The Sandbox and the Railroad

[Jacob Lewis]

Chris Perkins did an amazing job of skewing perspectives to see things his way. I don't subscribe. There is a lot more to consider than just such simplistic labels which marginalize everything that goes into something as complex as running an RPG. Whether it's a one-shot adventure or a campaigning spanning multiple worlds, there is more to be said about the execution and delivery of a good game experience, not the least important as the reception given from those who participate (i.e. actual players involved in any particular instance). "Sandbox" is not always greater than "railroads" just by virtue of category, and vice versa. As someone else succinctly put it (and beat me to the punch), a sandbox is just a box of dirt, and roller coasters are also railroads. Perspective is everything. But don't misunderstand my point: there are good and bad examples of both, which means neither is superior or inferior to each other. They are simply different styles of play.

A "good railroad" is like a roller coaster you want to ride. The ride is predictable, but thrilling when designed to excite the rider. Paizo's Adventure Paths are good examples of this. Their lengthy adventures follow a largely linear progression but typically have enough story depth, theme consistency, and character hooks to keep a group's interest for the entire length of the campaign. Some are more popular than others, and there are some points within that may be dull, but that is no different than how we would rank different roller coasters; some coasters are more exciting than others, and everyone has their favorites.

A "bad railroad", on the other hand, could just be a bland adventure module lacking continuity, interest, or otherwise unable to keep the attention of the players who are basically forced to see it through. Usually it is a result from a DM focused on showcasing his personal design or story without any consideration of the player's investment. For example, you're stuck in a tower full of traps and goblins and cannot leave until you clear all 50 rooms, each with a goblin and a random trap. This is a more common mindset of earlier 1st Edition (or OSR) material where the adventure modules were like amusement park rides where you challenge your character creations rather than story-telling. Still, a really good dungeon design or story can still be enjoyable.

Sandboxes, on the other hand, work a little differently. It only works if you have players invested in the sandbox to begin with. If you offer them an open map and just ask where to go next, that may be freedom of choice. But without motive, story, or incentive, it's just wandering. Players usually need some direction, unless they've already planned their character's career and story before the game even began. Then they're just fulfilling a fantasy, and the DM is just acting as the chauffeur.

A better sandbox has story or motivation which the players can invest, and choices (or the illusion of choices) for players to pick their own path, possibly altering the course of things to come. A sandbox can be any size between a small village with several sites of interest, to an entire world(s) free to explore. Ideally, the size should match the scope of the story, as well as stay within the DMs ability to provide reasonable coverage. Granted, some DMs are more capable of improvising material on the fly than others, but that's simply a matter of knowing your limitations and explaining reasonable expectations with your group.

Personally, I think this kind of bias portrayed by Perkins, whether intentional or otherwise, is more harmful than good. "Sandbox" and "railroad" are just two different styles. Knowing what makes good and bad examples of both is more critical to becoming a skillful DM. But how you really know isn't from what a bunch of strangers online say, but how your actual players respond.

 

[dragoner]

I think this is a reasonably fair high-level description of how "scene-framing" RPGs work (eg Burning Wheel, DitV, and one approach to 4e which has been widely discussed on these boards).

Thank you for the explanation; It sounds like more work than the normal sandbox-railroad, or I no longer have the neuroplasticity to evolve my play style. If it works for you, that's cool.

 

[pemerton]

Thank you for the explanation; It sounds like more work than the normal sandbox-railroad, or I no longer have the neuroplasticity to evolve my play style. If it works for you, that's cool.

It's actually quite a bit less work to play the game on the GM side as well as the player side, rather than to have to write material in advance, prepare random tables, etc.

Here's a link to an actual play report of a first session. No GM prep required.

 

[Bedrockgames]

I think people are muddying the term 'railroad'. The way I have always seen it used, isn't just a linear adventure. Railroad is a pejorative for when the GM pulls out all the stops to keep the session going in a particular direction. A railroad session would be one where the players legitimately should have killed a key NPC, but because he was important to 'moving things forward' the GM miraculously lets him live. It isn't just about the structure. It is about how the GM enforces the structure. So I think positing a spectrum of railroad to sandbox is a little misleading here. Lots of people like linear adventures and adventure paths. No one really likes a railroad. That is the point of the term: there is nothing you can do to get off the tracks.

