As is typical of practically any hobby, there are a couple of hot-button issues when it comes to tabletop RPGs and GMing: alignment restrictions, “rule of cool,” whether natural 20’s or 1’s should always be automatic successes/failures, when to say “no” to a player, and so on. Last week, I weighted in on one of these topics in our discussion on alignment and a different way of looking at it using an absolute approach. While it wasn’t my intention to start of this site discussing all of these controversial topics immediately, as it so happens, I played in two games this past weekend that ended up bringing one of these topics to the forefront, so today, we’re going to tackle the big one. Perhaps the most controversial of all these controversial topics.
Today, we’re going to discuss “railroading.”
What is Railroading?
If you’ve been GMing or playing tabletop games for any length of time, you’ve probably heard about the concept of railroading. To quickly sum it up for those of you who haven’t, railroading is a term generally used by the gaming community to describe a game where the GM forces the players to go down a certain path or make certain decisions. Depending on who you ask, there are a lot of things that a GM might do that railroads players, and the general sentiment amongst a large portion of the community is that railroading is a very bad thing. Why? Because it takes away the agency of the players, the sense that their decisions are meaningful and have an impact of the world. This reasoning isn’t entirely incorrect. However, like with most things, the concepts of railroading isn’t as straightforward as most would think.
A Deep Dive into Player Agency
To really understand the railroading debate, we first need to consider player agency. Agency, in a philosophical sense, is the capacity of an individual to act in a given situation. It is what we might call “free will.” From the perspective of an RPG, agency is the same thing – the capacity of the PCs to act freely in a given situation. Players should have the freedom to decide what actions their characters take.
In an RPG, though, there’s also a second aspect of agency we have to concern ourselves with, and that is the idea of consequences. It isn’t enough for a player to be able to act freely if their actions have no effect on the world. For an example of both sides of this, lets dive into the world of video games and look at the original Mass Effect trilogy from Bioshock. The original Mass Effect games became well-renowned for their world filled with strong characters to interact with and a sense that the player’s decisions were actively affecting the game world. Crew interactions, in particular, were one of the most prominent examples of this. How you interacted with a specific crew member in past interactions could affect affect future conversations with them. If you choose one crewmate’s side in an argument with another crewmate, for example, you would damage any relationship you had built with the other – perhaps to the point of them even just refusing to speak to you.
Your actions in the game would also have an impact in the game world. Mass Effect 2’s final mission is perhaps the best and most well-known example of this. After an entire game’s worth of preparation, you and your crew must carry out what is effectively a suicide mission against a superior foe with minimal support and with the fate of the galaxy hanging in the balance. During the mission, you have to choose specific crew members to carry out specific roles, and your decisions here will have an impact on what happens during the final mission. There are also tasks you can perform beforehand to help prepare for the mission, such as gathering resources to upgrade your ship’s systems. On top of all this, at a certain point in the story, members of your crew are kidnapped, putting you in a race against the clock. If you dawdle to get some extra preparation done, your kidnapped crew members might be dead by the time you arrive.
Once the mission gets underway, all of your preparation and decisions during the mission have a direct impact on the mission’s success. If you ignored the ship upgrades, damage to the ship could kill a crew member before the mission could even properly begin, and if you choose the wrong crew member for a specific mission role, they or another crew member might die. Since the crew interactions were such a large part of the series, these were all crew members that you’ve spent a great deal of time getting to know, some of them over two games, and even though the mission would ultimately be successful regardless of your choices, your decisions did have an impact in terms of casualties, and because of the relationships most players had with their crew, each casualty was a huge consequence.
This entire final mission is a great example of how make player decisions have meaningful consequences. Most importantly, the game also made sure to highlight that it was your decision that led to the casualty, and on the other side, if you made the right decision, it will highlight that your decision potentially saved your crewmate’s life. It was abundantly clear that the player’s decisions were having a direct impact on the mission’s results, and that was a huge part of what made that game so engaging.
