Top down, or bottom-up?

[Janx]

I use my patented "edges-in" methodology. Works well for software design and campaign world design.

Basically, you work out the points of contact with the players and flesh those out. Since those points of contact are both overview and local perspective, I design both at the same time.

So I start with a center point, say, the main location the players will be in.
I define that city, then country.
I then briefly describe other countries
I define where the races come from, so the players know where their elf or dwarf came from
I define the religion system, with emphasis on the one the Cleric player will need to know about
I then go back to the home country and refine its detail, I don't need much on other locations, but I need much detail on the aspects that affect players. If I don't have a cleric, then religion will be a paragraph at best, for example.

Janx

 

[ThirdWizard]

I like a mix. I've actually tried both ways, and here are the problems that I have personally experienced (I've made a good many homebrews).

If you concentrate too heavily on top down, you run into the problem that there are less details during actual play. There's a big picture, but it is best to convey a big picture in an actual game through events, npcs, and other things that are real to the players. You might have a kingdom to the south preparing for war, but you might not know how the area the PCs are in will react, what the town militia is planning, etc. You can have a wonderful pantheon, but the running of the local temple eludes you. Things like that are important to make PCs feel like part of the world, and if they PCs arn't part of the world, then what's the point of having all the info about it?

If you concentrate too heavily on bottom up, you run into the problem that the world seems to end where the PCs havn't been, or arn't expected to go. A town might be very well defined, but its relation to its neighbors, or the kingdom as a whole, might be non-existant. When those things spring into play, tension that should have been there the whole time just then is created, and the PCs are left feeling a bit out of place.

Generally, I like to start off designing a large area, kingdom sized perhaps, detail a bit about the surrounding areas, design some gods and genral religiouis beliefs, like a top down approach, but leave everything ambiguous enough that its there, and if the PCs want to know more about it, then it can be fleshed out fairly easily. Then I decide where the PCs will start and work on a town in detail. From there, I can decide where the PCs and where they live fit into the big picture and go back to modeling the top-down again, in a long and involved process until the two meet each other, theoretically somewhere.

It's been a while since I used a homebrew, though, actually. I run a Planescape campaign, and one of the things I like about it is that the top is already there. I "recently" started an Outlands based campaign, and while a lot of the big picture is done for me, I got to create an entire area, the city where the PCs were starting, all the surrounding areas, and all that fun stuff. Most of my PS campains are actually heavily homebrewed, come to think of it. Before this we had a Ysgard/Carceri campaign, and I had to create a lot for that as well.

 

[LathanM]

I went for the bottom up. My last campaign started with a small village next to a wall similar to the Great Wall designed to keep something out. For the first 3rd of the campaign there was enough stuff going on at a local level that there wasn't any real need to expand the world. And when I did start to expand the world we only finished the first provence based on things that happened in game. The players got to go through an entire 20 level career without ever having to leave where they started. The village just grew larger as need. Game time was about 15 years with the village going from an outpost/training school to a trade center then finally a major battle scene. Keeping the players focused on a smaller part of the world allowed me greater freedom to have characters and NPCs come and go as needed to build the plot. Since I had mobile NPCs the players had all the chances to branch out and go to different places and see other cities but they weren't forced to even after half the city was leveled and they had to go and find reinforcements.

There entire world was like DS9, a travel hub where you didn't have to find adventure it found you.

 

[Jdvn1]

I start from an idea and work there. If my idea is a bottom-up sort of idea, I do a bit like that. If the idea is a top-down, then I do a bit like that. I get a bunch of ideas and flesh it out where appropriate.

 

[Nifft]

I started with 12 planets and a thousand Planes.

Yet I picked "Bottom-Up".

For me, though, the "bottom" was the Core rules, especially the kind of wealth that would clearly be available at higher levels. That implied a certain kind of society -- not one with thousands of L1 Commoners running around, either. Society came "up" from what the rules implied.

So, kinda both I guess.

-- N

 

[Umbran]

Other. Specifically, a mixture of A and B. I started with some things at the top, and some things and the bottom, and worked both directions.

 

[Pielorinho]

It really depends on the game.

One long-running campaign started with a single small village and a nearby dungeon.

Another started with the idea, "What if the bad guy were slowly turning the entire world into hive-mindish zombies, and the game had a climactic scene in which the PCs' home city were completely converted to this type of zombie?" That home city never did get detailed, because the game fell apart after two years of travelling adventuring and I just couldn't figure out how to make the climactic scene work. My prep for the game consisted of figuring out in great detail the villain's plans and motivations, as well as the religions of the world that would make this work.

Another started with the concept, "What if I run a very political game 500 years in the past of the previous campaign, and have spirits be omnipresent?" My prep prior to this game consisted of coming up with a lot of detail for the starting city, where we played for several months before moving on.

I usually start with a hook of some sort. If it's a travelling game, I put my energy into mythology and spectacle; if it's a stationary game, I put my energy into politics and organizations.

Daniel

 

[demiurge1138]

Somewhere in between. For my biggest, most detailed homebrew, I started by detailing the concept, but in terms of geography, I detailed a corner of the world first, then moved outwards.

Demiurge out.

 

[sfedi]

I start with a strange mix.
Depending on the campaign.

In my current campaign, I started running some modules, while at the same time painting the big picture.
Now the top and the bottom, has converged, and I have everything defined for the current campaign.

On other ocassions, I started with the global concept, nothing specific, but then I run an adventure which involved neutral elements, things I knew would not be affected by the setting. As the adventure advanced, I completed the big picture and it's consecuences on the mundane and the rules.

 

[Galethorn]

Other! Ha-ha!

For my big homebrew, I started with the mountain valley of Elondale and the town therein by the same name...so, I started bottom-up.

However, after that, my next step was to figure out the nature of the world around Elondale in broad terms. I made a regional map, and from there I jiggered around details, and continued 'down' to the point where I had all of the nations relatively detailed, and had at least a basic idea about every last city and town. Then, I started detailing the places that the PCs were heading towards at any given time...of course, the first thing they did was go after the hook that led them to adventures at the edge of my map. So, I drew a continental map and started detailing the completely foreign civilizations.

Long story short, I use a strange, modular down-up system that comes in steps of top-down.