Robin Laws’ Revisited: Part 5, Creating your Adventure

This post is part of a series where I discuss Robin D Laws’ seminal work on GMing: Robin Laws’ of Good Game Mastering, written at the end of the 90’s . I compare it to my personal experience and opinions and I check how well the guide has ‘aged’.

The other posts of this series can be found here. If you liked them, I encourage you to purchase Robin’s book.

With a new edition of D&D around the corner, some GMs have new campaigns on their mind. Others, less inclined to play D&D are setting up their summer games or are playing various RPG products whose available pre-published adventures might be limited.

I bet that many from both groups have adventure design on the mind.

In fact, July is Worldwide Adventure Writing Month, so I think the timing for tackling the Adventure Creation chapter of Robin’s GMing guide is perfect!

Let’s dive in shall we?

Plot Hooks

Robin starts by saying that, be they fully detailed module, a few scribbled notes to help you improvise or somewhere in between, your adventure needs to have a clearly defined plot hook. A plot hook is a sentence (that starts with a verb) that lays out what needs to be done.


  • Escort the Duchess of Tropia through Orc-infested lands so she can sign peace treaty with the Kingdom of Clee-Shay.
  • Find the lost tomb/lair of El-Veess the Lich-Bard and stop the enchantment-infused music that is driving the whole continent mad!
  • Enter the cave, kill the Dragon and fence his hoard
  • Prevent the marriage of the warrior daughter of the Barbarian King to the half-human scion of the Ophidian Empire.

A simple plot hook makes it easier to players (who aren’t privy to your adventure plan) to know what is expected of their characters. It will gave them a clear point of reference to base their first actions on.

Of course, you have to ensure to convey this plot hook in a straightforward manner to your players (or their characters) as early as conveniently possible in the adventure.

Laws mentions that the plot hook must clearly set out the adventure’s ‘victory’ condition that defines a clear ending. Once those conditions are met, the adventure is over (except maybe for a last epilogue scene).

Structure vs Player Type

The next part hammers the core point of the whole book: The strategy you take to chose the structure of your depends entirely on your player types.

By structure, he means the type of scaffolding on which you build your adventure’s flow.

He argues that some player types are more inclined to appreciate structured adventures (Storytellers and Method Actors) while others (i.e. Tacticians) supposedly hate any kind of climactic buildup because it defeats the purpose of out planning the opposition to gain a complete victory.

However, I feel that the narrative structures you decide to build your adventure upon doesn’t have to take player types into account. That is, it doesn’t provided you include plot elements and challenges in them that will cater, at least in part, specifically to each type of players you have around your gaming table.

So when he says that tacticians hate an adventure whose structure leads him to a big boss fight, I disagree. Just make sure to put in a planning scene where the tactician’s decision will have a significant impact on the last scene and you should have a satisfied player.

In fact, very few players belong to one player type (thankfully), so feel free to adopt the structure you prefer (I’ll get to them shortly) when you design your adventure. In fact, I suggest that you mix and match them based on your campaign needs and the time you can spend prepping.

The (myth of the) Unstructured Dungeon

The first type of adventure structure is what Robin calls the unstructured adventure, represented by “”the plot-less dungeon” (his words, not mine), where players knock down doors, kill monsters and loot the room. The order of encounters is set by the whim of which area of the adventure the PCs chose to explore.

While Robin was careful not to bash this play style, he does mention that Storytellers find this form of adventure insufferable.

This is where I got my first point of annoyance with Laws’ tips. I understand that by the late 90’s, Storytelling games were on the rise and there was an emergence of new design philosophies around new RPGs (or new editions of current ones).

However, having played Adv. D&D 1e from quite a few years, I can attest that all published adventures (except possibly the tournament classics of the ‘S’ series) had plots, although it wasn’t spelled out in an ‘Adventure background’ section. You often had to hunt it down through the room by room descriptions.

