A few weeks ago in an update to the FiveEvil kickstarter (late pledges still open!!) I contributed a short post about something in the optional rules that I call Netherworlds.
It was originally a long post. I thought I might go back to the earlier version and finish the bits i didn’t get done, and release it here as a bonus feature. A special edition! In what follows, the bit between the ~tilde~ separators is Brand New Extended Edition Director’s Cut Extra Content. It takes the original 1,000 word post and turns it into a 7,000 word post, like if Jim Cameron released the Aliens Special Edition and it was sixteen hours long! Yes what a great idea I had to do this hurrah!
Here goes.
I’ve spent the last week revising chapter nine of FiveEvil: Fiendish 5E Horror, and it gave my brain a real workout. Getting every piece of the game working to a high standard is non-negotiable and this section threw down some real challenges. Solving these game design puzzles after much work was very satisfying!
Chapter nine is the rules options, a mix of rules pieces for different needs. Some of these rules are for situations that don’t come up in every game. Others are there to shift the feeling of a game this way or that. For example, the rules for improving characters are in here; be warned, they don’t look much like leveling up.
The bit I found a particular grind to get right was the section on what I call Netherworlds. It’s a decent chunk of writing (it might yet get broken out into its own chapter! We’ll see!) focused on those times in scary films and books where the main character enters the domain of the horror. Think about the expeditions into the sewers to find Pennywise in It, exploring Curwen’s tunnels in The Case of Charles Dexter Ward, the marines entering the atmosphere processor in Aliens, or what the entire Overlook becomes when the ghosts get active at the end of The Shining.
I’ve given these spaces special treatment, and special rules, because I think they are harder to represent in gameplay than they might at first seem.
~~
The reason why goes to the very core of how roleplaying games work at the most fundamental level.
So let’s unpack that, starting with an examination of one of the clearest examples of visible mechanisms in game design history: D&D 4th edition’s skill challenges.
Part 1: Skill Challenges
FiveEvil follows a model of RPG play that is often called “trad”. It’s an iteration of 5E D&D, which is of course the latest version of the game that started it all. Fifth edition D&D has moved away from the original form of the game in many ways (we’ll get to that) but at its core it operates on the same fundamental assumptions as the first version, and FiveEvil follows along.
Here’s part of that assumption set: there is a central figure, the DM in D&D and the GM in FiveEvil, who is in charge of running the world and determining what happens around some focal characters. Those characters are operated by the other players in the game like elaborate narrativised playing pieces.
And so, the interaction that sits at the core of trad play is this: the GM describes the situation; the players say what their characters are doing; game rules are invoked if necessary to determine how those character efforts turn out; and then the GM interprets those outcomes into the situation and describes where things are now. Repeat until the dragon is slain.
This interaction is the magical bit of gaming technology that was discovered-slash-invented by the people involved in Braunsteins and Blackmoor in the early days. It’s magical because it engages rules to deliver satisfying consistent outcomes, but the procedures involving those rules also transcend them, arcing events up out of the deterministic space where the game rules operate into the sphere of imagination and discussion, unbounded forces that influence the trajectory of their descent back into the defined parameters of game.
The archetypal situation of trad play is dungeon exploration. Your characters venture into a new room in the dungeon, the DM says what you see in there, you pick up rocks and kick over tables, maybe you find treasure, maybe a grell pops out of a hole to attack you, it all works great.
(Actually it doesn’t all work great. We’ll come back to that.)
There’s a reason for this archetype. It works so well for trad play, and trad play works so well for it, because the dungeon room is a perfect match for the affordances of the trad approach. There is a specific, contained environment in which play unfolds, one that can be marked out in advance in relative detail. The environment contains a discrete set of things with which to interact. You declare how your characters interact with those things, and then consequences unfold. You stay there, hovering over the character’s shoulder every step of the way, and find out what happens next in real time. It’s all immediate and you can hold it all in your head.
