All right, kids. I know some of you are disappointed that this appears to be another “Overview” thing where I talk about high concepts instead of getting down to the nitty gritty. But it isn’t. Not really. This is damned meaty stuff. There are some major, important ideas that don’t get shown off in any one area of the dungeon but rather drove the design and I need to lay them out. If you give any s$&%s about building satisfying, challenging adventures (and not just dungeons, remember, the dungeon is just a test-bed), these are the most important ideas I probably have to share. So, pay careful attention.
The party begins their exploration of the Underhalls of Maridia by having to secure an entrance to the Underhalls. They get a lead on a passage through a gate into the city’s ancient storm sewers. Somewhere within Via Draygara, there is an entrance to the Underhalls proper. The party must find it.
To make life easier, I divided the Underhalls of Maridia into several sections. Via Draygara is the first (named for Draygon, the boss of Super Metroid’s Maridia area). It serves as a self-contained, miniature dungeon. I’m actually hoping to build the entire dungeon as a collection of self-contained, smaller dungeons.
Here, the party has a simple goal: enter this section of the storm sewers and locate a passage into the Underhalls of Maridia proper. The party also has a secret goal they don’t know about yet: open a shortcut directly between the entrance to Via Draygara and the entrance to the Underhalls proper.
Goals and Progress
Exploration for its own sake can be a fun goal, but it has two problems as a driver of gameplay. It is interminable and it doesn’t ratchet or increment. Those are fancy ways of saying, you’re never really done exploring and there aren’t any major victories that mark signs of progress. As long as there is an unexplored room, there is exploring to be done. Exploring a room is a minor victory, but there is no point at which the players can “well, we’re halfway to being done exploring! Good job, guys!”
So, I decided that the party would never be able to wander around without a goal. They would always be working toward a goal, whether they knew it or not. It might seem odd that they can have a goal and not know it, but exploration is funny that way. See, exploration, of itself, can drive the party forward, but it can’t really provide victories (minor or major). Occasionally, there need to be things in the game that feel like victories and give the party the chance to cheer, but they don’t need to know they were coming. Opening a shortcut, getting into a new major area they had to bypass, dealing with a major villain, those things all provide the sense of victory the players need even if they don’t know they are coming.
There are a number of ways to present goals and I decided I would experiment with them. Some would be stated and the party would be able to work toward them actively. Others would merely be suggested by the environment. See, the nice thing about goals is that, when a party is flapping around without a goal, they tend to latch on to just about any goal they are presented with. A giant, ornate door the party can’t open turns into a goal, even if the party never vocalizes or acknowledges it.
The Three Tenets of Fairness: Consistency, Telegraphing, Agency
Homework time. Go watch Extra Credits S6E15: When Difficult is Fun. I’ll wait.
Consistency and Telegraphing are… I’m serious. Go watch it. I’m going to start talking about topics they explain so I don’t have to. Go. NOW!
Okay, fine. Maybe you can’t watch it right now, but you damned well better (and watch more of their videos too. They are brilliant). In the video, the Extra Credits folks discuss three concepts central to creating fair challenges. Applying the rules of the game and the world consistently (Consistency), allowing the players to have some idea of the likely future outcomes of their actions (Telegraphing), and giving the players the freedom to use their tools in a variety of creative ways (Agency). But that summary isn’t good enough. Go, watch that video!
Consistency and Telegraphing are not just vital to creating good challenges, they are vital to role-playing itself. Your players must know that the rules of the universe (game mechanics, flavorful rules, and the laws of cause and effect) will always work in consistent ways in order to make rational, logical decisions. Otherwise, they will not be able to make the decisions their characters would or should make.
But Consistency and Telegraphing also empower the players to solve problems. Imagine, for example, that the players use a magical fire to set fire to a villain’s stuff. The villain douses the fire with a bucket of sand. A few sessions later, one of the PCs is set ablaze with magical fire and the player decides to use a bucket of water to douse the fire. That player thought of that solution precisely because they learned that magical fire can be doused just like normal fire can (or assumed as much).
