auto-observational design

Playtesting is always full of opportunities. Let’s look at two different things and see how they come together because of play and offer a design method. Not the only one, not a whole methodology, but rather a tool for your toolbox.

In the current Sand Dogs playtest our intrepid heroes have left the world. So they are pretty much done with the Sand Dogs concept except as it impacted their character creation and development to date. New world! They are chasing the slaver Harrison who turns out to be a planewalker and since they freed his slaves and broke up his gang, he’s looking for greener pastures.

So I have to make a new world.

This is great because the next Soft Horizon book is to be the handbook which tells you how to make worlds. And I don’t know exactly how to do that yet. But I am an experienced referee of role-playing games and I’ve developed settings before. I have the intuitive talents to do this already.

The opportunity here is to design observationally. I have an artistic intuition about how to proceed but I need to formalize it so others can reproduce it. I could start by imagining a process but in this case there’s a more fertile possibility: I can just watch how I do what I do naturally and take notes.

I’ll work in cycles here — make a thing and then look back over what I did and turn it into a process. Now, I’m not done yet so I can’t tell you the whole process, but I can talk about the meta process — the process of developing the process.

The new world started with an image that I dropped on the table in a panic at the end of the last session as the characters arrived in a new world: it’s a jungle and there’s a huge ziggurat and it’s at the end of a long straight path of churned up earth: the ziggurat appears to have moved albeit very slowly.
Seriously that’s all I have to work with here.

But hang on let’s start there if that’s what we start with. We start with an image. How do you get an image? Imagine a place with at least two things that don’t fit together. When I think about how I got to that bit of loopy ad lib, I realize that’s what I did: invented a contradiction. A mobile building. A fixed structure that isn’t.

So now I have to wonder, what’s a good way to get to that? And how exactly did I do it? This is the meta-process: when you ad lib something cool, look back on your own process. How did you get there? Can you make that a procedure? Can you mechanize that, at least in part, so someone else can reproduce it? You did it. Tell other people how. That’s how writing game texts works.

Oracles of course.

My brain kicks out some random shit and I try to make sense of it. That’s the frame we build this house on: some random crap that you have to make sense of. That’s the heart of every great Traveller session: how the hell can there be a population of billions with stone age technology on a world with no air? Don’t start with “that’s stupid”. Start with “how do we explain that?” That’s where creativity thrives. Don’t block, as the improv folks say. Make it work.

So to start with I want to emulate the random crap generator that’s in my head. I talked before about using oracles and the random noun generator and that’s the model.

Start with a couple of random words. For the game text we’ll call them “elements” and we’ll have some mechanism to deliver them. Probably they’ll be printed in the margin of the book so you can flip to a random page and get an oracle. That’s pretty much an exact model for my brain anyway. But for now here’s the start of the process: get two random words. For my new world those are “wild” and “wander”. Everything about this world is going to be about either wildness or wandering or both. If there are fixed structures they wander. If there are civilizations, they are wild.

And there’s clue two: find the contradiction and develop that. If your element is “water” start thinking about things that can’t be water and make them water. Hunt the contradictions. That’s where the meat is.

A sartorial adventurer.

The next thing I did was draw. I want the players to immediately meet someone so that the world can be introduced through the eyes of someone who knows how it works. And whenever I do that I start drawing. So I drew this … person.

Obviously I’m not going to write a procedure that forces you to draw. That’s not practical. But at the heart of my instinct to draw were a couple of things: again I want an image. And I also want someone to meet, someone through whom the players can experience the world as a native. And so this is that person. But what can I tell you to do? And how did I get to this thing in the first place?

But maybe I’m missing the heart of this. Maybe the heart of this isn’t that I drew a cool insect guy. Maybe the heart is: create a non-player character to meet who illuminates the weirdness of this world. Yes, that’s what it is. That’s the next step. My personal process is to draw them first but that’s not the important part. Invent this character and then think about how they fit into this weird world. What does it imply about the culture? What does this character do from day to day? Is that typical (awesome, now we know a lot about the culture)? Is it weird? That’s awesome too because whatever it denies about the culture (making it weird) is just the complement of what the culture is. As long as the character is extreme one way or another, we can derive What People Here Do. At least some of them.

