I'm a first time DM (3.5e), and, being a passionate writer, I decided to create the entire campaign and campaign setting myself. It's been good fun, and I have about 12 pages of (elaborated upon) ideas for it. However, I'm starting to worry that I'm becoming too specific. My original plan was to get several dozen pages of ideas for my campaign setting (easy enough), then compile it into a formal "specification" that would basically establish the whole state of the world. (Originally I had planned to write it out normally, but I've been thinking now that I'm gonna use Obsidian Portal or something like that. Not too important.) Following this idea, I set some backstory for the world politics in general, and for the powerful nation that'll be the setting for most of the campaign. I decided it was a good time to draw a map of the continent, so I did. I ended up with about eight countries, including the main one. Within the main country, I drew in fifty districts. Then I look at it again.

At that point I realized that I would have to go through and name fifty districts.

Now, it's not impossible. But I've never done this before, and it doesn't seem necessary. I know that the PCs will need to know the names of three districts within an hour of play, and that's it. The others could come later. Do I need to worry about them now?

This is a bit more general, as well. What level of specificity should I write it at this point, when I haven't even had a game with my players yet? As a first-time DM, I really can't foresee what I'll need. Any suggestions?

  • 3
    \$\begingroup\$ All answers should be from the point of view of Good Subjective, Bad Subjective - something you've done or seen done and how it turned out. Speculation and "ideas" without experience backing them up will be mocked and downvoted. \$\endgroup\$
    – mxyzplk
    Commented Aug 26, 2015 at 0:56

8 Answers 8


I would say that you want to have a sort of reverse-donut shape for the detail-level of your campaign: Lots of detail at the top levels, lots of detail about the areas surrounding the players, and not as much in between.

Think about how the information will be used by the players. Obviously the lowest-level stuff (inside-out) is vitally important... That's what the players are experiencing now. The high level stuff is important too: when a player thinks about the world they're in, they'll be looking for a high-level overview (outside-in).

The stuff in between (the districts in your example), just isn't very important most of the time. There will be a few that are (locations of natural wonders, or important characters). There will be a few that the players need to interact with. Everything else can wait until they're needed.

Leaving this middle-tier somewhat hazy has the added advantage of allowing you room to improvise. If you need a town, you can add it to the blank areas of the map. The same for a forest, or a river, or a lake, or anything else your campaign requires. Role playing campaigns are much more fluid than most other forms of fiction, and having the extra room to improvise can be extremely helpful.

  • \$\begingroup\$ I like this, a lot. I was having some trouble with "only focus on important info for the PCs" because I've started writing the history of this world, and I really want to flesh out the history so it can be used in the story. I hadn't considered this, and it makes a ton of sense. \$\endgroup\$
    – Maulrus
    Commented Sep 14, 2010 at 19:48
  • 8
    \$\begingroup\$ I like the reverse donut analogy. Outside-in and inside-out design are both great, but the really go great together. Eventually, they'll meet in the middle after many months of play and organic development as its needed. \$\endgroup\$ Commented Sep 15, 2010 at 5:12
  • \$\begingroup\$ @Nitsua60 I guess I muddled the topology a bit. I was thinking of an area with the player at the center, and the "top-level" stuff at the edges. That gives you a reverse donut: filled in at the center, then an empty ring, then filled in at the edges. If you picture the players at the bottom of an area, top-level stuff at the top, and detail visualized as horizontal growth, you do indeed get an hourglass shape. \$\endgroup\$
    – AceCalhoon
    Commented Jan 21, 2016 at 15:36
  • \$\begingroup\$ Ahh... got it. =) \$\endgroup\$
    – nitsua60
    Commented Jan 22, 2016 at 3:04
  • \$\begingroup\$ This is also good because its more likely what the characters would know. They would know major cities and towns within travel distance, but not as likely the number of towns in x district 200 miles away or what to expect from it- even customs can be more difficult in a world with no cars or flight (and teleportation only for the magnificently powerful and rich) \$\endgroup\$ Commented Jan 18, 2017 at 14:48

There are 2 ways to build a world. Inside out and Outside in.

It looks like you started with an Outside in method. You create the map, major politics, etc. Then keep taking the world map and keep cutting just a portion and re-draw it bigger. Your world map with 8 countries, 1 country map with 50 districts, etc. The next step would be to fill in the fine details to what the players are going to interact with. Most players won't care about a world map (I'm assuming 1st level D&D characters here). They will care about their village, maybe 1 days walk in any straight line (more or less). Anything beyond that is "somewhere North of here. Take the [name] road to [name], then turn to the North to get there quicker".

Like you surmise, most players won't care about whether the district on the far side of hte continent is named Zibble Dee Biffle, or Shao Na. Their current mission is to go to the dungeon in the Misty Mountains (about 2 days ride from their hometown) to get the kidnapped villagers.

