Sword-Fighting on a Roller-Coaster: Railroading for the Best in RPG Play

The specter of the railroad haunts RPG adventure design. Every groaning bit of pipe and every rattling carriage has many players and GMs afraid the rail-ghost is near. When it strikes, it renders decisions moot and turns artistry to ash, revealing carefully cultivated choices to be just curlicues on a churning, infernal machine…

But whose choices are we talking about? Does “railroaded” play undermine the engagement of the player or does the notion of railroading impugn the prep that sometimes makes play possible? Both?

In tabletop gaming scenes, we’ve turned the metaphor of railroading into a two-speed toggle switch when really it’s a throttle with many degrees between full speed and a standstill. The train isn’t the only vehicle we can ride when talking about actual play, but it’s the one most soot-stained by years of use. We’re wasting some perfectly good metaphors because we’re afraid they’re haunted.

Let’s clean up the locomotive. Let’s see if we can’t renovate the train. Let’s maybe tip our hats to the engine drivers. And let’s put the metaphor to use as a force for good in gameplay and design.

“We’re wasting some perfectly good metaphors because we’re afraid they’re haunted.”

Rails Through Wilderness

Railroading, roller-coasters, rail cars, and the wilderness all around them. Players can get overwhelmed by open-world freedom in a world they cannot visualize or render in their imagination.

The insidiousness of the term is in its use as an apparently objective gripe where a subjective experience is real enough and important enough to make the point— because the experience of play is what’s really at stake. Feeling railroaded is what’s often not fun, and I feel railroaded when false choices are misrepresented as real choices.

Don’t let me go on thinking my character can rob a god if that’s impossible. Maybe you won’t let the gods be robbed because it would interfere with your narrative vision or maybe you just don’t want your world’s gods to appear weak — these aren’t problems if I know as a player that they’re just not going to happen. (Or that they won’t happen until I reach epic levels of power, maybe.)

If my character attempts to persuade the Queen of the Realm to avoid a war, and nothing will change her mind, is that railroading? If she is protected by psychic bulwarks and dead set, as a character, on going to war for revenge, is maintaining her decision in the face of my character’s intervention an act of railroading or is it upholding a consistent portrayal of the Queen as a character?

It’s about communication, in my experience. Showing that a mountain cannot be lifted isn’t railroading unless I, the player, have developed (or been given) an expectation that if I roll just right … maybe I can.

Players are railroaded when they’re rushed or pushed onto a train they didn’t want to board. Getting on a train that’s headed in your direction isn’t railroading.

“Getting on a train that’s headed in your direction isn’t railroading.”

The idea of railroading in RPG play is a pretty straightforward application of railroad as a verb: players and/or their characters get pressed into actions or travel they’re not interested in undertaking. The slippery, unhealthy aspect of the term comes from the combination of the railroad as a poisoned metaphor (as if everything that touches that third rail dies) and the act of railroading as a trust issue. Because railroading is something that’s done in secret, suspicion abounds.

For me, the sin of the railroad occurs when the GM supplants player decisions with his own decisions without player approval. This devilish offense takes many forms — players can board a train to one destination only to learn its headed somewhere else, for example — but it can be disarmed through honest engagement, whether the players are an audience or collaborators in that moment. RPGs are unusual in that the players get to be both.

Simply put, rails do not railroad if the players agree to board the train.

And rails are part of a great tradition of game-design technologies meant to focus play and save the developers and GMs from designing vast swaths of content that’ll never see actual play.

I like to hike and visit remote locations but I also enjoy a good road. I like to make little discoveries in out-of-the-way locales but I often drive to them. In a car. On a road.

But a nation’s worth of paved and gravel roads is a colossal undertaking. It’s more than many hobbyist GMs and professional developers and designers can manage, and understandably so! Winnowing down the potential areas of environmental and narrative design is a good way to get the most out of a GM’s or designer’s time — and a player’s. We use roads and rails to guide players toward the discoveries and adventures that are most worth their time.

Paths, be they roads or rails, are a form of creative restraint, too. Without barriers or limitations to focus play, a game is hardly about anything. The real world, for example, allows a huge variety of actions and results — it is the only actual full-scale world we have where anything that’s possible is possible — but the real world isn’t about something the way we want a game to be. Lots of boring shit happens in the real world.

We winnow options in gameplay all the time to great effect. Fiasco’s playset lists guide players toward choices that are on-topic. D&D’s encounter and treasure tables determine what you may find in the confines of a dungeon. Apocalypse World’s move results limit what can and cannot happen as a result of certain actions. When game designers put limits on our decisions, players are more likely to accept that as part of the designed experience.

That is, we sometimes honor the designer over the GM — though often “adventure designer” is a role the GM plays. I, as a player dabbling in the talk of thieves and art of wizardry, get to play at heroism and authorship. The GM gets to play at developing game levels and narrative design in turn. Let’s respect the work the GM does.