Agency is about giving the players the freedom to use their tools in logical, rational ways without having to overcome arbitrary restrictions. This is not the same thing as never restricting anything. But it does mean that if there is a restriction, a reason why they can’t do something, there needs to be a good reason that they can discern or discover.
In the original Tomb of Horrors, there was an invisible gem that could not be seen with a see invisible spell and a secret door that elves were not allowed to use their detect secret doors ability to detect. These are awful challenges. They simply take away a PCs abilities for no reason that is ever explained and there is no way to know it is happening. No Telegraphing, No Consistency, No Agency. In fact, most of Tomb of Horrors is an absolute bulls$&% challenge.
Agency, as noted, does not mean complete freedom. Players should not be able to spout bulls$&% and make it true. If the player is set on fire and wants to douse it, figuring out how to do so with the tools on hand is a challenge. But if they can simply conjure a bucket from the story ether (even if it requires spending some sort of ‘drama point’), much of the challenge is drained from the scene. Remember the end of Wizard of Oz? The Wicked Witch (to whom water was lethal, skin-melting poison) just happened to have a bucket of water lying around her castle that Dorothy used to destroy her? Yeah. Bulls$&%. If the WW was worried about fires, she could have had a bucket of sand. Just as good as water for putting out fires, but without the whole “it will melt all of your skin off and make you dead if you blunder into it while you are going to the bathroom one dark night.”
I won’t say Player Authorship removes all challenge. But unchecked, it turns all challenges into one specific type of challenge: Scribblenauts. Can you invent the one thing, the one fact, the one object out of the ether that solves your problem?
Agency also means you can’t let the rules of the game handle everything. Agency is mainly created by the fact that there is a DM with a brain who can adjudicate actions on the fly. If the player does something logical, rational, and clever for which there is no rule (or which the rules seem to contradict arbitrarily), you are creating a Bad Challenge if you refuse to break the rules. For example, the “Just Say He’s On Fire, Dammit!!!” incident.
I was a player (*gasp*) in a 4E game and a PC (not me) was taking ongoing 5 fire damage (Save Ends). On the PC’s turn, the player said he wanted to throw himself in the water (there was water nearby) to douse the flames. Examining the rules, the DM ruled that this would not help. A saving throw was required to end the effect. So, the question came up as whether you could use the Heal skill to grant yourself a save. And the answer, according to the rules, was no. You can only grant “an ally” a save using the Heal skill and you are not your own ally. Through the whole thing, the player kept using the phrase “since I’m on fire” and the DM kept rephrasing it with “you’re taking ongoing fire damage.” The DM did nothing wrong by a strict reading of the rules, but from a perspective of presenting a Satisfying Challenge, the DM had made a mess of things. There are a number of ways the DM could have handled that better, but none of them involved the phrase “because the rules say so.”
(By the way, I call it the “Just Say He’s On Fire, Dammit!!!” incident because that is what I screamed before storming out and never returning to that DM’s table.)
Agency is freedom, but not total freedom. It is freedom from arbitrary restrictions. The freedom to apply logic to come up with approaches to problems not covered in the rules or planned for by the DM. Neither giving the players total freedom to bulls$&% nor insisting on the letter of the rules every time will serve Agency and Fairness.
Teaching and Testing
Challenge is about empowering your players to deal with the obstacles you place before them. And part of that empowerment is allowing your players the chance to learn how things work and then formulate strategies. There are many video games that illustrate this principle, including Metroid, Megaman, and Portal. Personally, I think the Megaman series pulls it off the most artfully, especially given the difficulties inherent in doing it in its particular genre.
Megaman is a jump-and-shoot, 2D side-scrolling platformer about a robot who kills other robots to save the world. It did two things that are very noteworthy. First, it had rudimentary sandbox style, non-linear gameplay that was really effective back in 1987 by employing some really clever level and monster design. Second, it was very effective at building challenges and managing difficulty by training players to deal with its obstacles.
