For the sake of demonstration, let us construct a hypothetical game.

(Don’t worry, in this case, “hypothetical” means it won’t be a whole thing, and by “us” I mean me. You don’t have to lift a finger.)

This game contains, among many other components, the following snippet of rulestext:

“When you activate an Action, suffer its Consequences.”

This tells us a few things about the game right from the get-go. We know at least two defined components of the game: Actions and Consequences. We also know that each Action has Consequences specific to it, that Actions are activated on some occasion of the play, that Actions are enacted by individual players and their corresponding Consequences applied to them, etcetera. The game has denoted two important concepts to us, and in so doing also given us more insight into their interconnection, and existence within the broader context of the game. (Which we still know nothing about beyond this one line.)

Now consider a second game – similar in structure and goal, sharing much of the same DNA as the first, but still distinct in its specifics. This second game contains, instead, this snippet:

“When you Activate an Action, Suffer its Consequences.”

This tells us more than the prior one did, alongside everything it did. We have two new keywords, words designated as Important, to work with – Activate and Suffer. And precisely because they’re designated as such, we can’t just fold them into the presumed operations of the previous keywords. They mean something now. Can you Suffer other things, in other contexts? Can you Activate things that are not Actions? Or, perhaps, is Activating an Action a more specific thing than we might consider as “using” it – perhaps a given Action will Activate itself three times over the course of its resolution, prompting three instances of Suffering said Consequences. Or perhaps there are multiple ways to “use” an Action, of which Activating is only one – you could Gamble it, or Belay it, or Activate it, and only the last prompts you to Suffer the Consequences – ooh, but maybe Gambling it prompts you to Inflict its Consequences, a wholly different operation using that same Consequence field, and-!

Well, that’s enough of that for the moment. The point isn’t to actually build these out into games, as potentially compelling as some of those thoughts may be. Instead, it’s the simple note that, marking out words as keywords gives people insight on how the game works when they read it. Whenever you designate a given word as Important, in whatever format you use – it doesn’t even have to be capitalization, it could be italics, bold, or however else you want to make the word pop out – that tells the reader that it means something, and as they navigate the rest of the game they should keep an eye out for what it does mean, and what it can tell them about the rest of the game. And then, when they get to the point of sitting down and playing, they’re all primed and ready for exactly what Activating an Action means, and what they should worry about when they do it!

Of course, the fact of the matter is this is a question of presentation. “When you activate an action, suffer its consequences” is just as feasible a line for either game, and could mean everything the given lines meant, and it could also just as easily read “When you apply Operation X to Type A, apply Operation Y to Type B” and, on a mechanical level, mean the same thing. But I think you can already see the problem in both approaches. The former is, while fine, less indicative at a glance. Is suffer an important keyword, like it was before? Is action, even? We can guess, and are more likely to get it right the more we know about the game, but think about how much we got, both of explicit knowledge and of potential implications, from out prior examples. As first impressions go, it’s much easier if you get told which words are directly meaningful and which ones aren’t. And as for the Operation X approach, well, that’s all well and good if you’re plugging it into a computer, but the players will have to read that, sigh, flip to the glossary to figure out what Operation X does, flip back, then do the same three more times to identify the types of items you’re calling for and what you’re doing with them. A player can read “Suffer” and get the gist of what is supposed to happen even if they don’t know what it means in mechanical terms – and that’s going to help things run much smoother even once they reach the point where they do know that much.

That’s not to knock the Operation X approach entirely, of course. It’s certainly not presentable, and a headache to navigate, but as a design step, it’s honestly quite useful to sit down and write out all the processes you have for the game, and what, in mechanical terms, they do. That way you can see what components have the most hooks into them, which are relatively extraneous and could be integrated further (or removed altogether!), and the general flow of operations as you go through the game. Initiate Fight to Activate Action to Suffer Consequences.

That model of operation-flow, and what parts of the game are called upon by each step, is gonna end up looking like a weird engine diagram, and that’s precisely what it is. A game is an “engine of meaning” – a device that flows between concepts by using operations that are themselves also concepts, with tokens and progress marks and dice rolls and hit points exchanged to determine specifics of how it flows when. What you’re staring at, that diagram you’ve made, is a map of the engine you’re building, and what parts you gave for it. And you can see better that way if a certain part is top-heavy, or overstressed due to high traffic, or what have you.

