In this discussion article I’m going to weigh in on the age-old debate on sandbox v railroads. I was inspired to write this article after thinking about what I could have done better in the Minotaur Trilogy, and after reading a twitter thread from Mike Mearls. To boil it down, there are two ‘styles’ of campaign, or two ways in which you can run a game. They are, like most things, two ends of a continuum. Neither is inherently bad, but it seems that most people prefer a mix of the two.
A railroad, as I understand it, is an adventure/campaign/play style that involves little or no player agency. No matter what they do, the PCs are headed down one route that cannot be changed. Often, this arises because the DM has a definite story they want to tell, or one they have discussed with the players in advance. Railroading is often seen as a negative thing – people don’t normally play RPGs just to sit through a scripted narrative. When viewed as an adventure/campaign, a railroad often means that one part of the adventure is directly connected to the next, with no wriggle room to explore alternative options. I believe that this is, at times, unavoidable. By their very nature, dungeons are often linear, railroaded processes. You travel from room A, to room B, to room C and then D. Room A provides an intro, B and C provide challenges and tell a story, room D has a boss fight and some treasure. I believe railroading becomes problematic when it’s a play style. If a DM can’t handle player agency, and refuses to accept any route but the one they had planned out, games can get boring, and even become more meta-challenges of trying to guess what the DM had planned. An example would be refusing to allow non-combat solutions to a fight, or disregarding any attempts to break down a door which, in the DMs mind, MUST have its lock picked.
On the other end of the continuum is the sandbox – an adventure.campaign/play style in which ALL options are open to the PCs. Characters have the option to go anywhere and do anything that is feasible in the setting of the game. At its best, this occurs because a DM knows the world and the agents within it so well that they can adapt to most choices the PCs are likely to take. At its worst, this is because a DM has not prepared anything, so they throw down a map, improvise the location, but fail to present any real opportunity to adventure and change the world. In many ways, player agency is not the same as player choice, though it is often presented as such. Sandboxes illustrate this perfectly. The players can choose to go to the abandoned elvish temple, dwarven mine or orc-infested hills. This makes them feel empowered, and like they’re playing in a real world. When they arrive, and deal with the situation they find there, they expect it to change that real world, and to cause narrative developments. If this doesn’t happen, they have no agency – no ability to influence the world around them. As a campaign style, sandboxes tend to place you at location A, then allow you to explore locations B, C and D in any order you want, often with multiple interlinking options. As a play style, sandboxing allows characters to circumnavigate challenges not only by the most obvious methods, but by thinking outside the box.
For me, the best campaigns fall somewhere between the two styles of play. I think it’s incredibly important for both players and DMs to have a strong sense of the overarching narratives of a campaign. For this to be evident in play, it necessitates some form of railroad. In Chris Perkin’s The DM Experience (an essential read for all DMs in my opinion), he talks about ‘invisible railroads’, and how when characters stray too far from one rail, he can provide little nudges to place them back onto it over time, or slide them seamlessly onto another rail hidden just out of sight. I like to think my DMing style uses a similar approach. At any one time, there are several arcs progressing along their own independent timelines. The characters drift between these, slowly gathering information on each, building the mystery, before they start to commit more to one than another, and I guide them along to a conclusion. These form the arcs of the campaign, and each can be catered to be more or less suitable for the appropriate level. Having said this, once characters are on one rail, I try to keep them there, unless an entertaining detour presents itself. I know who the main NPCs are, where the main locations are, and what big events are likely to occur on the way. They are all subject to change, but are clearly set out in my mind (and notes) ahead of time.
On the other hand, in terms of play style, I like to be as sandbox-y as is reasonable. If my players come up with a wacky way around a challenge, I’ll normally let them get away with it, even if I know that it’s messing up planned events down the line. I’ve certainly over- or perhaps mis-prepared in the past because I expected the party to do one thing, and they’ve argued themselves right out of it within the first 15 minutes of the next session. What makes it possible to keep playing when this happens is what I talked about above. Knowing your world, and your arcs, well enough to keep the game alive. They’re leaving one invisible rail, but heading straight for another.
The above is all well and good in your home game, where nobody is paying to play. If you’re a 3rd party adventure writer though, you need to be a bit more careful. It’s extremely hard to write a good sandbox adventure for publication for several reasons. Firstly, D&D is increasingly a character-driven game, and I don’t just mean player agency. More than ever, narrative arcs are based around the backstories and actions of the characters, which makes for rich storytelling, but is impossible to write as you can never know who is going to pick up and run your adventure. Secondly, you can’t predict the actions of characters within the adventure. Even if you think you might have most of the bases covered, there’s no way you can cover ever eventuality. Just listen to the actual plays of adventure modules out there. Did Phil think that a party might cast enlarge on a familiar and use it to weight down pressure plates? I doubt it, but he’s written the module in a flexible enough way to allow it to work. Finally, you never know what kind of DM is going to run your adventure, or how they’re going to run it. Each DM is different. Some might tear it apart and re-purpose a few select elements, others might run it to the letter. The important thing to remember when publishing adventures is to keep it clear. Write a decent background, and an even better synopsis. Sure, things might go off the rails, but if you’ve provided enough information about the NPCs and events that are likely to happen, most DMs will be able to fill in the rest.
Ultimately, I don’t think it matters how you run your game as long as everyone is having fun. Each group will have its own preferences, and that’s a good thing! What I would suggest however is that you are cognisant of how you run your games, so you can make the most of your own style. If you run a railroad, make sure your links are tight, so players feel like the flow is natural. Make sure your BBEGs are enthralling and every-present, so they don’t get distracted. Make sure the finales are epic, so that victory is sweet. If you run a sandbox, make sure there’s always another option. Make sure you include a huge variety of potential stories to be told. Make sure you include a hexcrawl (we all know it’s the best way).
If you’ve got any topics you’d like to see me cover in future articles, just let me know on facebook, twitter or in the comments! If you want to read these articles before everyone else, get a chance to win free adventures each month and get access to even more D&D 5e content, join my patreon!