My recipe for starting adventures

I've got a little process that I use whenever I start a game of Dungeon World or Homebrew World. It's similar to the first session procedure that's described in the book, but different in some key ways. I've found that this approach reliably kicks off a new game quickly and with a lot of energy, in a way that makes it pretty darn easy to run and improvise.  

Here's the recipe:

1. Establish the adventure's premise with the group
     > Premise = a fantastic location + a grabby activity
     > Do this before anyone picks playbooks or makes characters
   
   
2. Players create characters, GM writes/updates hook questions, which should establish:
     > Motive: why are they here, doing this?
     > Stakes: what's on the line, why is this important?
     > Urgency: why shouldn't they dawdle?
     > Dangers: what do they expect to face? what do they know about them?
     > Detail: what specifically are they hunting/seeking/fleeing/fighting/etc.?
     > Complications: what's getting in the way? making it harder? constraining them?
   
   
3. Do introductions (by name, pronouns, class, and look).
     > Do not do bonds (or in Homebrew World, background questions) just yet. 
     > You're just establishing who the characters are.
     > Yes, you can ask questions, but keep it light for now.
   
   
4. Ask a few of your hook questions
     > Usually 1-3
     > Pick questions that elaborate on or clarify the premise
     > Address specific PCs, not the group at large
     > Ask follow-up questions; encourage the players to do so, too!
   
   
5. Do bonds (or in Homebrew World, background questions)
     > Ask follow-up questions; encourage the players to do so, too!
     > Use this to establish how they know each other, why they're working together
   
   
6. Finish asking your hook questions
     > Doing bonds/background questions often rolls naturally into this
     > Ask follow-up questions; encourage the players to do so, too!
     > Ask additional questions as they occur to you
   
   
7. Frame the initial scene, tightly
     > Start in media res or at least right on the verge of action
     > Who, where, when, doing what?
     > Give up to 3 strong impressions, ideally from different senses
     > Make a soft GM move
     > "What do you do?"

That's the recipe!  More about the background, details, and suggested prep after the jump break.

I keep hearing good things about Delicious in Dungeon,
but haven't read it; just seemed appropriate, y'know? Cuz recipes.

The "standard" Dungeon World approach

As I said before, the process I use is similar to the standard Dungeon World approach, but different in some key ways. In the first session section, the Dungeon World text tells you (as the GM) to...

1. Bring something to the table, even it's just a head full of ideas.
     > It's okay to bring a situation, or a mapped out location, as long as you leave blanks!
     > "The one thing you absolutely can't bring to the table is a planned storyline or plot."
   
   
2. Have them make and introduce characters, then ask questions like crazy during character creation—especially while the PCs establish bonds.
     > Use those questions to help build the world together with the players. 
     > The examples given are largely world-building type questions.
   
   
3. Start play already on the adventure: at the door to the dungeon, in the king's audience chamber being sentenced for a crime, with the lizardfolk attacking you, etc.
    
4. Continue asking questions, and turning questions back on the players, and using those questions to build the world and the adventure.

Then, throughout play, you're told to continue asking questions. "If you don’t know something, or you don’t have an idea, ask the players and use what they say."

The benefits of this approach include:

• Low prep. As the GM, you don't have to bring a lot to the table. 
• Creative collaboration. More brains working together will (usually) come up with more interesting, unexpected ideas than what you would have come up with by yourself.   
• Instant action. You should be able to get character creation done pretty quickly, then start the adventure in media res, and get some actual play in, all within a couple hours. 
• Inherent buy-in. If you do this right, you don't have to present a hook to the PCs and convince the players to go on your adventure. You start them on the adventure, and then ask them why they're there. Boom. They've hooked themselves.  
  
  

What can go wrong with the standard approach?

Honestly?  A fair bit. The biggest problems, in my experience, stems from a lack of structure, inputs, and constraints. 

There's a pattern that I've seen, both in my own games and when folks ask for help online. They did the first session, and asked a lot of questions about the characters and the world, but they run into these problems:

• One or more players were uncomfortable contributing, or just not good at it
• The details established were big, world-defining issues, high on concept but short on specifics
• The details established were contradictory, or at least jarring in theme in and tone
• It wasn't entirely clear why the PCs were together
• The group spent most (if not all) of the first session creating characters and talking about the world, and very little (if any) time actually playing the game

As a result, the GM is looking at this gooey mass of ideas and trying to make sense of them, trying to figure out what to do in the next session, how to make everything come together into an actual adventure, or how to make the details that were established early on actually matter.

This doesn't happen to everyone, or even happen to specific GMs every time. I think it's more common with large groups, and with less confident/assertive players.  A "good" GM can typically salvage the situation when this happens, but it'd be better if we can just prevent it from happening in the first place, yeah?

Creative Crystallization & Hard Edges

(I swear that I got this concept from Dan Maruschak back in the days of Google+, but if I did, it must not have been in any of the archived communities.) 

more cooking metaphors, coming right up!

Creativity, especially improvised creativity, benefits from hard edges to crystallize on.

If you're making caramel from scratch, you need to make sure that the pan you're cooking it in is extremely clean. Any impurity provides an "edge" for the heated sugar to glom onto, and then it crystallizes from there, the crystalline lattice building on itself and getting bigger and bigger, and now you've got a big chunk of burnt sugar instead of a soft, creamy, delicious caramel.

This is, like, the opposite of that. You want creativity to crystallize and start to grow and grow and grow on itself. And in order for that to happen, you need hard edges for the creativity to latch onto. Without those hard edges, the creativity is likely to be soft, amorphous... creamy. Not what you're after. 

For example, let's say that you pick a human wizard as your class & race. Now I ask "what's magic like in this world?"  I'm not giving you much to work with. No hard edges for you to crystallize on. I shouldn't be surprised if you stare at me blankly or give me something very bland and high-level like "um, it's really common?"

(Now, maybe you're the kind of player who wouldn't have any problem with this at all. Maybe you glom on to some detail in the playbook, or who just have a head buzzing with ideas, and your brain shoots out something specific and interesting like "I think all magic is demonic and corrupting and there's like an order of witch-hunters dedicated to stomping out spellcasters!" But unless we've played together a bunch before, I can't really count on that, can I?)

By comparison, let's say that we've already established that you're standing at the bottom of a ruined amphitheater, carved out of a mesa. There are tunnels and passages carved into the mesa, entrances visible at the upper levels of the amphitheater's seating. Ruins dot the scrubland all around you. The wind is ceaseless. You and the other PCs are here, trying to stop something bad from happening. You're playing a druid from the whispering plains. I ask you "What sort of destructive spirit do you think dwells in this mesa?"  

