Today, I’m working on that adventure I outlined a couple of weeks ago in the inexplicably titled Mr. Angry’s Wild Ride. The adventure that’s meant to be a quick, punchy, practical example of how to use my Amazing Adventure Building Checklist to build an adventure. And now I kind of regret not calling that thing the AABCs of Adventure Design.

Anyway, I tried to write this article using the same short, punchy delivery I’d used in the previous articles. But it didn’t work. The article didn’t come out the way I wanted it to. I wanted to burn quickly through s$&% and end up with a mechanical framework for the third scene of that adventure in one go. But I ended up talking a lot about something I’m calling mental predesign.

My problem is I’m writing this series for two different audiences. One audience just wants to see the AABCs of Adventure Design in action. They bought into the framework and accepted everything I said. And they didn’t yell at me for changing the definitions of my own jargon because my adventure design philosophy had evolved and improved. The other audience really wants to believe that I have a good way of adding stakes and loss conditions to games like Dungeons & Dragons, but they’ve been burned before and they don’t think RPGs can handle that s$&% and they don’t buy all the crap I’ve said about dynamic conflicts and visualization and scene structure. And they’re really mad that I keep changing my mind about things.

This time around, I ended up spending a lot of time talking to that second audience. I wanted to lead them through the thought process that precedes my mechanical design and to show them how my philosophy and my adventure design tools empower me to create more complex adventures. But most people won’t read more than about 5,000 words at a time. Not even brilliant words written by me. So I only got as far as mentally predesigning the third scene of my adventure. Now, I’m totally okay with that. It’s important. If you want to learn how to write good adventures, you have to come at them with a game designer mindset. You can’t just slap some encounters on a map and call it done.

To make this long story slightly shorter, if you’re expecting some quick, punchy mechanical design, you’ll have to wait for the next practical adventure design installment. This one’s about laying the mental groundwork for quick, punchy mechanical design. But, honestly, this s$&% is going to help you a lot more than watching me make fill out stat blocks and invent rules.

Shower Time Adventure Design

Welcome back to Angry Builds an Adventure for the Umpteenth F$&%ing Time, Hopefully to Completion for Once. In my previous entry in this series, I outlined a crappy little D&D adventure using my Amazing Adventure Building Checklist. In this entry, I’m going to start designing the actual adventure. Well, I’m going to start thinking about designing the actual adventure. So, let’s take a shower together.

When I started this series – AGAIN – I decided I wanted to come to each article as fresh as possible. I didn’t want to do any offscreen design work. I wanted you to see the whole process. Even the part where I had to pull ideas out of my a$& because I didn’t have any good ideas but still had to write an adventure that day. The problem is, when you’re working on a project like this, your brain tends to just start working on its own whenever you let it. And honestly, that’s really useful. So useful that I started writing this long digression about how to maintain a proper mindset for productive, creative work at all times and about the importance of boredom. But I cut that crap out. It doesn’t fit here. It’ll probably show up very soon as an article of its own though. I’m still getting a lot of questions about that s$&%. In fact, I’m getting more questions than ever about how to focus, how to be productive, and how to be creative. Especially when the world seems to be burning down around you. That s$&% needs an answer.

Point is, I can’t really come to each of these articles fresh and make s$&% up as I need it because my brain spends a lot of bored minutes every day thinking about the s$&% I’m working on. Both the purposeful bored minutes I make sure I build into my days and the accidental bored minutes that happen when I’m jogging or showering. When I took my shower this morning – where this morning is the morning of the day I wrote the draft of this article – when I took my shower this morning, my brain ended up doing a lot of predesign work for scene three of my adventure because it knew I’d want to actually sit down and design scene three today. And that predesign work included a thought process that is absolutely f$&%ing vital if you want to write a solid, engaging adventure instead of just Crappy Dungeoncrawl #17B.

So, today, I’m taking you into my shower and into my brain so that you can see how I get ready to design a scene for a D&D adventure that has a solid mechanical framework, reasonable stakes, and loss conditions that don’t rhyme with TPK.