At the same time, GMs must respect the choices that players make. They were made, after all, in the play environment that the gaming group has collectively established and in the game world that is built collaboratively. It’s up to the GM to clearly communicate the situations when players can make meaningful decisions and avoid letting a potentially cosmetic choice (“How do you slay the orc?”) be mistaken for a more meaningful option (“Do you slay or spare the orc?”).

Don’t disguise a stretch of road as an intersection. Instead, portray the potential outcomes or destinations down each path so the players can make decisions meaningfully. Sometimes hints are enough — one hallways smells of fresh flowers and moss while the other smells like rotting eggs — and how much information you give out to make a hard choice tough is one of the choices you get to make as the GM. And you can build on and remix that choice as you see the faces and hear the questions of the players!

“Don’t disguise a stretch of road as an intersection.”

Part of the freedom of play in RPGs is that players often have the power to invent their own options at a fork in the road. If you, as the GM, can’t or won’t create unique content for them on the fly in response to that sort of choice, be honest about that. “I really only have the game world designed for the trains that go east and west,” you might say. Take a break, remix what you have, or improvise as you’re able.

Players, remember that you are asking a GM to improvise or remix their work when you insist on striding off the train, into the wilderness. Part of the collaboration of play, even in a sandbox, is having fun with the toys at hand. Don’t send the GM off to the toy store or the workshop on a mere whim. Every player — including the GM — should get to add things to the game that are important or captivating for themselves.

If we accept roads and collaboration as a part of gameplay — though they supplant for the absolute freedom of the lone writer with a pen or keyboard — can’t we allow for rails as a similar creative restraint? If we can, then isn’t it a matter of how railroads are used that makes them welcome or verboten at your table?

So let’s reclaim and expand the railway metaphors as tools for good play.


A roller-coaster is a play experience in which the RPG players have a great time careening through a rollicking environment full of explosions and/or feelings while fully aware that the whole experience is designed or curated. Roller-coasters aren’t about if but how things get to the end of the track.

When the players are happy to be railroaded, we might call that a roller-coaster. If it doesn’t work out, it can be like someone describing the Dungeons & Dragons ride to you.

Often the GM works the throttle while the players work the guns and the dialogue. Like a rail-shooter sequence in a Call of Duty or an Uncharted, roller-coasters have their uses. These are dependable experiences that players can compare with each other if they all play the same adventure, because the commonalities between them are probably high.

“The GM works the throttle while the players work the guns”

Roller-coasters are great tools for avoiding two cardinal sins of storytelling: boredom and confusion. If the players are confused, it can be okay because the GM is driving the action forward the whole time. If the players are bored, then it’s not a roller-coaster at all — it’s unwelcome narration.

A roller-coaster sequence can be as brief as a few sentences that make the players feel great about their characters (“You dash through the flaming house, gathering every kitten and puppy along the way, holding them tight to your chest as you crash sideways through a shattering window into the airship outside.”) or it can be a number of rolls or dramatized moments on the way to an exciting destination where the players’ choices matter more — maybe even the choices they made while on the roller-coaster! And a roller-coaster can go on too long, of course, turning a rambunctious sprint through a battlefield into an all-night slog through endless waves of goblin hordes. A canned description of a violent parade through Goblintown is (even) less satisfying to play through than it is to watch.

I’ve had success with roller-coasters used to help players and/or their characters shine (since the outcome might not be dependent on their dice rolls) and to create a sense of harrowing, dramatic inevitability, like a stormy battle that the PCs can survive but cannot stop. I sometimes use a roller-coaster sequence as a practical tool to save real-world time during play.

Rail Cars

Rail cars are beautiful environments moving toward fixed destinations.

Players work the throttle and the brake, which can risk the whole train grinding to a halt or allow players to barrel through expected stops along the route. These are good for ensuring important themes or dramatic arcs get conveyed (even if they go by as blurs outside the windows). The GM usually gets to establish the scenery outside and the goal at the final destination, trading some pacing and narration to the players in exchange for more influence over the train stations and status quo. Although players might ransack, dismantle, or ruin the rail cars, the GM only has to design or populate one corridor through the game world — that’s a time-saver.

Friendly or collaborative players can build on the environments and the pace set by the GM, adding to the depth and detail of the cars instead of exploring the wide territories outside. Character creation helps establish what a player can bring aboard.

Think about how you can interface with the environments put forth by the GM, whether that means adding to the atmosphere or cooperative confounding through contrasts:

  • A beautifully constructed bar car with plush red carpets and brass fixtures makes a nice setting for dialogue or a conspiratorial soiree.
  • A glass-and-steel dining car lit by flat-screen monitors bearing news of an attack on the capitol is a good place to build on the backstory of the setting and portray your character’s feelings about it, however surprising they might be.
  • A private cabin iced-over by a magic spell might be a compelling place to investigate or show off your fire magic.
  • A dirty boxcar where spent shell casings roll out the open doors as the train wobbles may be apt for a fist fight.
Conwy Castle and its rail bridge.