I don't know what you'll say, but I bet you'll have an easier time coming up with something than if you were the wizard in that first example. Your creativity has so much more to build on: the amphitheater, the mesa, the ruins, the scrubland, the wind, the fact that you're here trying to stop something, the fact that you're a druid of the plains, the fact that it's destructive and a spirit... everyone one of those details is something that you can use, riff on, extrapolate from. 

• Maybe your creativity catches on the wind and the ruins and the scrublands, and you think that it must be a demon of dust and drought, responsible for laying this civilization low. 
• Or maybe maybe your creativity catches on the theater itself, and you think that this whole place is haunted or cursed by the spirit of a mad playwright. 
• Or maybe you're picturing this amphitheater as more like an arena, and the destructive spirit is some sort of extraplanar entity that this civilization used to summon and bind for gladatorial combat. 
• Or maybe you think about the scrublands and the whispering plains and think that maybe it's a locust spirit that's taken up residence here, one that you need to put down before it brings famine to the whole region.

The more details that have been established, the more patterns that are already in place, the easier it is for your brain to add new patterns, new details. You don't want things to be so packed with details that there's no room for creativity to grow (draw maps, leave blanks, right?) But it's easier for your creativity to crystallize if you've got a bunch of hard edges to grow on.

Side note:  some classes/playbooks come with more "hard edges" than others. For example, playing an Artificer says a lot more about the world and the things in it than playing the Thief. The more specific and niche the class, the more of an edge it brings. Likewise, I've found that classes with backgrounds tend to have harder edges than classes with "racial" moves. I first noticed this with Brady Lang's "humans only" mini-hack of the original DW playbooks.  Mechanically, the Path moves are almost identical to the Race moves in the DW books, but simply changing "Elf" to "Weaponmaster" makes the Fighter a much more specific, much more interesting character. That's a big part of why Homebrew World uses backgrounds instead of race. You can still choose between human, elf, dwarf, etc. but it's just a part of your look and the fiction, with no inherent mechanical impact) HBW also replaces bonds with a series of a "Which one of you..." questions that are tailored to the background, which provide even more hard edges to crystallize on.

The differences between my approach for starting adventures and the standard approach in the Dungeon World text is pretty much all about providing hard edges––about giving players stuff for their creativity to crystalize on. This keeps the benefits of DW's "official" approach (low prep, creative collaboration, instant buy-in, instant action), but makes it less likely that you'll end up with a gooey mass of ideas that you don't know what to do with.    

Let's take a look at the specific steps.

1. Establish the premise with the group

Establish a premise for the adventure with your group before they make characters.  That means you should bring something with you, or a choice of somethings. You should show up with a premise in mind, or a way to come up with one, plus whatever additional prep you feel will be helpful (more on that later). 

By premise, I mean: a specific, immediate, and evocative situation for the players to get excited about and interested in. Something that their creativity can crystallize on. I'm not talking about sub-genre or tone or even big-picture world-building ideas. I'm talking about where we are and what we're doing. 

For example, these are NOT good adventure premises (at least, not by the themselves):

• D&D in spaaaace!
• Medieval fantasy Western
• PCs as special forces in a fantasy World War
• Everyone works for the Library 
• Sea dogs! 

These are all good campaign concepts and I'd potentially enjoy playing a game using any of them. They tell us some cool stuff about the world, or what the PCs might be doing in general. But they're insufficient for the purposes of kick-starting an adventure. They don't establish enough of a situation, and that's the hard edge that we really need and want.

In my experience, the right level of detail for your adventure's premise is: a fantastic location + a grabby activity. 

The location should be mysterious, exciting, strange, and/or dangerous!  A sprawling necropolis, a city filled with thieves and cultists, an ancient temple, a floating city, a mysterious island, and vine-choked ruin, a storm-tossed ship, a great city under siege. It doesn't necessarily have be magical or fantastical, just fantastic. 

Hang Sơn Đoòng--a perfectly real, perfectly
natural, and utterly fantastic location

Your premise should also establish that the PCs are doing something, something that implies adventure, conflict, tension, excitement! They aren't just hanging out. At the very least, they're exploring this fantastic location. But you can do so much better! They could be fleeing from something, or chasing someone or protecting someone. They could be defending the place or assaulting the place. They might be here to kill someone or to rescue someone or to hunt and kill a monster or to put down a threat. They might be seeking an artifact or a hermit or a the location of a lost city/temple/vault. Etc. etc.

So, going back to those concepts above, if we were going to come up with a good adventure premise for each one, it might look like:

• Smuggling precious cargo through a githyanki blockade
• Chasing fugitives into a haunted, abandoned pueblo city  
• Hunting an enemy commando unit through the ruin-dotted swamps outside of town
• Retrieving a dangerous book from a noble's fortified and sumptuous estate
• Desperately seeking food and fresh water on an uncharted jungle island 

Notice how all of those premises imply their underlying concepts (D&D in space, fantasy western, etc.) without directly stating them. I don't think that there's much harm in directly stating the concept with the group up front—if you want to play D&D in space, you should tell everyone that up front and get their buy-in. But if you're not dead-set on the concept, I find that it can be quite rewarding to present the premise of the adventure (only) and then see what world-building concepts arise organically.   

This is one place where Homebrew World differs significantly from other hacks/drifts of Dungeon World. One Shot World in particular tells you to make characters (with a personal goal), then collaboratively create a map and world first (and the playbooks include questions to encourage and help with that worldbuilding).  Yochai expounds on his method for communal worldbuilding here. The tools he presents are great, and I know they work well for him. He's run like over 100 one-shots with his approach successfully, and says that he's usually able to spin up a game from scratch to the first "what do you do?" in about 30-45 minutes. Personally, I feel like you could easily fill 2-3 hours with all that world-building, and still not end up with a concrete, specific adventure.

So, in my experience and opinion: it's better to start with a tight little adventure premise—a grabby activity in a fantastic location—and build the world out around that.   

Example:  A while ago, I got together with three friends to play an impromptu game of Homebrew World.  I didn't really have anything prepped, except for an idea that'd been rattling around in my head for a while: they'd be seeking something in the ghoul-infested crypts beneath Kravenghast Manor.   That's all I had, really. I had some ideas of what might be down there, sure (lots of ghouls, an ancient Ghoul King, maybe an ancient artifact?). I think I had originally conceived the idea as a sort of "Indiana-Jones-beat-the-Nazis-to-the-artifact" sort of adventure, but I that's more detail than I needed.   So I said: "For tonight's adventure, you'll be trying to get something out of the ghoul-infested crypts beneath Kravenghast Manor. Go ahead and make your characters."

