I want to tell you about a structural tool I find extremely useful, but it is slightly more advanced than a lot of other DMing advice out there. I think a lot of my stuff is. This blog isn't a 100 or 200 level course, it's a 300 or 400 level course, so I usually assume you've already done a lot of the basic readings (you know, Dungeoncraft 101, Metagaming 150, Player Types 210, Random Encounters 215, etc). Most of the stuff that Matt Colville would have made a video about or the Angry GM would write about, I assume you're familiar with. So I want to take a very basic concept and start deconstructing it: Railroads vs. Sandboxes.
Revisiting the Fundamentals
We all learn about Railroading and Sandboxes/Open-World games early on as a pretty simple dichotomy to understand gameplay types and whatnot. Even when called by other names, it's still a fundamental every DM will hear about early on, usually described even in the DMG itself. It's a useful tool for organizing game styles, but it has its limitations. To explore those limitations, let's do some review.
For the purists though, let's do a useful exercise. Let's stretch the definitions to their extremes. This is usually an important analytical litmus test when discussing theory to see how well it holds up.
If railroads and sandboxes are the two opposing sides of a spectrum, what lies at each end of that spectrum? What would a 100% railroad look like? Something where the Player Characters don't have choices. Where they have no agency. Normally we think of choices being things like "where can I go? I go south." But a 100% railroad would have to be something where they can't affect the outcome at all. Even deciding what goofy commentary your character says is still a choice. As long as you can still have control over the character's dialogue or how they look or what their name is, then you are technically making choices that affect the game. So a game where the players can do literally none of those things is just... a book.
Like, that's just the DM writing a book and maybe reading it aloud to a group of bored people.
Conversely, what would a 100% sandbox look like? One where the DM doesn't do a single thing to restrict what the PCs can and cannot do. Where there is nothing forced on the party that prevents them from making a choice. There's nothing determined about the adventure's plot ahead of time so that the power to define it is completely in the hands of the players. Well... much like before, that wouldn't be D&D anymore. That would just be, like, a collaborative oral story-writing circle I guess. D&D without the DM. Because literally everything the DM says or does or defines about the game is technically at the expense of PC agency. When the DM says that there is a mountain on the north of the village, they are limiting possibilities. There are possibilities where the mountain wouldn't be there. In fact, when the DM said that there was a village at all, that was limiting against the possibilities of there being no village. The existence of every NPC, name, location, event, action taken, etc. comes at the exclusion of a version of the story without those things or where those things are different, and thus, a version of the story where the PCs may have made their own choices differently.
What to Learn From This
What my point is, what I want to illustrate with this exercise, is that technically all games are somewhere in the middle. No game is "truly" a railroad and no game is "truly" a sandbox. No matter how much you try to extol the virtues of the sandbox and its superiority over a linear game, I can, by definition of how D&D works, point to ways in which your campaign is limiting player choices somehow. The real debate then is about relativism. While neither extreme is truly achievable, nor are they even likely desirable, you instead may argue about which direction on the spectrum is preferable, and more importantly, how far in that direction you ought to go. Two campaigns can both be sandboxes, but one of them is slightly more sandbox-y than the other.
So if you are necessarily going to have railroading no matter what, just intrinsically because of the medium, then you need to understand the circumstances under which railroading is a positive thing. There is an incredible amount of effort that's been spent on people trying to understand and educate each other about "how to avoid railroading" or "how to railroad less." But because you're going to be doing it no matter what, you should be able to identify when and where it's best to do it. In a sense, knowing when to railroad and when not to railroad is another way of saying you know what to prepare before a session and what to improvise during a session. Or maybe, what to include when you write an adventure and what not to include.
Usually, the worst kinds of railroads are invisible walls. Things that are only justifiable with a meta reason, a Doylist explanation (warning: TV Tropes link). "Can we leave the city?" "No." "Why not?" "Because I said so." That's pretty much the worst. So instead you come up with in-story justifications for why these limitations exist. "Because the city has been quarantined." These limitations are a little softer, though. Players are not as likely to hear them as true limitations, but instead as challenges. "Ooo well maybe we can dig under the city walls and escape without being noticed..." If the DM really wants these types of limitations to not be overcome, then they have to be more creative than the PCs and account for many possibilities, and they need to do so preemptively. "There's a magical barrier around the city in all directions that will cause you to just pop out through the opposite side when you try to leave, like the edges of the screen in Pac-Man or Mario Bros."
