GM 101: Railroads and Sandboxes and Agency, Oh My!

A lot (and I mean a LOT) of digital ink has been spilled over the topics of railroading, sandbox play, and player agency in games like Dungeons & Dragons… but often a new GM comes into these conversations not fully understanding what that actually means for their own game.  Well, I’m going to try to break it down for someone who is newer the the GM chair.

First, we need to define terms in the conversation, because depending upon your point of view, the terms themselves may already loaded with bias.

Player Agency - This is the ability for a player to have meaning choices in the game world. This is portrayed as freedom from pre-destination. Freedom from an "illusion of choice" where a particular path is forced upon the players by will of the game master. In other words, if the player selects path A, it will be different than if they had selected path B. The two paths can end up in the same place, but they must be meaningfully different (I’ll try to explain that more in a bit).

Railroad - This term is generally used for when the GM is guiding the players along a path upon which the players cannot diverge (i.e. - you can’t turn a train off its tracks. The players lose the ability to affect the plot (and therefore, player agency) , because a pre-defined course or story has already been set. This is usually used as a pejorative. However, there are times where rails are useful as a GM tool (I’ll also get to that in a minute), as long at the players understand and agree why they are there.

Sandbox - This term is used when the GM presents a “open world” scenario where the players have the most freedom to move in any direction, follow any adventure hook presented, or abandon said hooks at any time for something else that captures their interest. There are few constraints to the players actions. Sandbox games supposedly offer the most player agency, but even a sandbox game can turn into a railroad, or can suffer from other issues that remove player agency.

Source: http://shamusyoung.com/twentysidedtale/?p=1144

Is railroading always bad?

Yes. No. Maybe… It depends.

Most of the advice about railroading has been along the lines of “Don’t ever do it. Railroading is horrible. You’re a terrible GM… just do your players a favor and quit now.” This is a bit misleading, as there are degrees or railroading.

In game terms, a railroad is any time the GM has set up a scenario with a certain pre-defined outcome in mind. It’s not just a plot, but also a set of assumptions that that the players will take certain actions and no matter what they do within the confines of the plotted scenarios, the outcome will always remain the same regardless of PC’s actions.

In other words, no matter what the PCs do, they will inevitably go to plot point A, followed by plot point B, then point C… etc until they get to the boss fight at the end of the fully predefined, pre-plotted adventure. This is a bit of an over simplification as even published adventures offer some choice as to how the PCs get to the end, but for the purpose of this conversation, this will do.

Is that a bad thing? 

There are those that state with vigor that Player Agency is the single most important aspect of a role playing game. The players only have control of their own PCs, while the GM control the whole rest of the game world. Therefore anything that infringes upon player agency is Bad, with a capital B, because it takes away from the one thing the players do get to control. Any form of railroading is expressly counter to player agency. Metaphorically, they are matter and anti-matter.

However, there is a middle road. To expand upon the matter/anti-matter Star Trek metaphor, if you mix the right amount of player agency with a certain small amount of railroading, you can actually warp to spaces of tremendous fun.  Ok, so it’s not a very good metaphor… but listen for a bit more while we turbo-lift down to Engineering.

If you think about an adventure hardback from Wizards of the Coast or an adventure path for Pathfinder, there is a certain amount of railroad tracks required to build the plot. Because the game designers can’t anticipate every action a party might take within the confines of the adventure (especially over 8 to 10 levels), the author must make certain assumptions about the order of events and how the story arc they will play out. This is the nature of published campaign-length adventures, Because not every action can be anticipated, the author plots out the likely course for the players.

If you don’t have the preparation time to build out your own homebrew campaign. Published adventures give a wonderful framework for a new GM finding their sea legs, or to an old GM, like myself, who just doesn’t have time (or sometimes inclination) to write a bunch of adventure scenarios week after week.

It is also useful if you are not particularly skilled at GM improvisation. GM improv is probably the hardest skill to master as a GM. It’s one thing to make up an encounter on the fly… anyone can do that. It’s much harder to make that on-the-fly encounter memorable, impactful, and meaningful to the campaign. A pre-planned adventure is also useful if you are not well versed with all of the storytelling tools within RPGs -- set ups, rising tension, story beats, climax, and payoffs. The published adventure builds all these things for you.