In Inside out, you start with just the surrounding terrain, and then as the players head out, fill in the details as htey are needed.

From the little research, there is a sort of coke/pepsi debate on the better approach. Each has advantages and disadvantages.

  • 2
    \$\begingroup\$ So you're saying I should go from here to focusing on the starting district and places they'll be visiting early on, and develop the rest of the world as I go? \$\endgroup\$
    – Maulrus
    Commented Sep 14, 2010 at 19:32
  • 7
    \$\begingroup\$ Players will mostly care about two issues: what do I need to know to create a sensible "in-setting" character, and what do I do with that character during play that makes sense. Everything else (or nearly everything else) for players amounts to window dressing. That doesn't mean the details won't be useful for you. So, if those two things are what the players need to know, what do you need to know to be able to effectively and smoothly meet those requirements for their benefit? That's how much prep you probably need to do before you start. \$\endgroup\$ Commented Sep 14, 2010 at 19:46
  • \$\begingroup\$ @Maulrus, I think you have a great start. Maybe write a few simple myths and other things like that, but really flesh out the area where the characters will start, where they need to go, and anywhere of significance in between. Give people some quirks. Maybe the people who live in the mining town go to church every day. Maybe the halflings eat more in an average day than any 3 hungry orcs. Give them the quirks, and then have some that break the mold. If possible, try to get your players' feedback on the world before starting the game. \$\endgroup\$
    – Pulsehead
    Commented Sep 14, 2010 at 20:25

I think you have enough to start drilling down to detail the region(s) that you will set most of your adventures in.

I did a couple of products involving detailing hex by hex various regions of a setting (Wilderlands of High Fantasy, Fight On #3, Points of Light I, and Points of Light II). I sat down and distilled my process into roughly 35 steps which are detailed here or here.

You did plenty of work for steps #1 to 3. Now what you need to do is pick a single region about 200 miles by 150 miles and start detailing that. Some of the process may feed back to the overall background as you answer more detailed questions about your setting.

The result will be an area that you can probably use for much or all of your campaign. My own personal fantasy campaign of 30 years is mainly focused on four such regions. The wealth of details I have for the rest of my Majestic Wilderlands has been largely unused in actual play although it served me well as additional background material for my main campaign area.

Also when you done make sure you distill that 12+ pages of setting material into a single page handout for your players. I generally follow a rule of one page of general information and one page of personal background. Much over that I get some eye glazing among my players.

One of the alternatives is what folks around here is calling inside out. The stereotype being a dungeon, a town and the wilderness that surrounds both. You focus only on those three areas expanding as the campaign progresses and the character venture out further. This allows you to get started very quickly. My 35 steps if done to the full detail take about 24 hours or two weeks worth of evening. Some moan about the time but most referees I know spend a month or two prepping a new campaign taking two hours there an afternoon there.

  • 3
    \$\begingroup\$ +1 for "make sure you distill that 12+ pages of setting material into a single page handout". It's heartbreaking, but true: Your players probably aren't going to read any more campaign background than that. And even that will likely only get read at the session. Gaming homework does not have a high completion rate. \$\endgroup\$ Commented Sep 15, 2010 at 17:13

It sounds like you've fleshed it to a proper depth, without going overboard. Just make sure you leave enough wiggle room to be flexible and you're good. That is, dont feel obliged to write down every detail of every district. Leave those for when you need them, so you can adjust things to suit your game. You might come up with a really good idea for an adventure a year from now, and you wouldnt want to be in a position where no place on your continent would be suitable for it.


Being a passionate writer helps. Get used to stretching those muscles.

Your main question is if you are getting too specific, and the answer is no. Honestly, there is no such thing. However, the real question you need answered is, "When do you need a certain level of detail."

Many early basic texts and primers on campaign creation focus on methodology that is somewhat flawed, as camapigns need background detail as well as local flavor. One feeds the other, and vice versa. The best campaigns generally are an outgrowth of background detail and local minutae, and their growth is done in the same way.

For instance, knowing the big-picture history of racial interaction and politics is things that PLAYER CHARACTERS WOULD KNOW. So no matter how small-to -large you want to build a setting, your probably doing your game a disservice by not knowing the basic big picture stuff from the outset. How people learn magic and who teaches it; the religions of the world, how the humanoid races interact with the civilized world...more of the stuff you have to know from the alpha point.

In your particular case, sounds like you have some of the framework set; i'd start working on details of the local area, work on personalities and the 'World In Motion' dynamics of that area before worrying about too much more framework.

As an added piece of advice, I'd go to OP as soon as possible, or whatever wiki system floats your boat, as soon as possible. Wikification of the campaign, with or without player interaction, allows for an incredible growth of both large-scale and local information, and with it all in one place, and with permisions, your players will have a better feel of what your world is about than ever before.

