# I'm creating an original setting for D&D—where do I start? [closed]

Creating a new fantasy setting for D&D is daunting but I don't want to use one of the published settings. I have a blank piece of paper in front of me and a whole world to create. My friends want to play as soon as possible. Where's the best place to start?

-

## locked by Brian Ballsun-StantonApr 6 '15 at 4:10

This question exists because it has historical significance, but it is not considered a good, on-topic question for this site, so please do not use it as evidence that you can ask similar questions here. This question and its answers are frozen and cannot be changed. More info: help center.

## closed as too broad by doppelgreener, mxyzplk♦Sep 12 '14 at 4:22

There are either too many possible answers, or good answers would be too long for this format. Please add details to narrow the answer set or to isolate an issue that can be answered in a few paragraphs.If this question can be reworded to fit the rules in the help center, please edit the question.

The answers below are awesome and daunting(and awesomely daunting), but one thing I don't think I see below is this. "What is the mood of your setting?" This has been the impetus for both of my settings. In each I decided what sort of mood or general feeling I wanted to convey to the players, and build the world to reflect that. For me the first mood I chose was post-utopic which lead to a very dynamic world, with a rich history. The second was post-apocalyptic, which lead to very foreboding campaigns and deep characters. For everything else, refer to below, they are awesome. – BBischof Sep 19 '10 at 8:04
– Wesley Obenshain Sep 16 '14 at 0:47

You've got a blank sheet of paper, and you want to play ASAP. Excellent. This is what I've done successfully:

1. Decide with your players what kind of setting it should be. You can skip this if you're playing in the assumed setting of your game. Otherwise, find out what elements your players want to explore. Should there be firearms? Political machinations? Should it focus on slice-of-life play, or an epic destiny to save the world? This step should probably be done prior to sitting down at a table with that blank sheet of paper—depending on what is decided, you may have some system hacking to do before moving to the next step.

2. Have your players make their characters. Make sure they chat about their characters as they do. If they ask something like, "Is there a place where lots of X live?" the answer is "yes" and encourage them to ask more such questions. They just helped you create a bit of setting.

3. Draw a map. This doesn't have to be pretty, just useful. Make sure there are features relevant to the characters that are currently developing. Lots of thieves? Have a big sprawling city somewhere. Rangers and druids? Don't skimp on forest and places where settlements clash and/or co-exist with nature. This is you making a "character" that fits in with the players' characters.

4. Flip over your sheet of paper and start making lists. For each interesting place on your map, make a list of things, people, monsters, and events that could be met/stumbled upon/spotted/hunting the PCs. Number these lists for easy random-encounter rolling, or use them as handy pick-lists for when you need inspiration.

1. Think about the characters more. What makes them tick? Build some plot hooks and keep them handy to inspire you during play.

2. Make your map pretty and add details. Your first map needed to be useful, but this one should be a bit more inspiring to you and your players. Now that you've got a bit of time, make your map make you want to play there more.

3. Refine and extend your lists. Rejigger the probabilities of your random encounter tables, add more interesting items, cross off things that you've used that aren't going to happen again, and add more tables to flesh out the world more. Not only are these tables directly useful when playing, but they're a quick and easy way to make the world come alive and inspire thoughts. Why are there a clan of gnomes living in the same hills as a ferocious ogre? How do they or don't they get along? Daydream a bit, and see what comes out.

-
Perfect. I know there's a +17 answer (at the time of accepting this one) and it's incredibly useful. However, I think that laundry list might overwhelm someone new to world-building. This answer captures a very simple procedure for starting quickly and building upon that beginning. They can always "graduate" to the other answers. Thank you, everyone! – Adam Dray Aug 28 '10 at 19:11

I would do the following. Note this is also found on my blog here .

