View Full Version : Building a campaign world top-down or bottom-up?

2011-01-05, 12:53 AM
When working on a campaign world, what would you say is the better method for creating people/places/things? Would you start with small-scale things, such as towns, dungeons or encounters, then building up from there, deciding on what would motivate them to do those things? Or is it better to start with a good picture of the major powers in the world, and from there go down into how those powers affect the smaller scale world that the PCs live in?

If you start with the small things first, it seems that you have more control over what the PCs experience and what is going on at the time. If you take the big picture though, you end up having a more cohesive world when you do get down to the small stuff.

2011-01-05, 12:58 AM
I make everything in small pieces, and then put them together. That way I can make self contained towns/dungeons that can be carried from one setting to another, in case the party (or me) gets tired of one. So I have a Wendigo area that I made 2 years ago that has never gotten play, but I don't have to toss it out because it can't ever be invalidated by overplot issues. Same thing with the Vampire fight I asked for aid for in an earlier thread, and a dungeon crawl about mist people I did 6 months ago. Basically they each have a theme to them that doesn't require a ton of back story.

2011-01-05, 01:01 AM
I do both and meet somewhere in the middle.

Generally, before making a setting outline, I have an overarching main idea I would like to use. Be it an extensive war, or a heaven crashing down on earth, I usually start with a "one big thing" that characterizes the entire setting.
Then, I work out the metaphysics, maps and political boundaries, philosophy and major historical events behind what's going on, in order to make that "one big thing" happen. That's the top down bit.

At this point, I usually have a few smaller ideas. NPCs I would like to drop in, a unique town or city. Perhaps a dungeon. I then tweak the big ideas to allow those to happen.

I've then got a scaffold to plug things into. After that it's just going crazy with the detailing pen and filling in all the little bits.

2011-01-05, 01:01 AM
I don't think there's one right answer, nor is it really an either-or thing. I like the idea of having the characters start in one particular place, and then both them and the DM exploring the world together more or less, but it's also pretty much a requirement for the DM to have an idea of big basic processes in the world just for the game to function - things like "what is the role of gods?" and "what is the nature of magic?" and "what sort of genre is the world?" - even if the answer is just the default generic D&D.
As I mentioned in that other thread, my campaign world started when my then-boyfriend/co-DM created a single city for a very short aimless sort of a game. The world spread out from there, but almost as soon as I started getting involved in the worldbuilding I started drawing up an full bleeding atlas of the world. We were inventing continents, nations, histories, concepts, planes - an entire cosmos - at the same time as we were still fleshing out a minor frontier city and its surroundings.
So, I guess, from an aesthetic point of view I favor bottom-up, from a personal enjoyment point of view top-down, and practically speaking a mix of the two.

2011-01-05, 01:11 AM
I start with wherever inspiration starts. I have two that were built top-down, Dragon Born and Mets Nchi Sals; both were inspired by an event or idea, and the smaller scale stuff came after to fit the theme.

Mets Nchi Sals started with reading a thread on here and noticing people really don't like elves. So I made them create a world shattering cataclysm that attempted to force people to give them the respect that they "deserve." It of course doesn't work, and another cataclysm happened as retaliation, so now there's three groups of races, natives of either part of the world currently under effect of one of the magical cataclysms, and the people lucky enough to live underground and not be affected. From that I got more specific, created a couple settlements and decided who lived there besides the elves.

Dragon Born, similarly came from wanting a post-apocalyptic rebirth of civilization vibe, it's been so long since the original writing that I can't remember the order of things, but it had Dragons worshiped as gods, and leading the only "built" civilization, and eventually tearing it down. There's really only one city for it, but it too was built from an event rather that a small place.

Then I have another major setting that was built from a small, sparsely populated island, about as developed as Mario (for it was for a video game, end-of-course project to be a throwback to Mario) but has spread from that to a huge setting spanning an entire galaxy. It's my only setting that started from the bottom, but it's my favorite.

Really I would suggest the same to you, if you have the idea for the town first, start with the town, then try and build out, you'll find a theme that your world tends to have pretty quickly. Elsewise, should you start with an event or theme, it'll be pretty easy to develop smaller areas that fit in, even starting from the top, you don't need to know all the countries at play, or every city of the starting country, or region even. If you know where your world is going, it's pretty easy to build the small areas on the fly.