  • \$\begingroup\$ Good advice. I actually was thinking that, once my players gave me their character sheets and a limited backstory, I would write like a page for each of them detailing what their character would know about the world depending on Intelligence, class, race, etc. All of that information would be based on the big-picture stuff, so I definitely need to have that. \$\endgroup\$
    – Maulrus
    Commented Sep 14, 2010 at 19:51
  • \$\begingroup\$ We try. This is speaking from experience. celtricia.pbworks.com/Igbar,+Capital+of+Trabler \$\endgroup\$
    – LordVreeg
    Commented Sep 15, 2010 at 1:37

You need to know what you need to know

Let me illustrate with a few questions about the world you live in:

  1. Where is Kazakhstan?
  2. What country is Libreville the capital of?
  3. What language do they speak in Paraguay?
  4. What currency do they use in Laos?
  5. Who is the king of Denmark?
  6. Does Australia use a common law or civil law legal system?
  7. Who are the United States senators from Kentucky?
  8. Are the sub-national governments of India states, provinces or districts?

Now, if you don't know the answers to those simple questions: how do you function in this world?!?!

Well, you function in this world because that information is irrelevant to your day to day life and, sadly, for a D&D campaign, so is most of the stuff you are worrying about.

If you like to write ...

Go into as much detail as you want. Most of it will never get used at the table but if you enjoy writing it then the effort is not wasted. Personally, I read many RPG rules and modules that I know I will never GM but I enjoy reading them.

As a side-effect, writing this stuff may trigger ideas that you will use at the table, however, that should not be the primary reason for doing it.

If you like to play ...

Develop stuff that has a chance of actually being relevant to the lives of the PCs and therefore of actually being used at the table.

That doesn't mean that only local things matter or that such things do not fit into a wider world, however, you can paint your world in low-resolution details and fill in the gaps as and when they need them. Knowing the names of countries is fine and if they are to the north, east, south or west but unless the players are going to actually go there, they (and by extension you) don't need to know that Bogrovia is ruled by Queen Kylie VI daughter of the late King Kevin III and that she has spurned all suitors including from the highly powerful Emperor Kzgthfkly of Plotz unless and until that rivalry is going to impact the PCs in some way.

If you really need to know ...

  1. Kazakhstan is south of Russia, east of China and north of Kyrgyzstan, Turkmenistan and Uzbekistan
  2. Libreville is the capital of Gabon
  3. In Paraguay the official languages are Guarani and Spanish
  4. The currency of Laos is the Kip [LAK]
  5. Denmark doesn't have a king: the head of state is currently Queen Margrethe II
  6. Australia is a common law country deriving its legal system from the common law of England
  7. Kentucky's senators are currently Mitch McConnell and Rand Paul
  8. The sub-national jurisdictions of India are states of which there are 29. There are also 7 union territories administered directly by the national government.

You've a good overview. Now start filling in where you plan to have players start.

You'll need the starting town, the neighboring towns and their distances, it's province's name and capital, and probably should name the surrounding provinces and their neighbors.

And some adventure locations nearby.


A slightly left-field answer that I've found fun, but which requires you let the players get behind the curtain a little bit: fill in details that are important to your current (and future) story arcs1, then let the players help you fill in the rest.

I don't like to do this in-game, as it tends to bog down adventures, but rather set up separate 'world-building' sessions for any players who are interested. A game like Microscope can guide you, or you can readily make up a system for a GM-less world-describing session.

The two most important features (in my experience) are that no player become attached to a region and no player contradict anything that's already been established. On the first point, I like to move people around so there's no 'ownership.' On the second point, it's improv: whatever happened, you've got to work it in.

Sample WB System:

Each player gets a country. Roll up some basic stats: pop density, dominant race, affluence, law level, gov't type, &c.

On a player's turn they describe the culture produced by the rolls. Next turn they each roll another country at the table to be in conflict with. Each conflicting pair of players describe the conflict. If neighbors, it's likely just your standard geopolitical fare. If distant, it may be that these are the only two countries that produce Unobtainium, so they're trade rivals. Next turn they roll up a country to be in harmony with. (It may be the same conflict-country, and that forces serious creativity!) Ditto, ditto, ditto.

Second round: pick a country. Each player picks a region within that country. Same as above, but at the regional level.

&c. &c. &c, zooming in and out, panning around.

You can pretty easily use this sort of system to do a snapshot (like above), generate histories (one roll represents ascendant vs. declining societies), towns, &c.

  1. If you end up liking this sort of system, next time don't even plan out story arcs: just do this exercise and then the major forces in the world will make clear what adventure-opportunities lie in wait for characters in any location.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .