What adventure structures (if any) do GMs use for their fantasy RPGs?

When preparing / designing your own adventure for a fantasy RPG, and the game doesn’t provide a definite structure creation procedure (PBtA games are examples that do), are there any well-defined and arguably complete structures that you use? Whether they are ones you have invented or discovered.

Some examples.
A timeline where events are going to occur regardless of player interaction, but certain spots on the timeline are intended to intersect with the PCs (as scenes) to give the players an opportunity to engage, if they want.

Justin Alexander has a series on node-based structures:

And an even more interesting article on using the structure to run mysteries:

The Hexcrawl and the Dungeoncrawl are obvious classic structures of the genre.

I’ve been curious for a long time if a form of relationship map could serve as a viable adventure structure. I hold no assumptions about what a “relationship map” means other than some map that connects significant people, places and things.

I also make no assumptions about how these structures are used by anyone, I see no reason why several could not be used within a single adventure. I’m just curious if anyone has any tools in their toolbag that they’re particularly happy with.


I apologize if this comes across as flippant, dismissive, or glib-- but I do not really design my fantasy adventures at all. I start with a couple of campaign ideas, then I seed the “adventuring region” with situations that… imply the existence of a plot. The possibility of a plot.

I give the players the opportunity to bite some of the hooks, and whichever hook they bite most enthusiastically becomes the campaign, while the hooks they ignored get a little worse-- the ones they bite on at this step become the important “subplots”, while the ones they’ve ignored twice fade away into the background. Once we’ve established which hooks they’re interested in… ignoring those has consequences.

From then on, everything else that happens is either the consequences of their actions or the consequences of their inaction.

There are a lot of little reasons I’ve adopted this playstyle, but it mostly comes down to the fact that I can’t even accidentally force the PCs to go along with my plan if I never have a plan in the first place. Fundamentally, as a GM, I’ve always found that the more of my “homework” I get my players to do for me, the more impressed they are with the results.


No worries, I don’t take your explanation to be dismissive at all.

When you seed these potential stories, what do they look like? Are we talking about a location on a map with a simple one sentence premise? “Cultists in the dark wood are raising something sinister” Or is it more than that, like a few sentences that cover a few high-level concepts behind the threat?

How do you introduce the initial hooks? Do you plop the PCs down into some kind of sandbox and respond to their preferences? “The thief decides to walk the street late at night, looking for opportunity so I have them overhear a few rumors. The priestess spends time serving at the temple and she has disturbing dreams as she sleeps there.” Or is there more preparation and planning to it?

Does this hook > hook thing just continue forever? So the cultists start as the highest level hook, then they investigate, so you add some more hooks in the woods. Then when they investigate one of those hooks, you add a few more… I’m curious how you arrive from the highest level (a few initial hooks) down to the nitty gritty details (NPCs, “encounters”, treasures and rewards, etc).

Maybe another way to pose the last question is, how much do you zoom in, and when do you know you have enough to start the next session? Or do you only ever prep high-level hooks and ad-lib literally every detail at the table?

Edit: I should add that what you describe here qualifies for what I intended when I used the word “design”. Maybe design is an overloaded term. I simply mean any original work performed by a GM in preparation for a gaming session. As opposed to running a session using a pre-existing “adventure”.


I love the idea of story maps you can find in Technoir transmissions, Rickard’s Fishtank articles (here and there), Eero Tuovinen Prydain chronicles railroad game, also there. Eero Tuovinen’s Zombie Cinema used a single track. You could also see, squinting, games with a heavy structure, say : Robert Bohl’s Misspent Youth or Jason Morningstar’s Fiasco, as following a track. “RPG design patterns” also describes some adventures structures. You will find loads of articles for video game level design and some on IF design that could be useful. Of course, geographical maps, Goblin Henchman’s Hex flower engine, themes maps (On mighty thews), turf maps (BitD) and Evil Hat’s Fate zones are narrative maps to a great extent.

Relative range could also be a thing (volley vs melee - Agon, pursuit - Psy*run), with a constricted enough starting setting, but that would be straying off too far from the Fantasy Adventure premise.

An underlying theoretical question is : does the setting move behind the screen, in which case new states are generated by what ?

edited to add links


I can work with as little as that, if I’m in a hurry. If I have time, I’ll usually flesh it out a little-- I am a Druid of the Circle of Economics. I know what the cultists are worshiping and what it promised them. I know the leader’s name, how many hit dice he has, what spells I want him to cast. I know how many cultists the PCs could find in the Dark Wood, versus how many are back in town with hostages/sacrifices. Most importantly, I know what the cultists need, I know who they’re taking it from, and I know who the cultists are afraid of.

