World Building for Wetware

What can we learn from computer science and games, and back port to our own LTS 1.0 wetware?

World Building for Wetware

Quick housekeeping: I am only just now enabling email newsletter functionality. If you receive this post via email and don't want that grave error to happen again, you may use the open door safety tool, and use the Unsubscribe link at the bottom. If you didn't receive this post via email but would like to get future ones straight to your inbox, first make sure you are signed up on the website. Either way, you can then view and change your newsletter settings on the website, by logging in and clicking the Account button.

This months' theme for the RPG Blog Carnival is "World Building." I have decided to contribute by offering a few small techniques I have enjoyed, inspired from computer programming and computer games. I will willingly preface that I am not someone who takes easily to setting creation - I do not hear the muse that leads people to write pages of lore, draw large regions or nation maps, nor create bespoke character options for my games. To gaze upon others' worlds reveals their temerity, and provides me with no small awe and inspiration.

Yet I have dabbled at times, more the seamstress taking the hem up or letting the waist out, rather than cutting from whole cloth. I want for game ideas and tools that once received, seem so obvious and self-evident, that I shall not forget them in the heat of play. Consider what follows as some tips for "world-building for the rest of us," if like me you do not wake each morning with complete flights of fancy delicately arranged on your pillow.

A final caveat: I do not believe there are any cardinal distinctions to be made between a designing a fantastical other world or setting, and designing situations or scenarios for play. These are the same nemeses, only refracted at different sizes and distances. Hence the following shall freely jump back and forth across that (wholly imaginary) line.

Push notifications

Rumour tables trace back to at least 1979, and were credibly pioneered by the beloved and recently departed Jennell Jaquays, in her seminal module The Caverns of Thracia. They have since demonstrated strong fitness within the hobby, becoming a fixture of published adventure modules and many setting guides. Rumours help referees to advertise parts of their setting that they would like to soon see in play, and for players they provide a sort of 'menu of adventure,' while also help providing colour and a sense of a 'living other world.'

Yet since their introduction, it has been non-obvious how best to facilitate players actually hearing of these rumours, and then following up on them. Even if players know clearly how to obtain further rumours, such as a procedure to spend a night carousing at the inn or visiting a weekly market, my experience is that they often keep putting this off, out of a misguided desire to "clear their slate" of all currently known leads first. Like the quest log in open world computer games, this becomes akin to an overflowing work email inbox you shudder to open.

Furthermore, many rumours-as-written provide insufficient motivation or interactivity to actually inform player choice. Some adventures are overly coy, reluctant to give up any real information, and others are happy to provide you with ample information - but about half of it is completely false. I don't mind rumours being misleading, but where they lead should still be interesting. If players are rewarded with dead ends or uninteresting red-herrings for their efforts, they will quickly learn your lesson: stop bothering. Good rumours should advertise SMART goals:

  • Specific. Who? What? Where? When? You don't need to give the game away and provide all of these of course, but the rumour needs enough detail that the players aren't left guessing how to pursue it further.
  • Measurable. Answer how, or how much? Quantify the timeline, number of people involved and their affiliations, or sum of an offered reward. We tend to make rumours terribly vague, but one or two points of specificity brings it alive, and seem perfectly natural.
  • Achievable. Ensure the rumour affords interaction at the adventurer scale, even if it also deals with matters of kings and nations. Absolutely have tales of an ancient red dragon be uttered in tavern halls, but perhaps a thief has extracted a sum and lived to tell the tale, and they could seek them out instead (for now).
  • Relevant. Why would this rumour be known in this place, to these persons, and why are they bothering to re-tell it? It's easy to focus on how the rumour interfaces with the players, but it can add to world building to consider the flip side as well.
  • Time-bound. Consider if the opportunity it presents should be time-limited (the answer should almost certainly be Yes), and then if they dally, iteratively re-write it to reflect how the world as dealt with it in their absence (nicely outlined by Brian Rideout of Welcome to the Deathtrap).