It follows that the default solution for trad play, when things get uncertain or complex, where people might wish for different outcomes, is to drop down to the level of direct character decisions, and hand every consequential choice to the player. Mechanical resolution tasks are held mostly at this same level, calling for over-the-shoulder tracking to sew them together into continuous narrative. You make every key decision at the level on which these tasks sit, and you keep resolving task after task as needed, until the whole situation has transformed into something that can be agreed as an ending point.
The challenge posed by Netherworlds is this: what if the scope of interest is too big for this solution? Staying in the moment to follow every step, accounting for every left/right choice, and resolving every possible task is not going to lead to fun play. How do you lift up out of those details to address a broader challenge, even though the game has a task resolution architecture set at that over-the-shoulder resolution level? How do you work out what comes next without playing through everything that happens first?
Abstraction is the answer, obviously, and abstraction is not an unfamiliar animal! While most tasks live down in the level of immediacy, some are more zoomed-out. The “Investigation” check in 5E covers a multitude of sins. The “Survival” check, even more so. You can sum up a lot of specific actions with one of those rolls, a whole montage sequence you might say, and spit out a definitive success/failure result at the end of it. Sometimes when you drop down to the level of your resolution tasks, you end up being abstract. Great!
There’s a further step, however. Abstraction within a defined task procedure, sure. But abstraction across tasks? Across an entire project? That gets trickier to deal with. If you’re not stepping through the precise sequence of consequential events A then B then C, how do you advance the narrative? If you’re skipping moments, which ones? If you’re handwaving some task resolution, which bits, and how do you account for that? And in fact even if you are stepping through it, how do you sum the pieces together as you go, are they done after the second check? Maybe a third one? Maybe a fourth? Is it just your gut instinct about when enough is enough? To the presumed delight of Dan Harmon, I think the ultimate challenge of trad play is this question: how do you know when you’re finished?
Which is why it’s time to actually talk about skill challenges.
Part 2: Skill Challenges
WAIT. Hang on. Doesn’t all this sound suspiciously familiar? This kinda resolution is based on tasks, and it’s kinda bottom-up? And for these Netherworlds you’re needing another sort of resolution that’s kinda top-down and has an eye on the overall challenge, or shall we say, “conflict”?
Cough cough number of years ago, this distinction was the hottest thing in RPG theory circles: Task Resolution vs Conflict Resolution. Task resolution (bottom up, focused on specific tasks) was one thing, but wait have you tried conflict resolution (top down, focused on goals and stakes). It was discussed at enormous length and inspired many great games.
It’s also quicksand. Those oppositions noted above, top vs. bottom, tasks vs. goals, those are pretty common shorthand explanations of what is being addressed here but if you have looked into this subject you will already be formulating your sternly worded forum post about why this is incorrect and the correct way of describing what’s at issue is [your own idiosyncratic version goes here]. These explanations are not just insufficient, I think they are actively inaccurate definitions, and in fact to my knowledge no-one has ever made a task/conflict definition that satisfied a majority of folks.
Better like this: task res and conflict res are a vibe, okay? One kinda about the specific thing you are doing and the other about the reason why you’re doing it. Now, this is the important bit: don’t think about it too hard, just design a game with that vibe! If the game works, that’s enough!
And so it proved. The games worked. The concepts that inspired them remained a bit elusive. (What about when your overall goal is… to complete the task? What about when the stakes are… do I complete this task? What about when the task is large and abstract and contains many diverse fictional actions within it, is it a conflict then even though I resolve it like a task? What about how all your tasks are always in pursuit of an immediate goal which relates to a larger goal which relates to a larger goal (and so on)? Isn’t a task just a specific variety of conflict some of the time/all of the time? Etc etc etc)
Here’s how I’d say it: it makes the most sense to talk about task resolution and conflict resolution in terms of design goals.
Design something with task resolution in mind when you want to focus on describing the set of activities the character is engaged in, and whether they do a good job or not at performing those activities, and you don’t particularly care why they are doing those activities.
Design something with conflict resolution in mind when you want to focus on what characters are hoping to achieve in the moment, and how they pursue that goal using some vector of engagement or personal quality.