The discussion of keywords is a question of how you translate that to language.

Most players, if you give them a flowchart of processes, are gonna find it rather obtuse. The more complex it is, the more useful it might be to see it presented in that way, but, readers like words. They want to be able to flip through the book and see, written out, what you should do when. And that’s why you get lines like “When you Activate an Action, Suffer the Consequences.” The flowchart is for your benefit first and foremost – it’s important for players to understand the shape of the game they’re playing and how it flows together, of course, but ideally in broad strokes. You can tell them that sometimes the game transitions to combat, and combat is a sequence of Actions and the effects of them, ramping up to a big Finale that ends the fight, and then it swaps over to the politics in the commentator’s booth mode of play, and really, that’s enough. The specific, moment-to-moment operations get to go in their own section of how you resolve Activating an Action, how you resolve Suffering, etcetera. By breaking down the concepts into small chunks that the players can get used to on an instinctual level, they can read a sequence of words that’s actually describing a whole chunk of the flowchart, and still get it.

There is, in some game design spaces, a sentiment I often encounter that this kind of thing is to be avoided at all costs. That it takes the reader out of the experience to read a suddenly capitalized word, to be told by the game that it is Important, that when it says “make an Attack” that means something mechanically specific. I understand this impulse! For people far less used to building mechanical things as their primary perspective on creation, it is jarring to see that. And writing things out in natural language that makes for easier-seeming prose can feel like the sensible course of action. But, the flowchart of play is going to be a truth either way, and writing it in unclear terms will make that harder for the readers to understand. If some uses of the word “attack” are not an Attack, and some Attacks never use the word attack, the player will struggle to have a concrete “here is the process of Attacking and I know when I should do that” in their head. And even when they do get that, the edge cases will make the worst effects of that fraying shine, and you do not want that.

The fact of the matter is, on an initial readthrough, a player will encounter “make an Attack” and will, inwardly, think “oh right, this is a game”. But the second time they read that, they won’t think that, they’ll think “oh, I know what to do when I Attack”, because of the clarity that a keyword brings. They can construct, in their own head, using their own mental infrastructure, their own flowchart, their own sense of the shape of the game and what each cog and bellow does when they call upon it. And once they’re clued into that, you can write things like “Resolve: Activate, Attack, Activate; Consequences: 1 Harm”! Seems like a string of gobbledygook, but to the player who’s already read that far, they can easily parse that that’s an action that makes an Attack and causes them to Suffer 2 Harm – 1 before the Attack, and 1 after. And then, when they’re down to 2 Harm left to take, and there’s an enemy that will go down to one Attack, they know they can use that power to win, and their thoughts are flowing precisely through how that flowchart wants them to.

It’s a handy feature of games that the substance that composes them, meaning, is so easily communicable. Each player can have their own copy of the device they’re toying with inside their head, and unlike with examining blueprints of a car or the code of a program that insists on crashing on launch, the engine can literally be inside of their head. For most players, showing the blueprint of the game and the flowchart of play won’t do that, but being clear about the components, the processes, and how things are expected to run, that can translate it much more effectively than it might seem.

That’s the power of keywords.

If you get squicked out by kinda-disgusting descriptions of anatomy, maybe skip this one. Speaking as someone who fits in that category, why did I write this?

“Mechanics are the bones of a game.” That’s a thing I say a lot. It’s pithy, and short, and kinda communicative. Enough for a quick remark that people can nod along to and move on. But it’s grown in my head to a whole perspective on how to build games, to the point where I think it’s doing it a disservice to not present the whole grisly image. Which is what this will be!

A game in motion, being played by people, is like an animal. It has ways it moves and wants to move, things it wants to go and do, as a product of the pitch and inner workings of the game and people playing it. Maybe it wants to go frolic down to the lake and sip up some of that delicious water of tragic-vengeance-quests-ending-in-a-hollow-victory, or it wants to munch on the higher-altitude leaves of the tree of killing-god-as-an-analogy-for-coming-out-as-trans. That’s great! We love to see a strange new beast taste of the fruits of this world and learn of good end evil.