1. Using one page sketch a world or continent map
2. Label important regions
3. Write one page of background giving no more than a handful of sentences to each region.
4. Pick an area roughly 200 miles by 150 miles
5. Grab a 8.5 by 11 sheet of hex paper.
6. The scale should be so that it represents a 200 by 150 mile region
7. Draw in mountains
8. Draw in rivers
9. Draw in hills using them to divide the region into distinct river valley
10. Draw in vegetation (swamps, forests, desert, etc)
11. Decide to place Population Locales note their race this includes social monsters
12. Decide to place Lairs (locales tht revolves around a home of monsters)
13. Decide to place Ruins (locales that revolves around a site)
14. Decide to place miscellaneous locales. (anything that doesn't fit a above.
15. Name your geography (don't forget islands)
16. Write a Half Page background describing the region and it's history.
17. Write a paragraph describing each named geography
18. Write a paragraph describing each Population Locale
19. Write a paragraph describing each Lair (you could get away with a stat block)
20. Write a paragraph describing each Ruin
21. Look at your notes and come up with two to four plots that ties one or more locales together. Write a paragraph or two on each.
22. For each population locale come up with three to five encounters. They should be a sentence each.
23. Come up with 6 to 12 general encounter for the region as a whole. Should usable in any area of the region. They are a sentence or two each.
24. Pick the 4 or 6 most important Population locales and draw a quarter page sketch map of the settlement.
25. Pick the starting population locale and draw a full page map of the settlement. This is the "Home Base"
26. Use Medieval Demographics to get an a idea of how many shops are in the town.
27. Pick or create 6 or 12 important buildings. Write a paragraph each.
28. Scan your descriptions for NPCs or noted monsters. Write a two sentence about each. The first a one line with minimal stats the second one sentence. This is your roster.
29. Pick the 12 most important NPC or Monsters
30. Write a paragraph describing each and fully stat them.
31. Pick the most six common encounter type. (City Guard, Border Warders, Bloody Hand, Orcs) Write a paragraph and fully stat them.
32. Scan your description for any regional organization and write a paragraph on them. Fully stat the most common encounters involved with them.
33. Make up a rumor chart with 10 to 20 items that feeds the players into the encounter and plots you created in above.
34. Identify major regions and create a random encounter chart for each (monsters, wildlife and NPCs) [Thanks Jeff I just plumb forgot]

This will probably run to about 10,000 words. You can do this in about 2 weeks spending about 2 hours an evening at a 1,000 words a evening and time drawing maps. Or consider it about 24 hours of work.

This is being expanded in the following posts.

A Fantasy Sandbox in detail Part I covers Step 1
A Fantasy Sandbox in detail Part II covers Step 2,3, & 4
A Fantasy Sandbox in detail Part III covers Step 5,6,7,8,9, & 10
A Fantasy Sandbox in detail Part IV covers Step 11, 12, & 13.
A Fantasy Sandbox in detail Part V covers Step 14, & 15.
A Fantasy Sandbox in detail Part VI covers Step 16.
A Fantasy Sandbox in detail Part VII covers Step 17.
A Fantasy Sandbox in detail Part VIII covers Step 18, 19, & 20.
A Fantasy Sandbox in detail Part IX covers Step 21
A Fantasy Sandbox in detail Part X covers Step 22
A Fantasy Sandbox in detail Part XI covers Step 23
A Fantasy Sandbox in detail Part XII covers part of Step 24
A Fantasy Sandbox in detail Part XIII covers part of Step 24
A Fantasy Sandbox in detail Part XIV covers part of Step 24
A Fantasy Sandbox in detail Part XV covers part of Step 24
A Fantasy Sandbox in detail Part XVI covers part of Step 24
A Fantasy Sandbox in detail Part XVII covers Step 25
A Fantasy Sandbox in detail Part XVIII covers Step 26

A PDF Collecting the above steps into a printable book can be downloaded.

-

You should establish a basic setting and campaign focus, depending on the themes you know your players will enjoy. Decide if its going to focus on combat, investigation, exploration, etc, and if the setting fits or inverts that style of play. For example, a campaign that takes place in Rome during the height of the Western Empire lends itself well to investigation, but less so to exploration.

As an example, I had a group of players that liked role playing and investigation. I started a campaign set in an alternative 950 AD Bavaria, where enemies could be fierce but were more likely to run away if they dropped below 50% hitpoints - meaning, they were always willing to talk first and cut a deal.

Knowing the above, you should have base culture/setting in mind. Add a starting base and and adventure location, plus 3-4 additional encounters and 2-4 NPC personalities. Again, focus on what your players like and don't like. Don't run an adventure after Sunday school that features a base at a brothel.

In the adventure above, I started with an inn located on a recently flooded river and a hidden fortress under a waterfall, complicated by a small tribe of starving goblins and a creepy harpy that never revealed herself.