These two points of focus absolutely do cover some overlapping space. “Skill X” can be a descriptive activity in task resolution and a vector of engagement in conflict resolution. A desired outcome can be accounted for by both a “good job at the task” perspective and a “hoping to achieve” perspective. But as you design, task res pushes you out of this overlap in one broad direction, and conflict res pushes you out of it in another broad direction, and if you use the push for momentum you end up in very different game system territories.
Mostly your task res will be more “bottom up” and concrete, while your conflict res will be more “top down” and abstract. Not necessarily, but mostly.
Mostly your task res will be specific and increment the fiction step by step, chaining inputs and consequences in sequence, while your conflict res will allow you to resolve a chunk of fiction in one hit, folding in the effects of various resources and influences and delivering a result. Not necessarily, but mostly.
These are useful differences for our purposes. One of the affordances of a conflict resolution vibe to our design process is that it allows us to shake off the necessity of sequential consequences in the fiction, in favour of definitively resolving chunks in the abstract. Lifting out of the details! (Although those details do still matter – fictional positioning is all the non-mechanical stuff that has been established, and that is real and that counts and things follow from that no matter whether your res is tasky or conflicty. But that’s a sidetrack.)
So with all that expressed as clearly as I can manage, the question remains: can we even do that using an architecture of resolution built for carefully interacting with a dungeon room?
Can you do a conflicty lift using tasky stuff?
Which brings us at last to
Part 3: Skill Challenges
To set the scene properly, let’s consider the relationship of the dice roll to the fiction in a tasky sort of game. Because if we want to build a conflicty thing out of skill checks we gotta know what we’re building with.
You’re rolling along in the game and something comes up that a character is trying to do and everyone pauses (or, sometimes, the GM says whoaaaa there hoss, before you can say that thing you just said we gotta roll some dice ain’t that right podner) because it’s not clear what’s gonna shake out.
So you figure out what task in your list of tasks in the game is a match to what’s up. In 5E it’s gonna be an ability check, based on whether you’re applying your Strength or using your Wisdom or whatever. And if you are proficient in a skill that helps you like being Athletic or being Perceptive you add that bonus on too. And then you try to roll high, beating the number that marks the line between success and failure.
What does this game system procedure encompass in the fiction? A buncha stuff.
- The set of actual physical in-the-world actions you take to attempt the thing. Using your arm and leg muscles! Directing your gaze! Synapses firing!
- The personal affinity you have for these actions, which shapes the detail of what you do and makes you more or less effective than someone else might be. Strong grip with your hands! 20/20 vision!
- Prior training and experience in doing the thing that leads you to more effective forms of the action. Climbing technique! Knowing what to look for!
- The inherent generalised difficulty of the thing you are trying to do. This wall is pretty damn smooth! This person is pretty subtle in their movements!
- Contextual external variables that help or hurt the possibility of doing this thing in this moment. It’s wet and slippery from the rain! The lights are flickering! Your ex-boyfriend is watching you waiting for you to slip up!
- Contextual internal variables. You slept too long! You’re feeling undervalued at work!
- Pure luck. If you started climbing 6 inches to your left, you would have found more handholds! If you were positioned at a slightly different angle you would see more of their face!
- Your motivation. How hard are you gonna work at this?
- And more, of course, lots more.
Some of those things are explicitly represented in the procedure, captured by factors like the difficulty class, advantage/disadvantage, ability, skill, but even then they can’t ever be fully represented. I think of it this way: the randomness of the dice roll also accounts for the summed relevance of each of these variables in a given situation. Yes you have climbing knowledge so you get +3 to climb, and if you roll really low, guess that climbing knowledge just wasn’t super relevant to the specific problems you face on this particular climb, huh.
With all that stuff wrapped up into the dice roll, you don’t need to account for them in your narration and action declaration. That’s fine. I don’t want to describe going through the motions of researching musty tomes and how I set up my desk and chair for optimal concentration. I don’t have the lexicon to describe how I respond to the challenges of climbing this particular wall and which muscles I engage and which specific risks and techniques I favour. I just wanna establish the fictional truth that I’m researching a musty tome/climbing a tricky wall, and find out if I do it.
