, and the longer the game goes on the more true this is.
Most importantly, rather than writing 50 pages up front about the world and setting that may never be seen, the story played out will add its own details as they become relevant, and the longer you play the more rich the setting becomes.
Try it sometime, you might be pleasantly surprised.
Until next time,