## Take notes during the first session.

Most adventures I run "create" future adventures through how players treat the NPCs. Attach facts to NPCs as they are played, and listen to what your players think about them. If NPCs develop personalities during the session, you should fill out their backgrounds and associated locations before the next session.

During the adventure above, an orc who was staying with the goblins became the focus of a bait and switch with a sword with a continual light spell on it. The players loved the NPC so much, they persuaded a new player to play the orc character. The players also were interested in where the goblins came from, and why the harpy was so afraid to show herself.

## Build out from the first session.

After the first session, add more surrounding details. Put that dungeon location and the base into a more defined region, and add a new dungeon location.

After the adventure above, I knew I had to locate the goblin's lair, and fill out the background of a ghost (from which the harpy was hiding) as well as a nearby tiny town where the players would eventually train.

-
This is very close to my favorite answer. =D – Adam Dray Aug 27 '10 at 20:55

What I would not do:

Try to create a campaign setting.

What I would do:

Make a list of all the ideas and concepts I want to use in the setting. Put the list away and go and make a cup of tea. Go through the list and make connections between the items. Leave all the dangly bits, because I'll find ways to connect them later.

That's it. Really. You don't want to try to create the whole thing from scratch or, and this is important, it will never happen. Start small, start with a bunch of ideas, build outward as the game goes on. Ideas will occur in play which would never have done so beforehand, and best of all, the players will come up with ideas that you can pinch and include.

Here's a (truncated) example of one of my lists:

Zombie apocalypse
Colony
Solo/iconic monsters

So what have we got? Well, let's say the zombie apocalypse rendered the original continent on this world uninhabitable, so there was a diaspora. One of the destinations was a small colony on the shore of a distant land, and now this colony is the main setting for the game. The new continent is vast and unexplored, and with many, if not most, of the fighting types having been killed in the Zombie Wars, probably in some kind of Dunkirk type rearguard action, the taming of this new land is left to the youthful heroes of the new generation (the players).

One particular danger in this new land is that it is ruled by vast beasts. These things stomp about causing all sorts of trouble, and generally make expansion rather difficult for the new arrivals, so a sort of Very Big Game hunting culture has arisen, with groups of heroes going out into the wilderness to take down these titans, and bring back trophies in order to improve their standings in an informal league of adventurers. Though it is serious business, this playful attitude is tolerated, even encouraged, as it inspires more youngsters to sign up to pacify the wilderness.

(In game terms, each of these giant beasts is a huge specimen of an iconic D&D monster, and there's only ever one of them. So there's one huge hydra, one massive manticore, one bloody big beholder, and so on. They're so big they cannot be easily or quickly harmed by simple sword strokes, and the heroes have to come up with more clever methods of attack. There may be an indigenous tribe of orcs or something, but for the most part, there's only one of each monster.)

So there you have a sandbox-type setting, with a base of operations, and a unique feel via the sport analogy and the giant monsters; all that from about fifteen minutes' work.

-

Rich Burlew at Giant in the Playground Games has an amazing series of guides to worldbuilding, including a side-along walkthrough of himself creating a setting from scratch.

What it boils down to is two general directions:

• Outside-in, where you outline and create civilizations, governments, geography and major history first, and then fill in the specifics and other details as you get to them.
• Inside-out, where you create a completely fleshed-out starting point (one city, one concept), and build the world outside from there. An example would be to have a home city, and put forests and deserts nearby because that's where you want the players to go next. Then you'd create outwards from there as necessary.

The main difference between the two is personal playstyle, in addition to what kind of game you're building it for.

-
inside out is usually better if you need to have a detailed world early on. outside in is usually better if you need the big picture early on. – Tsojcanth Aug 25 '10 at 16:20

I like to start with a village or town and then build and develop a few miles around it. This gives the players a base of operations, a place to heal and a place to upgrade their gear or blow all their money. It doesn't take too much time. Sketch some of the more important locals, a temple a noble or two, a couple guards and some other colorful characters. Oh of course you are going to need a tavern.

-

Step 1: Figure out why
Decide what you intend:
Is this a one-shot world? Is it a new setting for a series of several campaigns?
For just you? To be shared with friends? To be net-shared? Published?
Just your work, or you and others collaboratively?