But when and why do you need walls at all? Can't the DM just improvise? Well, improv is notoriously difficult. A good DM should be able to improvise, but it's no surprise that they'd like to have ways to avoid it if they can. The very concept of a dungeon in the modern fantasy fiction (especially gaming) context comes from precisely this need. In such a free-form, limitless medium of narrative building, where there are 4-6 people able to openly declare any action imaginable for their character, things can get out of control quickly. The easiest way to limit this is to put a literal wall in their way. And that's where dungeons come from. A pre-defined space for it all to take place within, where the DM already knows almost all the possibilities ahead of time because they drew all the rooms and hallways themself. Even with that much limitation on the story though, we still find that the creativity of PCs prevails and lots of wild, unexpected, and fun things come out of dungeon adventuring nonetheless. You can impose a lot of restriction on player agency without it actually "ruining" the game. Dungeoncrawls are easy to run and fun to play in. It's an enduring model because it reconciles the "Railroad vs Sandbox" dilemma so well.
There's another model to reconcile this that I call a "playground" adventure. "Alright, for this adventure you guys are functionally stuck in this city. But within the city you have no direction as to how you must achieve your goal. You do have a few goals that I'm telling you, though. You can approach it any way you like and there are dozens of NPCs and locations and items and countless ways to interact with them." A sandbox confined within a railroad.
Which brings us to the real value of occasionally railroading intentionally: the tyranny of the blank canvas. It is a well-known phenomenon that creativity is born of limitations. A blank canvas is intimidating. It's hard to fill when you have nothing to go on. But when part of the job has been done for you, even just giving you a one-word prompt to work off of, then it becomes much easier. You start having ideas that are reflected off what you were handed to start with. And in D&D, the story will start going interesting places when the PCs have been given something to play with first.
The Logic of My Model
There is a sandbox ideal that exists in the imaginations of countless players and DMs but which I've never actually seen before. I'm sure it's been achieved, but I'm certain that it comes with no small amount of difficulty or at least effort put towards accommodating it. The dream is that your campaign will be driven entirely by the whims and wills of the players, that they'll declare for themselves what their goals are and what they want to accomplish and what quests they want to do. That they'll have a cool home base and they'll start exploring out in every direction and discover all the awesome dungeons and landmarks you've been stockpiling for years, but they'll discover them because they went out and found it themselves.
Anyway, I have made the mistake quite a few times now of thinking I can set this kind of campaign up from session 1 and it simply doesn't work. I am telling you, if you say to your players, "you can do whatever you want! The world is your oyster!" then they'll just stare at you. I've had this reaction from a couple dozen different players at this point, with all the groups I've had in even my limited time playing this game. They don't know what to do. They don't have ideas for what they want. Or if they do, they are really vague and long-term. They might not be something with a clear path to achieve, like "I want to find my missing father." Sure, they showed up to the session ready to play, but they assumed you would have prepared the adventure for them. Believe me, once they're in an adventure then they'll begin breaking down walls. But without the structure to break through, they have no idea how to trail blaze. People are sometimes uncomfortable with the thought that you would ever dare begin a campaign with, "alright, you've all been hired to do XYZ" but I insist from much experience that there is literally nothing wrong with that. The DM will inevitably be defining things about the game and its world that the Players don't get a say in. All games have some railroading in them, and this is an acceptable and beneficial place to do it.
Rule 1: games benefit from a little bit of railroading early on. Something to get the ball rolling.