Obviously there will be more and more material but the important part is the meta part: if you do something intuitively then analyze it and figure out what a procedure might be for other people to replicate it. Maybe not your procedure exactly but one that gets close to it. Think hard about how you think hard and write it down.

completion vs. abandonment

Let’s see what the end of 2017 was like:

Canvas 1

Pretty bleak. Mostly I think I was looking at building a second edition of Diaspora. That didn’t pan out. And I had pretty much given up on Soft Horizon. Quite a lot happens in a year! And Elysium Flare was half done and halfway to the bin. So let’s look at 2018!

Canvas 2

Wow! Now Soft Horizon refers to the whole project or maybe the upcoming handbook and we see the new game, Sand Dogs running up centre field. Elysium Flare is out and I’m done with it. Diaspora is in the bin — it’s finished, it’s a good game on its own, end of property (as of now anyway). And The King Machine, a game I hadn’t even envisioned in 2017, is published.

No Contact and Navigator kind of switched places. Polyp is percolating again in my head, though it could rapidly head into the abandon pile. And nothing new is sitting ready to fire, which is fine because I have ideas about ideas that will go in there.

I thought this exercise was of dubious utility in 2017, but having 2018 to compare to is powerful. I attribute the change in pace and inspiration to Patreon and to my patrons.


I used to work as an alchemist.

No really. It was probably the coolest job I ever had, though technically it was called “fire assay” and not alchemy. But it has clearly alchemical origins since it apparently turns lead into gold. It doesn’t really, of course, but when it was discovered it sure must have seemed like it.

What it really does is extract platinum group metals (and silver, as it turns out) from a mixture. In our case the mixture was a “flux” of borax, lead oxide, silica, and flour; and a powdered rock sample. What happens is, the whole thing cooks down, the platinum group metals in the rock sample bond to the lead, and then you separate the lead from the other metals. Measure the recovered metal and compare with the mass of the original rock sample and you have the proportion of gold in the sample. It’s a great technique for surveying very large areas and looking for regional spikes of relatively low value, indicating a possible gold source underground.

The steps are great fun. Easily the most butch job I had.

Crucible. Or, as it is in the filename on my desktop, “curcible”. Apropos.

It doesn’t start that way. It starts more like cooking — take a pre-measured volume of flux in a ceramic crucible and add the powdered pre-measured sample. In my case this was just 10 grams since we were doing very broad survey stuff. Then test the sample with a drop of nitric acid. Does it smoke? Then it’s a carbonaceous rock and you want to add a little extra silica. No smoke? Silicaceous, maybe add a little borax. Then add a measure of flour. Flour is the critical reagent because it’s the source of carbon that will make the whole process work. Too much flour and you’ll draw out too much metallic lead. Too little and you get no lead. You want a very consistent amount of lead. So if you have carbonaceous rock, add a little less flour. If you have a soil sample (always a nightmare), no flour at all is probably best. Finally, plop in a very precise and tiny amount of silver nitrate.

Row upon row of very fucking hot.

Next you put your crucible in a 1500ºC furnace. In fact you put 24 in at once — this is a production line process! Let that cook for 40 minutes or so. While it’s cooking, the carbon will bond with oxygen in the lead oxide producing metallic lead. This lead will alloy with all platinum group metals (and the silver) in the mixture. So when it’s done you have a crucible full of molten glass and borax and a little slug of molten lead at the bottom.

pouring out
Mmmm, muffins.

Now you pour these out into an iron muffin tin. Seriously, it looks exactly like a muffin tin except the cavities are conical — pointed at the bottom. Let it cool and you have a bunch of glass muffins with lead tips. And then the fun begins.

When they are solid but still very hot, you put the muffin tin by your smashing station which has protective goggles, an anvil, and walls to keep everyone but you safe. You bang on one of the muffins with your cold ball peen hammer and it fractures from the temperature differential. Scoop out the lead divot with your giant tweezers and bang it into a cube on the anvil. This is mostly to get all the glass out of the sample. Now you have a lot of shattered glass and 24 little lead cubes that are allowed with valuable other metals.

A cube of lead! Now how do you turn it into gold?

Cute little poisonous cups.