Step 2: Planning
Start by making several key decisions.
Top-down, or bottom-up. Top down, you start with the world, and detail out chunks smaller and smaller.
Themes: Pick what themes and memes you want to impose.
System(s) pick which game system(s) to do it for. Determine as well how much modification to that(/those) system(s)
Who Writes What? If not doing it all yourself, who is responsible for what parts? make the decisions.
Set a schedule. pick when each chunk is to be done by. Leave extra room.

Step 3: Mechanic Writing and Building
If you need to make modfications to systems, that is the priority. Figure out the changes, and if collaborating, work out how those changes impact the setting. Some people think system doesn't matter, others let system trump setting unless the setting includes rules changes. Examine the impact of any magic, brainstorm those impacts.

Step 4: develop the environment
Bottom-up: If working bottom-up, pick your starting point, and start with the mapping and describing process. For home-use, this is often the preferred mode. Pick the place, and describe the residents, the government, the religions, the physical descriptions, even maps. Make note especially of routes out, which can later be used for dungeons or other locations.
Top-Down: Lay out the map, Picking natural boundaries and resources, and determining cultures and groups by natural boundaries and access to resources. Lay out the physical first, and place cultures upon that. You may want certain physical features for certain cultural issues. Start with a tag for each culture, then add details as you come up with them. (I've run several campaigns on a world where each culture of 30 had a paragraph, and then simply filled in extrapolations and details as my players brainstormed them.)
Both ends inward: Start with the big scale, then pick a spot, and detail it as a bottom-up, but knowing its place in the bigger picture. This is a pretty typical publication mode; several bottom-up locales in a top-down setting; later supplements often do a top-down on a subset, and several bottom up adventure locations.

Step 5: Revise and expand
Pick the parts which interest you, and expand upon them. Don't worry if they're out of scale to what's been done so far. Then, as you fill in, if you find something that feels wrong, revise it.

Step 6: Edit and Playtest and Edit
If you intend to publish it, rather than just play in it, edit it to a playtest document, then find some people willing to try it. Have them play it cold, let them ask a zillion questions. Revise it to answer the questions. Repeat. Then trim back to what is essential.

Step 7: Cut it into publishable chunks
If working for publication, odds are, you're going to either not have near enough, or have way too much. Too much is much to be preferred. Trim down to just a schematic of the largest scale, and pick a workable amount for the first chunk. Send to a new round of (preferably fresh) playtesters. If they can play with that much, and want more, but don't need more to play with, you've the right size chunk. Repeat the process for the other chunks.

In general, if you want to start "right off", for home play, start with a bottom-up, and let the players help design the central elements. You'll be skipping steps 6 & 7, and 5 will be done as you develop more adventures. Having run several campaigns this way, it works really well.

-

There is also http://www.io.com/~sjohn/demog.htm also known as

You can start either at the top or the bottom, I think its realistic to start at the bottom because starting at the top makes things top heavy I find (its all a nation of orks, a nation of gnolls, a nation of trolls ,etc when realistically the modern nation is not very old.).

Unless I was very averse to it, I would try to get a firm grasp of what feudalism is. Maybe read a bit of history.

You can also start at the very bottom which is game mechanics and build up from there, but that requires a bit more foresight.

-

In your context, friends who want to play right now this minute, I'd start with an initial adventure and starting point, seed it with a ton of ideas that the players can follow, and build the world from what they do.

In more concrete terms I'd come up with a basic scenario with at least three "sides", a map of where at least two are (the third side provides the "in"), and at least two personalities per side. Have side three get the characters involved and then grow from there. Steal a setup from a favorite movie or book to start.

Worked Example: A Fist of Seven Silver Pieces

South of the last Barony is a colony of people. The descendants of escaped outlaws they now are mostly simple farmers just trying to eek out an existence. However, they have found a small cache of old imperial silver they have used to trade for a better life. Now their mule trains to the last Barony are being raided. They've hired the adventurers to end the raids.

What the farmers don't realize is there are two groups raiding them, a group of orcs from some local caves and some of the young men of the village using a ruined Imperial border post.

Now, create the two young men leading the raiders and WHY, the orc shamen and chieftan, the village headman and his young daughter (who is in love with the lieutenant of the raiders instead of their leader).