2011-01-05, 01:27 AM
You start with the landscape, then add rivers based on the shape of the landscape, add trees where it rains or are rivers. Add town placed based on natural resources, make town fight over natural resources. Add culture to town. Make time pass, intrigue and war occurs. You've got backstory and a world that doesn't strain willing disbelief.

2011-01-05, 02:46 AM
I do it puzzle style!

Ever put a 5000+ piece puzzle together? Yeah? Than you know what I'm talking about. You don't start with anything pre-determined. You look at what you got(typically a few plot hooks, a BBEG concept, and the PCs' current agenda) and start looking for pieces that can connect to one or more of them. You keep building from these points, and whenever inspiration hits, you add that as another piece-cloud. Eventually, the piece-clouds will start connecting to each other, and that's when things get exciting.

2011-01-05, 05:55 AM
To be honest I've never had a good method for this, i just get inspiration and start drawing a map. the map could be a city, a river, a mountain, something under ground, a forest, an island, a continent. It could be anything. From there it'll unfold slowly over a month

2011-01-05, 06:03 AM
I've created two major ones. One (my earlier) is definitely bottom-up, the other top-down.

I really don't think one is better than the other. The players eventually started noticing that I didn't fill in the world map on the bottom-up world until they actually traveled there (how much I had really done there varies immensely), which... isn't cool enough.
On the other hand, I'm afraid the top-down version will by rather too rigid on the grand scale. :smallfrown:

2011-01-05, 06:10 AM
I generally go with the "macro" approach. I make broad, sweeping, generalizations, and let the players flesh out the world as they adventure in it. It greatly reduces my workload, and helps to improve my improvisation skills. So, definitely, top-down for me.

2011-01-05, 06:43 AM
People said here that there is no right or wrong way to do it. I dissagree.

If you want to build a campaign setting for a specific adventure, then you will need more things on a small scale - the town, the people in it, the surrounding, etc. Starting the work by thinking of how the world was created and the ancient history of the kingdoms isn`t important unless it will come up in the adventure.
If you want to run a complete campaign on that world, you will need to work on many big stuff too.

If you want to publish a setting, the top down approach is more fitting, since people will want to know the big things. I personelly will be more intrested reading about the gods then about a small insignificant town.

If you just feel like creating a campaign setting, then go where inspiration leads you.
I always started by building from top to bottom. The advandage of that way is that building a town that will have an intresting surrounding requires less work, as the surroundings are ready in general.

Ever put a 5000+ piece puzzle together?

Is there realy such a thing as a 5000 piece puzzle?
Or did you just mean 500 and another 0 slipped in?

2011-01-05, 07:26 AM
Is there realy such a thing as a 5000 piece puzzle?
Or did you just mean 500 and another 0 slipped in?

Plenty. (http://www.google.com/search?q=5000+piece+puzzle&hl=en&client=firefox-a&hs=itn&rls=org.mozilla:en-US:official&prmd=ivns&source=univ&tbs=shop:1&tbo=u&ei=QmMkTZW9OMuOswaW1-DGAg&sa=X&oi=product_result_group&ct=title&resnum=7&ved=0CEcQrQQwBg&biw=1680&bih=829)

On the original topic, I usually go for the top-down approach, starting with a world map. After this, I start considering the power of different nations, their diplomatic relations, etc.

Any large-scale unusual environmental stuff usually goes into the world map, and I consider its effects during this stage.

Then I go into more detail with most nations and leave most small-scale details for later.

2011-01-05, 01:55 PM
I do it top down mostly. Because I want over-arching themes and such. Much easier to apply it that way.

Bottom up also takes longer, and sometimes you have to retcon a lot which is annoying. I tried writing a bottom up campaign once and gave up in annoyance/frustration after the 4th page.

2011-01-05, 03:06 PM
Plenty. (http://www.google.com/search?q=5000+piece+puzzle&hl=en&client=firefox-a&hs=itn&rls=org.mozilla:en-US:official&prmd=ivns&source=univ&tbs=shop:1&tbo=u&ei=QmMkTZW9OMuOswaW1-DGAg&sa=X&oi=product_result_group&ct=title&resnum=7&ved=0CEcQrQQwBg&biw=1680&bih=829)

Your search - 5000 piece puzzle - could not be completed with the requested search options.

