RPG Game Design: Building a World that Works

Download PDF

I recently had a conversation with a role-playing game hobbyist who complained how difficult it was to create a game world that “worked.” She wanted to create an alternate history-type South Africa, one in which Zulu tribes, empowered by steam punk, dictated different outcomes with the invading European colonists. She liked the notion of that sort of colorful clash, two sides with equally powerful weapons but with different cultural and warfare traditions fueling their conflicts.  (This is what happens when game-geek girl plans to major in African Studies, apparently.)

Europeans vs Zulu: ready for steampunk?

But this intriguing concept began to break down the more she tried to develop the setting.Where did the tribes get their steampunk gear from? The colonists? No, to be on an equal tech footing (i.e. to have a fighting chance) against the settlers, the Zulu needed more steampunk than a few “tech-runners” could provide. It seemed they would need a manufacturing capacity of their own, at least for a certain class of production, to supplement if not replace what was brought in illicitly.

Things got more complex and improbable from there.

After trying to create an industrial base native to a pastoral tribal setting in a way that did not feel haphazardly grafted on, then came questions of specifics about the tech on each side, and what kind of interactions would likely happen. Besides “fight each other,” she was left with the question of what adventurers could do in such a setting. They could wander the wilderness endlessly in search of “bad guys” to kill, or live settlement-centric lives where they could…what?

Opportunities, motives, challenges in this setting seemed completely opaque to my friend. What sparked her imagination initially was the mental image of Colonials and Zulu warriors fighting and traveling with steampunk gear. But the landscape posed problems (what do adventurers do in a vast undeveloped territory with only a few enclaves of “civilization”, however defined?), the mashup of cultures and tech didn’t necessarily fit, and when it came to “giving the characters things to do,” she intuitively felt that the setting wasn’t working. She just couldn’t put her finger on exactly why.

I have a few thoughts about that.

Designing a Coherent Game World

The first thing that leaps out at me here is that what drove this “design” (or, more flatly stated, the cobbling-together of disparate elements) was a desire to achieve a certain “feel” in the setting and the game play.  An interesting idea drove the conceptualization, but it was just one central idea: steampunk tribesmen meet industrialized, steampunk Europeans, and entertaining conflict ensues. Everything else in the world design was built backwards from this desired end state.

And therein lies the biggest and most enduring problem in this entire process.

An ‘end state’ is exactly that: the end of the processes that preceded
it. It is the result, not the starting point, of good design.”

It is not difficult for gamers to imagine a setting that seems like it would be fun to play in.  We do that all the time, and then much labor is poured into creating maps, and towns, and characters and adventure hooks.  But it is challenging and often difficult to build backwards from a desired end point in a way that makes sense and truly supports that endpoint. All too often this results in exactly the kind of brainlock my friend went through in Steampunk Zululand: an unplayable setting that doesn’t click resulted, and it seems impossible to precisely identify or fix what is wrong with it structurally.

To resolve this dilemma, we need to think of that “desired end state” –  the flavor and challenges we want in a newly minted world design – in a different way, and use a very different approach to create it.

No Carts Before Horses

The basic rule in this approach is,  “Don’t Put the Cart Before the Horse.”  In other words, don’t write your end state in concrete and build backwards haphazardly from there.  An “end state” is exactly that: the end of the processes that proceeded it. It is the result, not the starting point, of good design.  If you are too married to one outcome, and invent underpinnings that force it to happen, you are sure to create a shaky structure that does not truly support (or give sound rationale) for the end state you desire.

* Don’t think, “X is what I want, and elements J and K seem to be part of X, so I’ll just say they’re present.”

* DO think, “X is what I want, so now I have to figure out how we got there from A, B, C…”

The first approach is the one my steampunk friend took. “I want steampunk conflict between these two cultures. It looks like we need an industrial base here in Zululand to make that happen, and lots of colonial soldiers and tribesmen in the wilderness so adventurers can have Things To Do.”

Notice the mode of thought here: plug in elements J and K so we can make X happen.  We’re pushing the process. The end state is fixed, and we build elements into the supporting gameworld structure to justify that state, prop it up if necessary, and serve our meta-game purposes (“give adventurers something to do”), regardless of the actual logical relationship of these things to end state X.

This way lies folly, my gamer friends, and is a sure-fire formula to creating settings that don’t work, or that feel “off” in some manner. That’s your hind-brain protesting about the lack of logical infrastructure, that is.  A milieu that doesn’t mesh right can be ailing in 100 tiny ways, and there is no point tackling each of them, because without a major adjustment in the underlying logic structure, you will only be applying caulk to the cracks in the foundation.

Much easier and more enduring (and I think, much more fun) is to build a proper foundation in the first place.

Getting from A to X, and even to Z.

The alternative approach I recommend, and which is illustrated by the second statement above, is to do a two-pass top-down, bottom-up design approach. (Yeah, like some kinds of software design and programming…).  This lets you start with your end-state vision, work down the chain of causality to identify some major elements critical to that state, then work back up from the logical beginnings of this development path.  You locate your hard-point “must haves” along the way, but also remain open to letting your culture and setting evolve as you do this.  If “A” is a given, how does that affect “K”, and therefore the end state X?

In this process you are inspired by the end state you’re striving for, but you allow the world design to find its own footing and grow in ways that are organic to it.  Put another way, you know the destination you are driving towards, but you allow the roadways there to develop along their most natural and logical paths.  This makes a more solid game-world structure, and will likely surprise you with the many sensible avenues it suggests for related development and future growth of the setting.

That’s all very nice in principle, but you are probably wondering now what that looks like in practice.  In my next post in this series, I’ll walk through an example to show you how this top-down/bottom-up principle applies to world building.  For the sake of illustration I’m going to abandon the Zulu example (which I am not intimately familiar with, anyway), and draw on my own experiences designing the campaign setting of Qua-lun. This “Asian-esque” land is the backdrop of my new fantasy novel Dragonsword (which you can download for free after Monday, August 1), and so the details of it have been on my mind lately.

So let me go collect some notes, and think of some good examples for you, and I’ll be back next week with Part 2 of this post on Building a World That Works.


Don’t stop by the blog regularly? You can subscribe to posts by email with the form in the bottom right sidebar. If you prefer RSS feed, click the orange RSS logo in the top right corner of this page.










 Subscribe in a reader

Did you enjoy this post? Why not leave a comment below and continue the conversation, or subscribe to my feed and get articles like this delivered automatically to your feed reader.


No comments yet.

Leave a comment