Some methods are better than others. You can give them a job to accomplish if you want. That works. But sometimes they'll begin to question the motivations you've accorded to them. Even if you can get them to initially agree to it, there may later come a point down the road, after they've had a lot of trouble with the job, where they ask themselves, "why would my character ever have agreed to this in the first place?" I firmly believe that any good player would be able to answer that question without too much trouble, because players are supposed to be creative, too. Not just the DM. As part of the meta social-contract of helping the campaign even function, players have to be ready to work with the DM to make sure their characters don't have too much friction working in the DM's campaign. But understandably, sometimes it's hard. So the DM would be well advised to make this early railroading be something really easy to accept. Here's an old classic: everyone wakes up in a prison (or a literal dungeon. Like, the historic kind). Motivation: you want to escape. No one can argue with that. All players would be on board with that and most rational (or even irrational!) people with this common goal and a lot of daunting obstacles would gladly work together to achieve it. Or maybe they all wash up on a desert island together, the only survivors of a shipwreck. Before, they were all passengers for their personal reasons, but now they have no choice but to be together and no outside goal can really trump the more immediate need here: get off this island and back to civilization alive. Maybe the group is a bunch of people who got lost in the Underdark and found each other. Maybe they were all kidnapped and made into slaves or gladiators. Basically, any situation where the goal is "survive" and the best option to do so is clearly "teamwork" then you can't really argue with that motivation.
Rule 2: try to make your railroading really easy to accept or difficult to argue with.
So let's say you do a little bit of railroading at the beginning. And that sets you up to have a really solid sandbox shortly later. Then, you enjoy the majority of the campaign as a robust, player-driven sandbox that was well-primed with some DM-provided foundations. Not just DM-provided locations and bad guys, but even DM-provided motivations and dramatic scenes. The rest of it follows naturally enough and the PCs start to take over. A lot of DMs would be satisfied right here and they would call this a good model to work with. The "Unfolding Story Structure" model or the "Broadening Story Structure" model or the "Paper Fan Story Structure" model or something. Shapes are intuitive. But I think railroading comes in handy at another point in the progression of a story: the ending.
No matter what anyone says, we all want our games to ultimately have satisfying conclusions. Or at least fun and memorable ones. Basically, anything is better than fizzling out with a whimper. Some OSR folks deride the very intentions of "Story Game" RPGs but the truth is that some of the best experiences we can get out of gaming come not just from the moments where the PCs made some wild and crazy shit happen, but also from when the DM found a way to bring it all back together under a brilliant vision. Chekhov's gun gets fired, there comes a payoff for every setup, loose ends tie up, plot threads resolve in a logical way, goals are achieved, and characters who want to retire get a clear exit strategy laid out for them.
It's a beautiful thing when this happens on its own. When it wasn't scripted, but instead happened in the natural course of improvised storytelling and emerged as a logical consequence of everything that had been building to that point. But this is really, really rare. When the DM improvises the next thing to happen, the next NPC to walk into the room, the thing found behind the next door, there could be a million options that cross their mind. They may not always choose the "best" thing that could happen, but as long as they know how to reliably choose consistently good things, then they are good enough at improv to feed the sandbox with good possibilities, to fill the player-driven clusterfuck train with high quality fuel. But if the goal is to come up with things that will lead to a desirable ending, then the bar is slightly higher. Being able to see the "best" thing out of all the options that occur to you is harder than being able to just see the good things, so it stands to reason that maybe the DM could take some extra time to sort through the options and piece out what will make a satisfying ending.
"I know the PCs have confirmed at least one NPC is a doppelganger. That was two sessions ago, lots of mysterious things have happened, it's about time they find out who it is. I don't know who the doppelganger is, though. Well let's see...
You know what? If the duke turns out to be the doppelganger, then... Well that would make perfect sense! And it would mean that..." These decisions can take some time to make, and I think that approaching a final session the DM should start having a few more railroads in their back pocket.
Rule 3: take the reigns back when the journey is almost over. You'll do a better parking job than the players, or at least it'll be better if you give them guidance.
Navette (or Marquise) Story Structure in action
(Warning: long Game Tale follows, but with annotations)
If you don't recognize the names "Navette" or "Marquise" as shapes, they are mostly used in the context of jewelry. It took me some time to find the right name for it. I almost called this "Inverse Hourglass Story Structure" which would have been terrible.
Here's an example of how I've used this. I wanted to run S2: White Plume Mountain using Knave with a group of about three people over the summer. We were looking at maybe 4-6 sessions to do this whole thing. I had one player who had never played an RPG before, too. I knew some guidance would be needed.
So to start, I gave them the context. I straight-up told them, "you're a bunch of refugees from the formerly Free City of Greyhawk. It got conquered by orcs so you jumped ship across the Near Dyv and are now wandering up north with not a penny to your name. The talk of the Flaeness has been of these three valuable magic weapons stolen from three super rich dudes. There's an open quest to retrieve them, you decided to give it a look, and now you're in a town just outside of their location: White Plume Mountain. If you get them, then that's your ticket anywhere. Like, it'll solve all your problems." If you know the adventure, you know the rest of the details. But the setup, while kind of specific, was something I just told them to run with and they obliged.