Contrast this, however, with the great controversy that was the ending of Mass Effect 3. All of the decisions that you have made across three games were supposed to come together into this one, final fight to stop the extinction of all the galaxy’s advanced, sentient life. You spend the majority of this third game gathering support for the war effort to aid in this final fight. But once the final mission comes around, its hard to see how all your work paid off. There are few consequences that carry the same personal impact as the suicide mission from Mass Effect 2, and the game doesn’t do a great job of establishing a direct connection between your decisions and their consequences (or lack of them) in the game world. Ultimately, this left many players feeling unsatisfied. For all of their decisions made over the past three games, the final payoff was one of three mostly-similar endings. It made it feel like all of those past decisions ultimately didn’t matter. The player had the agency to choose, but they never got to see that their actions had consequences.
Good Choices Have Good Consequences
When it comes to GMing, this is ultimately what player agency is all about. Not just the freedom of the PCs to make their own choices, but the freedom of the PCs to make decisions that have lasting consequences. Consequently, as a GM, maintaining a sense of player agency in your game isn’t just a matter of “avoiding railroading.” You need to ensure that the player’s actions have visible and appropriate consequences in the game world.
Conversely, it is actually okay to limit the options available to the players and guide them along a specific path in some situations, so long as overall, you provide a sense to the players that their actions are having a direct and visible impact on the world around them. To return to the Mass Effect 2 suicide mission example, the player’s progression through the final mission is pretty linear. There are a series of encounters throughout the mission, and once you’ve completed one encounter, you move to the next, regardless of your choices. No matter your choices, you will eventually reach the final boss, defeat the Collectors and save the galaxy. But despite this, the player still feels like they have agency because even though they can’t break off this linear path, they do still have the ability to make decisions and they can see the consequences those decisions are having on the world.
Since tabletop RPGs have GMs who can think up situations on the fly and who (hopefully) aren’t limited to whatever situations a group of developers programmed them to handle, we have greater flexibility in being able to handle a wider range of actions from our players than in these video game counterparts. Despite this, though, the basic principle we looked at in that Mass Effect 2 example still holds true: good games (and by extension, good GMs) find a delicate balance between player agency and the story’s need to guide the players along a certain path.
A Complicated Metaphor Involving Trains
If you don’t follow The Angry GM, a while back he wrote about the topic of railroading and player agency and described the balance between keeping the party moving along a narrative while allowing for player agency effectively as laying down railroad tracks while the train chugs along directly behind you. Honestly, this is pretty good way of looking at it. Some GMs and players like “sandbox” type games where basically nothing is planned and the GM just takes whatever actions the players choose to do and wings it. If that’s the type of game you and your players want to play, there’s nothing wrong with it. Every group will have different preferences. But at that point, it’s not really a roleplaying game. It’s virtually impossible to try and tell an engaging, cohesive story when everything you do is a reaction to the actions of your players. Depending on the group and the GM’s improvisational skills, the beginnings of such a story might start to develop out of this sandbox environment, for that story to develop into a full narrative for an adventure or campaign, some degree of structure will be involved.
To take Angry’s train metaphor and run with it, let’s pretend that your game is really just a long train ride from a specific starting point to some sort of destination, and you are responsible for building the tracks between your chosen starting point so that your train can take your players to wherever your destination happens to be. To help us get a sense of direction here, lets take this train metaphor literally. Say we’ve chosen Rome as our starting point. You think your players will enjoy visiting Venice, so you choose that as your destination. You think, “Well, why don’t I just save myself some time and get all the tracks down now?” and so you go ahead and build an entire track from Rome to Venice. Your players arrive in Rome and the train sets off.
Perhaps, after the train starts moving and makes some progress, your players realize the direction the train is heading in. Say the train has made it halfway and suddenly your players decide that they don’t really want to go to Venice. Instead, they want to head to Firenze and enjoy some fantastic food, art, and wine. Well, at this point, your train’s just passed Firenze, so you have to jump off the train and quickly rebuild the track to head to Firenze. Once you’re done in Firenze, your players might decide to get back on the track to Venice and you can just get them back to the track you previously built, but they might not. They might decide that they enjoyed Firenze so much that they want to explore more of Tuscany, and now you have to lay more tracks to take them to Siena or San Gimignano or wherever they want to head to. They might want to check out Napoli, forcing you to build a track all the way back to Rome and so you can head further south. You might never end up using that second half of the track to Venice, and all of that time and effort you spent laying those tracks were wasted.