So, you enter a room and there is a robot monster out of your reach. The robot monster can’t hit you from where it is, but it keeps lobbing attacks your way. You make your way up to the monster to destroy it, but the whole time, you get to watch its pattern. By the time you engage it, you have kind of figured it out. You can’t engage it until you’ve seen what it can do.
In the next room, there are two of those robot monsters. And once you move into the room, they can hit you. Sure, you already know what they can do and how to deal with one. But can you handle two of them? Avoid those attacks and take them out?
In the next room, you run into this passive crushing trap. When you move into its line-of-sight, it tries to crush you. You figure out its timing, you control the pace, and you get passed it. In the next room, there are two of those traps separated by bottomless pits. Trickier. Then, another room with three robot monsters. Then, a room with two robot monsters and the crushing traps separating them. And so on.
The game trains you to deal with its individual elements and makes sure you have to figure them out. And then it combines them in different ways to build more challenges. That means it can make challenges more complicated because it has given you what you need to figure them out.
Admittedly, Megaman doesn’t do this perfectly every time and the Megaman games are aggressively difficult at times, but if you’d like to try them for the first time, start with 4 or 5. They were easier than the others and quite polished and refined. 2 and 3 were more difficult, but also well put-together.
Empowerment through training is a concept I wanted to steal unashamedly. The adventure trains you to handle it. I want to let players figure out how to deal with the elements of the game and then combine them in different ways to build bigger, badder challenges. In that way, I could build an entire 10-encounter adventure with only three types of creatures and ramp up the challenge in later rooms.
And I didn’t want it to just be about creatures. Creatures, terrain, and traps, sure. But also concepts and ideas.I want to train my players to out think the dungeon by teaching them skills and then testing and rewarding those skills. Ultimately, it comes down to this: every encounter must either teach something (introduce a monster or demonstrate a design concept) or test something already taught.
The best part is, the players are subtly (perhaps even subconsciously) aware of the rising difficulty, so it feels like the stakes are ramping up and they are rising to the challenge. That feels like improvement and advancement and makes the players feel accomplished.
To pull this off, though, I had to be able to drop a number of ingrained DMing habits. First, I had to be willing to throw in non-obstacles: under powered fights, traps that aren’t threatening, puzzles that don’t require much effort to solve, and so on. Easy things the party could learn from in order to set up future difficulties. I literally would have to become okay with letting the players face two kobolds in a poorly designed room even though the players would trounce them.
Second, I had to stop relying on surprise. You know those pathetic Flash horror games you find on Newgrounds? The ones that are basically boring, crappy point-and-click adventure games but flash scary images at you every so often? Pop scares? Well, you can’t use those. And you might say “whoa, I never do that.” But you do. We all do. I do. We use them in the form of “things the party couldn’t have seen coming or planned for and now must decide how to deal with for the first time with no information.”
If a PC drops dead as a result of something they couldn’t see coming or couldn’t take steps to avoid or mitigate, that is Bad Challenge. You murdered the PC. You monster. I will have a lot to show off in regards to this when I get to the Green Slime room. But for now, I want to call attention to three specific ideas: Metagaming, Save or Die Effects, and Booby Traps.
With regards to Metagaming, I decided I would never make a challenge that could be broken by Metagaming. I would never require the players’ (or their characters’) ignorance to drive the challenge. If I want to use a troll, I have to figure out why that troll is challenging even if the party knows about its resistances. And then, I need to give them an opportunity to know about those resistances.
Save or Die Effects would never come out of nowhere. The challenge in dealing with them is figuring out how to circumvent them, avoid them, or render them moot. And that means the PCs need a chance to know what is coming. If the PCs do a good job dealing with a medusa, they should never have to save against petrification. That means, first, Telegraphing the effect and second, allowing the players to cleverly rob monsters of their iconic abilities.