Things that provide hooks for the future: all six characters, the raider's motivation, the ruins, something special about the Imperial silver coins (perhaps minted by a supposedly legendary Emperor-Mage who most don't believe existed but created a grand spell/item...is it in the caves) and so on. Plus, your players will invent connections you didn't think of...run with them.

Everything here is stolen from common items (two westerns, stereotypes, etc) and is easy to create (cave map, ruined outpost...all easily found online), and six characters.

That could get up you up and running Saturday even.

-

I've started a couple of times, years ago, by reading the Dungeoncraft serie by Ray Winninger. It explains and gives structure to the campaign building process. The second time I used it just as reference but it still was super-useful :)

-

I think you need to answer two questions before you follow any of the great counsel offered by other responses: 1) What kind of adventures do I want and like to run, and 2) What kind of adventures does my group like to undertake?

Don't set yourself up for frustration by over-fashioning your own world if, by their nature and temperament, your gamers are more enthused by hack-and-slash and don't require much depth, history, etc. Believe me, you'll find yourself trying to force them to care about the hours of work you put in, and everyone will be dissatisfied.

If every now and then they like to dabble in some history, but not much, buy someone else's world and adapt according to your good pleasure--that's your prerogative.

If your group loves geological and historical and magical histories and background and depth, then by all means, go ahead!

-

If you're like me, a lazy DM, then I suggest the following: Start with an initial adventure, an initial town, a handful of NPC names, each with a one-line description of what they look like and how they behave. I like hex maps so I'll pick six adventure ideas and place them around the starting area. I'll write a random encounter table. If you're good at improvising, just use a random monster table (and maybe some wandering NPCs?) that reflects the seven locations you have and the general area. If you need more structure, prepare a real encounter with each of the monsters and NPCs you picked. Then start to play. Keep adding just-in-time information and adventures as the game develops.

This approach makes sure that you don't waste a lot of time on stuff that won't get used. It treats actual gaming as top priority and assumes that you don't enjoy world-building per se. If you do, ignore this advice. :)

-
1. Talk to the (other) players. Talk to them again while you design.
2. Start from what you're excited about, and make it central to your game.
3. Skip whatever you're not excited about (chances are you'd make it boring anyway) and get to the exciting parts.
4. Start small. Don't design anything more than what you strictly, absolutely need for your first session(s). In Storming the Wizard Tower, Vincent has it right.
5. Leave some blanks at the edge of the map, so to speak, to allow for later growth. Do have a bunch of half-formed ideas, if you like, but don't plan ahead yet, don't commit.
6. Look at the rules of the game for inspiration. What is it that, in your mind, feels awkward and in need of a fictional justification? Provide that justification.
7. Look at the rules of the game for inspiration. What is it that, in your mind, feels too awkward for any amount of justification to redeem it? Throw it away and design around the gaping hole you're left with.
8. Pause for a moment and look at what you made: what sets this setting apart? Nice. Remember to reinforce that at every chance, except those rare occasions when you'd really love a big change of pace.
9. What you'd love the most to do, if anything, do it. Otherwise, just do the obvious thing.
10. Love what you're doing.
11. Love what you created, but don't wall it away. Let the player characters make a huge impact. If you discover you're walling it away, dig some PC-sized holes through the wall.
12. If you don't feel like you're loving what you made, let it drop and start anew.
-

Gather your players together and run a game of Microscope.

I've only done this for one game so far but it's generated a very interesting world history that the players have had input to and so know some background for immediately. I then expanded on this history and filled out interesting places and events hinted out by the Microscope gave.

-
+1! Microscope is tonnes of fun all its own, and as a bonus? You never have to worry again about the players not knowing the setting! One pitfall that's not hard to avoid but might need talking about before the game is that everyone has to be on-board with making the game of Microscope obey the fictional constraints of the system you'll eventually use, e.g., if you're playing D&D then keeping in mind how magic works in that system helps to avoid creating a setting that will require significant system hacking later. – SevenSidedDie Jun 27 '12 at 17:36
@SevenSidedDie A test game beforehand perhaps to get the players into the mindset might help, that and careful use of the palette to constrain the world concepts. – Rob Jun 28 '12 at 7:16

I have two rules for setting generation: Broad and Small.

A few sentences that tell you why this setting is going to be AWESOME (by whatever definition you use). These are broad strokes, designed to help you stay focused on what things to include or flavor to go for. "Dragons psionically holding together the Material Plane against Cthulhu things. Airships powered by sentient meteorites. Repeating design work on everything, interlocking like Islamic art."