That needlessly complex metaphor in GM terms is planning out an entire campaign in advance. You decided that you want your campaign or adventure to end at Venice and went and built the entire track. You planned out all the stops, all or most of the encounters, and chances are, you made some assumptions about what decisions the players would make for you to do so. Or, worse, you didn’t create any opportunities for the players to make any decision meaningful enough to change the course of the tracks. If your players try to derail the train because they want to go somewhere else, you become forced to either abandon the tracks you laid to go to this new destination or to refuse to change direction. The later situation is what most people mean when they talk about “railroading.” You don’t let the players change the direction of the tracks, and in doing so, you take away player agency.
On the other end of the spectrum, you could decide that since you don’t know where your players want to go, that you just won’t build any track except for the piece directly in front of the train. As the train moves and you learn where you’re players want to go, you lay down more and more pieces of track. This is the “sandbox” situation I mentioned earlier. You’re just letting the players decide where to go and reacting by putting down the next piece of track accordingly. The problem is, you have to find a way to make that next piece of track fit according to the train’s current direction and the surrounding terrain. You have to react and improvise to everything. As mentioned earlier, that’s not something that most GMs can consistently do well. We need at least some track laid out in advance to give us some breathing room.
Finding the balance between structure and player agency is effectively a matter of finding out the right length of track to build to give you that breathing room. Build too little track, and the train is going to run out of track before you have the chance to lay the next section. Too much track, and you might end up passing right by the players’ desired destination. The right amount of “track” will ultimately depend on the GM, the group, and the game – GMs more comfortable with improvising encounters and adventures on the fly will likely need less track, while GMs who prefer to plan these things in advance will likely want more. Personally, I find the right amount of detailed preparation work for most of my groups to be about a session in advance. For anything after one session’s worth of encounters, I will typically have high-level ideas as to what might happen or a long-term adventure or campaign goal to work towards, but unless I expect to get to a certain idea in the upcoming session, I’m typically not going to put in the work to flesh it out into a full encounter or scene in advance. This lets these ideas remain flexible as the players progress while keeping things moving towards an overall destination.
Preparing to Go Off the Rails
No matter the GM and party, if a game runs long enough, there will inevitably come a time when the party tries to go off the tracks. The players might latch onto a particular NPC or location that you describe in the world, even if you hadn’t intended them to be of any relevance to the main narrative. The players’ actions might cause things to move in a different direction than planned. They might try to resolve encounters in a way you had never considered. Or, they might just not be interested enough in an adventure hook to follow the tracks and find something else they want to do instead.
These situations are inevitable. The simple fact is that we aren’t capable of thinking the exact same way as one of our players will. We can’t foresee the thought process that leads a player to latch onto something you thought was unimportant, or find a ridiculous solution to an encounter, or decide whether a potential adventure sounds interesting or not. Depending on how well we know our players, we might be able to guess at these things, but at some point, we will guess wrong and things will go off the rails. Thus, if we want to be able to plan out a long-term narrative or put in any sort of advance preparation for encounters and scenes, we need to prepare for the possibility that things will go off the rails.
Now, if you’re a smart GM, you’ve held a session zero with your players and gotten an idea of what type of game they’re looking for. You should be able to come out of that session with at least a rough idea of what sort of destination the players have in mind or at least a direction, and you can use that to inform your structure and planning. Once your planning is at least in the same rough direction as the player’s desired destination, it becomes less likely that you’re going to be left with wasted preparation whenever the party diverts off the rails.
Oftentimes, these “derailments” are only temporary – the party has become distracted by what they see as a hook for an adventure, and once you have figured out what that adventure is and the party has completed it, you can often them guide them back on the rails towards their original destination. The closer that original destination was to the type of game your players wanted, the easier it will be to do this. If you picked a destination without figuring out what type of games your players wanted, however, your original destination might have been completely wrong, and the players may not have any interest in getting back on the tracks. If that happens, then you have to find a new destination to work towards, because if you don’t do so and try to force them back on the rails anyway, you are going to come directly in conflict with player agency, and your game will suffer for it.
Even if we nailed the planned destination for our game, there will still be occasions where the PCs get distracted by something else or go in a different direction than planned. There are, of course, many different routes to a specific destination. Although we can never plan for every possible route the party might take, we can intentionally design some of our encounters in a manner that gives us some flexibility to repurpose them as needed if things do go off the rails.
