Many game masters consider Sandbox Campaigns and Railroad Campaigns to be opposites.
Others say that they are the extreme ends of a continuum and there are many possibilities that involve mixing the two in varying degrees.
Railroad is a verb. Sandbox is a noun.
I don’t think about “Railroad Vs. Sandbox” as an either or proposition.
Applied to game mastering, I use the word “railroad” as a verb. A game master is railroading the player characters when the GM alters the outcomes of an encounter produced by player choice and game mechanisms.
Sandbox is a noun. I use the word “sandbox” to describe a dynamic open world campaign structure. The verbs that go with the sandbox structure are “simulating” or “modeling.”
A linear campaign is not necessarily a railroad
Many game masters conflate linear campaign structure and railroading.
The linear campaign structure allows the game master to direct the campaign a lot more than a sandbox campaign but that doesn’t make it a railroad unless the consequences of player choice are obliterated by the GM.
A game master avoids railroading a linear campaign when they let go of their attachment to the outcome they intended. The game master has to be open to events unfolding as player choice and the game mechanisms dictate. This produces an emergent story, not a contrived one.
Detachment is especially needful when adventures take a surprising turn.
The sandbox is a hostile environment for a railroading game master.
One reason why Sandbox Vs. Railroad seems like a dichotomy is because using the railroading method in a sandbox campaign structure typically leads to failure.
A sandbox structure is intended for players to go where they want and interact with the world as they want. A game master can railroad in a sandbox but it rarely lasts long.
Railroading a sandbox is like going to an amusement park. The players get to choose what attraction they want to visit but the game master controls the path between attractions and the activity is constrained to whatever the game master has contrived. It feels like danger when you are riding the roller coaster, it seems like a real social interaction when you meet the saloon keeper but it’s all a constructed illusion for your edification, not a living world.
That is OK if that’s the sort of thing you like but it lacks the spontaneity and uncertainty of a sandbox.
A “hub and spoke” structure is not a sandbox
One of the falsehoods around sandbox campaigns is that they are railroad campaigns where the players get to choose which track they ride on.
In this false sandbox, player characters begin at a hub of the “sandbox” and are given several options to choose from. They make a choice and ride that track to it’s end.
I call it a “hub and spoke” campaign structure. Once they’ve finished the first spoke, the player characters can go back to the hub and pick a different spoke or start from a new hub at the end of the first spoke.
This is not a sandbox campaign structure.
You could run a hub and spoke campaign without railroading it the same way you’d run a linear campaign. The game master allows each adventure to go where it goes.
The difference between this and a true sandbox is that the adventures in a hub and spoke campaign have only one connection point. The sandbox is more complex.
A dynamic system made up of dynamic systems that interact with each other
To me, a sandbox campaign is a dynamic model of a fantasy world.
The milieu is a complex system consisting of interconnected and interdependent subsystems.
This is why a megadungeon can provide hundreds of sessions of engaging play. It is constantly responding to player actions.
Each time the party interacts with the dungeon’s denizens, its doors, it’s traps… the dungeon changes. It isn’t the same thing from session to session.
It’s a dynamic system that answers a question, “What happens when the player characters do this?”
The change can be small or gargantuan depending on what the players are interacting with and how much of a disruption they cause.
The party destroys the lich on level 9. The lich was a counterbalance to a demon in sublevel 9b. The lich and it’s undead soldiers out of the way, the demon can send its minions to upper levels since the demon doesn’t need those forces for defense of its lair. Once the demon subjugates all the other monsters in the dungeon, it can use it’s new minions to launch powerful attacks on the surface world.
A minor element of the sandbox such as an ordinary wool merchant doesn’t have much effect on the whole.
A major element of the sandbox, such as a legendary monster, a 20th level wizard, or the mercer’s guild in the capital of an empire will be connected to many other factions, NPCs, locations, and monsters.
Player interactions with those major elements can disrupt the entire model. These events tend to be major epochal moments in a sandbox campaign.
Alternatives to the Sandbox vs. Railroad comparison
Because I consider Railroading to be a method and the Sandbox as a campaign structure it doesn’t make sense to differentiate between Sandbox and Railroading.
More useful comparisons would be…
Sandbox vs. Linear structure
Sandbox vs. Branching structure
Sandbox vs. Web structure
Sandbox vs. Hub and Spoke structure
If I want to compare railroading methods to other methods of running a campaign then I would look at methods of play like Railroading vs. Simulating or Railroading vs. Player Narrative control.
That is why I consider Railroad vs. Sandbox to be a false dichotomy.
Grumpy Wizard 
I agree!
LikeLiked by 1 person
Thanks- as always- for your thoughtful post. So here’s a question, what do you do between game sessions? Do you have some kind of outline or checklist for each faction? eg: Well, the heroes arrived at Village X and now the thieves guild will spy on them. Or are you asking the players directly like: “Hey you guys have arrived at Village X, how would the local thieves guild react to your presence???” And then you go home and prepare notes based on the players’ replies? Also, since you are ever reacting, how do you know what notes to prepare for a gaming session? I mean you could write up a whole dungeon and then the players just decide to ignore it. How do you adapt to that? Maybe you have already addressed this in a previous post, in which case feel free to refer me to that. Cheers!