But other times this gets problematic. The dice roll procedure steps on some of the stuff that would normally live in the descriptive fiction. Consider the eternal challenge of social skill checks! Your character is chatting to a senator NPC, and you and the GM are enjoying a bit of in character banter, great. Then you decide to persuade the senator that she should come along to your community fundraiser. When do you roll the dice on that? Do you just start making your case and see what happens? Do you go to the dice only if the GM thinks the actual conversation goes somewhere uncertain? If you know you just said the most amazing persuasive stuff and then roll poorly, how does that work? Do you roll up front and then tailor your roleplay to what you rolled, knowingly tanking yourself if you rolled terribly?
Or, living in the D&D of it all a bit more, consider searching dungeon rooms. There is, and has always been, a “searching for secret doors” roll and a “find traps” roll. But also there is an expectation that you describe how your character interacts with stuff in the dungeon room and the DM says what happens. How can these both be true? Some OSR games encourage you to ignore searchy/findy rolls and just run entirely off description: I search the walls and look for cracks in the mortar! Which sounds great and vivid, a real test of player ingenuity, but gosh if you ever actually try to do this literally, prepare to have your brain turn slowly to mush as you spend hours on fruitless searching until the players just stop searching stuff entirely and fall into a pit trap and smile gratefully as they are finally released from having to describe looking for cracks int the mortar! (I said it doesn’t always work great.)
Okay enough about the inputs. What about the outputs! What does the GM say when you make the roll? Kinda… whatever they want? Like, assuming the GM isn’t being a dick about it and is acting in good faith, they will assess the details and the context and declare what being successful looks like. And sometimes that’s clear! You wanted to climb, you rolled good, you are at the top.
But sometimes it’s not! You wanted to persuade the senator, you rolled good, now what? The Senator… agrees with everything you say? Tells your people to speak to her people? Finds you persuasive but isn’t going to sign up to anything because Senators have very full diaries and don’t just cancel stuff to go to a random fundraiser no matter how persuasive you are? The kind of answer you get ultimately comes down to the GM’s take on what’s actually happening here, and that is about their vision of the fictional situation, which includes… all those details we’ve established, or failed to establish. Success or failure always needs to be interpreted back into the fiction, and that can go all kinds of ways, but it reliably follows on from what is known.
(The failure state of this when the GM takes advantage of this indefinition and just puts their thumb on the scale of what’s gonna happen so the dice result provides nothing more than scenery for a sequence of pre-written events. Make the roll if you want, dude, I’ve already got the next three beats of my story I mean scenario written down and I’m gonna make them come true.)
The bigger point is just this: a tasky roll lives in the details. It doesn’t do all of them, it can’t! It always abstracts a bunch of fiction stuff! But still, even then, despite the abstraction, it declares that details matter in a concrete way, and represents that as best it can without bogging everything down. And then those details, such as they are, inform how the dice roll becomes more fiction stuff.
And that, in turn, means tasky res gives each interaction in the game its own power and weight, because it is something that arises from the specific details of the moment, and then gives a resolution that lives in that context and develops it further.
So you get what was being talked about above, how each thing you do shifts the ground and sets the context for the next thing you do, and this accumulation of if/then logic generates the narrative. That’s how you get to find out what happens: you resolve each step of left/right decision making and each moment of tasky uncertainty along the way. And you can’t skip one because maybe that was the one that would change everything that follows after. In fact, if your system isn’t capturing all the right details, maybe you’re already unacceptably shortchanging the player experience????
This is why there was an entire branch of game design in the 80s and 90s (one that still survives in small pockets, much like the brave heroes of Living Steel who sleep in their cryo chambers until they shall one day wake and save us all) about getting more and more detailed in your tasky res. The only way to make games authentic was to save the GM from having to exercise their fiat about what makes sense next, and the end result of that is giant tables of modifiers: +1 if you’re kneeling when you do the thing! -2 if you’re prone! -1 if you ate a hearty breakfast this morning! ~and intensely articulated resolution systems that spit out highly precise outcomes: your bullet (roll) grazes off his (roll) left temple and then (roll) ricochets off the (check diagram) brickwork to (roll) strike the jeep in the (roll) fender doing (roll) 8 points of structural damage to the front left defensive fortification value… This was not the way. All that precision was a heavy load, and few people wanted it.