The Against the Giant series of modules, while very hack and Slashy, had surprising politics, stories and mysteries hidden in the few pages each of these 3 adventures took.

I won’t dwell too much on this as I’ll probably make it a different post, but I really prefer the more modern term of ‘Site-based’ adventure which can include mindless dungeon crawls but are certainly not limited to such type of gaming.

I strongly believe that the site-based adventure has evolved in the last 10 years, incorporating most (if not all) of the other adventure structures Laws’ talks about in this chapter.

In fact, I invite you to have a look at Paizo’s adventures (From the moment they got stewardship of Dungeon Magazine up to and including Pathfinder and Gamemastery modules), the Goodman Games Classic adventures and especially Wizards of the Coast’s Dungeonscape to see how ‘the Dungeon’ has been treated.

The Episodic Adventure

This form of adventure is a series of unrelated (or thematically themed) encounters presented in an order set by the GM. Similar to an episodic campaign (where one adventure has no obligatory link with the next one), an episodic adventure is basically a linear dungeon where “rooms” can be switched for “scenes”.

The archetypal episodic adventure would be a trek where static unrelated encounters are met one after the other.

Using the 5 room dungeon adventure model as an example I could get:

  • Plot Hook: Bring back recently freed slaves from mountains back to civilization.
  • Scene 1: A Group of slavers ambush party in order to recuperate ‘stolen goods’.
  • Scene 2: Avalanche!
  • Scene 3: Trolls requires an outrageous toll per head to allow passage.
  • Scene 4: Wolves attacks at night.
  • Scene 5: Hungry Dragon Family fancying a free meal jumps on ex-Slaves.

While technically well suited for exploration-type games for world-building DMs, I really don’t picture myself doing such an adventure often. I could be tempted to use this structure when I need to create a transition adventure between two larger plots.

(However, I would use this as a sub-part of an adventure…)

The key characteristic of such an adventure structure is that the party’s performance in a prior scene has no impact on later ones (except total party kill, unless players can create new PCs). Each scene is a self contained Roleplaying/Combat/Puzzle challenge.

I heart the Set-Piece!

The Set-Piece is based on a technique used (among other things) in movies/TV shows where the action revolves around a few successive scenes leading up to a big finale (the Set piece). All scenes are built around common areas and NPCs.

The first Matrix movie or some of the episodes of the 1st season of Heroes are great examples of the Set-Piece approach (The Save the Cheerleader or the season finale episode comes to mind).

In such adventures, you have a central plot idea and a set of characters and ‘props’ on which you build your scenes. Each scene thematically leads to the next one.

Contrary to the episodic structure, what characters achieve in one scene can help or hinder their progress in later scenes. So here’s an example, using the same 5 room dungeon/scene approach.

  • Plot Hook: Rescue kidnapped children taken by a Chaos cult who consider monstrous mutations as signs of divinity (with thanks to Monte Cook’s Chaositech).
  • Scene 1: Character witness and possibly interrupt a kidnapping attempt at an Orphenage
  • Scene 2: Character Investigate clues to find out who cult is and where it operates
  • Scene 3: Mutant monsters and cultist minions ambush nosey party
  • Scene 4: Party finds HQ and Rumble against Mutant Mooks (or possibly sneak in without being noticed)
  • Scene 5: Boss Fight against Ogre-Sized, Acid-breathing, Wall Climbing Boss (and assorted failed mutant turned into oozes) in a multi-storied underground aqueduct.

This is my natural adventure design philosophy. By combining this with the 5 room/scene approach, I can churn out an adventure in about 2 hours.

I even sometime nest an ‘unstructured’ 5-room dungeon in this (The Cultist’s HQ for example) to flesh out the adventure into a 2 sessions arc.

While players have choices that can affect the outcome of current or future scenes, the adventure itself remains pretty linear.

In fact, since the adventure is linear, Laws’ mentions that such adventures must not have choke points where a failed skill roll or missing piece of information prevents the adventure from moving forward.