But once they finally got into the dungeon and started having choices before them, I did a whole lot of "Yes, and..." in response. The dungeon has three major wings, with an intersection early on to decide which of the three you'll venture down. I don't try to push or prod them to do one wing before the another, even if I could plan a story where one wing should come first because it's easier and another should come last because it's more epic or something. None of that. They decide where they go. Now we're in the sandbox.
And as they go, they take a hefty amount of damage and decide to leave the dungeon, rest up, and come back when they're in better health. Alright, I won't try to come up with a reason why they can't leave. The story can still work as long as they don't give up on the dungeon entirely. BUT I am going to make a logical consequence for this. The next day, some more adventurers have shown up outside the dungeon with ambitions to go get the weapons themselves. After all, this quest hook is pretty well known at this point. Every time the PCs return from the dungeon, the number of other adventurers grows. By their second time coming back out, there is a huge encampment with tents set up and an ad-hoc market and some armorers and whatnot. A small community grows around the base of White Plume Mountain. But to keep the focus on the PCs, I don't let the other adventurers interfere too much. They're all too scared to go in, but too proud to admit it, so they say, "once your group has finished exploring then we'll go in, fresh and ready to face the challenges that stopped you. We'll push past where you could not!" In reality, these cutthroats and dirt bags are just going to wait for the PCs to emerge with one of the magic weapons and then slit their throats for it, going to collect the reward for themselves without lifting a finger to retrieve it. These are all ingredients driven by the PCs which feed further into the potential for the PCs to make choices and shape the story and the conflict. But they are also going to be ingredients for the ending.
Now, as the PCs go through the dungeon, they make choices about how to interact with things. They first retrieve the war hammer, Whelm. They can't make use of it because none of them are dwarves. They don't feel satisfied that they've really completed the quest, even though they could still go cash in this war hammer and be done. From day 1 the thought had occurred to them that, if these weapons are powerful enough, they could just retrieve them for themselves and never return them to their owners. The weapons could be their own reward. The setup I offered the players at the start was to get them through the door, but that doesn't mean we need to stick to the original plan once other, better possibilities are created. So they decide they're going to keep pursuing the other weapons, i.e. continue adventuring WITHIN the confines of this railroad but still taking advantage of its sandbox elements to keep brewing more good story.
But they've also figured out that the other adventurers have pretty nasty intentions, so the guy who claimed the war hammer for himself (he earned it, trust me) also decided that he would keep it a secret. If he brandishes it about, then he puts a target on his back. But again, I improvised a logical consequence of this: after 3 or 4 delves into the mountain with nothing to show for it, the other adventurers start to get suspicious. Rumors circulate that the party has ulterior motives, that maybe they work for the villain who runs the dungeon and stole the magic weapons to begin with. Some of them really believe it, while others are merely opportunists who need an excuse to make the PCs look bad so they can later kill them.
The PCs continue making choices that I always say "yes" to. A werewolf sorceress and her boyfriend work for the bad guy, but the PCs convince them to abandon their evil ways and leave this place. In fear of persecution, they need some major persuasion to return to the outside world, but the PCs convince them to hide their identities and make camp with the other adventurers. Now the PCs have to explain to the outsiders why they've somehow emerged from the dungeon with more people than they went in with. And this time, one of them (a fighter type) has the trident, Wave, and he doesn't hide it. He makes a big show of it, because he can use it and it's extremely powerful so he can reliably defend himself from backstabbers (for the time being). Plus, the weapon is incredibly loud and religious and so the PC decides to run with it and try converting people. He cultivates a social following that, by no fault of his own, contrasts against the party wizard who is now looking more and more suspicious by his distinctive lack of a magic weapon, or any indicator of progress. The cutthroats are trying a "divide and conquer" strategy on the PCs' party. Why the wizard and not the other guy who actually does secretly have a weapon? Well, I thought it makes sense because the wizard has been acting cocky, taking the lead a lot, and can't actually prove he's made any progress if the pressure gets to him, unlike the other guy.