As designers, we aren’t pet owners, or caretakers, of this creature we seek to gift life. We aren’t even gengineering it, really. We’re building it. We splay out a bunch of diagrams of bones and muscles and “how the fuck do lungs work” onto a table, draft up some plans, and spend the rest of the time frequently panicking and redrawing the plans when we realize that the poor thing needs to vastly increase its oxygen intake or it simply won’t survive.

I say mechanics are bones because the big thing they do is define and support the shape of the creature. Can you make something with very few bones at all? Sure, we’ve all oohed and ahed at Lasers and Feelings, and if we consider its one mechanic to be like a backbone, it can be a pretty nice snake-worm-thing. I love snake-worm-things! But if they try to grow out, or up, or too big, that one thing isn’t going to support all the meat it wants to have. Unless it has other strong supports, it’s gonna collapse. There’s only so intricate a game of Lasers and Feelings can get, and it’s not trying to be intricate.

Suppose you want it to be more than a snake-worm-thing. You wanna give it a lump. Okay, fine, a limb, something useful, whatever. If you don’t extend any bones into its new fork-tendril, it’s kinda gonna struggle and flop there, and be limited in how it can move. It needs strong musculature to compensate, which, I’ll get to that in the analogy a bit later. So, you go, “okay, I want some space politics intrigue in my Lasers and Feelings, and I want it to work with the mechanics, let’s figure something out for that!” And you make a little faction sheet each with their own laser-to-feeling ratio and maybe a little token economy they spend on influencing the PCs to do things, and hey presto! The snake has an arm now, and it can finally fulfill its lifelong dream of punching a rock.

Unfortunately, the players don’t really care about punching rocks. Or, the space politics. So they just don’t interact with that, except for when the mechanics absolutely force them to. Whoops.

So now there’s no meat on that bone! It kinda just dangles there, while the snake-worm-thing slithers about like it would normally, and occasionally its bone-protrusion bangs into a passing rock and hurts it a bit. The thing’s worse off than if the bone wasn’t even there, and, realistically, for a game as simple as that, sooner or later the players are gonna talk and agree to just snip that part off.

See, when a designer’s done with their mad biology and maniacal cackling and “I’ll show those fools at the institute”-ing, the end result isn’t a fully living creature. It’s just a game. Like I said, a game in play is a creature. ‘Cause the players, and the specific dynamics they’ve got, bring another important component a designer can’t make all on their lonesome: muscles.

(Yeah, I said it. No game designers are swole. Not a one.)

A muscle is a way for the creature to move. It’s a way it wants to move. Specifically, it’s a way the players contributing the muscle want the game to move. They all want it to move somehow, since a game that’s still is a game that isn’t doing anything, and if the players wanted to sit around not doing anything, they could just do that in a room or discord call or some such, put on some music, hang out, and call that a day. And honestly? That’d be pretty nice! If you just wanna do that, go for it! We’ll be here grumbling and swearing vengeance against whatever part of the rpg space we imagine to be the foolish academics who cannot comprehend our brilliant but dangerously amoral designs.

But if the players do want a game, they want a game that moves. They want Lasers and Feelings to slither about through the grassy plains. And maybe they want the space-intrigue-arm to reach out and punch rocks! Or maybe they don’t want that at all. If they don’t want it, like I said, they don’t bring the meat to the bone. The arm hangs limp, and does nothing beyond get in the way when it moves.

But what if they do want it, and the bones aren’t there? That is, what would happen if they’re playing Lasers and Feelings normally, with no modified skeletal structure, but wanna focus on space intrigue? Well, bones indicate to players where the muscle might want to go, but players are gonna do what players do. They’re gonna push out a lump-arm, bones or no.

This can work! Kinda. Octopuses have tentacles and elephants have trunks that are supported all by musculature, and don’t need any bones. And there are always microbones – established tone, the frustratingly-unreliable “what would make sense to happen here”, even as simple as the play-by-post etiquette of “player one posts, then player two, then repeat”, all of these are mechanics! Freeform roleplay is not truly ruleless, even if the rules are often left to emerge from community norms. But those aren’t enough to support a whole arm unless the muscles are carrying a lot of the work, like a tentacle. The players have to really go “wow, let’s examine a lot of the power struggles between these space factions and dedicate our focus and thought to that” to grow the snake an arm with no bones in it.