2. Make characters, write hook questions 

Have the players pick their playbooks and start making characters. Do all the logistical stuff: answer their questions, clarify the rules, "there might be other wizards out there but you're The Wizard," etc. 

If they ask questions about the world, or start making grandiose statements (like "I think the gods all died in a cosmic war"), respond enthusiastically but try to rein it in a little. Encourage them to hold these things as ideas and not as truths just yet. "That's cool, but let's wait to establish stuff like that until we're all ready to do so together, yeah?"

Meanwhile, you should be writing hook questions. Or, if you brought some with you, you should start thinking about who you're going to ask each one to.  

Hook questions are leading questions that help you flesh out the adventure, the world, and the characters all at once. You want questions that will establish all the following: 

• Motive: why are they here, doing this?
  
• Stakes: what's on the line, why is this important?
• Urgency: why shouldn't they dawdle?
• Dangers: what do they expect to face? what do they know about them?
• Detail: what specifically are they hunting/seeking/fleeing/fighting/etc. 
• Complications: what's getting in the way? making it harder? constraining them?

If you ask those questions, framed the right way, you'll immediately have buy-in for your adventure.  You won't have to meet in a tavern, present rumors, or have a mysterious stranger offer them a job.  Instead, you get to start the adventure at the point where it gets interesting.  It's automatically interesting to the players care because they made it up, and its automatically meaningful to the characters because the players just told you why they care!

The trick is, you don't want to just straight-up ask the players "hey Bob, why are you trying to smuggle something through the githyanki blockade?" You want to craft the questions a bit. I could write a whole blog post about the art of asking questions, but here's the "quick" version:

1. Address the characters, not the players:  "Krikor, what sort of beasts are you worried about encountering in these badlands?" is better than "John, what monsters inhabit these badlands?"
   
   
2. Assert as much as you ask:  If the premise is "we're seeking something in a ruined manor," then don't just ask "what are you seeking?"  Make an assertion with your question. For example: "What treasure is rumored to be buried in the caves beneath this ruined manor?" asserts that there are caves and buried treasure. Each assertion gives the player another hard edge for their creativity to grow on. 
   
   
3. Assert things that you want/need to be true:  Frame your questions so that they assert any details or ideas that your prep relies on, or that you just really want to be true. (You might want it to be true because you just think its awesome, or want to head off certain shenanigans, or whatever. You're allowed to have an opinion, too!)  
   
   For example, if you've prepared some moves and dangers and maybe a map for exploring a cramped, chaotic, twisting sewer system, then don't just ask "Vigo, what makes these sewers so dangerous?"  Vigo's player might tell you about the huge chambers with caustic chemicals and implacable machines that relentlessly process waste, creating a very different place (organized, big open chambers, magic-tech vibe, etc.) than the crumbling horror-show you prepped for.
   
   Instead, ask something like "Vigo, so... you and your previous crew were hired to map the city's beehive of a sewer system a few months back... who hired you, and what horrible fate befell the rest of your crew?"  This asserts that the sewers need mapping, and establishes that they are vast, unknown, labyrinthine, probably kind of cramped. The assertion that they met a horrible fate establishes the danger, with a tinge of horror. The fact that someone hired them establishes some interesting NPCs and likely leads to follow-up questions like, "Oh, who are they? What do you suspect their interest was?"
   
   
4. Ask for meaningful contribution: Asking "What's the name of the merchant you know in Hightown?" is kind of weak-sauce. You're not asking them to contribute anything that matters. It's just a name.
   
   Better, ask "Who's the merchant you know in Hightown, and why do you trust them so much?"  That prompts the player to create some backstory, and gives them a large amount of say over the merchant NPC's personality. (And maybe you really wanted/needed them to trust this NPC—and now the player is telling you why they do!)
   
   
5. Be specific, but not too specific:  Even better, ask:  "Who's the silk merchant in Hightown that you're planning to visit, and why do they treat you like family?"  This still invites meaningful contribution, and it establishes something you want/need (a close connection to an NPC merchant), but it gives them more specific details—hard edges!—to crystallize on (a silk merchant, going to visit, treats them like family). 
   
   Careful not to go too far, though. If you ask "...and how did you save their life, so that they treat you like family?" then you're kind of painting them into a corner. Unless it's really, really important that they saved the NPC's life and that the NPC is grateful for it, leave the why up to the player!
   
   The line between "too specific" and "not specific enough" is blurry. It's mostly a matter of taste, and/or judging how a particular player might react. Play around with the level of detail you assert, and try to develop your own feel for it.  
   
   
6. Get personal:  ask questions (and make assertions) about the PCs' past, their feelings, and their relationships. This is actually a bit more powerful after you've gotten to know the characters a bit, but a little dash of intensely personal Q&A right up front can really help define a character. "Who exactly are you here to murder, and why are you so sure that they deserve it?" That establishes that this isn't, like, an impersonal political assassination, or justice, but a rather a murder, and it raises the possibility that the PC's motivations aren't exactly sound. 
   
   Or "What treasure is rumored to be buried in the caves beneath this ruined manor, and how would finding it change your family's fate?"  That asserts that the PC actually has family, that they're doing this for them, and that their fate is somehow in question.  Juicy stuff, no?

As you can imagine, your hook questions benefit from a taking time to write them in advance. But with experience, you can create some workable questions in the time it takes the players to make their characters.

If you did prep your hook questions in advance, then take the time during character creation to think about who you're going to ask which question, and what order you'll ask in. You don't have to commit to anything just yet; heck, I often write down a choice, like "Fighter/Ranger" and then pick one of them to ask once we get into it and I see who hasn't had a lot of input. Or maybe I just figure out who I'm going to ask the first couple questions to, and feel my way through the rest.  

Back in that game of Homebrew World: the players choose the Wizard, the Thief, and the Paladin and start making characters. I start thinking about hook questions. I wanted to start with a combo of motive, stakes, and detail, and for whatever reason, I wanted to ask the Wizard. So I jotted down: "Wizard, who or what are you seeking in the crypts below Kravenghast Manor, and why is it so important that you get them/it out?" To add a complication and a sense of urgency, I also jot down "Who else is looking for __, and why?" (A rival 3rd party is always a good complication!) "Ghoul-infested crypts" already implies plenty of danger, but I think I'll ask the Paladin about something that's changed. I jot down "Paladin, why are you/your order watching this manor so closely?" So there we go: motive, stakes and detail for the Wizard. Complication and urgency (probably for the Thief). Extra danger and some details from the Paladin.