Setting the Scene

The plan for today is to start designing the third scene in the adventure I started building a few weeks ago. The one I’ve decided to call Kobolds Stole the Thing; Get the Thing! Of course, that’s just a working title. It’ll probably end up with some cool name like The Sword of Saint McGuffin or Lost Mine of Fangdeliverer or Five-Talon Discount or something. Having now outlined the whole thing, it’s time to design all the scenes and encounters that actually make up the adventure itself. And I’m starting with the third and final scene of the adventure partly for reasons I already explained and partly for reasons I’ll explain below.

Remember what this adventure is about? There’s a dragonborn thrall of Tiamat who sent his kobold minions to attack a town and steal a thing while the heroes are conveniently in town. The heroes help defend the town and then when the theft is discovered, they head to the kobolds’ lair in an abandoned mine to retrieve the thing. The third scene – the one we’re working on first – is the one in which the players enter the abandoned mine to recover the thing. If they get the thing back, they win. If they fail, they lose. And if they die, they die. So it goes.

At this point, most people would just map out an abandoned mine and fill it with kobold encounters. And they’d pick one room deep inside the mine for the final fight with the dragonborn. And that’s why most people suck at designing adventures. Because that’s crap design. And that sort of crap design is precisely why I define – and design – scenes the way I do now.

When I design a scene, I treat it like its own little miniature adventure. Or like a complete game of its own. And that’s why I start by figuring out how the game is played, won, and lost.

How is this game played? Well, the heroes have to search the mine for the dragonborn. Once they find him, they have to either kill him and take the thing from his cold, dead fingers or they have to steal the thing from him and escape. That’s how they win. But how might they lose? Well, they might get killed. That’s always possible in D&D and it almost always should be. But it should almost never be the ONLY way to lose. Otherwise, all the players have to do to win is survive. And, in this case, explore every room in the dungeon. The problem with that is that surviving and exploring every room in the dungeon are things the players are always trying to do anyway. If the players can win merely by surviving and exploring every room in the dungeon, they never have to resolve any internal conflicts or make any meaningful choices about what to do next. All of their goals always align.

That’s why you always want another way to lose.

So, how might the players lose THIS game? Well, they’d have to lose the opportunity to retrieve the thing. It’d have to be taken beyond their reach or it’d have to end up in some unknown place. If the thing vanishes from the dungeon, the players lose the adventure. Obviously.

I realize I’m opening up a whole semantical can of bulls$&% about where one adventure ends and another begins when I say this – and it IS semantical bulls$&% – but such failure states will almost always present opportunities for future adventures. Like one in which the heroes have to track down the lost thing. Or thwart whatever evil plan the dragonborn enacts with the thing. For now, though, just accept that once the thing is beyond the players’ immediate reach, they have lost THIS adventure.

How can the thing end up beyond their reach? What if the dragonborn is more concerned about using the thing to do evil than he is with fighting the heroes? Once it becomes clear that the heroes are a significant threat to his plans, he will probably try to escape with the thing rather than fight the heroes.

Given that, I can rephrase my idea about how this scene should play out. The players must search the mine for the dragonborn and recover the thing before the dragonborn escapes from the mine with the thing and disappears into the wilderness. And now I have a win condition and a loss condition, but I’m not quite done yet. Because, in a real game, the players must have a way to affect the outcome. There must be something they can do to affect their odds of winning or losing.

It’s unlikely that the dragonborn will flee from his cushy abandoned-mine-cum-kobold-warren unless it’s clear that the lair is a lost cause. He won’t try to run the moment the heroes attack. Instead, he’ll stay put and put his faith in the lair’s defenses. If too many of those defenses are breached, then he’ll try to escape.