LikeLiked by 1 person
Great questions Armand. I’ll give some quick answers and I’ll include a few links. I may expand some answers into full blog posts and videos for YouTube.
For a long time, a lot of this was very much in my head. In the last few years I’ve gotten more disciplined about keeping notes because of screwing up continuity.
What do I do between game sessions?
In my binder, I have a sheet or section for each faction or major NPC. Any session notes that impact that faction are recorded in a log. I note the session number and date it occurred; both real world session date and the date in the game. A lot of what I do between session is a sort of solo role-play. When I create important NPCs I give them a long term goal, problems they are facing to reach that goal, and the resources they have. My NPC creation process is the thing that drives a lot of my campaigns.
Based on what they want and their resources it’s not too difficult to create a response. If I am the Thieves Guild master, I have my own abilities and magic items, the guild’s wealth, influence, hideouts, safehouses, bought off government officials, contacts and the junior members of the guild. If the guild master gets a report from a lower level member about something the players do, I write down the response and put in my prep material for the next session. I often don’t detail those resources until I need to. They are just a note in the entry on the NPC/Faction that I bring to life when it comes up in game.
I don’t know where every NPC and every faction is at at all times but often make some common sense rulings in the moment unless there is some on going conflict with the PCs in progress.
This is one reason why I came to agree with Gary Gygax that time keeping is very important in a sandbox game. I have a timeline of events that helps me reference what has already happened, what the players are working on and any salient events that I need to take into account.
“Also, since you are ever reacting, how do you know what notes to prepare for a gaming session? I mean you could write up a whole dungeon and then the players just decide to ignore it. ”
A lot of this is addressed when I create the campaign and before we even start playing. Even though players can go mess with anything in the setting that doesn’t mean that everything in the setting is worth messing with. We know that PCs get experience by acquiring treasure and killing monsters.
I create locations with monsters to kill and treasure to find and then tell them where they can find them with hooks, rumors, events etc. Assuming that players want to advance their characters, they’ll pick one of those locations. I generally keep about 3 to 5 location adventures ready to go. I use a lot of one page dungeons, and an occasional purchased one. My homebrew location adventures are NOT big blocks of text. They tend to be general ideas, lists, and then a roster with stat blocks. Super simple and focused on usability at the table.
When I create the starting area or a new area players move into create or recycle several small monster lairs or adventure locations and micro adventures that take an hour or less to run. I use these as micro-exposition about the campaign world. https://grumpywizard.home.blog/2021/05/20/micro-exposition-for-game-masters/
Players discover those little “side quests” as they are exploring a part of the campaign world or on their way to another location. I’ve never written about how I place these micro-adventures in the setting so maybe I’ll do that as a separate post but it is also tied into how I map out a sandbox (also a separate topic)
Early in the campaign, I have to work more to figure out what the players are interested in doing. The beginning of the campaign is introducing players to the location, telling them about opportunities for XP and treasure through quest givers, local rumors and gossip, events they can get directly involved in and other hooks. As the campaign progresses I have a much easier time because players start developing more specific goals for their characters as they learn about the sandbox.
I talk about in a YouTube video I put out a few weeks ago. https://youtu.be/03ov3PJ0c1E?si=qNO6gAtgEbfn7e3Y
Sometimes players ignore stuff but it doesn’t happen as much later in the campaign and even when it does, that’s OK.
Does that answer your question?
LikeLike
Also this post is good to keep in mind…
If you have these things in your sandbox, a lot of the problems of players not biting on hooks and adventures are reduced.
LikeLike
Yes- this is a fabulous trove of info! I skimmed through it and plan to go through it more carefully again. I’m working on a campaign idea for 2025 using sandbox design (which is relatively new to me. I am realizing now that I was in a Vampire the Masquerade campaign in the 90’s that used sandbox design, but at the time, I didn’t realize the difference). Another thing I’m realizing (for me) is that it’s really hard to move from linear planning to this sort of open concept design.
Hope you have a nice weekend! I’m painting the wall behind my kitchen fridge today which is one of the least rewarding activities in life, but I do have my Ravenloft game tonight. That one is definitely a hub and spoke campaign with the feel of a very sophisticated board game, but it’s a great experience in its own way; great people too!
LikeLiked by 1 person
Another link… Sorry. Last one.
This is a session report.
What I want to draw your attention to is that there are several factions mixed up that mostly only interacted with the players individually. The players actions over the previous 20 sessions or so drew them all to this one point in time in the sandbox and that was not planned by me. It was response that came from NPCs reacting to the situations created by player interaction with the sandbox and its elements.
LikeLike
I really like the idea of the 3-ring binder with a page/ section for each NPC faction and logging the event history (or timeline) of events and then following up with a check in on the NPC’s progress. I also like the idea that each major NPC has an intention/ goals. What I am registering is a movement away from using a map to keep track of things to something that’s more like a TV show storyboard.
LikeLiked by 1 person
I guess there’s also the option of having *someone* play as the factions in a separate game…
Though it might end up with more work for little gain.
LikeLiked by 1 person
There are some groups doing “patron play” where someone who isn’t playing in the weekly sessions plays a major character but asynchronously and only gets to give orders to minor NPCs or use their character’s personal resources. I’ve heard it can go well if it is managed and coordinated well. It’s definitely an interesting idea.
LikeLike