Small

A town, a few interesting NPCs. A dungeon or place of adventure, a few interesting NPCs. Within this situation, put in a few rumors, hooks and potential directions things can go from there - if the players show interest, you know what section to develop next.

The players will make up backstories too, which also helps develop your world.

Broad covers the fun unique parts, small is the parts the players regularly engage with. Don't build stuff unless it's very likely to see play - no point in making 200 pages of setting if you only play 5 sessions.

-

Building a fantasy setting is hard, damn hard. You have to think a lot of things and make them coherent. Also, you don't have to have dark spots. The best to do is to create a "small island" in a well-built, pre-made setting.

That said, if you really want to develop a setting all by yourself, you should define the following things first:

• Type of creatures you want to inhabit your world. Only humans (such as middle ages Europe) or not ?
• Level of magic. No magic at all ? Scientific-based magic (such as alchemy) ? Very low power magic ? Or a magic feast like in Forgotten Realms ?
• The terrain. Define the world map, together with geographic features, such as mountains and rivers. If you want to be picky, try to follow proper rules for orography. An example of notable offender is LoTR: under no circumstance Mordor can arise from proper geology.
• Define the countries/kingdoms, normally following geographic features (rivers/mountain chains), give them cities, villages, resources (mines/crop fields), temples, religions, preferred races, language and coins, a king/steward.
• Define the details of the cities, such as their approximate map.

You can stop here. Once you know these things, you choose one city where things start. That city and kingdom must be well defined, so you have to concentrate a lot on its NPC, buildings and problems. Normally, a campaign involves traveling and interacting, so you will eventually need to refine the outside world later on. Having a rough idea of what's outside the borders of the country the characters start in allows you to define a plot, and from this plot, you can plan what kind of details you will need later on.

For inspiration, I strongly suggest you to read European history, and Italian lords families such as Medici, Savoia, Estensi, Gonzaga. There's plenty of inspiration in there.

-
Building one isn't hard at all. It's just time consuming. I've built a dozen... all of them well enjoyed by my players... – aramis Sep 10 '10 at 23:43
I agree with Stefano, I think it is hard. And frankly, I am jealous that you have build so many nice ones. :( – BBischof Sep 20 '10 at 17:26
The trick is to prioritize what you need. My best games start with a big map, and comments like "Nihon: Pseudo-japanese here" and "Widdish Islands: pseudo european 12C", then letting players build characters that define more. That same campaign world has the "Fallen God Islands" which is mapped to look like the Hero System logo done in small islands; it's a thaumatocracy... with 4 ranks of citizens (Apprentice, Journeyman, Magus, Archmagus) and their minimum requirements, the rest of its details arose in play. – aramis Sep 24 '10 at 7:30
Likewise, the Burning Wheel and Burning Empires both start the typical campaign with world building; a number of other games do as well; the focus is on building just enough framework for everyone to be on the same page for character building. From that common point, just run with what you know or develop in play. I'm doing similar with Blood and Honor... it's a big paradigm shift from the very montessori-esque prepared sandbox mode I started with for several less successful worlds. The best worlds grow by making it up as you need it, not by pre-game prep. – aramis Sep 24 '10 at 7:34

Here's what I did:

• Have a bunch of adventuring locations ready (the more, the merrier).
• Have players create their characters. Listen carefully.
• Start drawing a sketchy map. Add whatever the players mention or their characters require to make sense, as well as some safe village or town or metropolis or inn.
• Play the adventure. Don't worry about the characters getting there or getting out.
• If some of their characters survive, ask how they will spend their loot, if any, and whatever else they are doing when not adventuring. Develop some nearby safe settlement based on that.
• Have them return to the same adventuring location or go to a different one. Continue until the session ends.
• After the session, but before the next one, flesh out the setting: Add random encounter tables, draw a detailed/hex map (the sketchy one is a player map; don't feel compelled to stick to it), add more adventuring locations and think a bit about the safe places. Add tables of rumours.
• Session after session, continue elaborating like this on whatever draws the attention of the player characters. Add a new adventuring location with rumours on the table to guide PCs there every now and then, but especially when the PCs start widening their area of influence.
-
The character creation normally needs to come after at least a rough idea of the setting. I would definitely have a different backstory and possibly very different characters for Forgotten Realms than I would for Eberron or an Underdark session. – TimothyAWiseman Apr 12 '12 at 16:26
A player can build their character around the setting, or the GM can build the setting around the player characters. Both approaches can work. (And there are other options.) – Thanuir Apr 13 '12 at 9:10
It is difficult to build the setting around the characters unless there is a high degree of coordination between the players making the characters, and at that point the GM has handed over many of the setting decisions to the players. – TimothyAWiseman Apr 13 '12 at 16:02
Yes, the players have influenced the setting much. – Thanuir Apr 13 '12 at 17:01