Now, the fact is, the simplest answer to the players who would play Lasers and Feelings and Space Politics and not care about the space politics is “hey, what are you doing, literally just go play Lasers and Feelings“. And it’s important to note that, a lot of the time, that won’t need to be said, the players will do that on their own! Mechanical literacy in players can vary quite a bit, so it’s not something you can always consistently rely on, but mechanics are a way of signalling “this is a thing that is Important to the game”, and players will usually pick up on that enough to go “hmm, this intricate flower-arranging minigame with hooks into every other mode of play doesn’t exactly fit what I want out of Doom Tabletop, I’ll try something else”. (This hypothetical player has terrible taste, and I want to make Doom Tabletop Featuring Flower Arranging now exclusively to spite them, but, so it goes!)

As an aside – this is the core of my dissatisfaction with a lot of the OSR design ethos. (With the disclaimer that this is my understanding of what the OSR design ethos is, and I know that’s somewhat of a contested topic.) In short, the idea is that combat and health-attrition have the most hard mechanics, with other interactions favoring a “rulings over rules” prioritization of making the GM do all the actual design work and arbitration. The thing is, combat is implicitly to be avoided – it’s generally punishing, and pretty much always viewed as a punishment for failing to figure out a way to bypass the combat. Which is fine in its own right, but it means, if the players actually do play “correctly”, it’s a big ol’ ribcage floating in the open air that contains nothing! It’s the snake’s fleshless arm banging into rocks again. The upshot is that either it’s a mechanic you’re never supposed to interact with, or the game is supposed to be set up where the players can never actually do it “right”, and then they get tossed through the ribcage of you-fucked-up in a way they couldn’t avoid, so what’s the point.

Right, grumble over.

Designers are, like I said, beast-builders. Bodybuilders? Wait, but I already said none of them are swole… whatever! The point I want to emphasize is what I didn’t say – that they’re skeletoncrafters. That that’s all they are. ‘Cause it’s not! We do make skeletons, make mechanics, and slot them together, and hoo boy do we grumble and yell at each other about them, and go “ooh, that bone shape is real neat, I gotta remember that one”, but a game is more than just a jumble of mechanics. And a thing to remember about muscles, however insistent the players may be about what they want – muscles need blood to get to them to work. If they don’t have that, they can’t move.

The veins of the creature are its tone. Which, you may note, is a thing designers totally set up! When you write your paragraphs about the Vampire Queens and how the scars of their Profane Right have afflicted the stars to this day, such that every PC has a constellation in their eyes that shall eat their heart in a year and a day, yeah, that’s setting things up for what people expect! People are gonna aim for broody vibes, maybe some angst about impending mortality, probably some edginess, and then if they find out the actual mechanics are about courting your princess of choice for the upcoming ball, well, that might not be where they expected those veins to go. (But, that’s a matter of how you present the prose! You could totally lean into the angle of a temporary romance between those poisoned by the blood rites of the past, and then the tone would be communicated a fair bit more. It’s all in the presentation!)

Except, that’s a lie. It’s only some in the presentation. The rest is in the potluck – the handful of veins the players brought to the table, to stick into the creature as it is being made. They bring the muscles and you bring the bones, but both of you give the veins. Which is why the communication is so important! If you say “broody vampire game”, but don’t tell the specifics of what the point and focus is, well, they’ve got their own ideas of what broody vampires are like, and that’s gonna affect what veins they bring. You can’t control it, either. The one who reads a lot of Blade comics might bring a lot of veins to push towards the cool combat and gratuitous violence wing, and when they find out later that the game doesn’t have a combat wing, it doesn’t have any wings, it can’t even fly, well, that’s a thing to catch before they’ve gotten themselves worked up about the veins they plucked fresh this morning to bring to the potluck.

But hey, even with that example, you can present the whole schtick, and players will still have their own bits of tone to bring! Maybe their thoughts of romance trend light-hearted-adventure-swoop-them-off-their-feet-y, and the impending mortality of the player characters gets sidelined. Or maybe the players really wanna play up the angst, and how their courtship is inherently temporary! Maybe the cutthroat intrigue between the Vampire Princesses sounds most compelling, or maybe they just want a nice moment for all of them to share before half of them die. The beast template you provide them will nudge in directions, to be sure, but it won’t control which ones they take, which veins they bring to the table.