Given that, the players can affect their chances of success. They can strike decisively, plowing through the defenses as quickly and efficiently as possible. That way, they can reach the dragonborn before makes his escape. Alternatively, they can strike quietly and try to prevent word of their attack from reaching the dragonborn. If he never realizes he’s in danger, he won’t try to escape. On the other hand, if the players take their time or dick around or raise too much of a ruckus or retreat for too long, the dragonborn is likely to run before they get anywhere near him. He might be long gone before the players even discover he’s fled.

And this is where my scene structure becomes really useful. Obviously, I need some sort of thingy to keep track of how much of a ruckus the heroes have raised. Some kind of alert tracker or something. And when I start thinking in terms of an alert tracker, I then start thinking about how the encounters in the lair might change along with the alert status. That’s just a side note, though. Because I still haven’t finished thinking through how the scene will play out.

So far, I have a goal for the scene, I have a central conflict, I have a win condition and a loss condition, and I have some ideas about how the players can impact their odds of winning or losing. But there’s something else I need to worry about. Something a lot of GMs forget to worry about. How do the players know the rules of the game?

Think about it: when the players get to the mine, they only know that a bunch of kobolds stole a thing and that the kobolds are somewhere in the mine. The players will probably assume that to win, they just have to search every room in the mine carefully and thoroughly while also not dying. But that strategy is actually really slow and inefficient. It’s the exact opposite of the right strategy. Because they don’t know the actual rules of the game they’re really playing, they can’t implement the right strategy. And they can’t also come up with unusual, outside-the-box strategies.

I mean, the strategies I mentioned above – attack quick or attack quiet – are pretty basic, obvious strategies. They’ve got a reasonably good chance of working. But clever players who know the rules can come up with other ways to improve their odds of winning. Which is precisely the sort of s$&% you want them to do. If the players speak draconic, for example, or have access to a comprehend languages spell, they could capture and interrogate a kobold and find out exactly where the dragonborn is and take a more direct route. A very clever party might take a chance on splitting their party, leaving their main force at the entrance and sending a small strike team in to raise enough of a ruckus to flush out the dragonborn. Risky? Yes. But very efficient.

And that’s why, when I’m designing a scene, I always take note of what the players know at the start of the scene and what they need to know to play the scene properly. You know, in addition to all that other stuff I take note of like win conditions and loss conditions and conflicts and strategies and s$&%. I take note of a lot. That’s why I take long showers sometimes.

In this case, the players only know that kobolds stole a thing and that they’re hiding out with it in an abandoned mine. But, to play the scene properly, they need to know that a specific creature – the kobold leader – has the thing and that he’s likely to flee if things get too hot.

When there’s a difference between what the players know and what they need to know, it’s called an information gap. And it’s your job to either close that gap or to allow the players the chance to close the gap for themselves. And that distinction is really f$&%ing important. Because the former makes the adventure easier while the latter adds a mental challenge to the adventure.

If the players end up playing the wrong game in the final scene – methodically searching the dungeon instead of going quick and quiet – they’re much less likely to win. If I want to make their lives easy, I can just tell the players the right game to play. I can feed them the information directly and unambiguously. But if I want to make their lives more challenging, I can plant clues throughout the adventure and make them figure out which game is the right game to play. If they find the clues and put them together, they can build a winning strategy. If they miss the clues or put them together wrong, they may not be able to adapt in time when they discover the situation isn’t what they thought in the final scene.

And this is why I sometimes build – or at least predesign – the last scene of the adventure first. Remember how, in the first scene, the players are going to be protecting townsfolk and buildings and s$&% from the kobold attack? And the people and things they successfully protect will reward them with useful resources? Well, if I want to hide some information to make winning the adventure easier, I can hide some of it in that first scene.

Maybe someone in the town knows something about what was stolen. Or knows that the kobolds are under the thrall of an evil dragonborn with a plan. Or knows something else that’ll help the players close the information gap in the third scene.