Yes, a pre-planned adventure can limit a player’s game world impact if you let it. However, you don’t have to be a slave to the plot as written. If you are a flexible GM, you can let your players put that adventure through the blender and still come out with a tasty smoothie in the end. It may take a little more work, but you can give them a fair amount of player agency within the confines of the overall plot arc, and still be able to use much of the content within the adventure book(s). You will likely have to alter some of the later plot points or scenarios to align with the ripple effect of their earlier actions in the adventure, but it’s certainly possible.

Many of the newer adventure paths also offer options for more sandbox play. In Storm King’s Thunder, it is true that in the end there is an inevitable face-off with the major villain. However, there are several paths toward that goal, and the GM is given a lot of latitude to allow the players to roam within the setting and affect one of more of the various sub-plots within the adventure. Similarly, Curse of Strahd also is a little open about the order in which the characters achieve certain objectives. At some point, they will have Strahd’s full attention… but how they get there can be altered and manipulated by a GM that wishes to give their players a more open-ended campaign.

Is a sandbox game “better”?

To say a sandbox is “better” may be introducing some bias into the discussion. A a roller-coaster that is on rails can be a blast, as long as everyone on the ride has the same expectations and understands that there are certain story beats that will happen. That’s a perfectly normal, acceptable play style. The group (including the GM) is there to experience the ups and downs together, and though they will likely end up in the same place at the end, their journey may be very different than another group playing the exact same adventure.

Sandbox games offer a different kind of fun for the group that likes to explore with fewer constraints, but even sandboxes will have some walls. In a sandbox, the GM may come up with several small plot ideas, lay out the hooks, but doesn’t necessarily have a fully formed plot laid out. Part of the sandbox style means that the GM has to adapt to the player’s actions from week to week.

It may require more prep time, or at the very least, a heightened skill at improvising scenes and plot arcs as the players move around the map. This can actually create a really unique game experience that differs from the pre-planned story. Is it better? Well, that depends upon what you and your players want out of the game, and how skillfully the sandbox is maintained by the GM.

A sandbox can have its own pitfalls. Players may feel lost in a sea of possibilities, or conversely, feel directionless because there is not a strong sense of purpose or obvious objective. Most players prefer the metaphorical sign-post that reads “Here be dragons >>”. Even in a sandbox, you need to provide the players a list of actionable, meaningful choices.

Also, sandbox campaigns are often not truly wide open. If you are exploring a mega-dungeon, you are pretty much constrained to that single environment for most, if not all, the campaign. There may be lots of different things to do and even numerous mini-plot arcs to explore, but if you are just not into staying in the dungeon for multiple sessions on end, you are pretty much stuck unless you can convince the GM to move on to a different scenario. This is less of a problem with a wilderness hex crawl where a range of different environments and activities can be offered as the players roam around the map. Wtih my preferred play style, the mega-dungeon has very little appeal, regardless of how "sandbox" it is, whereas a sandbox wilderness campaign would suit my tastes.

Lastly, if you are looking for a Hobbit or Lord of the Rings style heroic adventure with a main plot arc and specific end goal with many obstacles in between, sandbox play may not be the style you are looking for since its nature doesn't support the model of a pre-planned epic story. As I noted, a certain amount of plot railing is required for a published adventure. As a GM, you certainly have the freedom to use the books only for its ideas, and play a more free form game letting your players explore all the corners. Many GMs already do this with published materials, especially experienced GMs… Many of us actually live the the gray space between following the book and letting the players sow their oats outside of the rails.

So why is railroading considered so bad?

When the players are agreeable to the plot rails (when they know there is a story to follow and they are willing to ride along) a great campaign can result. The players and GM can explore the story laid out together and the players are still given the freedom to play out individual scenes and vignettes in many different ways. The enjoyment comes from the plot reveals and twists and turns as the players move through the story.

The problem exists when the players have one idea about what they want to do, but the GM already has a preconceived notion of how the story should unfold. This can happen regardless of whether you are playing published material or playing a home brew adventure. A GM is actually more likely to cling onto a plot they have developed for their own game rather than someone else’s published ideas.