Trad systems learned this lesson and managed the challenge a different way: they limited opportunities for detail load. These games can cover anything, but in publication form, they optimise to something specific and contained, something about which the GM can reason without too much recourse to abstraction or overload.
Remember that riff way above about how the dungeon room is perfectly suited for the affordances of trad play? This is what that was about. In a dungeon, you have a door here and a door here and a list of items of furniture (never too many), and the GM can say what happens regarding all of this stuff without needing to get abstract. And you can still get rich, unexpected interactions: I put the caged chicken from room 3 in the pot of magic potion in room 29!
But what you also get, what’s hidden in here as a consequence, is that you end up with weird, denuded, unreal dungeon rooms, because the more things you have the harder it is for the GM to track every element and spool out consequence in the moment. You get point and click worlds with a limited set of fun things to interact with and not much else. Or, environments where the background isn’t unclickable exactly, but it’s definitely marked as unimportant, greyed out into blandness so as not to draw attention to itself and inadvertently make the GM’s role harder. Or you get this:
(Time to share once again my favourite dungeon room, from the Temple of Elemental Evil…)
Oh cool, okay so rugs and tapestries and then i’ll find out about these two servants who are in the room, I bet they have an interesting take on things, although I don’t know that #metoo was a thing when this was written so who knows how sensitively this will be handled, so somewhat guardedly i’ll turn the page,
Huh okay well good to have all these details, D&D is about looting stuff sure, now that’s done those servants will get their focus in the next column,
okay well we’re hearing about his sets of different-coloured slippers, now, such a wealth of detail! I can’t wait to see the same attentive focus turned to these two servants who spend all their days with this deceitful master of evil, any paragraph now-
WTFFFFFF)
Let’s look in a different direction, at Vampire the Masquerade. This game felt incredibly new and fresh in 1991 because it prioritised character, and embedded personalities in a social web. There were a bunch of social skills with clear tasky resolution approaches. But the game very quickly figured out that it was a big lift for GMs to infer sensibly from one social interaction outwards through a densely packed network of NPCs. So they invented the relationship map, a graphic representation of who links to who and how they feel about each other. By tracing lines of consequence across these maps, the GM could follow the reverberations of a social event as it spread through society. It was still a big lift, and the stereotypical Vampire game almost immediately became about being a badass fanged superhero in a trenchcoat anyway, but nevertheless, this gave you a tool for turning a dice roll into a broad-scale fictional consequence.
Still, it only went so far. There was no way to generalise out from this. The promise of tasky res and detailed choice tracking was that it could cover everything, but the reality was over and over again the GM just had to say what happened, and the only way that worked was if they tracked all the details they could, and accounted for them as fairly as possible.
And when you’re dealing with a bigger challenge, too much to track every detail? When you’re exploring a netherworld, when you’re dealing with a really big environment where lots of left-right decisions and tasky uncertainties just lead to more left-right decisions and tasky uncertainties, what do you do? Crawling through every one of them isn’t going to be fun, but when any particular choice might be the one that starts a domino effect of consequence, when in fact the only thing your task mechanism can do is tip over that first domino and discover what happens next, how can you chuck any of them out without throwing away the precious quality of player agency in an undetermined narrative?
Which brings us now to: Skill Challenges.
Part 4: Skill Challenges
In the 4th edition of D&D, this problem of accounting for complex situations without negotiating the detail was met head-on with a procedure called Skill Challenges. They used, clue is in the name, Skills, a list of things the character knew/did not know how to do.