 

3. Introductions (by name, pronouns, class, and look)

Not too much to say here. When everyone is ready, go around and I have each player introduce themselves.  If you're playing Homebrew World, have them include their backgrounds as part of their class. You want something like:

"I am Hawke the Fighter, a veteran of the wars! I'm a grizzled human with a thousand-yard stare, a giant frame, and an oft-broken nose. My pronouns are he/him."

It's cool if the players get a little more involved here, but try to keep anyone from hogging the spotlight. Ask for clarification ("you're grizzled? like how old are you, actually?") and maybe ask some obvious follow-up questions ("how long ago were the wars you fought in?"). Keep it simple, though! Show interest, be enthusiastic. But don't spend more than a few minutes on each character.

Also, encourage the other players to ask questions.  "Anyone have any questions for Hawke?"  Again, try to keep things moving and don't spend too long an any give character. But if you set the precedence now that everyone can and should ask questions, it'll go a long way towards creating a collaborative spirit.

Don't do bonds (in DW) or background questions (in HBW) just yet. I mean, it's okay if it happens, but hold off on doing them formally. You're just figuring out who the PCs are, not necessarily how they know each other. You want to flesh the premise out a bit before you get into that!  (See next step.)

Continuing that game of Homebrew World: the players introduce their characters. We've got:  Roberta (she/her), a fae-touched Wizard of indeterminate age and bloodline (maybe human? maybe elf?), crazed eyes, boney-limbed, draped in amulets and talismans. Thadeus (he/him), a paladin and paragon of virtue, young & innocent human with eager eyes, all shine & polish, lit with an inner fire. Incredibly earnest.  Peter (he/him), an operative (Thief), older human but he's still got it. Thin as a whippin' stick and moves with no wasted movement. A bit of a silver fox.  I remember asking Roberta's player if she was half-human and half-elf, or had some elven ancestry, and she was like "eh? who can tell?"  Okay! I recall that Thadeus's player volunteered that he was a "a paladin of the Light, not really a god, but more like a non-personal source of all life and all hope."    "So is this an established religion?" I ask.    "Oh, totally. Pretty big, fairly old, kind of corrupt in the way established religions get."    Huh, okay.  "And do you hold like an official role or position in the church?" "Oh, yeah, I'm part of an ordained order of paladins. But I'm, like, the only true believer in my order, certainly the only one with any powers. I weird the others out."  Well that's cool. I think I must have also asked Thadeus about his vows (from the HBW Paladin's "Bound by a Higher Law" move). He can't cheat/utter falsehood/deceive with words, he'll always offer mercy, and he'll protect the weak and give aid to any innocent.   I think, when Peter's player introduced his character, he might have jumped ahead and asked one of his background questions.  "Which of us has a... complicated... past?" Roberta's player chimed in with "oh, definitely me." No harm done, but I asked them to hold off on asking any more questions.

4. Ask a few of your hook questions

Once you've established who your PCs are, ask 2-3 of your hook questions. Your goal here is to flesh out the premise. Elaborate and clarify it, until the situation really starts to snap into focus.  

Often, this means asking whatever hook questions you have that establish motive and details, plus maybe one other element.  For example, if the premise has them chasing fugitives into the haunted pueblos, I'd start with hook questions that establish what these fugitives did (details) and why the PCs are chasing them (motive). I could also ask about the haunted pueblo (danger), or about the other NPCs who are part the PCs' posse (complication), or why it's so important that they catch these guys here and now, before sunset (urgency).  But I'd probably only do one of those, and leave the rest for later.

Address each question you ask to a particular character (and thus a particular player), rather than posing the question to the group as a whole. This prevents confident, assertive players from hogging the spotlight, and tends to draw out less confident, less spontaneous players. It also helps prevent that sort of "dead space" that can happen when no one wants to answer, for fear of stepping on someone else's toes. (Yes, I'm from the midwestern U.S.—this is a thing here.)  

Addressing your questions to a specific character (and thus player) is particularly important when playing online. Audio/visual lag often results in people talking over each other, and this approach cuts that off. 

Jot down the answers they give you, to whatever level of detail you find useful. Ask follow-up questions. Be curious about the answers they give you and what those answers imply! Encourage other players to ask questions as well. 

If a player seems to be flailing or unsure how to answer, you can:

• Try rephrasing the question, to help clarify ("Sorry, let me rephrase that... I'm wondering what your personal stake in this is, Krikor... like, why are YOU part of this posse?")
• Try adding some direction or options to choose from ("...like, did they hurt someone close to you? steal something from you? were you hired for this? something else?")
• Tell them to think about it, and ask the next question to a different player, coming back to the other player later
• Ask them what they're considering; invite them to think out-loud. Probably address this invitation to the player, not the character. ("What are you considering, Bob? You want to talk it through?") 

If they give you an answer that just doesn't make sense, then ask them to elaborate and/or rephrase and clarify your question. Try to identify the disconnect, and get on the same page. For example:

GM: "Krikor, who are these outlaws that you're chasing, and what did they do that you're so hell-bent on bringing them to justice?" 

Player: "Oh, these assholes betrayed us during the last robbery and made off with the loot!" 

GM: (thinks to self: "what the hell?", but says:) "Wait, what? What do you mean by 'betrayed you?'"

Player: "Like, we were doing the robbery together, and we pulled it off, but these jerks grabbed the loot and ran and left us to deal with the Law.." 

GM (lightbulb clicks) "Oh. OH! So you're not trying to bring them to justice, you're out to get REVENGE."  

Player: "Yup. Well, and to get that sweet loot." 

Sometimes (like in the example above), a player's answer will turn the whole premise on its ear, or at least take things in a direction that you didn't expect. When that happens, you've basically got two choices: run with it, or try to rein them in.

In general, I recommend running with it when an answer throws you for a loop. Like, in the example above, the GM clearly expected the PCs to be the good guys here, part of a posse trying to bring the outlaws to justice. But when the GM realizes what Krikor's saying—that the PCs (or at least Krikor) was in cahoots with these outlaws, and they're out for revenge and money—well, the GM just accepts it.  