Building Flexible Encounters
For some encounters, flexibility can be as simple as the ability to change one group of NPCs with another. Let’s say, for example, that you’re planning out your next session and you intend for the party to run across a crashed caravan along the road which then leads them to a group of cultists that have been kidnapping travelers to be sacrificed in some sort of ritual. In the actual session, the party sees the crashed caravan, but decides not to investigate and instead just carries on along the road to the next town, which you originally didn’t intend for them to reach this early. Though you could just wing it and try to work out a new encounter when the party reaches the town, you could also look for a way to repurpose the original encounter to try and get things back on track. You could even adjust the encounter slightly in a way that ties it back to the party’s choice to ignore the caravan along the road.
Perhaps once the party arrives in town, they hear rumors about strange things happening near the outskirts and of townsfolk who have mysteriously vanished in the middle of the night. If they investigate, they find that some group has been kidnapping townsfolk under cover of night and follow the trail back to the cultist’s ritual site. Once at the ritual, they have an encounter with the cultist’s as originally planned, but perhaps there is now an additional element to the encounter. If the party found the ritual after investigating the caravan, perhaps they would get there just as the ritual was starting and could stop it before any of the victims could be sacrificed. But if they skipped the caravan and found the cultist’s after investigating in town, they’ll arrive and find that the ritual is already well-underway. They might find the bodies of those sacrificed to the ritual already, and might notice that some of the bodies have the same sigil as a symbol that the party noticed on the caravan along the road. The encounter might also now have an additional mechanical element where the party needs to disrupt the ritual before the cultist’s can summon an even greater threat.
If you do this right, you can both get things back on track and help bolster the feeling that the PCs’ actions have tangible consequences in the world. The players have no way of knowing what the encounter might have been like had they chosen to investigate the caravan. For all they know, they might have been able to stop the ritual before it began, subverting the encounter entirely. The one caveat, however, is that, like with many things when it comes to GMing, it can be easy to overdo this. Most of the time, you can’t just move the original encounter and pretend like the decision the party made to avoid the original encounter didn’t happen. If you do that, then you are channeling that Mass Effect 3 ending where the players’ have the freedom to make choices, but those choices ultimately feel inconsequential to the overall world. Sometimes, the players might be none the wiser that you just moved an encounter to a different location, but if you keep doing this over and over again, they will eventually start to catch on. Even if the encounter is just one that you repurposed, you need to create the perception that the party reached the encounter as a direct result of their choices and actions.
Complete Derailment
In some cases, the party might reject whatever adventure hook you had planned outright, or do something else that derails the game to a point where repurposing your scene or encounter is simply not feasible. When this happens, you need to have a backup plan – a backup adventure hook that you can use so that you’re not completely winging it if the party has completely rejected the path you where expecting the session to follow. You don’t need a fully planned out alternate session with stat blocks and meticulously planned encounters – just a hook that you can use to present an alternate adventure and a rough idea of what types of encounters you might use should you need them.
Ideally, one or two of these encounters are ones that you can create just by taking one of your original encounters and swapping around NPCs and other elements to suit the new adventure. To continue to use the example of the cultists, lets say that when the session starts, the party is in a major city, along the road we expect them to travel on. Once the session starts, however, decide that they actually want to just stick around and investigate the city further. Your alternate adventure hook could be a crime organization in the city that the PCs might run into should they stay. One of the possible encounters under this path could be the party running into a group of the organization’s thugs. In that case, you might be able to alter the stat blocks of the cultists somewhat to determine the stats of the thugs. Instead of a ritual, the party might be instead trying to stop the thugs from carrying out a hit on a city official who is trying to weed out the gang’s corruption from within the city’s government. The crashed caravan investigation might instead be a crime scene investigation where the PCs find the body of a messenger who was killed by the gang while trying to find and warn the official of the planned assassination.
Between the alternate adventure hook, repurposing encounters, and some improvisation, you should be able to use the alternate adventure to get you through the session, after which you can then revisit your planning and figure out whether your final destination for the game has changed. If it has, you can figure out the new destination and plan the next session accordingly. If it hasn’t, you can figure out how to get the party back on track towards their original goal once they have completed the alternate adventure.