A Branching structure is, according to Laws’ the hardest one to pull. It is the true, non-linear adventure.

The DM creates his plot hook and the initial scene and tries to work out in advance the most likely outcomes and plans possible (usually 2) scenes that can develop from there. Each scene can then lead to more, branching out the adventure in multiple directions.

As you can imagine, this can get out of hand pretty quick, even if 90% of the scenes are improvised by the GM.

That’s why the various published adventures who try to emulate a branching out structure will either be very light on details, leaving them up to the GMs, which often makes for frustrating reading (been there!) or actually have scenes branching out and then converging back to the ‘trunk’, creating an hybrid type of structure.

You may have guessed already that I think that hybrid = good in my (yet unwritten) book.


Puzzle-Piece adventures are a way of creating apparently non-linear adventures while limiting the ammount of work.

As the name entails, this is the structure for investigation-type adventures. The GM sets the plot hook, defines the adventure’s victory conditions and breaks up the required conditions to achieve victory in bits and pieces.

Such bits and pieces are then spread over various locales and NPCs that PCs have to investigate in order to solve the “puzzle”.

Dante of Stupid Ranger has already tackled this and I invite you to checj it out as I’m close to hitting the 2000 word mark!

Suffice it to say that investigation adventures can become some of the most unfun RPG sessions when they start running in circles or when Players start pixel bitching each and every scenes.

GMs using such an adventure structure are encouraged to keep a sharp eye on his player’s enjoyment level, pacing and should really hand out far more clues than what one usually expects to solve a puzzle.

In the end, whatever the structure you chose, Robin Laws hammers repeatedly that

Whatever the Structure you use, Tailor the adventure’s content to your player’s tastes.

Yeah, it’s like the take home message of the whole book.

This post has run long enough. I’ll post a follow up article soon that distills my ‘Adventure Prep’ series to extract my personal adventure designing philosophy as I tend to disagree with some of Robin’s approach to adventure design.

Bottom line is that you should pick and chose any structure, including combining them in the same adventures in order to be able to create the adventure that you feel like prepping and that your players will enjoy the most.

Any other adventure structures out there that have evolved since? Feel free to share!


  1. Neat article. I’m more intrigued by this book the more you talk about it.

    “Dante of Stupid Ranger has already tackled this and I invite you to checj it out as I’m close to hitting the 2000 word mark!” <– This link appears to be broken.

    – Brian

    Brians last blog post..Why I Can’t Play the Newest Branches of the D&D Tree

  2. This is definitely a worthwhile purchase. I wasn’t disappointed at all…


    Bens last blog post..WTF: Excerpts from the new DMG

  3. @Brian: Thannks! It’s totally worth it. Link has been fixed.

  4. I’m a bit weird with my opinions on this.

    The “game theory” of RPGs says that most games fall into the category of Gamist, Simulationist, or Narrativist (with hybrids possible). I’m not going to go into what that means here, because a) I may well be preaching to the choir, and b) does it better.

    Much of my distaste for the way D&D is “typically” run (read “by the guys I game with” rather than “by everyone”, since clearly I can be an authority only on the former) is that it is fairly strongly Gamist, and that later incarnations of the game (2nd edition+) are “incoherent”. It is the combination of the two rather than just the former that bothers me. 1st edition AD&D was “about” killing things and taking their stuff, which rewarded you with experience points that made you better at killing things and taking stuff, and so on – an almost purely Gamist implementation. (I’m not saying that I’m a 1st edition grognard – far from it; the 1st edition of the game was terribly unbalanced and poorly executed – but the overall design philosophy was coherent). 2nd edition introduced the idea of “story based XP”, which essentially translates as wizards that research spells get better at killing things and taking stuff – the coherent (if bizarre) “practising at X means you get better at X” is broken. 3rd edition does not have SPECIFIC awards for anything other than “overcoming challenges” (which translates, often, as killing things and taking their stuff) but the house rules for Story Based Awards are so prevalent as to be “virtually” part of the core.