Running with it usually means asking follow-up questions, or might require you to adjust your prep on the fly. Maybe that GM's next question was going to be "Who else from town insisted on joining this posse, and why are you so sure they're gonna cause you grief?" Well, now that they know the PCs are outlaws themselves, that doesn't make sense. So maybe they switch the question to "Aside from the other PCs, who else from the job is with you, and why are you so sure they're gonna cause you grief?"  But they might also drop that question entirely and ask something different, to take advantage of the new details.  "How did you give the Law the slip, and just how close are they to catching up with you?"  or "Who are the famous bounty-hunters on your trail, and why does the thought of facing them make your blood run cold?"      

So, yeah, you might have think on your feet when the players give you an unexpected answer, but... that's kind of the point. You're playing to find out what happens, right?

So, when should you rein them in? I'd try to limit it cases where their answer...

• ...violates the other players' (or your) expectations or boundaries, or...
• ...contradicts or invalidates details/choices that have already been established by others
• ...undermines your prep, in a way that you can't adapt

So, like, if Krikor's player says these outlaws raped and murdered a bunch of villagers, when you've already established that sexual violence is behind a Line (as in Lines & Veils), or if you just noticed that another player feels uncomfortable with that detail, or you realize that you feel uncomfortable with that detail... then you should rein Krikor's player in. "Uh, we yeah, I'm not comfortable with sexual violence being a thing in the game, even in the background. Sorry, I'll add that as a Line. Can you come up with something else?"

Likewise, if Krikor's player says something that's counter to the tone you've all agreed on, like "these guys stole a bunch of prize gerbils from the local gerbil ranch," and everyone was expecting some gritty fantasy-western stuff, then... yeah, rein that in. "That's hilarious, but I think gerbil thieves is a bit gonzo. I'd prefer to play this straight, yeah?" 

If Krikor's answer contradicts an established detail, point out the detail they're contradicting. Often, the player self-corrects with something like "oh, crap, yeah, sorry... how about instead __."  But sometimes, the new detail they added will be good enough that everyone wants to change the previously-established details to accommodate it. Or, maybe you can figure out a way to make both details work. "That's really cool, yeah, but... Jin is playing a Paladin... That might not work with y'all being part of the same outlaw crew that you're pursuing." "Oh, sure... well, let's say that I'm part of this posse because they betrayed me on the last job. But maybe no one else knows that?" "OOOH, yeah, cool!"

If you think that a player's answer undermines your prep, stop and think before you rein them in: does it really invalidate your prep? Can you maybe just adjust it a little bit and still use it?  Do you really need this now-invalidated prep, or can you go off-script and run without it?  If you really, really can't—or just don't want to—then, yeah, rein them in. "Ugh, I'm sorry, but that actually undermines a lot of my prep. I'm sort of expecting that you guys are here with a posse from town..."  But also: maybe next time, make your prep a little looser or your questions a little tighter! With experience and foresight, this is a problem you can avoid.   

Anyhow:  remember that your goal right now isn't to ask all of the hook questions, but just a few. Ask enough to flesh out the premise into a full-fledged situation. Once the situation starts to become clear, do the next step. 

Continuing that game of Homebrew World: we've got Roberta the fae-touched Wizard, Thadeus the young-and-eager paladin of the Light, and Peter the operative Thief (who has a complicated past with Roberta).   My first hook question is the most obvious one: "Roberta, who or what are you seeking in the crypts below Kravenghast Manor, and why is it so important that you get them/it out?" "Ugh," she says in this gravelly, entirely put-out voice, "my TA. You know, my teaching assistant. He stole something from my office and disappeared down into the crypts."  I don't know what I was expecting, but I was not expecting that. I'm briefly tempted to rein it in, or challenge her ("wouldn't that be more appropriate for the Formally Trained or Steeped in Lore backgrounds?") but screw it, let's see where this goes.   "So, you're a professor? At, what a university?"  Yup.  "And what did this TA steal?" "Oh, it was some weird hunk of starmetal. It's been in my office for ages, in the back, collecting dust."  Of course it has, her fae-touched background says that starmetal it anathema to her magic, so this is already all sorts of interesting.      Roberta's player didn't really answer the question of why retrieving them was so important, so I shift to the Thief and get him involved. "Peter, why is rescuing this kid so important to you, specifically?"   He grabs Roberta's idea of the university and builds on it.  "I'm like a fixer and troubleshooter for the University.  Technically part of the campus security staff. And when students go missing, its bad PR and there is so much paperwork.  It's way, way easier to go into the ghoul-infested crypts and rescue him than to deal with that. Or, at least recover the body. Then it's only like a J55 form. But when they disappear, it's, like, UGHH."  And, boom, there we go. We've got a pair of extremely cynical university employees delving into ghoul-haunted crypts in search of a TA they don't really care about and an artifact that the Wizard can't really use, mostly because it's easier than dealing with the bureaucratic headaches involved if this stuff goes missing.   A few follow-up questions reveal that it's not just a university, but The University, preeminent place of learning in all the civilized worlds. There are others, sure, but this is like the big one.    I would never have come up with this on my own, and I am loving it.     I feel like we should get to the background questions, but I kinda want to know a little more about the paladin's involvement. "Thadeus, why did these two come to your order for assistance on this mission?"  (This isn't a question I had prepared, just one that seemed obvious to ask. Also, notice how I'm asserting that they came to his order, and implying that they're working together, but inviting him to answer the question of why.)   "My chapter is charged with monitoring and containing the darkness of Kravenghast Manor. It's like, shrouded in a pillar of perpetual darkness.  Our chapter house is right on the edge of the property." WHAT?!? Awesome. "So where is Kravenghast Manor, exactly? How far away from the University is it?" "Oh? It's on campus."  Yaaaaasss.

Honestly, I pictured something more on a hill, but good enough!

5. Do Bonds/Background Questions

You've established your premise and your characters, and you've fleshed out the the situation. Time to figure out the PC's relationships with each other.  

If you're using standard Dungeon World, now is the time to choose bonds. It's been a while since I've used these, but I seem to recall it working best when one player would read out a bond, filling in another PC's name. Then another player (often the one mentioned in the previous bond) would read and assign one of their bonds, and so on. It's all pretty organic and messy, really—in a good way.  

Bonds should be done in a spirit of collaboration. The players are establishing bonds together, in conversation with each other, rather than writing in names in their bond slots and then revealing them to each other after the fact. Sometimes, the conversational aspect means that they "declare" bonds by asking questions.  "Hey, Krikor, I think that maybe you owe me your life, whether you admit or not. Does that sound good?"  Or "Hey, which of y'all do you think I've sworn to protect?"