Railway Stations

A railway station is a finite but dynamic area of the game world where players and their characters have a lot of freedom to explore play.

Lots of video-game levels are railway stations — guaranteed stops on the way from inciting incident to end credits — where the game can be learned, played, expanded, and explored. If you can shoot, sneak, or talk your way through a game level in whatever direction or manner you like, that’s a nice, robust space even if it’s linked to the same follow-up level every time. (“Next stop: the snowy infiltration level!”)

When the GM uses quick summaries or cinematic language to cut to a new dungeon to explore — whether you choose the order of the stations visited or not — that can be an example of railways stations in action. The time between railways stations (in and/or out of the fictional world) is usually in the GM’s hands, not the players’.

I have a space-fantasy game world called Orbita that I use for focused dungeon crawls through a weird medieval/industrial mash-up universe of cathedral asteroids and zero-gravity castles and the whole thing is a massive collection of railway stations with magical teleportation linking them together. Want to explore the smithy sinking into the coolant of the water planet? Activate that teleporter. Want to unlock two new dungeons to explore? Turn on the magic engine in this nebula-choked monastery. It’s an overt implementation of this style of play where the space between dungeons is instantly traversed for the players and the characters. If the players want to slow down and get their characters a rest, they need to clear out some monsters and make the space to do that.

That’s a highly gamey way of enacting this kind of space. More cinematic options abound through the magic of sentences like this: “You buy a ticket and fly to Hong Kong.” The point here is that we don’t dwell on the rails so that they don’t take up a lot of our time.

Contrast this with…

A single path (in green) taken through the high-speed rail network in Europe hearkens to linearity in actual play.

The Railway Network

A railway network sets out clear and defined paths through the wilderness of the larger game world so players can take whatever trains they like to whatever destinations they like, when they like — even if the GM has a hand in making the train schedule.

Rail networks are fine examples of paths toward engaging play. The GM can neatly convey where exciting things can happen in the game world without totally dictating when and how things happen. This can help create a practical balance between player and PC freedom and GM prep. In a game that demands a lot of prep from the GM — and in prep-heavy campaigns played through prep-optional games — visiting destinations designed by the GM can be a straight-up friendly and fruitful thing to do. You might not know what is in store at that destination but you know that there are things to do there. When you leave the network, you may be risking boredom or confusion by asking the GM to invent stuff for you on demand.

A stop in a railway network might be a single short scene or a whole dungeon to explore, and the size of each stop on the network can vary within a single campaign. It might even be a roller-coaster or smaller local rail network!

The point here is that players are aware of when they’re boarding the train and when they are free to explore.

The stations and rails and dramatic Act structure of Always/Never/Now.

Hybrid Structures

All of these metaphors can be easily remixed, built up, stripped down, and otherwise worked with to create new structures — if we take more of the stigma off of them. The best way I know to do that is by creating game adventures that use these structures to deliver satisfying play that respect and reward the time spent on them.

My cyberpunk-action adventure, Always/Never/Now, uses some railway network design to link railway stations into a storyline that highlights forward momentum and a dramatic three-Act structure. As you move along the rails, time progresses and the NPCs prep actions of their own. There’s no going back, except in flashback. When played as a campaign, A/N/N gives players some pacing control over the space between stations, too, for some railway network functionality. For shorter play or even a one-shot, I trim out the rails and transition directly from station to station on the network.

For my purposes, these channels between action scenes were important for momentum and for supporting the GM with ready-to-play content so she could deliver high-octane action for the players in ways that intersected with the pre-gen characters directly. I’ve heard A/N/N pejoratively called a railroad, too, which is fine as far as opinions go but wasn’t particularly helpful as a critique. Of course it has certain set paths or rails in it — obviously it does — because it has things to say about how rails can be used and retrofitted for play. To me the “criticism” of the railroad there was rather like saying “this book has chapter titles!” Accurate, I guess, but I don’t know what the motive was for telling me that.

Swift passage through grand mountains.

Lots of new techniques and approaches to designs using the rail metaphor are out there for us to explore. They don’t have to take whole new games to make them work. A single adventure can riff on these ideas in compelling ways. Some great ideas only yield one or two sessions of engagement, rather than whole campaigns’ worth, and that’s okay! But if we surrender to a fear of the rail-specter we abandon ideas that we can use to make play accessible, digestible, and easier to talk about for new and established gamers alike.

Let us fret less about “railroading” as a label, now, and spend less time arguing what the term stands for in any faux-empirical way. Let’s talk more about the instances of play, of the example in question, when we can. If someone tells us that this adventure or that choice by the GM feels like railroading, let’s ask “When did you feel that way?” or “What could we have done instead?” and focus on the practices of play and where they take us.

Let’s drive the conversation forward, through the mountains, toward new playgrounds.