I'd start with a megadungeon, and build the first few levels (about twice as far as you think the players will get in the first session, just in case), simply because they're a great way to keep a party occupied. Then fill in some general terrain around it, a village/town/city or two.

Have some areas around your core, where the various player characters come from. Do this after the players have decided their race and general background.

Build out from there with different types of landscape and monsters. Make a handful of NPCs and give them various problems based on their surroundings. Some the PCs can help, some they can't. Try to make the NPCs be more than quest-givers. Some background is good, and makes the world seem more real.

Place some dungeons occasionally, with various justifications as to what they use to be. A few crypts or cemeteries, some castles/keeps, a handful of caves, and an abandoned town should generally be enough for the starting area.

You'll need some explanation as to why there isn't a local ruler getting rid of the monsters. Various methods I've used or seen are:

1. There isn't a single local ruler with enough power to clear the dungeons, and the rulers don't co-operate.

2. The local rulers are busy with something else (war, famine, plague ect.)

3. There are no local rulers - the area is pretty savage or on the frontier of civilisation.

4. The power in the region is in the hands of people who, for various reasons, have no strong motivation to get their hands on the treasures in the dungeons. Maybe they don't use metals as currency, or they are nomads who never stop in one place long enough to go for a dungeon delve. Maybe they believe the dungeons are haunted, or simply respect the monsters enough to leave them alone.

5. The local ruler is getting rid of the monsters - he's hiring the PCs! This also explains why no-one bats an eyelid when the players haul some sacks of gold out of the old deserted castle - it's their job.

Make sure that as the players explore, you generate new areas of the map for them. Also, let them find new 'base camps' from where they can explore. This should come naturally from how you built your core. Make sure to have accurate maps for where all the new areas are.

After that, you have to decide where to take it. You can develop governments and nations, or wilderness and dungeons.

One important piece of advice: Don't go into detail for a region the PCs won't reach in the next few sessions. By the time they get there, your underlying situation may have changed a lot. Or they may simply keep going in the opposite direction.

Also, you don't have to balance everything. If the players know that there are some things out there they are no match for yet, they are more careful but also more motivated.

-

I would start with what your players want. What do they want to play, and what story do you want to tell? When you get their input, then you can move on to more specific setting creation.

-

Start with the God Mythos that you want to use. Is it one that exists already, or is it your own, or a combination of the two? This then will give you the greatest culture upon your planet.

Now draw a map of your new planet, with the continents, land masses, and oceans named. Now decide how many cultures are on your planet, and then place them. Also create one unique culture that you and others won't recognize where you got it from.

Now how magical of a world do you want? How many suns and moons do you want?

Don't do the earth, reason being that it's too restricting on what you can and can't do, for history has already done this for you!

Now you are on your way and the rest will flow from you. Remember the old books you read; use them as examples to stay away from. You can use a small percentage from foreign works — just have your own ideas so outnumber the borrowed ones that it shows. It shows you liked that concept and thus why you included it; but don't do this to often, otherwise it looks like all you know how to do is borrow from others. (The book Dinotopia did this too much, and that cheapened it.)

Also create a list of rules alterations to give to the players from the start.

Now have fun and create.

-
Here, I've edited. Click this link to see how I changed it. I added paragraph breaks between ideas, I used English punctuation and spacing (tip: if you think you need a semicolon (;), stop: you don't need a semicolon), and moved related ideas so that they're together in the same paragraph. I fixed run-on sentences by using periods where sentenced ended, commas and conjunctions where they continued, and semicolons, em-dashes, or colons where there are related but separate sentences. I used a spell checker. – SevenSidedDie Sep 16 '14 at 1:39