If you're playing Homebrew World, you don't establish bonds. Instead, you go around asking questions associated with your backgrounds. For example, here are the questions associated with the Ranger:

 

Pick someone to start (or ask for a volunteer). They choose one of their questions, read it off, and one of the other players says something like "yeah, I think that's me!" Then have the next player go, and then the next, until everyone has asked (and gotten an answer to) at least one question from their list. 

It's possible that one player will ask a question and no one will want to answer. That's fine. Talk it out. Maybe you (as a group) can come up with an angle that makes it interesting. But if, seriously, no one wants to answer "me," then just ask the player to ask a different one of their questions. 

Once everyone has asked their first question, check if anyone wants to ask another from their list. Make sure that folks don't think that they're supposed to ask all 4 questions. They're not! They're meant to pick and choose the ones that seem most applicable or interesting. 

If you're tight on time (like, it's a convention game or a one-shot), then encourage folks to stick to just one question. Obviously, play this by ear. If the tomb-raider Thief asks "which of you put this job together" and that was basically already answered as part of the hook questions, then, sure, let them ask another one!

Regardless of whether you're using bonds or background questions: take notes, ask follow-up questions, clarify, and be curious about the PCs and their relationships to each other. Encourage the other players to do so as well.  Think about what these questions and answers mean, and what they suggest about the PCs' pasts. Try to make connections to the current situation as already established.  

With that said, leave some questions unanswered, at least for now. Maybe write a little list for yourself: "I wonder... what does Jin actually think about Krikor? What did Jin do to get shipped off to this frontier? What was the first tomb that Felix raided?"  Leaving some questions to answer in play can be a powerful tool. It gives you a hanging thread to pull on later, often in a very rewarding way!   

Before we move on, I like to try and summarize the web of relationships that we've just established and how they fit into the adventure.

Continuing that game of Homebrew World: we know that Roberta and Peter are searching for Roberta's missing TA, who stole a starmetal artifact and disappeared into the crypts of Kravenghast Manor. They're searching for him because it's easier than doing the "missing student" paperwork.  They've gotten Thadeus's help because his order, the Paladins of the Light, watch over the manor, and the manor is shrouded in eternal night.   That's a solid situation, so it's time to ask the background questions.  Peter has already asked "which of us have a... complicated... past?" and Roberta said it was her. I go back to that and ask about it. "Like, were you two lovers?"  The two players look at each other and Peter's player says "Eh, probably not? I think its just that we've both worked for the University for a long time, and that we've had plenty of other screwed-up situations like this where we've had to work together."  Roberta's player adds that they don't really like each other, but she at least respects Peter's competence. Peter is... less forthcoming about his opinion of Roberta, but I can already tell that there's sort of a weary camaraderie between them.  Roberta asks "Which of you has the most beautiful soul?" and Thadeus's player jumps in with "I think that's obviously me."  I ask Roberta's player what she finds so beautiful about it and she says "He's so naive. It's precious." Thadeus then asks "Which of you is a better soul than you let on?" There's a little discussion, but Roberta eventually decides that, no, she is not actually a good soul.  So Peter it is. "How so?" I ask. Peter says that, as curmudgeonly as he pretends to be, he really does try to do the right thing. Just... not that hard.    They've each asked one question from the background, and I ask if anyone wants to ask another. They do not!     I quickly recap:  "So, Peter and Roberta know each other from a bunch of similar shit-show situations like this. Thadeus, you've joined them because your order told you to, I guess, and your order watches over Kravenghast Manor. And you know that Peter really is a better soul than he lets on. But not much. That sound right?" It does. We move one!

6. Finish asking your hook questions

Go back to your hook questions and ask whatever's left, and whatever new questions occur to you.

You might find that some of the hook questions you had prepared no longer make sense. That's fine, you're not, like, committed to them. Drop them, change them, or replace them.   

Remember, your hook questions are meant to establish at least the following:

• Motive: why are they here, doing this?
  
• Stakes: what's on the line, why is this important?
• Urgency: why shouldn't they dawdle?
• Dangers: what do they expect to face? what do they know about them?
• Detail: what specifically are they hunting/seeking/fleeing/fighting/etc. 
• Complications: what's getting in the way? making it harder? constraining them?

I find it helpful to look over the list again and consider whether we've established all of them, or if any of these elements feel "murky" still. If any haven't been established to my satisfaction, I'll ask questions until they we're done!

It's also somewhat inevitable that your players will have introduced some ideas that need fleshing out. Ask questions about these, too, until you've got a clear enough picture to use it in the game.  (Remember, though, that it's sometimes good to leave yourself things to wonder about and answer in play.)  

As always: ask follow-up questions, be curious, look for connections, and encourage the other players to ask questions of each other.  

Once you've got the elements above established to your (and your players') satisfaction, then move on!