You put each lead button on a porous ceramic cupel, a little cup with a very thick base. Then those go back in the furnace but — critically — with the vent open. As the lead melts it oxidizes away thanks to the air from the vent, disappearing up the reclamation system and hopefully not into the atmosphere. But the platinum group metals do not oxidize and the silver won’t oxidize much (and you’re not measuring it anyway). After a few hours your cupels are yellow-orange from absorbing all the lead oxide and each contains a little bead of silver — the silver from the silver nitrate you put in at the beginning. Also gold and platinum but mostly silver — you put the silver in in the first place so you get something practical to analyze since the volume of gold is usually very very tiny.

See that little bead? That’s not even the gold. That’s the drop of silver you put in at the beginning. Once I saw actual gold in there and the geologist was arrested shortly thereafter for spiking his samples.

Then you give these to a real chemist who dissolves them in acid and fires them through a spectrometer of some kind to get the final results.

What’s not to like? Furnaces, molten lead, broken glass, and cooking. Best job ever.

oracles for everyone

Okay here’s something I’ve done since forever. So long now that it seems obvious but it’s not. And it’s an RPG superpower if you’re the kind of ref that’s doing a lot of creating. I’ve touched on it before with respect to the tarot, but it’s far more general than that. I also embed it into games now.

Often all you need to get going is one kernel of an idea. Not a good idea and not a whole idea. Just an idea. And if you give yourself two, that’s even better. Once you have a couple of things (let’s call them nouns) you’ll find that your creative brain can take over — why do they work together, how do they clash — and build something interesting.

So let’s say you want to create a new monster. As with anything you want it to be about something, to have a theme, to have some cohesion. What I find (and find in others) is that if I start with a blank page I fall back on good ideas I had in the past. That’s fine, those ideas are still good,  but I really want new ideas. I want things I haven’t thought of before. And the best way to do that is randomness.

Out in the interverse there are tons of noun generators. They just pick from a random list of nouns. This turns out to be the most useful thing ever. Now, keep in mind that not all nouns are created equal. In fact only about one in five is worth your attention. But if you set that generator to make ten at a time and click a couple of times you’ll find two that excite you together.

These are oracles. Random words that trigger more creative work. You start with a context (monster, plane, city, disaster, whatever), you generate the oracle (or oracles) and then create.

Let’s make a monster for a wrecked space station (always have a context). This is not pre-cooked so it’s a genuine test of whether this works or not. It doesn’t always.

First run of ten I get:

  • Begonia
  • Chit-chat
  • Choice
  • Cost
  • Lion
  • Motion
  • Oval
  • Production
  • Vacation
  • Welcome

Lion is the obvious one so I’ll skip it. I’m torn, weirdly, between “welcome” and “motion”. “chit-chat” and “choice” are luring me too. Let’s take “welcome”.

Second run I get:

  • Cylinder
  • Diploma
  • Drum
  • Glut
  • Jam
  • Karate
  • Membership
  • Plot
  • Somersault
  • Tsunami

I like “cylinder” here, though I’m also thinking cool things with “tsunami” and “drum”. But let’s stick with “welcome cylinder”.

This is a pretty straightforward one. I’m picturing a hovering metal cylinder that’s designed as a kind of maitre d’hotel, a greeting bot. It’s smooth, featureless, because it’s designed to be deployed in a formal setting and it can be decorated as needed. Its sole job is to make people comfortable in their surroundings, to anticipate their needs, and get them situated.

IMG_20170907_131718But this is a monster. So it used to be a greeting bot.  Now it’s out of place in this context I’m trying to populate with monsters. Let’s say it’s a wrecked space station. So one of the things that still works in this damaged and decaying space station is the greeting bot. It really wants you to take a seat in the dining lounge. It’s very powerful, with effectors that can push and pull you with great strength. And the dining lounge is formal attire only — black tie, no space suits.

But the dining lounge is in vacuum.

Well that worked out! I got more than just a monster, I got a whole story. Oracles are magic.

I use these a lot in Soft Horizon games — there are a bunch of tables that are designed to just give you cues. For example, in Sand Dogs, one of the conceits is that the gods who sleep in the tombs have left their garbage around and it’s all intensely weird, unknowable. All anyone can know about a piece of godjunk is what it looks like and how it obviously interacts with its surroundings. Yes this is directly from the Strugatsky brothers but it’s not new to them. And this is still in flux but here’s what we have right now — there’s actually more but we start with rolling two descriptors, an adjective and a noun. Roll for each:

  1. Empty * Container
  2. Blue * Spindle
  3. Full * Page
  4. Jittering * Fuzz
  5. Golden * Ring
  6. Round * Cover
  7. Black * Limb
  8. Dead * Eye
  9. Ceramic * Engine
  10. Pliant * Tube

So if I roll 3,9 I have a strange artifact best described as a “full engine”. What is that? What does it look like? What does it do when it’s just sitting there.