    I think D&D SUFFERS by trying to introduce plots and stories; I think that doing this tries to make D&D do something it’s not particularly good at, and that other games might be better at achieving. This is obviously just IMHO, and lest it be said that I am accusing D&D players of never having plots or stories, let me unequivocably state that I am NOT saying that – I am just saying that gamers who introduce those elements into their D&D games are doing so largely in spite of rather than because of what the system supports. That doesn’t mean that they aren’t successful by any stretch of the imagination – a good enough GM makes the rules set largely irrelevant – but in the best of all possible worlds, your game would directly support the types of activities that you intend to have within it. All D&D directly supports – I claim – is the structuring of challenges with the intention of rewarding successful overcoming of these challenges with increased ability to overcome similar challenges in the future (by means of experience points and, at least currently, better equipment).

    But Gamist RPGs are not (IMHO) the best way to tell stories; Narrative games are. I use the term “games” here rather than “RPGs” because there is considerable dispute about whether games like Capes actually qualify as roleplaying, but there are relatively uncontroversial examples as well – HeroQuest, for example, I believe is largely accepted as a genuine RPG. HeroQuest plays very differently to D&D; the things that are important in D&D (hit points, power level, equipment, mortality, and so on) are much less important in HeroQuest, while the things that HeroQuest deems important (connection to the community, dramatic importance of actions, mythic significance, and so on) are much less important in D&D. Narrative games are “better” for telling stories but often have less character “improvement” in the sense that D&D has; success is not necessarily better than failure in a narrative game, and sometimes you literally do have more fun if you mostly fail at your actions (to the extent that in some games like Capes or octaNe the mechanics determine not success or failure, but rather who gets to narrate – as that ability is far more important to these sorts of games).

    Now, there is the possibility there that someone will take offence that I’ve claimed narrative games are objectively “better” at telling stories than (say) gamist designs, but understand that this is NOT saying that narrative games are better in any absolute sense; to derive that implies that storytelling is what RPGs are about, which is – in many ways – what Robin is saying above, and what Chatty is disagreeing with.

    I think the entire argument is moot, because I don’t think storytelling is INNATELY what RPGs are about. It’s what narrative games are about, no question, but I think D&D is best enjoyed as a game with a very loose storyline where you don’t really try and conceal the fact that, while you might SAY you’re rescuing the princess because it’s the “right thing to do”, you’re REALLY doing it because you figure that the adventure will give you enough XP to reach level 10. Such games are fun, and no matter what anybody tries to tell you, it is not innately “wrong” to play D&D this way. (However, understand that gamist RPGs, of the three types, are the ones most susceptible to automation – you will see far more MMoGs that play a reasonable game of D&D than you will something like Amber. That’s not a bad thing – just an observation).

    But it’s when you consider the third category of games – the Simulationist type – that things really start to unravel. When you’re playing a Simulationist game “properly”, you’re really trying to force a square peg into a round hole if you try to impose a plot. When you’re preparing for a session of something Simulationist (and GURPS is probably the most well known example of a game system that tries to do this), you should be designing the world, not creating a story. Stories will undoubtedly occur, but they will do so organically – things will happen for reasons internal to the game world, rather than because the GM has decided that NPC X is the subject of tonight’s adventure. These types of RPGs are hard to do well – lots of players tend to drift in the absence of the clear goals that a Gamist adventure has, or the dramatic necessity of a Narrative adventure – but they can be a lot of fun because of the tremendous flexibility and exploration possibilities that are afforded. (And these are the easiest of all, theoretically, to automate – indeed, many MMoGs simulate a world using rules that would be far too tedious to use in a tabletop game run by a “mere” human).