While everyone was making characters, I wrote three hook questions: "Wizard, who or what are you seeking in the crypts below Kravenghast Manor, and why is it so important that you get them/it out?"  (motive, detail, stakes)   "Who else is looking for __, and why?" (complication, urgency) "Paladin, why are you/your order watching this manor so closely?" (danger) I've already asked the first question (that's how we learned about the TA and the missing starmetal artifact). That has established motive (avoid paperwork) and stakes (paperwork, the TA's life, the missing artifact), and some details (TA, starmetal artifact). We've also established that Thadeus is here because his order is watching Kravenghast Manor (shrouded in eternal darkness, on campus, ghoul-infested crypts).  The second question is still valid, but given Roberta and Peter's rather cynical motives, I decide to crank up the urgency and stakes a bit.  I ask the Thief:  "Peter, who else is looking for this starmetal artifact, and why is it important that you find them first?" Peter's player tells about an agent/troubleshooter from a rival college (I think he named him "Sparrow") who was just... unprofessional. "Kind of guy that doesn't care who gets hurt. Never hesitates to put lives in danger. He's just... not okay."  I consider asking more probing questions, but it's clear that Peter really dislikes this guy and I've got a good handle on how to portray him (ruthless, wreckless, selfish).  I do ask Peter how he knows Sparrow is here, and looking for the starmetal artifact, and Peter says (mysteriously) "I got people. They told me."  The third question, about why Thadeus's order watches the manor so closely, has basically already been answered. They're Paladins of the Light, and the manor is shrouded in a pillar of eternal night. Seems pretty obvious why they'd watch it so closely.  So instead of asking that, I ask "Thadeus, why has your order been so worried about the Manor recently? What’s changed?” (Asserting that they’re worried, there’s something wrong, but I don’t know what.) “Oh, the pillar of darkness that enshrouds the place? It’s been growing. Quickly.” Ooh, nice. I think he also tossed in a prophecy re: a force of primordial darkness breaking loose.   I've got motive, stakes, urgency, details, dangers, and complications all tied up. There are just a few things left I want to know.   I ask Thadeus why his order picked him (specifically) to help Roberta and Peter ("Oh, I’m the only true believer here, and everyone else is like HAVE AT IT, KID.”).   I asked Roberta what her academic studies focus on ("portals, mostly") and how the startmetal artifact was related to those studies ("well, I think it's some sort of key, but I can't make use of it because it's anathema to my magic, right? so it's just been collecting dust").  I also ask Roberta for the TA's name ("pfft, I dunno, Simon I think") and why she thinks he stole the artifact and brought it here ("oh, I think I found some of his notes about some cosmic portal beneath the manor? I think he was trying to get a good paper out of it this, cut me out of his research").    Okay, so we've got: An ambitious (maybe bitter?) TA named Simon who stole a starmetal key and disappeared into the ghoul-haunted crypts beneath Kravenghast Manor, possibly to use said key to open a "cosmic portal.'  A pillar of eternal night enshrouding said manor, growing rather quickly of late.  An order of Paladins of the Light that are supposed to be watching over this place, and who have some sort of prophecy about a force of primordial darkness arising from within, but who can't really be bothered to do anything about it and are like "eh, kid, go check it out."  A professor of portal-magic (Roberta) and an agent from the University's security department (Peter) who are honestly more worried about paperwork than they are about getting Simon the TA back.   A rival of Peter's, Sparrow, who's also on campus and looking for this starmetal artifact, and isn't above letting some folks get killed in order to get his hands on it I'm loving it. I feel like this is almost going to run itself.

7. Frame the initial scene, tightly

Maybe take a quick break. But then: time to start playing. 

Pick a specific place and situation to zoom in on. They aren't on their way to the adventure, they're already there. 

If I'm NOT worried about real-world time constraints, I'll typically start things at the threshold of the adventure. Like, in that "pursuing outlaws into a a haunted, abandoned pueblo city" scenario, I'd start with them riding their horses into the ravine and seeing the first of the cliff-dwellings.  For a more traditional dungeon delve, they'd start at the entrance, or at least overlooking it. I might ask some more questions—about the journey, the weather, the NPCs who are with them (if any), what they're planning to do, etc. 

Sometimes, I'll have them Undertake a Perilous Journey (in standard Dungeon World) or Venture Forth (in Homebrew World) to inform exactly how close they've gotten and what, if anything, it's cost them. (For UPJ, we'll talk about the danger they encountered on the way, and figure out what happened based on the scout's roll. On a 10+, they got the better of it, a 7-9, maybe someone's down a few HP or bandages, and on a 6-, maybe they all are!)

Alternately, I might throw together a quick custom move, like: 

"Fighter, tell us how you got past the lizardfolk who inhabit the swamps around this ruin. I'll tell you what STAT to roll. On a 10+, cool, it worked. On a 7-9, it worked for now, but they're either actively looking for you or they're working themselves up to follow you; on a 6-, they're in hot pursuit!"

If I AM worried about real-world time constraints, I'll push things even further along and start them out in media res, in the middle of a dangerous situation: a fight, fleeing something, a tense standoff, fording a raging river, scaling a sheer wall, etc.  Basically, jump straight into the action, and then fill in the details later (asking questions and using their answers). 

Either way:  I'll describe the environment, the weather, the time of day. I'll try to give three impressions from different senses. We'll establish where everyone is and what they're doing. I'll answer any questions that folks have, and ask questions as appropriate ("What are you using for a light source?" "What's going through your mind?" "What here tells you that this place once teemed with life but was suddenly abandoned?"). 

Then, I'll make a soft GM move, something that prompts a action (either by giving them an opportunity or by forcing them to react). I'll ask "What do you do?"  They'll respond. We'll play Dungeon World!

Time to get this game of Homebrew World started!   This is a one-shot, so honestly, I probably should have started them in the middle of a dangerous situation, maybe already in the crypts and fighting off some ghouls. But I was more interested in building up tension, so I framed the first scene on the threshold of the adventure, with them entering the pillar of eternal night surrounding Kravenghast Keep and trying to find Simon the TA's trail.     "Okay, so it's mid-morning when you pass into the pillar of darkness, a disquieting experience to be sure. The manor house, once a grand edifice, now slumps on the hilltop and a massive cemetery spills out around it. Everything is white and grey and pitch black in the moonlight. It's quieter in here, too, stiller. The sounds of the daytime world outside are muffled, and even the wind seems to avoid this place. As you make your way, what's something that you each notice that tells you this place hasn't been touched by daylight for dozens if not hundreds of years?"    We get some good details (dead, cracking trunks of once-mighty trees that long ago lost their branches; thick, pale lichen growing on every stone; mud and gravel everywhere instead of grass).  They're not carrying a light source yet, and they're trying to be quiet (at least for now).    The scene is set! I offer them an opportunity, with a cost (danger, time): "Roberta, Simon's notes indicated a particular mausoleum, or maybe like the markings and carvings on one. Seems likely that he'd go there. But there are well over a dozen mausoleums in the place, and it's dark, and there are ghouls sniffing all about. What do you do?"    And we're off!

Using This Stuff in Play

Okay, so you've followed this recipe and now you've got all this stuff established: a grabby activity in a fantastic location, personalized motives and stakes for the PCs, a sense of urgency, details and dangers and complications and relationships between the PCs.  You've framed a scene and made a move, and you're playing.  

What do you do with all of that stuff you've established? Mostly: you let your creativity crystallize on it. When it's your turn to say something, say something that builds on one or more of these established elements. Look for connections. Ask yourself (or the players) questions about them.     

But also: drive play towards answering the adventure's implied questions. The adventure you just set up asks questions about what will happen. You also probably still have some questions about what has happened or what is true about the world. For example: 

• Will the PCs bring the outlaws to justice?
• Will Krikor get away with the loot?
• Will the others (especially Jin) find out about Krikor's past?
• Will Constance and the other NPCs in the posse survive?