So the next morning, an "impatient" NPC adventurer named Jeff has apparently stolen the wizard's spellbooks and ventured into the dungeon himself to retrieve the last magic weapon. This pisses the wizard off beyond belief and their conviction to see this quest through is stronger than ever. We have a villain now, one who the PCs care about and they want to fucking murder him. They go into the dungeon and try to track down the thief. They get attacked by an Ettin named HunkyDory but they completely dominate him and beat him into submission. Now they have a pet Ettin with one surviving head. Dory was killed in the battle. They push on and they get ambushed by some ghouls. This time, they lose. The party member with Whelm, named "Gregg with 3 Gs" (or just Gregg for short), gets knocked unconscious and is bleeding out. The party retreats and makes Hunky carry Gregg back to the entrance as fast as possible, taking advantage of his greater speed. They are trying to get to the Sphinx at the dungeon's entrance. They know Sphinxes can cast great magic, and they've previously established a quid pro quo that, for every riddle they can solve, they can get one free answer about something from her. So they have to rapidly answer a riddle so they can ask her to save Gregg's life. Unfortunately, it takes long enough that Gregg is dead. But she says she knows of some beings with magic greater than hers who can save Gregg still. Sphinxes also get time-travel based powers, so she sends Gregg back in time to be healed among an old lost Elven clan in the land of Geoff (opened up the Greyhawk Gazetteer, flipped to a random page, and chose that country. A fucking country named "Geoff"??) that had restorative secrets and who can bring him back from his ghoul-induced wounds. Yes, my improvisation got a bit weird, although I think it makes a fair amount of sense. How can a sphinx save someone killed by ghouls? Send them back in time to a now-dead elf clan, duh. Elves are immune to ghoul attacks, so they must know something the rest of us don't.
This ended a session, and it was evident that they'd complete the quest in probably only one more session after this. Meaning that we're approaching the ending and it needs to be good. I now have the task of looking at everything that's happened and seeing if it can come together in a satisfying way. There's at least one possibility that is too good to pass up: Jeff is Gregg with 3 Gs. I knew it instantly. When Gregg went back in time, he was brought back by elves but was bitter and angry. His wizard friend failed to save him. And Gregg still wanted that last goddamn magic weapon. He felt gypped for getting a war hammer he couldn't even use . So he'd train for years and years so he could return to White Plume Mountain on that fateful day and finish what he started. He couldn't remember his name, all he could remember were the words "3 Gs." But he woke up in the land of Geoff and thought that must be a sign, one of the Gs, and so that must be his name. It was too perfect. I never spelled out the name "Jeff", after all. He also remembered Ghouls, and he remembered Geysers. He knew he had ghoulish wounds from what the Elves told him, and when he heard the poem about White Plume Mountain and heard mention of the Geysers, it all came back to him.
Furthermore, I was hoping there could be some way to make the party confront the intended final boss, the evil wizard who owns the dungeon. As written in the adventure, when the party tries to leave they're supposed to get stopped by his cronies. They either fight their way out or they lose and get taken to the wizard to be brainwashed. This didn't really work with my party's constant in-and-out approach to this dungeon, but I could now do something like this and have there be a risk of the wizard emerging from his lair to corner the PCs in his dungeon and finally do away with them. But to make it a dilemma, I wanted a reason to keep them in the dungeon and maybe force the encounter. So when they emerged from the dungeon after the death of Gregg, there was a new party at the base camp. One of the three rich assholes who owns the last magic item was down there with his full retinue and many knights. He has been told all about the PCs and has heard the accusations. He asks to see if they've made any progress past Wave, but they can't prove that they got Whelm now because it was on Gregg's body when he was sent back in time. They're fucked. They look really suspicious.
The rich guy is going to give them a loyalty test on pain of death. They have to go back in the dungeon to get his sword, the last weapon, accompanied by two observer parties: 1) his most loyal knight, and 2) one of the many cutthroat adventurers who has been waiting at the base camp. In this case, a dwarf (this was to give Gregg's player something the do during the session so he could still participate. But I think it also makes sense in-story. And being a dwarf specifically will help with something later). If they try to keep the weapon or harm the observers, the knight and dwarf will flee and tell about it. If they return without the knight or dwarf, it'll send the same message. And the rich asshole and his army will be waiting at the only entrance to the dungeon. The party has 24 hours to get the weapon or the army will charge in, spread out, and find it themselves, leaving nothing alive.