Know When to Use Encounters vs. Scenes
The last topic I want to touch on when discussing player agency and “railroading” is the specific topic that led to this article in the first place. As I mentioned in the introduction, there were some situations that came up in games I played in this past weekend that brought these topics to the forefront for various reasons. Ultimately, the underlying issues all of these situations had in common came down to the difference between encounters and scenes.
To look at it in very abstract terms, a scene is one particular part of a sequence of events that combine to tell an overarching narrative. Every scene has a specific purpose in terms of the overall narrative. That purpose might vary in importance – one scene might just be a humorous scene intended to help lighten the mood after several tense scenes, while another scene might be the final battle between good and evil that serves as the climax of the narrative. Those two scenes have very different degrees of importance to the actual story, but they are still both scenes.
Encounters, technically speaking, are a specific type of scene. Encounters are scenes where the end result of the scene is unknown. There is something that is preventing the protagonists from accomplishing or making progress towards their goal and there is a chance that the party might fail and incur consequences.
The key difference between an encounter and normal scenes is that an encounter has some sort of conflict. There is something that the players’ have to overcome in order to progress, and that implies that the players have to take action. More than that, the final result of the scene is in some way linked to what actions the players’ take. Scenes, in contrast, don’t have an end result that is in question. Think of them as cutscenes in video games – there is no conflict, no question that needs to be answered, and – most importantly – no actions that need to be taken. A scene will typically be able to progress without the players needing to make a decision or a single roll.
The problem that sneaks up on a lot of GMs is that all of us tend to draw inspiration for our scenes and encounters from other sources of media. Video games, in particular, is a big inspiration for a lot of GMs I know, though books, film, and other sources of media certainly have an impact as well. None of these other forms of media, however, have the same level of player agency as tabletop RPGs. In a video game cutscene, you choose what actions to take to affect the scene unless the developers have programmed in the ability for you to do so – and even if they have, the decisions you can make are likely extremely limited. Often, we might see a cool cutscene or moment in a video game or in some other sort of media and try to incorporate that in our game. The problem is, when we do so, we sometimes forget that the players have a level of agency that is not present in the media that inspired the scene.
Static cutscenes do not exist in tabletop RPGs, and you cannot make them exist without denying the players’ freedom to make choices. If your scene has a dramatic question and any form of conflict, you do not have a scene – you have an encounter. If you think that a particular scene is a non-encounter scene and your players’ think that it is an encounter, then your scene and player agency is going to collide, and one of the two is going to have to give. And if you ultimately decide to sacrifice player agency for the sake of your scene, you’re going to cause frustration in your players. You cannot risk this for the sake of a single scene, no matter how cool or dramatic you think it will be. If your players start treating your scene like an encounter, you need to figure out what the unanswered question, conflict, and possible outcomes are and run an encounter, because regardless of whether you think it is a scene or not, your players think that it is an encounter. And if their actions can do nothing to affect whatever sequence of actions you’ve decided happens in this scene, you will break player agency and your players’ will be led to believe that their choices ultimately don’t matter.
Villain Scenes are a Myth
The single area where I have seen this problem occur the most, by far, is in scenes involving the villain of an adventure or campaign. In most types of media, it’s a common trope for the protagonists to have an early encounter with the villain, or perhaps multiple encounters. They might try to stop the villain and fail, they might arrive the room holding the Secret Artifact of Plot Importance just in time to see the villain leaving with it in his clutches, or the villain might just be there to give the standard Bad Guy Monologue. Most of the time, these scenes exist to help give the villain character, provide insight as to his motivations and plans, or just as a barometer to show how far the protagonists have come once they finally manage to overcome the villain later in the story.
In tabletop RPGs, though, your players aren’t just going to stand by and let the villain do what they want. They are most likely going to try and kill or otherwise stop them – i.e. an encounter. If you decide that all of the actions the players attempt are impossible or near-impossible because that’s how the scene goes, your players will be left frustrated and with the impression that their choices don’t matter. The only way to avoid this is to let the scene turn into an encounter, which means that the players have a chance to succeed or somehow change the outcome of the scene in their favor. And that means that there is a very real chance that they might kill your villain, even if you didn’t want this to be their final confrontation.