Anyways I found this: http://www.worldslargestpuzzle.com/hof-001.html

Also, as a note, it always takes a while for me to draw a map for the world, and defintly not the thing I start with.

2011-01-05, 03:21 PM
Is there realy such a thing as a 5000 piece puzzle?
Or did you just mean 500 and another 0 slipped in?

I meant what I posted.

2011-01-05, 06:22 PM
I'm usually fairly top-down when it comes to building a world. It's usually a small number of specific places or idea I want to set out (nothing too complicated, just "port city to the far west with a strong merchant alliance", "massive kingdom to the south", "tribal lands to the east", "blank edges of the map to the north", "theocracy in the center"), and let things fill in naturally, possibly with a map. Maybe not though. It's sometimes better for the players to have no idea and actually explore.

My friend and I also tend to do silly things like play Dawn of Worlds and then work with that so we have more of an idea of what's been going on in the world. If a map is to be made, it's usually done in MS Paint to make editing easier. On that, though, I don't like using continent-spanning maps, or world maps, for that matter. I prefer to generally make a concentrated area (Like a peninsula and an associated island chain) and focus on adventuring there. Fill in a few rivers, mountains, maybe get an idea of where some mines and natural resources would be, and base civilization on those spots. Nothing absolutely set in stone, but also only extremely important areas.

Although, once all that is out of the way, I build villages/towns/cities/whatever bottom up to the point where I feel they're "livable". Start with key people, decide their positions in life (merchant prince, priest, etc.), and then get a vague concept of organizations. If you do start really small, and flesh out each individual person in a hamlet, it can turn into a place the PC's can feel is like home. I don't usually think about culture until this point, where I'll start editing previous details, unless it was set from the start (my current campaign is an isolated region of Vikings cut off from the rest of the world, with the only other culture a splinter group tired of fighting, I made them more like Enlightenment Germany).

On the subject of 5000 piece puzzles:
Of course they exist. I used to do them all the time with my grandfather :smallbiggrin:

2011-01-05, 06:43 PM
there are even 10.000 piece puzzles for you to buy in regular stores, Im sure there are puzzles with even more pieces though I think those you will only find in specialized toy stores or online ^^

As for the world building I use a mix of both and there is no better way in general there is only a way that works better for you :smallwink:

2011-01-05, 06:45 PM
I design the big picture first: continents, then go into countries and geological features, and then towns and dungeons.

2011-01-05, 06:46 PM
I do a combination. I have the legends and epics of the world first, them make a world map with political boundaries. After that, ut's all bottom up.

2011-01-05, 07:05 PM
Top-down. I made a map and started filling parts in by DMing campaigns.

2011-01-05, 07:47 PM
I think the real question is, if you want to present the players with a well fleshed out background for the world, or if you want to start with a small area where the campaign starts and develop it further as the campaign progresses.

If you already have an entire world, or at least continent in mind, I don't see why one should use a bottum-up approach, though.

I do it puzzle style!

Ever put a 5000+ piece puzzle together? Yeah? Than you know what I'm talking about.
Pick out all the edge pieces and start with only those first. There's not that many of them and after that you have a frame to work with.
I think this analogy translates very well to setting creation. (As a top-down fan. :smallbiggrin: )

2011-01-05, 07:54 PM
I start small. With a town or village, and the lands surrounding it. I then wait to see what direction the players want to go, and build out that way, occasionally adding on to the other areas so the NPCs can talk about "The distant eastern kingdoms" or something.

2011-01-05, 09:21 PM
I would build the town it starts in, and a few possible adventure hooks, and have things progress naturally out of that.

EDIT: At the same time i'd also have a few ideas for who the final mister badguy would be, but i wouldn't get too attached to those ideas. After all a party whose family was killed by satyrs wouldn't necessarily want to hunt down the vampire just 'cuz he's evil. It has to be personal.

2011-01-05, 10:31 PM
I build both ways.

I make the driving characters and the starting area, be it a big city or a town and the surrounding hills. I decide the politics and the environment. Who's in charge? When do they step down? How do you select who's going to be in charge next?

Then I make the continent it's on. Tectonic plates, elevation lines. From there, I get rivers and mountains, hills and valleys, and some deserts-by-proxy. I decide what latitudes the continent spans, and get a climate.