Skills were introduced late to D&D. Not that late, if you count thief skills, which came along in the very first supplement to Dungeons & Dragons in 1975. Later on there were non-weapon proficiencies in some AD&D 1e hardback supplements, while Basic D&D got general skills in its Gazetteers, and finally in 3rd ed D&D skills became a structural thing on the character sheet that you used a lot. Until then, lots of people played with no skills, apart from thiefy ones, and the gap was filled by, well, vibes. By GM lift. 3E was a turning point by mechanising and regularising a lot of that GM lift. (That’s the getting away from the original assumptions I mentioned way back there, told you I’d get to it.)
The 4th edition skill list was a shorter, tighter version of the 3E one, for a game that was well aware of what was going on in the indie game scene with all its conflict resolution shenanigans. They wanted some of those shenanigans for themselves! And so (really late in the design process) a hasty two-page chunk of rules was written for a little minigame using skills. It was the first use of conflict resolution in D&D. (Or so everyone said at the time. Put a pin in that.)
Skill challenges worked like this: the DM says “this is a skill challenge!” and (referring to a table that got revised multiple times just across the short lifespan of 4E) figures out how many successful skill checks you gotta make to beat the challenge, and how many failures mean you’ve lost it, and what skills are part of the challenge. There’s further details but basically that’s it. You wanna convince the Duke to send his guard with you to the mountain pass to fight the giants that might not be just a legend after all?
- Make 6 successful skill checks before you fail 3
- Everyone has to try something
- Someone must try persuadey stuff
You can use other skills too if you have some kinda justification (“I’m going to use my Basketweaving skill to weave a Basket that shows the Duke’s heraldic mark, flattering him into trusting us!”)
Later updates revised and adjusted the processes, but that was the basic rhythm. You go around the table and each player describes a dislocated moment of fiction based around whatever skill they want to use, and you roll your dice and figure out how that moment ends, and then go to the next person. Across the lot you assemble a kind of montage which concludes when the success/failure count determines whether the whole thing is won or lost, in which case the GM intervenes to say how it’s all sorted.
Compared to the standard modality of trad play, this is a really (Really) (REALLY) (REEEEALLY) different way of determining the cumulative arc of fictional events. As a comparison, consider inductive vs deductive reasoning. It’s not a perfect match, but I think it points at something.
Usual trad play is inductive. You use procedures to take what happens in the fiction and then derive from that an appropriate update to the game state; the induction in question being one part skill check and one part good sense and understanding of causality and probability. You’re inferring what follows.
This approach is deductive. You identify an overarching question, use a set procedure to determine the overall outcome of the question, and derive from that some of the fictional moments that might have contributed to that result.
Skill challenges were, shall we say, controversial. Some DMs and groups loved them! Many others bounced off them hard. Partly it’s because the difficulty settings were pretty wonky because of how 4e was set up. But I think another important reason is what this deductive approach took away from play groups who were accustomed to the open-ended possibilities of inductive play. Why do these moments in particular add up to the success or failure of the conflict? What other possibilities are closed off by not considering the whole challenge so each moment occurs (or doesn’t!) in the full richness of continuing context? When you take a conflicty process and overlay it on a bunch of tasky moments, then take away the richness of those task resolutions, it feels bad. It’s making them add up to something bigger all right, but only by stripping out all the undefined but real aspects of what they already do.
So as conflict resolution attempts go, this is not exactly a success. But y’know there is another conflict resolution system in D&D, one that definitely does a better job, because for many players the whole dang game revolves around it. I speak, of course, of combat.
Part 5: COMBAT
Vincent Baker was on this last year in a post where he dove in deep on task and conflict resolution (and I think his take is compatible with mine? but, like, Vincent is very good at thinking about this stuff very hard and well, and I am me stream of consciousnessing through a riff on skill challenges, so this is just to say i’d probably trust him over me in the first instance)
In that post Vincent identifies D&D’s combat system as a conflict resolution system. Like, the entirety of the combat encounter is a conflict resolution mechanism. The to hit roll, the damage roll, saving throws, initiative, all of it, are pieces of the conflict resolution.
And those pieces… are… tasky pieces? So is this an example of how you turn tasky stuff into conflicty stuff?
