• Don't force yourself to create more than you have to.The first rule might make a lot of sense if you've ever started creating and burned out before even getting to a single play session. Don't feel the need to create more than what you need for some basic background and the next few sessions or so of play. Not only can you later fill in details that better match the players as play progresses, but you can get started much earlier without having written the equivalent of the Forgotten Realms book before you start. That doesn't mean that there isn't some frontloading of the work involved, but minimize this as much as possible and rechannel that energy into things that are practical. You don't need to develop prevailing winds, trade routes, 1000+ year histories or the like. Although it won't hurt, it likely won't help your campaign either, at least not for a long time. The second rule is obvious as well -- it gives you hooks right off the bat that can drive gameplay by giving mysteries for the players to gradually uncover. Not only that, this gives the illusion of depth. Your players may not ever feel as if you are working on a campaign setting that is only developed a little bit beyond their next sessions, they will feel as if the campaign has a rich history with all kinds of things going on. Be careful about the pace at which you dole out these secrets, and they can last for months if not years. But they should be developed, at least as a core idea, when the campaign element they match is developed. That way you can drop hints and foreshadowing long before the secret actually is revealed.
• For every campaign element you introduce, create a secret to accompany it.
Of course, I don't necessarily follow these rules to the letter all the time. In fact, much of my noodling with setting design is done during my "off seasons" when I'm not actually running the game, so I can feel more relaxed about development and work on whatever strikes my fancy. But when you need to run soon and need a setting right away, focus on what's important right away and ignore everything else for the time being.
The main idea of this installment, though, is to create a "hook" for your campaign setting. The hook is something that can be summed up very quickly, and which differentiates your setting apart from other settings "at a glance." There's no reason you have to have just one hook, but too many of them, and your game will feel very different from traditional D&D and might also be difficult for your players to digest and understand. The whole point is to be able to describe how your setting is different from "generic fantasy" in just a sentence or two. If you've got to describe more than that, chances are you've gone too complicated. The guys at Paizo have pitched the idea of the "Hollywood pitch"--and then described their first novel, Prince of Wolves in terms of calling it "Indiana Jones meets Brotherhood of the Wolf in Transylvania." That's the level of simplicity you should be able to reduce your hook(s) to.
Hooks can be classified by type, as discussed below. One thing to keep in mind about whatever hook you develop, is that it should suggest campaign action. A hook that doesn't suggest stories, adventure seeds and the like isn't very useful except as flavor. If that's all the hook adds, that doesn't mean you can't still incorporate it, but it can't be the only one you use.
The first type of hook to develop is a cultural one. Standard D&D assumes a pseudo-Medieval Western Europe in terms of culture, but by changing this base assumption, you can automatically make the world feel very different. These types of changes also typically suggest character ideas and things for the characters to do, as well. As an example, look at the published setting Rokugan which is based on an Imperial Japan instead of Western Europe. Clearly, character types are different, with an extremely sharp division between noble and commoner, samurai as character classes, lots of politics and intrigue suggested by the campaign setting, large cities and many other things that are different than what you'd expect normally.
A unique environment is also a campaign hook that can strongly change the feel of the game. Dark Sun for instance, is strongly influenced by the harsh, unforgiving desert that makes up almost all of the landscape of the setting. Characters struggle for even basic necessities of survival, rather than marching from one friendly tavern to another, making pit stops in dungeons along the way to gather treasure.
A campaign that focuses more tightly on a class or race can be interesting, although it may also be difficult to pull off. Some players are enamored of playing a certain race or class, and not having the option doesn't feel like D&D to them. However, especially as a more short-term game, this might be lots of fun. In general, though, if you make restrictions, you should probably diversify your options in some other way. As an example, if you want to have a wilderness focused campaign that uses rangers and druids, you might want to consider using some other alt.ranger type classes, or something along those lines, so your players still have choices. If you're doing an elf campaign, you might want to consider opening up subrace options such as gray elves, wood elves, wild elves, etc. Also, be aware of the potential problems if you eliminate access to arcane spells, healing spells, or the like. This doesn't mean you can't pull it off, you simply have to be careful and account for those types of things up front. If there is a monster in the Monster Manual that assumes you will have access to certain spells, and is very difficult to defeat without them, for example, you need to be aware of that before you send that monster against the PCs. Still despite the potential hazards, the built in story potentials make this type of hook an intriguing one for experienced GMs who don't mind the challenge.
Another interesting campaign setting hook is opposition that the players will regularly face. A campaign setting ruled by dragons, for example, or over run with undead, (two of Ray's examples) offer immediate story hooks for both you and the players to dig into and get interesting gaming started almost immediately. The only thing you have to watch out for are opponents that don't suggest anything -- as Ray said, a world dominated by stirges suggests… what exactly?
A situational hook is one that works really well. This means a world in which a major story is clearly already in progress and the PCs are caught up in it. Midnight by Fantasy Flight Games is one such setting. In that setting, which is like an alternate Tolkien in many ways, the Dark Lord won the final battle. The PCs are typically the resistance fighters trying to preserve what small amount of light they can. Scarred Lands also has such a hook -- it produces a mythology much like Greek mythology in which one generation of gods (the gods) overthrows the prior one (the titans) and begin to remake the world in their image. The hook here is that this just happened a generation or so ago, and remnants of the titans and their armies still cover huge portions of the world. Outside of the d20 movement, White Wolf's World of Darkness games all feature an impending apocalypse of some kind that the PCs tend to get caught up in trying to prevent.