There are more tables to help figure out what it “does”. Maybe it heats up, maybe it repels things, maybe it just hovers there. The point is, it’s weird, it’s inexplicable, and maybe it’s useful. Even if I can’t think of a use for it, I bet the players can.

In the Soft Horizon handbook there will be more general oracles. Certainly each plane will have an element, a single overarching concept that determines everything else about the plane. In fact since its inception this has been the one thing Soft Horizon has always had — planes have a theme, an element. An oracle.


People who love and play setting-heavy games are a whole ‘nother culture to me. It’s not for me, I don’t get it, and of course I acknowledge that they exist and that they are having fun and that’s awesome. I don’t like football but I don’t think you’re stupid if you like it.

But what kinds of setting are there? I can only talk about what I’ve seen and what I’ve made, but I think it covers the gamut. Shout out in the comments if you’ve seen something else.


There are games that present a full history. There’s a timeline of events that have happened and there are places and people (specific people) defined. The world is well detailed (though necessarily not remotely as well detailed as the actual world so there is still room to create though personally I have trouble finding it). Many licensed properties fit into this category.

This kind of game suffers if the players are not at least a little invested in understanding at least their corner of this world. And the ref of course needs to have an even greater understanding. This is essentially why I can’t game like this: I have a no homework policy* and this kind of setting demands homework. Like right off the starting line: you need a certain amount of information to create a credible character.

If you don’t do the homework or (worse, you’re like me) don’t care about the setting details you can still have some fun — lots of people are thick about their surroundings and current events, so there’s no reason your character will have to get everything right — but the other players will carry you. And they need to be prepared for that and willing to tolerate it.

This sort of game is perfectly inaccessible to me but when it works it creates wonderful rich narratives because of all the built-in touchstones to grab on to. It does not suffer a lot of player-created input. You pretty much have to be in love with the fiction.

shitloads of maps

There are settings that are well elaborated with lots of fiction and essays but are nonetheless deliberately incomplete. There are places, perhaps, but not people. There are past events but not an explicit timeline (though look out: one may develop over time as the property develops). These games provide you with a lot of rich material that you can pull together to make your own variant on the writer’s intended world. There’s lots of creative work that can be done by you and yet there are also solid touchstones to lean on.

I’m thinking of games like Reign and Hârnmaster here (though Hârnmaster maybe got a little out of control as it developed the Encyclopaedia Hârnica).

These are fun. Often you can throw out the setting if you like since it’s rarely integral to the system. You can do only as much homework as you want and fill in blanks for yourself. And maybe most importantly, since there’s not much of a timeline, the other players (the not ref players) don’t need to do a ton of homework. Sure, it’s probably a good idea to read up on the dwarves if you’re playing a dwarf, but it won’t be critical. You’re unlikely to make some horrid social gaffe because there’s just not a lot of detail to get wrong.

These give you a lot: they give you a map. A sense of place. Without that gift being a burden. This kind of setting is a great place to start, in my opinion: a lot of hard work is done for you and in enough detail to get you excited about play. A good map is worth real money. They often come with a burden of fiction — usually bad fiction — but you don’t have to read it. It’s atmospheric, illustrative.

implied place

Traveller subsector, generated at randome with the rules. A whole universe of new places.

Ah, my personal preference. There are games that imply a place — whose system generates or is just perfectly consistent with some sort of platonic ideal of the setting. You don’t need to read a lot of essays about how the universe works nor do you need to understand your place in a timeline: the system, used as directed, ensures that anything you do is consistent with the setting.

While I think for sure that Dungeons & Dragons fits in here, I think the first game to really score a direct hit in this category was Traveller. In its first incarnation it was settingless — you were encouraged to play Starship Troopers inspired games or Star Wars inspired games. Anything with lasers and shit. But there really was a core setting hidden in there — the character generation made middle-aged people with careers. The starship creation system created very specific kinds of space ships. And the world and subsector generation system established a specific kind of place — but not a specific place. My own ideas about what the Traveller universe should be are consistent with the rules and to some degree generated by them.