But wait wait those tasky pieces are all very direct and bottom up and you don’t end up abstracting out hours of combat into a few rolls. The opposite! You end up hyper-detailing seconds of combat into many many rolls!
NOT SO FAST! Look at the length of the D&D combat round. In early editions, Gygax declared a combat round was one minute in length, and the to hit roll was a summary of all the back and forth and maneuvering within as you sought to find one vulnerable moment in which to strike your foe. In later editions, the D&D combat round shrunk to 6 seconds, and that to hit roll now represents if not one swing of your sword then one momentary sally of sword-swinging. Important point being: the mechanism is the same, even though different amounts of fictional stuff are being abstracted into that one roll to hit. So in principle, in principle, we can do this… right?
D&D combat is a procedure that musters a bunch of “task resolution” moments and allows them to keep all their inductive glory, but also tethers them to an organising mechanism that is pure conflict resolution, one that answers the “how do you know when you’re finished” question: hit points. OH YEAH HOT TAKE Hit points are conflict resolution!
And hit points are the organising mechanism but not the only thing that catches that inductive output. There’s conditions, resource consumption, and i can’t think of a third thing so I’ll say resource consumption again that covers lots of stuff.
Like, you can imagine a D&D combat without hit points. Instead, after each successful backstab or fireball, the DM has to just make a decision about whether the combat is over. That would be perfectly doable – and in fact, many DMs ALREADY DO THIS WITHOUT REALIZING, secretly fudging the hit point total of the baddie to ensure a battle isn’t over too soon, or to drop them early if a total party kill seems about to happen. What they are doing is judging top down, deductively, that it feels about right that this combat is finished.
And many people hate that way of playing! It seems to fly in the face of the fundamental value of inductive logic in games – you generate specific moments and find out what happens based on that! How can it be task resolution if you’re putting a thumb on the scale for the outcome?
Unfortunately that’s the only way it works. As soon as you lift up the abstraction of your task resolution to a point where the GM needs to apply ‘common sense’ or ‘logic’ or whatever, your task resolution system is officially in the realm of its dark twin: GM fiat. And that’s where a lot of these bigger challenges live. The GM is having to assemble the bigger meaning of lots of task res moments.
And I want to do better!
D&D combat is a conflict system that swallows up some abstraction and spits out a result while honouring the specificity of the moments within. Nice. Can we apply this to what we’re doing with FiveEvil? Maybe I can take some of these lessons to refresh the other conflict resolution tool from D&D, and make it work better? I am referring of course to
Part 6: Skill Challenges
In FiveEvil I present an optional encounter format called Challenge Encounters, which is my reimagining of Skill Challenges. It takes a bunch of cues from the structure of 5E combat rounds to good effect. At least I think so.
The GM sets the difficulty of the challenge (notably not something the GM does very much during FiveEvil play), and then we go into challenge rounds that give everyone a turn to take some kind of action and maybe contribute to the goal. Then there’s a checkpoint moment where the GM assesses consequences for what has happened so far, and a potential trouble that the characters must face. All of this puts the montage of chosen actions into a group context, part of a back and forth with trouble, and where the GM is reminded to assess consequences of what is depicted in the encounter. It’s still pretty abstract, allowing you to get to the end, but it is grounded in the realities of ability checks and skills, the task resolution baseline of the 5E system.
It works well! But there’s one other thing to add into the mix to drag this into a tool for managing the challenge of a netherworld: choice.
Remember, the task res level of engagement with the fiction is handy because you can track all the decisions the characters make as they interact with the dungeon room. Every choice they make has consequences and none of it is wasted. Lifting up out of the details isn’t just about abstracting a sequence of tasks, but also about handwaving these intimate choices into a kind of high-level average of character agency, where you fill in choices made retrospectively just like you figure out how your chosen actions resolve the fictional situation retrospectively.
To mitigate this, in FiveEvil I’ve taken the approach of listing out step by step procedures. And I’m sure for many in the audience this is teaching you to suck eggs. (What a saying that is. I don’t even know how to suck eggs? Is that something most people used to know?) But I walk through the whole sequence: instead of just declaring an abstract moment of skill resolution, you first make a call about your general direction of travel, and then the GM responds; then you make a decision about how you handle what you’re up to, and the GM runs what happens; and then you make another choice about how you engage with where you end up, and the GM responds with all that stuff.