As a general rule, if you are going to have a villain be in a scene involving the players, you need to assume that the villain is not going to come out alive. Every once in a while you can pull off a “last-ditch escape” where the villain teleports out or pulls off some other sequence of events to survive. But you can only do that once. Once the players have seen the villain use some BS to escape, they are going to do everything they can to prevent the villain from doing it again. And if you turn around and find some other BS escape method the next time, they’re going to resent you for it and, once again, will feel like their choices don’t actually matter.
A Player’s Perspective
One of the situations I encountered as a player this past weekend was one of these villain scenarios. It wasn’t even the “major villain,” either. Our party was trying to rescue two NPCs we’ve previously worked with after they were kidnapped by a woman who was hinted to be a major figure in the main villain’s forces. This investigation led us to a cave, outside of which we found one of the two NPCs, who had just enough life in him to give some dramatic last words. Naturally, any magical healing or attempts to stabilize automatically failed. We entered the cave eventually saw the woman performing gruesome magical experiments on the other captive NPC.
This was where our GM had planned to have a scene with this newly-introduced villain. The problem was, we were already in this cave expecting an encounter, and we already had been given the objective of rescuing the captive NPC before the villain could complete her experiments. Thus, our natural response is to try to engage the villain before she can finish the experiment. Any player who approached had to beat a DC 21+ Will save just from being in the villain’s presence or be magically restrained. This Will save restrained all of our party except for our skald and alchemist. The skald charged at the villain with her legendary axe which she had just been gifted only a session or two prior, which resulted in an automatic disarm and a impossibly-high Reflex save or be restrained. Our alchemist then attempted to shoot the NPC to put her out of her misery as she was clearly not going to survive the experiment, which had no effect and resulted in an automatic restrain.
The villain then proceeded to taunt and monologue until a second future villain teleported in to lead her somewhere else while the captive NPC transformed into the boss monster we were actually capable of fighting. It was blatantly obvious that we only had the illusion of choice and nothing we actually did or could have done would have stopped the transformation or done anything to the GM’s newly introduced villain. And when that happened, we were just left there wondering what the hell was the point of our decisions, since they would clearly be overridden whenever they did not fit whatever narrative the GM had for the game. Player agency was shattered, and as a result, we were left frustrated and unengaged.
Ultimately, what caused this was the fact that the GM was treating this entire sequence as a non-encounter scene, while we as players were treating it as an encounter. When the GM failed to recognize that his scene had turned into an encounter, it became apparent that we really only had the illusion of choice. Once that illusion was broken, player agency followed suit and we were left feeling “railroaded.”
Consequences are the Key
As I mentioned back at the beginning of this article, the concept of railroading isn’t as straightforward as many make it out to be, an as a result, we’ve covered a pretty wide range of topics throughout this article. We discussed the basics of railroading and how it is really just a way of expressing a game with a lack of player agency. We used a very over-complicated train metaphor to look at some of the problems that can occur when players have a different goal for the game than we do. We also covered some ways we can create flexible encounters and alternate adventure hooks to be better prepared to respond to unexpected player actions and decisions, and also looked at the question of scenes vs. encounters and how mistaking an encounter for a scene can lead to us inadvertently harming player agency.
Though some of these topics might seem unrelated, there is a thread that ties all of these things together that we briefly touched on earlier. Player agency is about much more than “free will” and the ability of players to make choices. Ultimately, the most important part of player agency is not the choice itself, but the consequences. Having the freedom to make choices means nothing if those choices have no meaningful impact on the game or the world. Sometimes, those consequences will affect your whatever planned-out encounters or story you have, and when that happens, you have to be willing to be flexible. If a PC takes an action that might derail your scene or story, and that action is both physically possible and capable of potentially producing the player’s desired result, you need to let them perform that action and resolve it fairly as you would any normal action. No automatic failures, no impossibly-high DCs, and no “you succeed but it doesn’t actually change anything.”
If the action succeeds, you must make sure that it has an appropriate impact on the game and the world, even if those consequences come at the cost of your scene or encounter. Adapt to the new circumstances and see where things go, whether that means adjusting your previously-planned encounters, running an alternate adventure hook, or even improvising something entirely new. For all you know, whatever new adventure or story your players end up finding or even creating might be better than the one you had planned. No matter what ends up coming of your party’s journey off the rails, you will have a world that the players feel like they can interact with and have an impact on, and that sense of agency will do more for your players’ immersion and engagement than any scene or story you create ever could.