And it shouldn’t! You don’t know the tastes of every group of players who will pick up your game. They might like parts of it you consider only ancillary, and not care about the things you’d say are the point above all else. But if what you’ve got works for them, and with their tones, and their muscles, they can make it live, and set it loose to run and frolic and live, and that’s a beautiful thing.

It’s sometimes hard to remember, when measuring bones down to the millimeter and debating which vein shape will transport our brilliant new hemolymph cocktail and shaking our fist at the ingrates who dared to mock our grand vision, that what we’re making is gonna be out there, eating fruits and basking in sunlight, but we can’t be the ones to bring it all the way there. It’s the contribution of players that makes it live. And that can be a beautiful thing! But, among other things, it’s an important design constraint. If the shell of a beast you make can’t align with what players will think they need to bring to it, the best it can do is limp along. If a bone is jutting through a muscle, or the veins lead out to nothing at all and the blood spills on the ground, that poor beast will suffer, and won’t truly get to live.

And you wouldn’t be showing those fools at the institute that way, now would you?

(Postscript: oh no, I might actually want to make that vampire game. It sounds neat. This is the danger of giving examples, folks!)

(Fair warning, this one’s pretty much just an angry rant)

So there’s a refrain I’ve heard a bit too often in game design spaces. “You’ve convinced me that it’s bad, but I’m still convinced that it’s fun.” Usually in regards to, like, “hey, this power gives the players a giant headache and screeches the game to a halt as they have to cross-reference several things and completely change how they approach the game”, or “hey, there is literally no way people have any impact on what happens here, so instead of supporting any interesting gameplay dynamics, you’ve essentially just made a gambling machine where the reward is how much you get to play the game”. So, y’know, serious, structural issues with the thing they built.

Thing is, the “fun” defense is… nebulous, at best. “Fun” pretty much means anything people can enjoy, ie, anything, and more specifically when making a game, it’s “the specific kind of enjoyment that this game is designed to induce in its players”. Now, the mechanical framework and flow of the game is how you induce that joy, so, a disruption to it is at baseline a disruption to that, but, you can’t really conclusively prove it won’t produce the brain chemicals, y’know? It might not enable any interesting player incentives, or it might enable actively perverse ones, but, what if that still feels good despite the structure of the game you’re making?

The trick is, that’s not really a useful response to feedback, nor does it actually help give a sense of where to go with a thing from there. If a thing doesn’t give much room for any interaction, or the interaction it incentivizes is fundamentally not good for the dynamic you’ve set up, even if that is fun for players, that means you’ve got a new headache of figuring out why. You’re trying to induce a certain experience in players, and the mechanical framework you’ve built up is supposed to do that – if you make a thing that doesn’t have interesting hooks in, and players enjoy that, that means you’ve gotta shift your framework so that this thing does hook in, and potentially change how the hooks work entirely, ’cause oh no, people aren’t enjoying that as much as this thing that’s a disruption to it! But, the thing is, I rarely ever hear this in response to genuine play feedback. This is from the ideation phase, when trying to think of ideas to fill out a content roster before giving them any testing. And usually, it’s said in resistance to the realization that an idea is not worth pursuing further.

(To be clear, this isn’t just the concept doing the defending. Designers are stubborn lots, and it always stings to have an idea that seems cool and then have it pointed out that it’s fundamentally not gonna work for the setup you’ve got. The impulse to reach for a defense is always gonna be there! The trouble is, it’s decidedly one of the easiest to reach, and it does bad things.)