Booby Traps are a little more forgivable, but not much. I differentiate Booby Traps from other traps because they come out of nowhere and the only way the party has to avoid them is to succeed on a random die roll or think to search the correct one of ten thousand doors in the dungeon. They should be used sparingly, if at all, and they should reward foresight or cleverness by being avoidable. I will talk more about Booby Traps in the Overflow Sluice Room.
So, do not call out ‘metagaming,’ do not rely on surprise, and allow the PCs to cleverly destroy a monster’s best abilities. I know some folks reading are balking at these ideas. Hell, I’m half balking at them even though I know what I’m saying is true. We think we need those things. But we don’t. And we hurt our games with them. Don’t create challenges that can be broken by metagaming. And if the players manage to trash a monster’s best abilities, they won. Let them do it.
Incidentally, if you are interested in the deconstruction of Megaman design, Simply Simon (a YouTube poster) has done some great Let’s Play Videos of various Megaman games with a really critical eye for level design. Check out the Let’s Plays of Megaman 3, 4, 5, II, III, IV, and V (the numbered ones are NES titles, the Roman numerals denote the Gameboy titles) on his YouTube Channel. This is a great way to see what’s going on if you don’t want to play the game yourself.
And if you have the patience for something shoutier and cursier, but ultimately also very brilliantly deconstructive, check out Egoraptor’s Sequelitis: Mega Man for a comparison of the design of Megaman and Megaman X. It takes a little while to get to the good parts, but it is worth it to see some of the subtle, wordless tutorials built into the games’ levels.
Overdesigned for Satisfaction
Since I brought them up already, I will mention something else extremely notable in the Metroid, Megaman, and Portal games: the games are designed very, very carefully. Everything has a purpose and nothing that doesn’t serve a purpose is left floating around. This is most evident in the first Portal with its extremely sparse design.
I wanted to be just as careful with my design. I want things to be flavorful because I am a role-player at heart and I am not running a video game. I’m just trying to crib from video games, and build on what they’ve done, not reconstruct them on graph paper. But I want all the game elements to be placed carefully and serve a purpose. That means, I worked harder on designing some individual encounter spaces in this dungeon than I have on some entire dungeon delves I’ve run in 3E and 4E.
At the same time, I never want to rob the players of Agency. I want them to feel as if they are free to explore and interact with the environment, to play with it. I want them to feel as if it responds to them. I don’t want to not run an RPG. It is a tough balance to strike, but it is easier if you remember this: players do not need all the Agency, they need exactly as much Agency as they need to feel as if they have all the Agency. If the players feel free to make any decision they want, but they make only the decisions you want them to, everybody wins.
So, in places, I have designed areas specifically to nudge players toward certain decisions. When running the game, I still won’t force any decisions. Ultimately, if the players break the design, that means I failed in implementing my ideas and I need to know about that. Getting in their way in arbitrary ways would not serve my purpose at all.
But that means I’m trying to eschew randomness as much as possible in favor of tighter design. I won’t put 1d4 kobolds in a room, I will put three kobolds in a room and I will place them in optimal positions. Everything will be carefully designed toward building satisfying and effective challenges.
This is again something that a lot of DMs. And while I understand the reasons, they are not doing themselves or their players any favors. It does seem artificial when a group of creatures are in a room designed perfectly for their abilities and just standing around waiting for the players. But it only seems artificial to us, the DMs. As long as the players think it is natural, it doesn’t matter whether it really is or not. The truth is, players don’t want “realism” or “verisimilitude” or whatever. They want a fun, challenging, interactive storytelling experience that doesn’t seem artificial or unrealistic or whatever. They want the illusion.
We’re not participating in the world, as DMs. We’re the people under the stage who maintain the animatronic figures. We shouldn’t complain that, from the waist down, Singing Abe Lincoln looks like a bunch of servos and pistons. All that matters is that the audience only sees him from the waist up. So, I had drop that habit and so do you. Stop worrying about what is real and natural. Worry about what looks real and natural from the waist up.