You can pick up the game, play according to the rulers, and the implied setting is delivered to you. For the players this is a zero-homework game. For the ref there may or may not be some preparation required. In the best examples the prep is mechanically directed to some degree. This is where Soft Horizon games sit.

no place

And then there are games with no place, with no implied setting. These are games that, rather than try to give you a place to live, give you a playstyle to play. Anywhere you want to enjoy that play style will work. Since the burden is on you to build these places, these games often have substantial support in the form of supplied settings. The two richest lines have to be GURPS and Fate.

These games are usually described as “universal” but they aren’t. Yes, the setting space is infinite but the play style isn’t — each has a very specific style of play delivered by the system as these things must be. Certainly there is some flexibility — there are often ways to manipulate the system to vary somewhat from the expected norms of play — but if you play the system you get the game, and these systems deliver particular kinds of play.

But not setting.

So if you play a game set in Middle Earth using GURPS, I guarantee you will get a very different experience of play (and a very different narrative) than if you play it with Fate.


As with any categorization scheme, this one is bullshit. These aren’t strict categories so much as they are modes in a distribution curve — there’s plenty of fuzzy space, gray areas, and overlap. But they do help me identify what I like and why I don’t like what I have come to dislike. For me, the games I keep coming back to and inventing are those with implied place.

Do your favourite games fit into this scheme? Is there a space I’ve missed that’s not clearly an overlap or an outlier?

  • I have a no homework policy: I never expect the players to go away and do stuff on their own between games. I’ll elaborate on this later but basically it comes from the experience that no one does it anyway, when they do it’s not to my liking, and basically that trying to schedule other peoples’ time for frivolity away from the actual frivolity is bullshit.

2018 in review

2018, despite being a shit year for pretty much the entire planet, was a good year for the VSCA. For me.

On the home front, though Jackie’s mobility and other functions continue to decline thanks to Multiple Sclerosis (you want to help cure something horrible? Send your Christmas money to the MS Society of Canada) other issues are largely under control. My place is still a mess, my time soaked by pretty much everything and Jackie unable to assist, but that’s really a minor issue. There were no major hospitalization incidents and mania and psychosis are under control. That’s pretty good. That lightens my load a little and is a big factor in the rest of this review.

This project also got me drawing again. That’s a Really Big Deal.

Elysium Flare got released! And it’s my first full-colour project. We even managed to bring in some art from Colombian art-genius, Juan Ochoa. It’s a really lovely book and a tidy little Fate-based science fiction game. It features a lot of loose setting material out of my own head and represents my best effort at grabbing some enthusiasm for space opera as a genre. Now look, I had fun making it and I think it’s a great little game. And it’s beautiful, especially in hardcopy. But I’m not likely to do space opera again unless it’s pretty psychedelic and I’m not likely to do another Fate game.

This project also got me to figure out the Drive Thru RPG POD system. It’s convoluted and frustrating but it works. As a publisher experience, Lulu is much nicer but you can’t beat the integration and the storefront of DTRPG. Both make nice books.

We started and then abandoned a Diaspora second edition. That’s probably not ever going to happen.

I started the Patreon to keep this new effort running and it’s been very powerful for me: feeling like I owe people regular product makes me make it. I can’t emphasize enough how well this works for me. Thank you to all my patrons: you have made this a great year.

lazer-owl smallI almost secretly released a little zine-sized product called anomaly digest. It’s for sale in hardcopy only and it’s at cost — there’s no profit in it. Add it to your cart when you buy everything else! It contains a number of little adventure hooks and mini-dungeons created as part of the RPG Talk (that’s a Discord invite link) regular content contests. I don’t participate in these any more since it kicked me into making games and that’s where the energy goes now. This went our as a physical reward to some patrons.

bonobo anthropologistAnd finally the Soft Horizon gelled and I got out the first issue, The King Machine. A game I’m really proud of, but I’ve talked about it elsewhere.

And now the second issue is out for playtest (really a textual playtest — the game itself is already heavily tested but now I need to see if the text works). So Sand Dogs could be out in January.