• Who were the pueblo's original inhabitants, and what happened to them?
• What is the Tear of Heaven, really?
• What exactly is the thing the dwelling in the High Cave? 

Basically, now that you've started the adventure, what are you playing to find out? These are the implied questions posed by the adventure's setup, and you want to drive play towards answering them. 

Not sure what scene to frame next? Frame a scene that gets the PCs closer to catching up with the outlaws, or that puts Constance in danger, or that reveals something about the pueblo's original inhabitants.  

Someone just Spouted Lore about the pueblo's inhabitants? On a 10+, make a decision (or ask the players to) and answer the question. On a 6-, answer the question but make the answer bad, something that indicates the PCs (or the NPCs, or the world) are in serious danger (probably related to the thing in the High Cave).   

You're well into the adventure, someone rolls yet another 6-, and you're thinking "ugh, what do I do this time?"  Well, look at the outstanding questions and answer one of them—firmly, absolutely, and to the PC's dismay.  "Well, as you're approaching the High Cave, thinking about traps and ambushes, THUNK, there's suddenly an arrow in Constance's chest. It's weirdly quiet. She just sort of stumbles, then sits down, red seeping into her shirt, eyes glassy. Another arrow flits by you, what do you do?"   

The more of these implied questions you answer in play, the more satisfying the adventure will feel. In a one-shot, you'll rarely be able to answer them all, but try to answer the big ones. In an ongoing game, the adventure is done when the implied questions are answered or are no longer relevant. At that point, switch from "normal" play to denouement and epilogue. Wrap things up and start thinking about the next adventure.  

I can't say that I actually write out these implied questions; they tend to be pretty self-evident to me, and I just keep them in the back of my mind and maybe think about them a little during breaks or when I'm stumped for something to do. But if it helps you to actually write them out, then do so!  

As we start playing this Kravenghast Manor adventure, I obviously use a lot of the elements that they've given me. We already made the pillar of eternal night an important detail. When Peter uses his Operative background move to declare that he's got a student on work-study monitoring the manor grounds for the Security Department, I run with it, because it makes perfect sense based on what they've already established. When they biff a Discern Realities, I reveal that Sparrow is already here, in the crypts ahead of them. When they finally encounter some ghouls, the tone we've already established leads me to make them "talky ghouls" (scavengers that absorb the memories and personalities of the corpses they consume) rather than ravenous predatory ghouls.     There are a number of implied questions:  Will they rescue Simon? Or at least recover his body? Will they recover the starmetal key? Will they escape with their lives? What the hell is Simon up to, anyhow? What is Sparrow's interest in all of this? Will the primordial darkness be released?    As the one-shot progresses, the PCs naturally try to rescue Simon and escape with their lives, and at first my hard moves introduce challenges and obstacles to make that harder, like revealing that Simon was captured by the ghouls and taken to see the Ghoul King. As things progress, I push play towards answering some of the other questions, revealing that the Ghoul King is using the starmetal key (and Simon's life-force) to open the portal and unleash primordial darkness on the world.   In the climactic scene, a number of the implied questions get answered. Roberta manages to open a fae-portal back to her office (while Peter distracts the Ghoul King by fighting him). Roberta drags Simon through, and Peter follows close behind, but they have to leave the key because it's starmetal and won't go through Roberta's portal. The entity of primordial darkness starts to make it through the portal, but Thadeus draws its attention, holds it at bay with his righteousness, and engages it in philosophical debate. He ends up binding it to himself, creating a perpetual diode of Light and Darkness deep in the crypts below Kravenghast Manor, turning the pillar of eternal night into a yin-yang like pillar of light and darkness, slowly swirling over the manor.      We never did find out what Sparrow's interest was in all of this, though. Peter got him captured by the ghouls before we could find out. But as we wrapped up the session, I did one last "post-credits scene" where I described the Ghoul King on his throne, the eternal diode of Thadeus and the Darkness still swirling about, the starmetal key laying in the middle of the chamber. Some ghouls drag Sparrow before him. "So," he says to the Ghoul King, glancing at the starmetal. "Looks like you won't be needing that anymore." And we end play there!

What to Prep

I was going to write up a bunch of stuff about what to actually prep when you use this approach, but this is getting long enough.  I'll probably turn that into its own blog post.

The short version is, I recommend prepping the following, in this order:

1. At a bare minimum, prepare an adventure premise, or a choice of premises, or find a procedure you like to generate the premise with the players. 
   
   
2. If you've decided on a specific adventure premise, and you still have time, then prepare your hook questions. If you've still got time, arrange them in the order you intend to ask them, maybe indicate where you plan to break for doing bonds/background questions.
   
   
3. From there, prep as many of the following as time allows (and by "prep" I mean: create, find, steal, re-purpose, recycle, etc.). The 7-3-1 method is pretty good for this!
• Map(s) of the location (probably un-keyed)
• Picture(s) of the location
• Key areas (descriptions, contents, questions you might ask the players) 
• Monsters (stats, descriptions, behaviors)
• NPCs (descriptions, "voice," motives)
• Treasure (nature, value, placement)
• Custom moves
• Countdowns (impending dooms, grim portents)

Whatever elements you do prep, consider it and your hook questions in light of each other. If you prep something that you really want to use, make sure that your hook questions are phrased to assert that your prep is true. Like, if you've prepared a particularly awesome monster (like the glass-eating Jabali boars), then don't ask "What do you know about the dreaded Jabali boars?"  Rewrite that question to something like: "How did you manage to escape from that first encounter with the glass-eating Jabali boars, and who or what did you lose in the process?"

Remember: your prep is for you. Only prep things that will benefit you at the table. Don't waste time on stuff that's obvious to you, or that you can improvise easily, or that's unlikely to be relevant.  Better to spend your time accumulating tools to help with improvisation (like random tables or a monster-generating cheat sheet)

Addendum May 2022: 

Here are some different scenarios that I've prepped that largely use this format and approach:

• The Fallen Sky-City of Kitasa: pretty bare-bones prep... an adventure premise, a series of setup questions, and a rough unkeyed map of the adventure locale. 
• The Obsidian Forests of Yend: a bit more elaborate... a premise, a series of setup questions (with some flavor text), a starting scene, some impressions, =some dangers, some discoveries. 
• Under the Dark of the New Moon: the first adventure I ever wrote or ran for Dungeon World, actually! It doesn't explicitly state the premise, but it's there: you're in a cesspool of a city, searching for "Valleois, that dog," shortly before the new moon when all locals stay locked up at night.