You (my dependable straw-man) might argue that in reality, many rumours are rubbish and do not lead to SMART goals, and you would be technically (the best sort of) correct. But we are playing an elfgame, and rumours are the sine qua non of helping players find adventure pies to get their fingers stuck in, so quit your ballyhoo and help them find the fun.

Now for the real tip: try pushing these (well-formed) rumours and leads onto the players, rather than waiting for them to 'pull' them. Any person they meet (including friendly 'monstrous' people) is an opportunity to relay a complete or part of a rumour. Small talk and gossip are grist to the rumour mill. Try keeping your rumour table(s) ready to hand, and feel free to just scan down it and provide whichever jumps out as most salient in the moment.

This approach is related to Mike Shea's advocacy of prepping about ten secrets for each session, and also can be seen as the information equivalent of Numenera's plot keys. There is a whole dead horse I'll be stepping over now, that felled equine having once borne the unwieldy epithet: "Quantum plot keys are anti-blorb." I'll take the stance that flexible rumours and secrets, as they are inherently creatures of ephemera, should do less to get one's hackles up.

The second part of this is if we are pushing anyway, we can go harder: the rumour can become an intrusion. Think about what could be happening locally that doesn't just wait for the players to find it - it comes looking for them first:

  • If the original rumour is increased crime after a change of the head of the thieve's guild: the party get robbed overnight, but the thief leaves a clue to catching them.
  • Rumour of a man-eating giant in the woods? He's just chased a woodsman back to town, and the town is presently in die straits trying to drive him back off.
  • Delectable night-tomatoes grow in the warrens beneath a nearby oak? A pushy peddler accosts the party to buy the last one of the season (he claims) at an usurious price.

I want my players in a sandbox campaign feeling pushed and pulled every which-way, entangled in debts and obligations and opportunities. If they drop the reins and stop actively driving towards their goals, they shouldn't just come to a gentle stop. Rather, they should get swept up in the eddies and whirls of local currents. But this is by no means should constitute a railroad: related to what Chris McDowall has called primacy of action, we still keep player agency aforethought. We should be fair, yet generous in the players' favour, when ruling the outcome of their actions. Even in a riptide, we should afford them every opportunity to free themselves, and take back the reins of their destiny once more.

Just-in-Time compilation

Imagination gaming runs on the operating system of human consciousness, it's capacities in some sense far greater, and in others far less, than any manufactured computer. We excel at intuition and pattern recognition, but are poorer at retaining and manipulating large amounts of data. This disparity suggests that the way we conceive of our worlds can and necessarily should differ from the open-worlds of computer games, despite the clear legacy those games owe to the tabletop hobby.

Fundamentally, we do not need to conceive of nor write down our imaginings with the same specificity and exactness as a compiled computer program requires. We have great faculties to take just a few prompts or loose concepts, and in real-time flesh those out as play touches upon them. Though stretching the analogy, our approach is more like a just-in-time compiled programming language. We can resolve a high level outline of the scenario or region of the world into fine detail at the moment it comes into play: just in time. If you are otherwise looking to emulate published game materials, your perceptions become warped, as they must necessarily be more verbose and detailed. When building for your own table (the only table that matters in this mixed-up hobby) you can take a mere few lines of hastily scrawled concepts and get a solid evening of gaming out of them.

A useful instantiation of this idea was given by Felix Isaac during an interview with Dave Thaumavore about his game The Wildsea:

Make big pillars, make small details, leave everything in between in question.

The big pillars are easy to understand, to grasp, and they should spark imagination. The sea is made of trees, that’s a big pillar. Flame is almost entirely forbidden. Ships cut through branches to make progress, but those branches grow back incredibly quickly. Mountaintops are islands.

These big pillars help bring people in, because they’re easy to grasp. That’s the entry point.

Then you add small details, ones that only matter if the characters are going to be interacting with them. Are slaughtermelons full of the vindictive souls of drowned sailors, that hold a grudge against those that eat them? Yeah, and if they feature in your game that’s important, but if they don’t then that just doesn’t matter.