Now, I don’t want to blow my own horn – especially not while I’m in the shower with all of my readers – but it’s this kind of s$&% that my adventure structure lets me do. When I started talking about stakes and loss conditions and complex non-combat encounters, lots of you got really excited. And then, when I mentioned that I’d adopted a completely new and weird scene structure, I got a lot of pushback. People were literally pissed off at me for changing my own definitions. Seriously. I got A LOT of blowback over that s$&%. Same with that stuff about dynamic conflict.

I hope some of you are starting to see what I’m playing at now. The scene structure allows you to break an adventure down into a series of separate games – each designed as a complete game – whose win and loss conditions are interconnected. It also gives you a place to build overarching mechanics – like an alert tracker – and extended conflicts – like the hunt for a specific villain in an evolving dungeon. And it’s also a great information management tool because it helps you identify information gaps and rectify them. Or build challenges around them.

Maybe next time, you’ll f$&%ing trust me.

Now, this isn’t just me digressing to say, “I f$&%ing told you so.” It’s important to understand that THIS is where the real art of adventure design lies. Any idiot can follow the instructions in the DMG and drop some encounters on a map and end up with something playable and fun. But that’s not really game design. It’s just assembling building blocks. It’s like a kid with an electronics kit. He can make a cool circuit that goes BEEP or even a radio receiver if he follows the instructions, but he can’t design a bridge rectifier on his own or even know why he’d need one if he wanted to build a digital device that he could plug into a wall outlet.

It might seem like running through all this s$&% for every scene in your adventure takes a long time, but it really doesn’t. That’s why I keep pointing out that all of this got figured out in the span of one shower. An hour before I sat down to draft this article. Sure, it takes some practice and brain training to get to that point, but not as much as you’d think. You can do it. You just have to actually do it until it gets easy to do.

Digression over. Back to scene design.

So, I can plant information needed to close the information gap in earlier scenes or even at the beginning of the current scene. And I can either just feed the information directly to the players or I can feed them clues and hope they make the right deductions. And that’s what I’m going to do. That second one. Partly because it makes the adventure more fun and partly because it gives me a chance to talk about information management later.

For now, though, I just need to… oh… s$&%… it happened.

Click!

Until this point, I’ve been pretty vague about my adventure details. There’s an evil dragonborn with an evil plan and he stole a thing he needs for his evil plan. And he has some kobolds working for him. And they live in a mine. And I’m not being vague because there’s some weird advantage to being vague. I’m being vague because I just don’t know anything else about my adventure except those vague facts.

Most people think you need things like a backstory or a narrative to write an adventure. Truth is, you don’t. You can just start writing an adventure and fill in the backstory and narrative details later. It’s no different than leaving some empty rooms on the map to fill in later with encounters. Backstory and narrative details aren’t special. The only thing you absolutely need before you start designing an adventure is a goal. You can fill in all the other blanks pretty much any time you want.

But what often happens – not always, but often – while you’re designing is that your brain will suddenly, spontaneously fill in a bunch of blanks. Usually, because it comes up with some brilliant way to connect a bunch of disparate bits of your design together. Basically, sometimes, things suddenly just ‘click’ in your head.

I have this information gap, right? To play scene three properly, the players need to know there’s a chance that the adventure’s villain might up and leave at any time. That’ll get them thinking in terms of working quickly or working quietly. What they really need is a feeling of urgency. Something to make them realize they have to work fast and that they probably need to maintain the element of surprise. One way to help create that feeling is to let the players know the villain’s already got one foot out the door. That’s he’s getting ready to leave already. But why might he do that?

The dragonborn has an evil plan. And to do his evil plan, he needs the thing. Now, he’s got the thing. Why hasn’t he done his evil plan yet? Because he’s not in the right place to do it. He came to this place to find the thing, discovered the town had the thing, aligned himself with the kobolds, sent the kobolds to attack the town to cover the theft of the thing, and sent other kobolds to actually steal the thing. He must have known that the town would want to get the thing back. He must have known they might send a mob of angry townspeople or guards or adventurers to recover the thing. He was probably hoping the attack would cause enough chaos that the townsfolk wouldn’t realize the thing had been stolen for a little while. And even if they did, he’s hoping enough able-bodied people were hurt or killed in the attack that the town couldn’t strike back right away.