That means I know what the cultists are going to do next if the PCs don’t bother them, and depending on how exactly the PCs bother them… I can work out what the survivors are going to do next very quickly.

Economics is the ecology of civilization-- it is the circle of life for intelligent creatures. If you know what a creature of economics needs, you know what it’s going to do when it’s cut off.

That snippet up there about the Cult in the Dark Woods that you just made up? I can answer all those questions in under half an hour, written up, and fairly easily bang out half a dozen in a week. (You could probably do a lot more, but I’ve got really bad untreated ADHD.) All of them are potential jumping off points for a whole campaign or potential game-altering complications for the campaign the players decide to play.

Just like that. Priests and magicians have visions, mercenaries and thieves hear rumors… pretty much anything that tells the players “something’s going on here” and gives them a little hint what it’s about.

No more consideration or effort than rolling on the “random rumors” table in an old-school adventure module.

The cycle continues, but it slows down as the game progresses-- the PCs stomp out a fire, and then I start a bigger fire, but there’s also the repercussions of their actions. They stomp out those brushfires, they stomp out the bigger fire… they start getting ideas on putting those fires out for good.

Sometimes you have to do it a couple of times before it catches, especially if your players are used to being fed a storyline, but when the engine finally turns over the players start generating their own hooks-- they start feeding you the storyline, and all you have to is make their plans more difficult and complicated than they expected.

Preferably via the enemies they’ve made along the way, because this looks an awful lot like something you’d planned all along.

This is where I spend my prep time. A cult needs a leader, a cult leader needs a name. The cult controls some of the people in town, which ones? (Economics.) Deciding where the treasure comes from tells you what it is. As much as possible, the answer to “what should be here?” is “what logically must be here”.

I don’t prep as much as I should. I rely too much on my ability to ad-lib and my real-time games suffer for it-- especially in terms of continuity, over repeated encounters. I actually tried to run a Groundhog Day scenario to try to improve my capacity for taking notes and staying consistent.

Generally… I feel like I’m ready for the next session when I feel like I know what everyone my players are (probably) going to encounter wants, what they have, and how they’re going to respond if they’re tempted, pressured, or threatened. That’s the stage where I might worry about their (abbreviated) combat stats, or their access to treasure… but that’s also the stage where I’m more likely to skip that.

Not perfect, by any means, but I can identify where I have room for improvement.


I used to write incredibly structured adventures broken down into scenes with ways in and ways out, and an inferred sequence that amounted, with hindsight, to a storyboard for a film. The height of my pretension in how I felt GMs should run my stuff is here: https://www.lythia.com/series/dark-rethem/, for reference.
These days I’m am at the polar opposite. My PbtA game in development of Dark Age Heroes protecting the innocent from the supernatural begins with creating a settlement and a handful or so if key NPCs from a series of pick lists along with some nearby locations. Then the characters establish who they know in the settlement, and that influences how they are received, generating an ‘inciting incident’.
Then we play to find out.


This is a great question, and one that’s not asked enough. How do we structure play? How do we structure prep for play?

I don’t believe that there is any kind of “one size fits all” approach here: the answer will depend a great deal on a number of factors, like the ruleset you’re using, how many players you have, the genre, the expected playtime, and the style of play. (We can do a dungeon crawl with a good set of random generators which draw a few rooms and put treasure and monsters in them*, but we probably won’t get an intimate personal drama the same way - we need different tools for that.)

*: Take a look at The OSR in Vivid Colour: The Fellowship of the Bling, for example.

Some games have intricate rules structures which require a certain type of prep - or make a certain other type of prep impossible. If my game is a series of combat encounters, and the rules require a lot of work to prep stat blocks for those opponents, that restricts what I can do: I probably need to prepare opponents/monsters in advance, which means I have to structure the game so that the PCs will fight those monsters. That means a very specific kind of prep is necessary.

As a general rule of thumb, though, the more players there are at the table, the more I want the game structure to get them to interact with each other (either as PvP or relationship drama). In traditional GM+players play, the GM forms a sort of “bottleneck”, and this can start to slow the game as you add players. (Old-school D&D players have sometimes used a different solution, of course, by appointing a “caller”, who makes decisions for the group when necessary and conveys information to the referee [GM].)

A relationship map can be a great format for a fantasy game, and I often use something like that.