Wanderers and Respawners
Okay, I admit it: I am not eschewing all randomness. See, there are two things I do want to play with: wandering monsters and respawns. Both serve important purposes.
Wandering monsters are surprise encounters that might pop up every few minutes when the players wander around or appear when the players do something to attract attention. They have fallen out of favor lately, but they do serve some useful purposes. First, they keep the players moving and on task, which is an important pacing tool. Second, they create the sense of a living, breathing world instead of a static pile of encounters waiting for the PCs to wander by. Remember what I said about Singing Abe Lincoln? Well, wandering monsters are a neat way to put a stovepipe hat and topcoat on your dungeon and make it look like a living person instead of a collection of servos and pistons and a voice synthesizer.
But they are also a pain in the ass. See, random encounters are, by their nature, random (surprise!). And I said I want to stay away from that for now. I want a careful design, not chaos. And random encounters are chaos. So, do I take them out?
*No. I can have them, but I need to minimize their impact. I want them drive the behaviors I want them to drive without screwing things up too badly when they do but I don’t want them to break any of the other principles. So, here is the plan: wandering monsters can only consist of monsters the PCs have already encountered and dealt with. They must already be familiar. I will not use a random encounter to introduce a monster. Ever. Additionally, wandering monster encounters must always be “easy” encounters. They should, essentially be a nuisance and a minor drain on resources. Wandering monsters will never establish a goal, serve as a goal, or be a part of the story. They are random tiny nuisances, just enough to annoy the players and make it seem like the dungeon lives.
*I have decided there are no wandering monsters in Via Draygara. The beasties in there mostly keep to their lairs and camps, but the players don’t have to know that. I’m playing with one idea at a time. In future areas, there will be.
Respawns are trickier. I decided early on that I wanted respawning to be a mechanic. Respawning means that if the PCs leave the dungeon (or an area therein) for any length of time and then return, they will find areas they had previously cleared of danger occupied by new dangers. Usually monsters. Like wanderers, respawns serve purposes. They discourage the party from leaving the dungeon to rest unless they absolutely need to and drive them to keep making progress. They also provide the sense of a living, breathing dungeon that is changing in response the party’s actions. Ideally, I’d carefully design the respawn encounters to fill the areas the party clears, but, truth be told, that means designing the same dungeon multiple times. There is a certain point where the benefits of doing something no longer outweigh the cost of doing it. The energy required isn’t worth the payoff.
*Thus, I am striving to minimize the impact of respawns. They are there to drive behavior and create the illusion of life, but nothing more. If the party leaves the dungeon to rest or leaves an area for any length of time, I will repopulate some rooms and halls with beasts or other challenges. Like with wanderers, the encounters will be nuisances, mostly vermin, and they will be creatures the party has already dealt with. I want them to be an inconvenience, but little more than that.
*Respawning is especially important when I talk about Gating. See, the secret is that Respawning is useless unless the players can do something to avoid, prevent, or mitigate it. Otherwise, there is no real progress. So, we will have to build in the idea of opening up shortcuts or clearing areas completely.
The Things I Can’t Talk About Yet
I would really like to talk about how I designed the overall layout and map of Via Draygara, but I can’t for a couple of reasons. First of all, I would have to spoiler the whole damned thing. Second, it was built around the concepts I’m going to reveal as I deconstruct the place. So, I’d basically have to introduce all the concepts, then describe the design, then break it down, and talk about how each room accomplishes what it is designed to do.
So, I’m going to reveal the areas as the players explore them and talk about the concepts in isolation. And then we will go back and talk about the top-level design. Think of it like exploring the dungeon with a critical eye and revealing the map one room at a time before you can put everything together.
As for my players, they will soon discover there is really only one sure-fire way to survive and thrive in the Underhalls of Maridia: by not reading these spoilers. Seriously, if you are looking at this, I will kill your motherf$&%ing character. But we’ll let them discover that on their own.