Either way, he was probably planning to abandon his lair pretty soon after the attack. But he’s also got this useful tribe of kobold thralls. And you don’t abandon good minions unless you have to. So, he’s probably getting the kobolds ready to move out. That’s not a fast process, especially if they’re getting ready for a long trip. The kobolds that didn’t participate in the attack are probably all gathering supplies and equipment, foraging for food and water, stockpiling ore and lumber, and doing other moving-day s$&% like that. Meanwhile, they’re waiting for their warriors to return from attacking the town. And waiting for the strike team to come back with the thing. Assuming the heroes set out for the mine within a day or two of the attack and that they don’t get lost in the hills for too long on the way, the heroes are probably coming in right behind the strike team and the few surviving warriors. And probably while the kobolds are still packing up and getting ready to move out.

When the heroes arrive at the mine, the kobolds’ numbers are greatly diminished. Many perished in the attack on the town. And the survivors are all busy getting ready to move on. The heroes should immediately recognize the kobolds are getting ready to abandon their lair. And they should recognize the urgency of their situation. If they don’t, well, it’s the players’ own faults for not figuring that s$&% out.

And those details also suggest a few things about the lair alert system. When the players arrive, the lair is on low alert. The kobolds are all doing work things. When the kobolds go on alert, they shift some of their resources from travel prep to defense. But many of the kobolds keep working. Mostly, the defenders want to buy time for the rest of the kobolds to finish packing for the journey. And the defenders are probably hoping they can repel or kill the attackers which are likely a small force of untrained people from the town. The dragonborn won’t panic initially either. He’ll prepare his own defenses, but he’ll trust the guards and keep the remaining kobolds focused on packing for travel. It’s only when the defenses get overrun that the kobolds go into full-on panic mode. And that’s when the dragonborn will cut his losses, abandon the kobolds, grab his thing and a bug-out bag, and run.

That little bit of story provides a lot of gamey details. It suggests there’s only three alert levels. And there’s no reason for any more than that. And thus, it suggests that each placed encounter might have three different versions, one for each of the alert levels.

Some GMs – by the way – would use those details to come up with some complicated-as-f$&% way of simulating how encounters move around the map or how the kobolds shift from room to room to bolster their defenses or whatever. GMs and adventure designers love doing s$%& like that. S$%& that adds absolutely f$&%ing nothing to the game except a bunch of extra bookkeeping and tracking.

The players will never know how the kobolds are moving off-camera. Hell, they have no idea how many kobolds actually live in the lair. All they know is that they’re either encountering a bunch of distracted kobolds who are loading up a cart and whose weapons are stashed nearby or they’re encountering a bunch of alert kobolds crouched behind an overturned cart ready to fight to the death. There’s no reason for the GM to keep track of anything more than that.

I can also deduce from this story that the dragonborn likely visited the town at some point. He found out where the thing was and realized he didn’t have the skills to steal it himself. And that’s probably why someone in town might be able to tell the players that there’s probably a dragonborn with an evil plan behind this whole thing.

For no other reason than it’s what popped into my head and it works, I decided the dragonborn wanted a sword that was buried with the remains of a sainted paladin in the catacombs beneath a temple in the middle of town. A place it’d be hard for him to rob – especially if he isn’t a rogue – but a place that isn’t particularly well-guarded. It’s hard to rob because there’s always someone in the temple, because the entrance to the catacombs is probably locked, and because it’s in the very center of town.

A long time ago, the paladin used the sword to kill a powerful, evil dragon in some distant land. Years later, the paladin founded the church in the town and retired there to live out the remainder of his days. The dragonborn needs the sword because it’s infused with the dragon’s born and it can be used to raise a dracolich from the dragon’s remains. And he traced the paladin’s life in some ancient texts and then visited the town to confirm the story and find out where the paladin was buried. He pretended to be a student of history or something. And while he was visiting the town, he happened to learn about the abandoned mine and the kobolds that were an occasional nuisance in the area. He made contact with them and capitalized on the fact that kobolds are predisposed to worship dragons and evil dragonborn
because they were created by Tiamat as a slave race to serve her children.