A GM can easily get caught up in their own storytelling. They have a notion of where the story should go, which special encounters they want to run, and their plans do not allow for flexibility. Instead, the GM guides the player’s actions in the direction they want. Sometimes this can even result in the GM purposefully thwarting or diverting good ideas from the players because they are not following the script the GMs sees in his own mind.

This is an absolute show stopper. This will destroy your game since the players will feel powerless to do what they want as a character -- the only thing they have real control over in the game. It is one thing to accept the visible confines of a published adventure path, but it is wholly different when the GM dictates what you as a players must do next. This is a player agency killer and your players will resent you for it.

One example is the “Prisoners” scenario. There have been many examples over the years where the PCs are taken prisoner, just because that is the set up for the next scene in the adventure.

GM: “You wake up in a cage with all your gear gone.”
Players: “Wait, what? When did this happen? How did they subdue us? I have the Alert feat…” etc.

Source: Christopher Stevens, Deviant Art

If you want the PCs to be taken prisoner, play it out. Set up an ambush. Have the enemies use sleep, hold, or other subdual methods, but give the PCs the opportunity to fight back or escape. Narrating the “you’ve been captured” scene is a cop-out unless all the players are willing to be in on it. If you want to suggest it as an option outside of the game to hand-wave a capture scene, that’s ok if they all agree to that bit of chicanery. Basically, they need to be willing victims. But don’t just take their character control away just so you can set up a plot point scene.

The second example is the unplanned Villain (capital “V”) death. Sometimes the PCs will get the jump on the main bad guy before you actually reveal the full breadth of the plot and he or she dies much sooner than anticipated. Let it happen. Don’t steal the PC’s thunder with some shenanigans. Let them have their victory and then figure out what the impact on the plot is going to be and take the campaign in that new direction. Sometimes a villain already has a near-death escape contingency plan. If you planned for that possibility and have valid justification, then ok... But don’t retcon the villain death because you didn’t expect it. Your players will be cheesed off and the game will suffer.

Another example, as described by Courtney Campbell, is the “Quantum Ogre” (or illusion of choice), where the GM describes two different paths that have distinct pros/cons to their selection. However, the GM is intent on running the exact same encounter regardless of which direction is chosen. In this case, the GM is giving a choice that isn’t actually a choice. He actually does a much better job talking about player agency. I’m trying to just summarize. So check his blog out.

For me, the "Quantum Ogre" lives in a bit of a gray space. I totally understand that sometimes there isn’t a scene or encounter the GM has prepared for every single direction the players might take. There is a desire to use content you already have and just move it around as needed. As long as the players are having fun, what's the harm?

I’m personally on the fence about this. If there are rails, trying to hide them from the players will likely annoy them and cause issues down the road. I well understand the limited time a GM has to prep a game each week and not all contingencies can be anticipated. I don’t think there is always a clear cut answer, but the ideal situation is that if you give the players a choice, make that choice meaningful. There should be a difference between the two options, or they should not be presented as two options. (again, go read the Hack & Slash blog. You won’t regret it).

In one of my own games, my players did take an unexpected turn such that I actually had to time out the game and just be honest. I didn’t think I could improv a scene that would be satisfactory, so I told them “Hey, I didn’t plan for that option… Can we come back to that next week? In the meantime, I have this fun little side trek prepared over here.”  They were actually fine with that. I did not want to limit their freedom of action, but I was caught flat-footed, so I offered some filler so that I could catch up and redirect my efforts to their desires for the upcoming sessions.

Final Thoughts

There’s actually a lot more to say about this, but this post has run really long, so I may follow up with another post based on the comments this post evokes. The long and short of what I’m saying is

1. A little bit of railroad is ok if the players are willing to go on the ride with you. 
2. Don’t try to hide the rails and create an illusion of choice. 
3. Give them as much freedom as you can within the confines of the story if you are playing a published adventure path. 
4. Alter the plot as written as much as you are able to allow the characters to direct their own narrative, but don’t stress about running the adventure as written if you are not great changing things on the fly, or your prep time means you are less able to divert from the main plot. Be honest with your players if you need time to make adjustments.

Stay flexible and just make sure everyone is having fun. If the players seem to be wanting to go in a different direction, try to roll with it, even if it means abandoning a lot of content you wanted to play out. You might be able to repurpose unused parts of the adventure at a different point during the campaign.