One type of structure which hasn’t been discussed yet is the concept of Threats/Fronts, from Apocalypse World (and later, Dungeon World). We establish where the PCs are, and what they care about, and then the GM preps “threats” or “fronts”, which are bad actors who have (horrible) endgames in mind. An invading army, a curse, an advancing prophecy, a serial killer… they are all written up to threaten things the PCs care about. In play, when the PCs do not act to forestall their actions, the “threats” advance towards their endgoals, one step at a time (in some variations, the steps are written ahead of time).


I agree and is the main reason for this open-ended call for peoples’ personal experiences. I want to learn about as many approaches as possible. I have a set of rules I’m working on and I’m simultaneously doing a lot of thinking and taking notes on how I will want to prepare content for sessions (not one method, but a toolbox of options that resonate with me).

You make a good point that the context of the rules matters; not every method applies for every possible fantasy RPG. I want to run an OSR style game and in this game, PC death and re-rolling new characters will be a fairly common occurrence. That might explain why I didn’t list Fronts, because that particular method ties into character motives, IIRC. Since the kind of campaign I plan to make adventures for will have a built-in party role, the PCs will have a shared motive that ties them in to the campaign’s adventures (which share that theme). This feels necessary for a game with a revolving door of PCs; it would be fatiguing to create motives for each new PC and disappointing to see them killed when you had exciting plans for them.

Characters that survive and succeed can of course develop more interesting individual motives as play happens, if their players wish for them to (it’s a sandbox campaign). I should read up on Dungeon World again to see how different Threats/Fronts are from the classic RPG timeline; maybe there are still some ideas there I could mine that would be applicable or at least inspire a similar idea.

Knowing that you are approaching this from an OSR-type perspective narrows things down quite dramatically, actually. A lot of my prep techniques, like character-based thematic resonance, just won’t be of interest.

I would recommend that you read about Threats or Fronts, either in Apocalypse World or Dungeon World (I can’t think of any major differences between the two, off the top of my head), as they could be a useful techniques. If your characters are all tied into a shared motive, all you have to do is to make Fronts which relate to that motive somehow.

They are different from a “classic timeline” in that you don’t know in what order things will happen, or whether they will happen at all. You just know, basically, who the antagonists are, and what will happen if no one interferes with them. (You could, of course, combine Fronts and a timeline by deciding ahead of time which will happen and when.)

A Front basically says, “There’s an invading Orc horde. They plan to (a) enter the valley, (b) dig up their dead chieftain and bring him back to life, and © burn down the fort.” In play, we’ll find out how far they get, and whether and when these things happen.

A timeline might say: “Day 2, the Orcs enter the valley. Day 5, they bring their dead chieftain back to life. Day 9, they burn down the fort.”

It’s a subtle difference (especially if you only have one Front, as in the example above), but makes for different gameplay. With multiple Fronts in action, it can get very dynamic, but ultimately it’s just an organizational tool for the GM’s prep.

I think the former is better suited to drama and action and suspense, and gives the GM lots of tools for driving the game forward. The latter might be better for a “sit back and watch” kind of GM style (where it’s acceptable, for example, that nothing might happen at all between Day 2 and Day 5), like an old-school referee.

However, hybrids are possible. Maybe some of your threats are Fronts and some have a distinct timeline?

For extra OSR flavour, you could randomize the advance of your Fronts. For instance, if you have 6 Fronts, you roll a d6 at the beginning of each day or session or turn or year (depending on the scale you’re operating on) and move that Front/Threat one step closer to completion. Or, a more dangerous variation: roll a 6 for each Front still in play, and on a 5 or 6, it advances closer to its endgoal.

For a roleplaying game, I generally start with just a circumstance, one possible outcome, and one interesting path, with details only barely sketched out. Almost everything is improvised on the fly and dynamically rationalized based on current and past events, or (very rarely) retroactively rationalized.

Anything beyond that has proven to be a waste of my time because players don’t follow scripts, and constraining creative minds to tracks is generally counterproductive.

In story games, I generally start out with even less because all the details are negotiated during play.

At the moment, I am running a non-gonzo game with no character sheets, no plot, no clear paths, no expectations and no published goal or expected outcome. Failure is not shunned. Death is not possible. There is a foundational theme, but all activity is driven by circumstances and emergent events.

I think improvised play is a great paradigm, but I’m sure there are structures or principles at play beyond “we just make it up”, at least if your play is consistently good.