The dragonborn told the kobold strike team to kill the priest and to lock the entrance to the catacombs on their way out. He hoped that would keep anyone from figuring out that the sword was stolen from the paladin’s grave and going looking for it. They’d just assume the priest had been in the wrong place at the wrong time during the attack. But someone in the town remembered the dragonborn, remembered his interest in the paladin’s history, and remembered his interest in the stories of the kobold attacks. On hearing the priest had been killed, he put two and two and two together and immediately investigated the catacombs after the attack. So he’s able to tell the players what they need to know to get them moving from scene one to scene two and three.

And that’s the click clicked.

…And Scene!

Where we were? Oh yeah. Predesigning a scene. See, that’s what I call what’s happening now. Predesign. Figuring out how the thing you’re about to design works. It’s mostly mental work, though it usually results in a few scribbled notes. It’s tremendously important. And most people stupidly skip it.

Thanks to my predesign shower, I pretty much understand how scene three of my adventure plays out and what mechanics I’ll need to make it work. There’s a few little odd details to fill in, but I find they tend to fill themselves in if you just take a minute to summarize. Which is what I’ll do right now.

Scene three starts when the heroes arrive at the entrance to the abandoned mine that the kobolds have been using as their lair. Only a fraction of the kobolds remain. The paladin’s sword, infused with the blood of a dead dragon – has been delivered to the dragonborn leader of the kobold tribe. The kobolds are preparing to abandon their lair and follow the dragonborn to a distant land where he – the dragonborn – intends to raise a dracolich using the stolen sword and the bones of the dragon slain by it.

Initially, the kobolds and the dragonborn are unaware of the heroes. But as the heroes move through the lair and engage with the kobolds, some of the defenders will attempt to escape to warn the dragonborn of the attack. The heroes will have to deal with the runners swiftly to prevent them from raising the alarm.

Once the alarm is raised, some of the kobolds will switch to a defensive footing while others continue their work. Consequently, some of the encounters will become more difficult and the heroes are likely to encounter patrols as complications. The defenders will hold their ground against the heroes, but if the guards at a particular location are overrun, one kobold will attempt to flee and alert the dragonborn that the kobolds’ defenses have been penetrated. Once again, the heroes have to deal with such runners swiftly. Once word has spread that the kobolds’ defenses are overrun, the kobolds will arm themselves, and hold their positions to protect the dragonborn and cover his escape.

After the initial alarm is raised, the dragonborn will fall back to a defensive position with his personal guard and prepare himself for a fight. But once word reaches him that the lair’s defenses have been overrun, he will abandon his guards and flee. The GM will have to keep track of his position on the map as he attempts to escape the lair. By chance, the heroes may encounter the dragonborn as he flees, perhaps while they are already engaged in an encounter. The heroes will have to shift their focus quickly to capturing the dragonborn or else they will lose him. Once he escapes into the wilderness, if the PCs aren’t right on his heels, he’ll be lost and the sword will be lost with him.

The only special mechanic I’ll need in this scene is a three-state alarm tracker. Apart from that, I’ll need a map of the mine and it’ll have to have a lot of branching paths and loops to allow the dragonborn to escape. The tunnels between the chambers should be long so that sound won’t easily carry from one combat to another, thus requiring runners to actually escape a ways to raise the alarm. The tunnels should also be windy and branching so as to break lines of sight. It should thus be easy to determine when a runner has sufficiently ‘escaped’ from the fight to raise the alarm. I’ll need to design the encounters and figure out how each one changes based on the current alert status. And I’ll need to figure out some tension pool complications for each of the three alert levels.

And I’ll start doing all of that in my next entry in this series.

Now, get out of my shower and let me dry off.