“Fun” is, in game design terms, doing one of those linguistic tricks where it hides more specific and useful analysis. I’m a big proponent of analyzing player incentives with everything you make – including the incentives you’re relying on that drew players to your game in the first place. In a game where the pitch is being haunted by the ghosts of your past murder victims until you succumb to your guilt, it might sound “fun” to let you have a power that has a random chance to double-murder a ghost, eliminating it for good, but if players came for a story of clinging, eternal guilt, it’s not gonna be. And if it is fun for them, that means what “fun” is isn’t aligning with the original pitch, and you should reexamine that! If players do want the ability to fight back, then that’s now gotta be a part of the game concept, and one power with a random chance nobody has control over isn’t gonna cut it. Can you pivot the game to now being about managing to forgive oneself for one’s past misdeeds, and give the ghosts defeat conditions? Sure, if that would still fit what you want from the game! (Remember, sometimes the answer to players wanting something different like that is “that’s not what this game is going to be, it’s not what you’re looking for, you should go play something else for that”. Make a game to match a concept, but the players have to be willing to align with that concept, too, in order for them to play it right. Figuring out that concept and building for it is what this art is all about!)

What do they want, why do they want it, how can the framework be modified so that something like that can fit with it, should that be a thing they can get considering how it can change the gameplay and themes of what this game is. Those questions are very important to ask, and all are buried under the word “fun” in a most insidious way. When analyzing parts of a game, and when building them, you want as much perspective as you can, to see how the parts fit together and what can be discarded or turned. If you think a thing will be fun, but it doesn’t play well with the framework you’ve got, why will it be fun? That’s not a rhetorical question, it’s a serious one! Analyze and dissect the concept of fun until it’s splayed out in front of you on the anatomy table, and figure out what path of incentives and effects you want the thing you’re building to send your players on. If you conclude the worst, you can just set the thing aside and have it be the seed for a new project in future. Just ’cause you’re committed to the Silent Hill ghost-guilt here, doesn’t mean you can’t make a cool psychopomp-with-a-shotgun romp later down the line, if that still tickles your fancy! It’s just important to recognize that that should be a different thing.

One final note, which a friend brought up – the corollary to this, that because “fun” is a concept that’s more harmful than helpful to consider without elaboration, so is “not fun”, isn’t true. Much of the time, it can be – if you get feedback that an ability isn’t fun, your first step should be to examine what it incentivizes and how that plays with the framework at large, but, sometimes it isn’t that! Lancer, which is a D&D 4e-derived mech combat rpg, has one of the most elegant examples of this. There are two defensive ability types that come up I wanna analyze here – Resistance, which is just “whenever you take damage, halve the amount”, and Invisibility, which is “whenever you are attacked, there’s a 50% chance the attack misses”. There are many foibles and interactions that make these not quite analogous in expected result, damage that doesn’t come from attacks, attacks that deal damage even on a miss, attacks that don’t deal damage but do do nasty things on a hit, etcetera, but the short version is, hey, if you’re getting shot for X damage, the expectation value from having either of those effects is X/2 damage. Different dynamics, similar oomph. Sounds good, right?

Well, it turns out, no. Because everyone complains about Invisibility. Invisible enemies? Utter bastards, hate them every time. Invisibility-granting PC systems? Bane of every GM’s existence. Does Resistance have this same ire? No, of course not! The reason is very simply, if Invisibility does work on you, you lose everything you worked for. Did you unload a gun, or overheat yourself, or move out of position, to try to attack a target, only for Invisibility to mean the attack doesn’t happen? That’s all loss, no gain of any kind. You took a gamble, and you didn’t really have any control over that gamble, and you lost, and that just doesn’t feel good to people. If you’d gotten a half damage hit, that wouldn’t feel anywhere nearly as bad – because it means you get something. Brains like getting things when they try things, and get immensely upset when they don’t.

Now, is this a problem for Lancer? Not necessarily. Every attack has an attack roll alongside it, and while the math is intentionally shifted down so those hit more often than they miss, it still means every attack, Invisible or no, is something of a gamble where the loss condition is just “no forward advancement occurs from this action”. Having played Panic at the Dojo once, I have become infatuated with tactical rpgs where that isn’t a risk, and were I to make something like that myself, I would take cues from Panic in that regard, but Lancer isn’t doing that, and, that’s more than fair! It’s a game where that kind of gamble is part of how a lot of effects work, and that’s central to how it’s built to function. When you find out that something like that is getting consistent “this isn’t fun” responses from players, that can be your answer! But, if you’re making your own Lancer, and don’t know if misses actually add much to the game, and then get feedback like that? It’s something to consider!