It’s a useful method, breaking down the cycle of play into its core moments. Each stage you get player choice as a fundamental element, steering the entire action, and your tasks are embedded in key moments instead of just floating around in there, and you know for sure definitively when you are finished. But how about that GM lift? How about GM fiat? How do you work out what comes next without playing through everything that happens first?
Answer: you have a procedure that tells you “these are the bits that matter”, but doesn’t tell you how they matter. That’s what skill challenges failed at. That’s how tasky stuff can function in a conflicty context. You use abstraction to get the selection of moments, but keep it open so you don’t know what the ending of each moment is going to be. Play to find out!
(oh yeah you know what that sounds like HOT TAKE: wandering monster rolls are conflict resolution! Wait what)
~~~
So let’s loop way way back around to the starting point. Horror game. Bad places. Netherworlds.
The trick to these environments is they are scary and threatening. But the things that make them scary and threatening make them a poor fit for the kind of play you’d use for a dungeon crawl. Netherworlds are places of anxiety and suspense. That means potential danger everywhere… but *actual* danger waits for its moment. The more interactive stuff you put into the netherworld, the more likely your players get caught up interacting with it, breaking the tension. The more you withhold stuff to interact with, the more likely your players get kinda bored with yet another gloomy room with dripping water and ominous sounds. The more you dungeon out the rooms, the more likely your scary netherworld just becomes a dungeon.
So you don’t run a netherworld like a dungeon! But that’s easier said than done, right? If you’re not asking questions of your players, and giving them chances to interact with the netherworld, how can their fate be a result of what they choose? Aren’t you just gonna be making up what happens and telling them?
I’ve devoted a whole section to netherworlds because I want FiveEvil GMs to be confident about running them. In this chapter I give you the tools to run a sequence of play in which characters explore a scary netherworld, while avoiding both Scylla (describing every room like a dungeon, which gets tedious or overstuffed) and Charybdis (just declaring what happens, which takes away agency).
I’ll even spill the secret here, because it’s not really a secret, it’s likely something you’ve seen elsewhere before. Calling it out clearly in a procedure is just giving a very strong prompt to actually use it here, because it works so perfectly with the assumptions and logic of FiveEvil. It’s this: prepare, then play, each step of the way. Each time you prepare, you’ll either check your notes or roll some dice to populate the path ahead with obstacles, or hazards, or nothing but the omnipresent bad vibes. And then you play: narrate things until they collide with what you prepared, and see what happens. Then you come out of the encounter and finish that piece of the journey, and go back to the start to prepare again.
If you’ve seen travel or journey rules in other games, you’ll recognise this logic. Yep! Same same. The trick is in how we’ve aligned our rules to fit in with the horrific nature of the netherworld, and how we provide ways to ensure you have meaningful player choices that have consequences. Those choices are part of what makes horror work: your predicament is on you. Getting all that right, so it all lines up and works and reinforces the important parts of the experience, that was the hard part.
The netherworld exploration toolkit is optional. You don’t have to use it! You might have your own way of running sequences like this, something that works perfectly well for you! But I know a lot of GMs out there don’t have a procedure like this at their fingertips, because it’s not something they’ve had to do very much. It’s part of the horror genre, though, at least the kind of horror stories we want to explore with this game. So we’re doing it. And I’ve sweated hard over getting it right.
The very first horror movie I saw, the one I was way too young for and my friends and I watched on a sleepover, was A Nightmare on Elm Street 2. (It’s a weird one! And much better appreciated now than it was at the time.) That film ends with heroine Lisa having to go into the factory that Freddy Krueger haunts, the centre of his power, because she wants to save her friend. She goes into a netherworld. That final sequence was, for the young me having his first experience of horror, absolutely terrifying.
I really hope this section of FiveEvil helps you to unlock that same emotion when you sit down to play.