What's the concrete tip? For each loosely sketched region in your world or the next unmapped floor of your megadungeon: write 3-6 big and 3-6 little such ideas. For a whole campaign, about six big truths is enough to sketch out a unique and memorable other world. For a city, wilderness region, or dungeon floor just 3 might be plenty. Likewise, though you will ultimately come up with countless little ideas in prep and play, having just 3 to hand for when the players push into a new area and catch you with your pants down, is enough to seem prepared and give you enough juice to reach the end of the session when you can regroup.

All the actual play, where players encounter a situation and ask clarifying questions and take action and repeat, will happen in this middle undefined layer. You don't need prepared encounters though. Your short lists of big and little ideas sandwich the field of play, giving you the necessary hard edges upon which your creativity can crystallise, as you compile just-in-time.

Then, between sessions, you can further flesh out the immediate region of play, and make sure any new horizons have just this minimal 'big and little' detailing, to ensure you are ready for whatever the next session holds.

Feed-forward state

I have only just discovered and devoured the Internet Office Hours podcast, regularly featuring Alex Shroeder and guests. A topic pitched, but not further discussed, has stayed with me - how can we make our after-session reports more useful for further play? Having just recently written a few such reports, I have quickly concluded that they are indeed not the optimal way to maintain world state for future play.

Once again using the analogy between tabletop and computer games, the written setting or adventure module is like the defined game level at the start of play. Then as soon as the players start acting, and time advances, there are state changes that need to be tracked. During a session, like while the game remains running on your computer, the game state is tracked in our working memory or RAM. When we close the session or computer application, we want a way to save that state to longer-term storage. We might entirely rely upon our mind's longer term memory, but experience quickly shows the cracks in this method.

The writing of after session notes (that need not be brushed up for dissemination) is the usual approach, yet they are still imperfect. It is time-consuming to write an account of any real detail, and equally the longer each report becomes, and the longer the campaign runs, the more burdensome it is to review these summaries before each session. In computer terms, this is like a running text log that we can only concatenate to - every growing, never distilled.

This computer analogy leads to a different approach though - which I must confess has not been properly battle tested yet - though I find the idea intriguing enough I want to present it anyway. What we wish for our future selves is data locality (important in modern computing with multi-level caching and the significant speed penalty of needing to read RAM or heaven forfend, the SSD) - we should track state change as close to the point we will next need that information as possible. This implies that our time at end of session might be better spent going through and writing little notes and updates throughout our prep, right where it will next be needed.

Obviously jotting down defeated enemies and claimed treasures in a dungeon key is nothing novel, but I hadn't extended that idea in this way before now. To make this explicit: the note doesn't go where the actual play happened, but instead we are 'feeding forward' the game state to the place in our notes where we predict it will next be needed.

  • Did the players let slip to the blacksmith of their new found wealth, who is in cahoots with a local group of highway robbers? Note that down where the robber random encounter (and/or their terrorised highway) is recorded.
  • Did the players pull a lever that caused a chamber deeper in the tomb to become flooded? Note that in the margin of the now-flooded room, and if that changes the surrounding hallways, they need a quick note as well.
  • Did you have to improvise some characters who are now likely to recur, or the political unrest of a remote nation? Record that in your roster of characters, and in the entry for that nation, straight after the session.

I'm sure some people have already realised this and consider it obvious gaming tech, but it was a revelation to me, and I can't think of a game text that clearly explains to do this. Most adventures provide no acknowledgement they are actually to be played, with notable exceptions including the hit boxes in Basic Fantasy RPG and the checklist in Nightmare over Ragged Hollow. In general, a lot of the best practices on how to prep and run games are not written down plainly in game texts, which is at least a 'fruitful void' left for the blogosphere, I suppose.

Next month I have the honour of hosting the RPG Blog Carnival, so stay tuned!