    Sorry for rambling, but my point is basically this: RPGs are done a disservice by the “modern” view that they are “supposed” to be about telling stories. It’s not that you can’t do that in D&D or whatever – it’s that you shouldn’t feel that you HAVE to do that. “Pure” gamist campaigns where the objective is a very thinly veiled “get to level 20 and then attain godhood via a series of killing things and taking their stuff” are not inferior to structured epic campaigns on the order of the Iliad.

    All IMHO, of course. πŸ™‚

  5. Wow that was one hell of a comment Gazza… πŸ˜‰

    Your comment about simulationist game seems to be the exact way I was GMing my GURPS games back in the day. Branching/simulationist as always been my default GM style.

    It requires heavy preparation (defining the world) and serious improvisation skill (running the game). The major problem is to guide your player to avoid stagnation. Since you don’t have a well defined plot. My usual ploy was to have something huge fall in the hand of the player and move on from there.

    “You’ve find some CD and when you try to read it’s content it’s heavily encrypted… (Unknown to the player it’s classified military secret that multiple agencies will do anything to get)”

    Or use the capacity of one of the player to feed plot hooks.

    “You’re part of a private investigation team driven by the vision of one of your member who is somewhat precognitive”

  6. Yeah, by the time I was about half way through that I realised I’d probably rambled a bit, but I couldn’t really see what to edit out. Sorry about the length. πŸ™‚

    I love to tinker with things like economic systems, small scale battles, and stuff even when I’m playing D&D, but like I allude to – it’s not really designed for that, and rather than try to achieve “one game to rule them all” it’s arguably better to try different games.

    On the other hand that’s highly dependent on whether your group shares your enthusiasm. Sympathise with my plight: there are four of us, and I prefer narrative games (though I’m partial to simulationist as well), one of my buddies loathes narrative style (but loves simulationist), and the other two loathe simulationist (but like gamist). So we’re always forced to pick something that will annoy the least number of us, and we switch around a lot. πŸ™‚

    But having labels for these things at least allows me to understand WHY my enthusiasm for game X won’t necessarily be shared. (REALLY disappointed that my simulationist-fan friend didn’t like Capes though, as the other three of us had a ball).

  7. Well, I don’t disagree with much of what you said, GAZZA, except that while each type lends itself to a certain style, nothing prevents a game from “dipping” into other types, such as a narrative in D&D, or gamism in Capes. Certain games are predisposed to making a certain style easier, but there’s no way to play D&D “wrong” by introducing the other styles.

    But it doesn’t really matter, because GNS pisses me off anyways.

    Or, rather, many of the proponents of GNS piss me off.

    GNS is a thought experiment, and a design guide. Some of these people? They just take it way, way, WAY too seriously.

    They’ll talk down about “gamism” as though it was a cardinal sin, they’ll tell you you’re playing your game “wrong”.



    I think I’m done now.

    In any case, my favourite game theory is, of course, Jeff Rients’ RSP model, followed closely by Cheetoism.

  8. Whoa, I leave for a doctor’s appoint ment with my son and I get a face full of GNS on my way back?

    I completely get your points Gazza and I’m happy that the labels help you. Like Graham, I too am not too keen on the more vocal proponents of the theory and I really don’t like discussion on coherence or lack thereof (and fail to see the relevance, and please don’t try) of a RPG. I’ll stop there lest I start ranting.

    I think that what really counts is player preferences (including the GMs) as pointed out by Gazza. Since I’m tackling this in chapters and I already

    Now for the subject at hand, my main points of divergences with Robin Laws are:

    1) Unstructured Adventures aren’t all that common (or at least tend to gain structure) as a playing group explores the possibilities of RPGs (either within the same family of games, or by switching systems)

    2) For adventure design, I really think that you should pick any structure and mix freely so you the GM/DM gets the adventure in the best possible form that you pictured.

  9. I should clarify, as well.

    None of that hatred was actually targetted at you, GAZZA, or at anyone who finds the system helpful.

    Hell, I find the terms helpful sometimes.