Finally, a hook that Winninger didn't mention but which is a popular one at least with me and the gamers I know is a genre hook. For the most part, D&D assumes some middle ground between sword & sorcery and high fantasy, but a lot of new writers in the field have been making hay with splicing that concept with other genres. What about a game that's steampunk fantasy (a la Iron Kingdoms or Perdido Street Station?) How about caper/heist fantasy, like Scott Lynch's "Gentlemen Bastards" series? Horror fantasy is a popular one, and D&D itself has dabbled in that with the Ravenloft setting, which was classic gothic horror spliced with traditional D&D.
As I said earlier, more than one hook isn't necessarily bad; it can be very good, especially if you find a way to tie the hooks together, which gives the setting more coherency. For my sample setting, I'll start with more of a singular hook, but some of the other types of hooks will come out in play, particularly the kinds of plots and adventures that the PCs will encounter, which are reminiscent of thriller mainstream novels by folks like Robert Ludlum.
Last time, I discussed what I wanted for flavor, and a lot of that suggested (if not outright dictated) what I was going to have for my campaign hook as well. Of course, my Dark•Heritage setting's actually been around in various forms for a few years, so it's evolved and developed in my head for some time, but putting it through this Ray Winninger methodology should hone it to a very useable and practical state. My first incarnation was primarily environmental, and inspired by space opera set on Mars, such as written by Edgar Rice Burroughs, Otis Adelbert Kline, Leigh Brackett, and even Ray Bradbury, I had a harsh, cold desert world full of alien creatures. This hook started to feel a bit too "one trick pony" for my taste after a while, plus whenever I started trying to describe it to other gamers, they immediately thought of Dark Sun (which was completely coincidental and based on drawing on the same sources, I think. I'm actually not very familiar with Dark Sun at all.) So my concept and my hook evolved from a more environmental theme to one of genre. Now, using James Sutter's "Hollywood pitch" style summary, I describe Dark•Heritage as "Pirates of the Caribbean meets Charles Dickens with a strong dash of Lovecraft as directed by Sergio Leone." What kind of action does this suggest?
First, it suggests a great deal about the environment, even though it seems like two incompatible environments at first blush. Is it a sailing type setting with a strong nautical theme, or is it a harsh frontier setting with small settlements of amoral tough guys surrounded by hostile bandits and savages? Why not both? In my case, since I was tinkering with the setting during my off season, I've spent some time developing a bit more detail and a broader scope than I might otherwise have done. I really liked both of those themes, and while it seems unlikely that both will feature at literally the same time and place (some areas will be more swashbuckling mayhem while others are more gritty cowboy like in tone) PCs could come and go from one region to another over the course of the campaign. In fact, setting the campaigns' start at the tail end of the "cowboy" region; a southern peninsula that serves as a port city and outlet for trade from the fur, logging and mining industries in the interior, and a haven for ships and sailors, often of dubious morality, allows me to get the best of both worlds to some extent, and easily go one way or the other as my whim strikes me.
Secondly, it suggests a great deal also about antagonists. In many ways what it suggests is that many antagonists will be human or at least humanoid. Pirates, bandits, "indians", and hostile, head-hunting or cannibalistic natives on the islands… that's the kind of pulp/adventure vibe that I want to cultivate for much of what the PCs will face. But there's also the Charles Dickens---urban criminals and mafia-like organizations, crooked politicians, and worse. And, of course, a strong dash of Lovecraft is where the fantasy comes in: evil sorcerers, dangerous cults, and strange and terrifying monsters and daemons (as opposed to the frequently--and sadly--routine monsters of a normal D&D campaign). These various suggestions are enough to build many campaigns; heck, I could run stuff non-stop in Dark•Heritage for the rest of my life and probably not run out of interesting things to do.
The only thing to do now is create a secret associated with these hooks. Because the hooks are kind of vague at this point, that will also help further define them and make them something that really "pops" during play, as these secrets are gradually revealed. I'll keep my secrets somewhat vague too, until I have something a little bit more detailed to develop a secret about. As I develop my home base area, with its political factions and whatnot, that'll be ripe territory for secrets. For now I'm going with a vague "big picture" secret, and I'm not going to be subtle or even surprising about it. In fact, I'll mine familiar territory with a Lovecraftian secret history of the setting. Buried under the vaguely Mediterranean-like Mezzovian Sea that sits smack dab in the middle of my setting are the remnants of the pre-human inhabitants of my world. And while they seem to be on the decline and mostly quiescent… they're not really, of course. Isolated coastal regions are vulnerable to attack or corruption, not unlike Lovecraft's own Innsmouth.
Further to the north, the landscape has an even more sinister secret history. The Mezzovian Sea is the remnant of a once much larger ocean, and the land to the north was once submerged. It is, in fact, the semi-legendary lost continent of Muuh, sunken below the ocean when its inhabitants delved too deeply into the secrets of the prehumans, and then raised again as the ocean waters receded. Those few occult experts who suspect this relationship of the lands of Baal Hamazi to the north of the Mezzovian Sea and Muuh, suspect that it also might be linked in some way to the rise of the hellspawn (i.e. tieflings) in that region.
Come back next time, and we'll have a look at politics and government, as well as developing a homebase for our first group of PCs. As we get more into the details of the setting, we can sink our teeth into secrets that we can start dropping clues about immediately as well.