I often like to think in terms of “how would I teach someone else to run a game this way?” That can be quite enlightening.

Sharing examples can a good way, for people who find that easier than dissecting their play process in an abstract fashion (never an easy thing!). I’d recommend that, for anyone interested in sharing their techniques. (It can actually be really fruitful and really fun to say, “I did this, and then this,” and then to have an observer say, “Ah, so what you’re doing is that…” - I’ve been on both sides of that conversation, and it is often really illuminating, since so much of what we do ourselves is opaque to us.)

Fair enough.

If I had a single driving principle, it would be:
Choose a single player, find out what their character is really after, and make their challenges just complex enough to hold their (the character’s AND the player’s) interest.

The first corollary would be:
To better engage the other players’ characters, shift focus to different ones in different sessions. If you don’t do this, your group will have to establish them solidly as “sidekicks”. It’s not a bad thing. It’s just good to recognize what’s going on.

1 Like

I like that.

I think the whole “sidekick” concept is a potentially really useful one that doesn’t see a lot of discussion or design, but is entirely legitimate for certain games or certain groups.

Where does this all start, though? How is the setting, situation, and so forth established?

Can you give an example?

For all RP and some story games, the GM tends to have more control over the “playground.”

Their starting point (inspiration?) is always triggered by a constrained randomizer–a “seed number” based on the whatever catches their interest. It seems obvious, but it has to be said. Your inspiration comes from your experience and no one else’s. If you’re inspired by someone else’s idea, it’s because that idea has come into your personal experience. Could be a system, a character from a film or book, a bit of natural architecture, a great meal, another game.

From there, I’d say establishing cause and effect is fundamental, because your scenario has to make sense in some way. Cause and effect is a pretty basic approach, and it just requires that you answer the basic questions: how, what, when, where, who, why. Theme is another approach, where you hew tightly to a specific concept. There’s still cause and effect, but that’s not the main driver of ideas.

A concrete example:

I recently played some background effects (downpour) over a series of sessions for my players because they were in an environment of unrelenting rain that accompanied loss and other discouraging circumstances for their characters.

It had a profound impact on them and I started thinking about how to use sound more fully in my games. Turns out it’s a LOT of work. (More than I care to do on a regular basis.) You can talk about sound, but until you actually hear it, you won’t feel it. Still, the feeling wouldn’t go away.

And one day, I thought, “A flute is just a piece of wood with holes in it. So couldn’t a skeleton with hollow bones create piped music as it moves?” So I decided to put a little effort into creating a true musical dungeon. Not a playlist or background music for a dungeon. But a dungeon that was designed around music. This meant I had to answer questions about what other hazards could manifest in a musical way.

There’s no backstory yet, but I’m imagining a music-based wizard-turned-lich (who!) at the center of it all. Stalactites and stalagmites that generate tones like a xylophone as they’re shattered, releasing poisonous fumes or frozen creatures that eventually thaw and attack when least expected. Gravel-covered steps that create the sounds of maracas as you move across them. Tunnels of different lengths and diameters might generate different notes as wind howls through them. Tripwires that strum as you fall across them and get sliced to shreds. Skin stretched across a frame is the definition of a drum. Imagine having to leap between stretched and curing skins of some unknown creature to get from one point to another. Puzzles might involve selecting the correct instruments or notes to play a tune. Maybe if you have the howling tunnels, choosing which doors to close and which to leave open generates specific notes you need.

Perhaps different monsters would have attack styles based on musical pieces, styles or techniques, such as staccato for quick, light attacks or vibrato for an extended, jarring effect. All my major NPCs have their behavior keyed to songs that guide their personalities anyway, so this probably wouldn’t be too hard to design.

Anyway, the next steps would be to finalize encounters, establish a flow to the climactic encounter (crescendo?) and then build in a strong enough backstory to why it all exists, and for color, when and how it was created.

Why are the characters involved?
Find a magical song or instrument
Slay the lich or convince him to do something because (reasons)
Thrown into the dungeon for their crimes (maybe all the PCs are hapless bards)
Rescue a previous victim of the dungeon
Destroy the dungeon and it’s nightmarish contraptions completely

So yeah, concrete example. Due to it’s nature, it requires a bit more specificity and a few more rails than most of my creations. But that’s only because I’d want to stay very tight to the musical theme.

I’ll add to the mix Francesco Zani’s Pachidermi :

It’s in italian, but you can read the scenarios and see how the structure of the game can go from freeform LARP to adventure and exploration map. Brilliant !