I find a location that fits the town, and plop it there. Around the town, I make a country. The country gets a capital, and the politics are grafted on. Add more towns, then maybe another country, and so on so forth.

Kinda got out of hand this time, when I'm making a world atlas for a campaign that hasn't left the city I basically made from an Arse Pull. But 4 other continents gives me plenty of room for more adventures here, in different systems, in different times, maybe in different realities.

It'll be fun.

Mark Hall
2011-01-07, 12:30 PM
Generally, I prefer to have a broad framework (i.e. how deities work, outer planar travel, metaphysics, overview of major races) then build upward from a small scale, filling in and occasionally changing the framework.

2011-01-07, 12:35 PM
I always start with a continent, then I pick the different countries (if there are any), then I pick the different duchies/provinces/states/whatever (if there are any), then the towns. From there I do people and classes and flavour and other stuff. So I'd have to say from the top-down.

Duke of URL
2011-01-07, 12:49 PM
Kind of depends on the goals of your setting design. Are you just creating your own geography in a fairly well-established theme (e.g., it's FR, but with different places and people)? Or is there something fundamentally different about this campaign world than other established settings?

In the latter case, I think the most important thing is figuring out what makes the setting unique. That implies a top-down start, at least, as you have to establish the basic parameters of the setting that everything has to work within. From there (or in the former case), I think top-down or bottom-up is a matter of preference, and need not be exclusive. You could start by specifying a single town or city, then expanding on that as to the bounds of the civilization, going back to fill in the details.

2011-01-07, 12:51 PM
Whichever. I've done both. Im not sure if there's really a big difference. Whatever strikes you as interesting is fine. Flesh out from there.

2011-01-07, 12:58 PM
In the latter case, I think the most important thing is figuring out what makes the setting unique. That implies a top-down start, at least, as you have to establish the basic parameters of the setting that everything has to work within. From there (or in the former case), I think top-down or bottom-up is a matter of preference, and need not be exclusive. You could start by specifying a single town or city, then expanding on that as to the bounds of the civilization, going back to fill in the details.

Sounds about right- the "top" would be the planes of existence, and the basic layout of the world, the "bottom" could be the details of the starting town for the adventurers, and any nearby sites of interest.

Everything in between, can be filled out as the adventures progress.

Duke of URL
2011-01-07, 01:16 PM
Part 1 of Rich's New World (http://www.giantitp.com/articles/YPgbz2j3PckGjjviJU5.html) article series covers a lot of these questions.

As a personal example (and an indirect plug), as part of my Invocation Magic project, I wanted to a create a campaign world as a supplement (or possibly included in the main book) that would showcase invokers by having them fully replace "traditional" magic.

How we chose to do this was to start from the basic premise of: take a "default" (Greyhawk-style) D&D setting, and decide what happens if magic suddenly disappeared one day ("the Cataclysm"). The choices made for the "history" both before and after such an even are likely different than what another designer would have made, but the post-Cataclysm events at least follow logically from each other, creating a new world. Then, some hundreds of years later, magic returns, slowly and hesitantly, in a new form.

Such high-level details point out the needs for mechanical changes to the default rules as well as further setting specifics (e.g., deities and how they react to being "cut off" from the material world). With magic having disappeared for a time, how did this affect inherently magical creatures? Etc.

The "history" led to the creation of several regions, civilizations, and empires, and from there we needed to create a geography and distribute the factions meaningfully. The current state (suspended as revisions to the core Boundless Horizons system takes precedence) has two of the main regions fairly well specified with a variety of locations, but those locations need more detail, and the history timeline is pretty well solidified, with some very specific incidents described, though our "fluffmaster" is still working on making them read better than a pack of notes. :)

Now, in this case, working top-down was pretty much required because everything flows from the basic assumptions made about the world. Now that we're into the details phase, we can work top-down and/or bottom up at our leisure, because we've already established the basics and the parameters.

2011-01-07, 01:50 PM
I work top down. First I create a map. Then I create large regional dynamics, like what industries would probably be big in which locations. Then I do the kingdoms or nation borders and then I narrow down focus like you would do if you were on gmaps and continually develop details at each level as I go with the final 'street view' being where the characters are going to interact.

I don't worry about anything else until the players are actually going there.