And then there was the impending for real death of G+ which forced me to make some plans about where to communicate and get communicated at. That bore this blog, a remake of my old one (and I’ll repost some stuff from there occasionally), and some exploration of other spaces. So far I can be found at:

soft horizon: the basics

Capuchin scout
In The King Machine you can play a capuchin monkey. I mean seriously, what else do you need to know?

At the heart of the games The King Machine (available now) and Sand Dogs (available soon) is the Soft Horizon system. This system is designed just for me: it eases my stress as ref.

The heart of the resolution system is very simple, and very amenable to play by text chat (which is how we playtest). Every character has nine stats (or skills or whatever) called methods. Each is assigned a die, either a d6, a d8, a d10, or a d12 (but you need to advance quite a bit to get one of those).

When the conversation of role-playing reaches a logical conflict that merits resolution, the dice come out. Based on the narration first (please don’t search your character sheet and tell me “I use Socialize” — tell me what you do and we’ll work out the method together) an appropriate method is selected. The player gets that die to bring to the pool.

Then the ref will set the risk. This is stolen entirely from Rob Donohue who is a genius. Don’t worry, I told him I was stealing it. It’s not the first time I stole it but it’s the best time I ever stole it. The risk determines what’s going to happen if things go bad. It’s nice for everyone to know this up front.

Then we figure out who else can help. Does someone else have a story to offer that merits addition of one of their methods? Maybe someone has some loot to deploy in the situation (loot has dice too).

All the dice are rolled. The highest die determines the result (this is a lie — actually the player who started all this chooses the die because there are sometimes reasons to choose a die other than the highest). This is simple:

1-3: fail and the risk is realised

4-6: succeed and the risk is realised

7-9: succeed with no strings attached

10-12: succeed legendarily

Realising risk drives the game forwards. It creates new plot twists (revelation, for example, reveals some new information no one expected including the ref). It changes or adds motivations (harm gets you wounded and that becomes a priority to resolve since it reduces the die on half your methods). This is just for me: I like to ad lib but I need a cue and this system keeps the cues flying. If you are the kind of ref that plans a lot in advance, this game will not work for you: the system itself will drive off your rails.

Because the system drives the narrative, the ref’s preparation is simple: list some ideas for what to do in a lull. Here’s the actual cheat sheet for the ref:










Start some shit is fed by preparing a few simple fronts, a streamlined version of the same thing found in, say, Dungeon World.

Set a deadline triggers a countdown clock towards some event. When play addresses the event on the horizon, it doesn’t tick. Whenever someone answers “What do you do?” with something that doesn’t address it, the clock ticks down. Until it happens.

Create a hazard invites you to simply demonstrate how dangerous the world is. Sandstorm, someone slips on the mountain pass, whatever. If the players had their characters in a dangerous environment, this is when you show them how dangerous.

Call in a bond triggers a feature on the character sheet: a bond is a connection between a character and someone else. And that someone else needs your help.

Make a scar a problem triggers a different kind of character feature: when characters heal their wounds they get scars. These can be used to advantage, but the ref can also stir up trouble with that alien artifact you use as a prosthetic hand.

Introduce someone interesting is your chance to bring in that NPC you love. The cheat sheet is where you make that note so you don’t lose your cool character idea. It may never happen, may never be the right time, but this sheet is your quiver and that mechanic who loves to gamble and knows where the rocket launchers are kept is your arrow.

Dry up a resource is any idea you have for making things scarce. Being out of gas or water or food in the middle of the wilderness demands attention.

Recall a missed hook is really just for me. I have a tendency to leave weird shit on the table. An explosion destroys the house you’re all in and although you are largely unscathed the place is levelled. Except for a very fragile looking vase. Now that’s a hook in plain sight but players often miss them. Leave them there, don’t press it. But use this to bring it back into play if things get slow.

Make it night is there to avoid Endless Day Syndrome. Sometimes a game that propels itself just keeps going: there seems like no good moment to break in and get some sleep. So force it in a lull. And maybe that’s when some shit gets started.

The point of these is to reduce my stress. I need a palette of simple one sentence or even one word ideas to draw from in a lull. I have to say though that the system generates so much ongoing movement that I rarely have to pull one out. Mostly the players generate all the plot I need to keep things moving at a very fast pace.

Obviously there’s more to these games than this and each has its own setting-specific variations and oracles to help get into the feel of the game. But this is the core and if you know this you can sit down and play. You can pretty much sit down and run it.