The party is trapped between a rock and a hard place, and their ultimate intention has a major snag in it. They need that weapon, and their wizard wants his goddamn spell books back. This time, they go through the dungeon, make a brilliant plan to carefully eliminate the ghouls, push Hunky through all the remaining traps, bully the knight the whole time while befriending the dwarf (they've got a lot in common with this fellow cutthroat!), and get the knight killed when he tries fleeing but is tricked into instead heading towards certain danger. They've narrowed things down to the final part of the last wing: the inverted ziggurat, the dungeon's most famous location. When formulating an ending, I already decided to replace the monster guarding the last weapon with, instead, Geoff/Gregg with 3 Gs + a dwarf he picked up a few years back to wield Whelm for him. A great final boss battle, more dramatic and appropriate than what the adventure planned. I even let Gregg's original player still control "Geoff" in the battle while also controlling the good dwarf. He had a ton of fun.
The big reveal happens, they see "Geoff" remove the hood from his cloak and reveal himself to be a very old Gregg with 3 Gs. He explains how he turned evil and why he did this for revenge and pride and whatnot and that now he is going to work for the big bag evil wizard in charge of this place to have the party killed or kidnapped for brainwashing. They have an awesome battle. They get Geoff on the run, fleeing up the ziggurat. The wizard had been saving a potion of flight for the entire campaign up to this point and he finally decided to use it so he could catch up. From up in the air, he shot Geoff with his bow and arrow and got a critical hit, killing him off. The corpse was in the right position to fall directly into the sea lion tank and get devoured. Some elements of the ending are so perfect and satisfying without being something you can plan for or script. Don't forget that.
But now they have three party members and three magic weapons. The wizard has his spellbooks back and a fancy sword, the fighter guy has his trident, and the new dwarf party member who replaced Gregg has the war hammer and can actually use it. This is all essentially something I railroaded to happen but it felt very natural to them. But how do they resolve the final challenge? The problem of being stuck in between the Evil Wizard and the Rich Asshole? That one came down to player-driven improv, I'm proud to say. In my version of Knave I allow my players to roll for random new abilities when they level up. They leveled up from their last fight, so to capitalize on that they needed to take a long rest and make camp in the dungeon overnight. Well wouldn't you know it, the wizard player rolled to get a new spell, any from the list that he wanted. He scanned it thoroughly and found the answer: "Gate: a portal to a random plane opens." That was their exit ticket. They did it right in front of the Rich Asshole just to add insult to injury. And I decided that they wound up in Hell, because whether this story ever continues or not, that's the funniest way for things to end.
I hope, from that example, you can see the principles of Navette Story Structure in action and how they helped make it a better campaign. Had the beginning, middle, and end all been a railroad or all be sandbox, I think the whole thing would have suffered a lot. You can kind of say that the rule of thumb is, "sandbox by default, railroad when needed" and you should learn how to recognize when you've reached a point where some railroading is beneficial to maintain a clear direction for things.
This was a pretty small example, only for one adventure, really within just one dungeon (which is itself a railroad structure). But you can see how these same principles can apply to an entire campaign. If you use the Navette structure in cycles, you might get a campaign that is 90% sandbox but with occasional chokepoints where you can exert more control and direction on things.
You need to be selective about removing player agency, but don't be afraid to do so if it's appropriate. We like it when the DM says "yes" as much as possible. Oftentimes we see "no" as a necessary evil, but I think it can occasionally be a virtue. The DM taking the reigns back at the right time for the right reasons can sometimes be the best possibility in a Sandbox.
This was a pretty small example, only for one adventure, really within just one dungeon (which is itself a railroad structure). But you can see how these same principles can apply to an entire campaign. If you use the Navette structure in cycles, you might get a campaign that is 90% sandbox but with occasional chokepoints where you can exert more control and direction on things.
You need to be selective about removing player agency, but don't be afraid to do so if it's appropriate. We like it when the DM says "yes" as much as possible. Oftentimes we see "no" as a necessary evil, but I think it can occasionally be a virtue. The DM taking the reigns back at the right time for the right reasons can sometimes be the best possibility in a Sandbox.
-Dwiz
No comments:
Post a Comment