    It’s just the vocal jackasses that bother me.

  10. I think that too much story leads to railroading, something that I’ve realized was wrong with my own games. The characters need to be able to tell the story by their actions and what decisions that they make. If there isn’t any decisions to make, or if what your PC’s are doing isn’t going to change anything which the DM has planned, then we DMs are failing to set up our encounters properly.

    When ever a story is involved, some amount of railroading is going to take place, and it should! PC’s generally need directions on where to go next, thus we need to give them options during the game on things to explore further, sometimes these are meaningful to the story that is taking place around them, and sometimes it is a total red-haring. The ultimate purpose of this is to let the DM know what he needs to focus on during his prep time, of course, the PC’s could chose to simply ignore the planned adventure, this could be on purpose but most of the time it is on accident, which when this happens, it separates the true DM’s from the people who have just read the DMG a couple of times.

    Another thing to think about is that the players want to play D&D, that is why they are showing up, and it is our duty as Dungeon Masters to provide THAT game, vs. our own. Not to say that we need to depend on modules and settings and never create our own worlds, this means that we need to stay true to the rules, and use established monsters over our own fabrications. . . well, most of the time. It never hurts to put ourselves into the game by creating a new monster or religion, but as long as they stay true to the spirit of Dungeons & Dragons, then we aren’t changing the game to suit us, but enhancing it for our players.

    My main point is, to much of anything is always a bad thing. To much story can kill a game, just like a dungeon that is nothing but traps and puzzles, or a game that is nothing but combat. Any session needs to be balanced, it needs conflict to keep all of the characters moving, puzzles to slow the game down, strategy to avoid potentially lethal errors, role-playing to give information to the PC’s, and, of course, rewards. Though, I think that randomness has a very important place in the the D&D game too. Randomness keeps the Dungeon Master in line, and allows the world to really come alive, vs. total railroad jobs. It also allows for mystery and danger which the DM can’t properly establish unless he is willing to allow the gods of the dice to have their say.

    Good Article Chatty!


  11. Man, it never would have even occurred to me to take any of that personally. πŸ™‚

    I’ve not encountered “GNS Nazism” before; I used the terms purely to illustrate a (belaboured, possibly) point that the point of an RPG wasn’t necessarily “story telling”. I agree 100% that few games are “pure”, and that most D&D games dip into narrativism domains (via a structured plot) – most of MY D&D games dip into simulationism as well. And I’ve yet to run, play, or observe a White Wolf game that wasn’t played in a fairly gamist “superheroes with fangs” fashion – though I’m perfectly prepared to believe they exist.

    I apologise for use of the term “coherence” there Chatty; I used the term because I dislike saying “realism” for games that involve magic and fire breathing winged reptiles. Would you prefer a different term, or is it just that you think complaining about things like that is largely unimportant? (I’d probably agree with that; the inconsistency doesn’t necessarily detract from playability in practice, because most groups have a social contract to prevent that sort of thing).

  12. It’s quite all right Gazza… I had to wait 2 hours at the hospital only to be told that there was a snafu in the appointments and that my son only needs to be seen next week. I was grumpier than usual.

    In regards to the term incoherence, it’s just that I’ve seen it used in so many conflicting ways just to try to prove how D&D was badly designed that I grew tired of the term in general.

    I’m not trying to censor you,. You just caught me off guard with your tangential rant a bit is all.


  13. On GNS: The “only one agenda at once” happens due to arbitrary definitions or is untrue or means that there are agendas missing (take your pick). It is not advisable to focus much energy on that part of Forge theory, anyways, the others are generally more interesting.

    There’s a difference between focusing on story and flavouring the game with story; the former and D&D take some serious work, while the latter works just fine. Gradual transition may or may not work well but is worth an experiment.

    Rip; I think that it is both easy and fun to run games with story that have no railroading. I do it all the time. This may be just a case of different definitions, though.

    Tommis last blog post..There shall be war.