 

[dragoner]

It's actually quite a bit less work to play the game on the GM side as well as the player side, rather than to have to write material in advance, prepare random tables, etc.

Here's a link to an actual play report of a first session. No GM prep required.

It seems to me there had to be some sort of prep, as Hardby, the Green Dragon Inn, the Gynarch sorceress, other NPC's, etc.; all did not exist ad lib. Honestly, it looks like a sandboxy style too, with the players picking the direction they were going. Myself, I usually play sci-fi, so things have to make a little more sense, not for just science 101, but that the setting depth is usually more involved. I think some of this discussion dovetails back into having an established setting, and player buy in to that setting; that lowers prep a bunch. Some players want a more tabula rasa approach where the characters can forge their own destiny.

 

[Jacob Lewis]

I think people are muddying the term 'railroad'. The way I have always seen it used, isn't just a linear adventure. Railroad is a pejorative for when the GM pulls out all the stops to keep the session going in a particular direction. A railroad session would be one where the players legitimately should have killed a key NPC, but because he was important to 'moving things forward' the GM miraculously lets him live. It isn't just about the structure. It is about how the GM enforces the structure. So I think positing a spectrum of railroad to sandbox is a little misleading here. Lots of people like linear adventures and adventure paths. No one really likes a railroad. That is the point of the term: there is nothing you can do to get off the tracks.

That is an excellent point! Somewhere along the path, the lines got crossed and became synonymous. RPGs are a really complex topic!

 

[Tony Vargas]

a railroad is:
The main events of the shared fiction are determined by the GM, either in advance, or perhaps by improvisation in the course of play.

a sandbox is:
the GM authors the main elements but the events (both which events occur and what sequence they occur in) are initiated by the players.

In any event, we can now see what a RPG that is neither railroad nor sandbox might look like, because it's not as if those two things cover the field of ways in which fictional elements and fictional events might be generated.

I tend to think of railroad as linear, and sandbox as unstructured. Maybe that's a little broader than the usual, especially as the usual would be in the context of D&D, as the big coelacanth in the small pond. In that D&D (dominated) context, though, the DM coming up with everything in advance (or feigning 'in advance' via adroit improv) can be pretty safely assumed. But, it also seems like 'railroad,' in particular has some negative connotation, so getting away from that should be part of any analysis....

So, if we remove that assumption, I think specifying where the authorship comes from is superfluous to, lets say, more, clinically: linear vs unstructured.

For instance, we can imagine the players generating the elements but the GM establishing events. ...It's not a sandbox because it reverses the authorship roles: the GM rather than the player initiates events, but the elements of those events are not authored by the GM in advance but are established either by the players or by the actual outcomes of prior play. And it's not a railroad because you can't author this sort of series of events in advance, because you can't have elements that are generated from play until play actuallay happens.

Sounds like it's mainly not GM-authored.

 

[innerdude]

One thing I've noticed too is that the player's perspective on the level of freedom available to them can change if a player determines that a character is "bought in" to whatever stakes are at hand. When this happens, then the player and character are "invested," and plot linearity (if any) fades to the background. The character is doing what the GM "wants" because the character's fictional positioning would indicate that they would act in that manner.

This is, I think, the rationale for long-term "adventure path" play --- sooner or later, the players/characters will be invested in the stakes at hand, so even if events are linear, it "feels" organic to what the character would be doing in the fiction.

The problem comes in when a player has a character idea that doesn't quite mesh with the fiction as authored. I had this problem in the last campaign where I was a player. My character, as envisioned, had VERY specific backstory and active plot "hooks" that were germane to the concept of the character, but in play were never addressed in anything but token fashion by the GM.

At one point I was ready to drop that character and bring in a new one, but by then the character was so integral to the party's strategy and planning that it really wasn't tenable to do so.

All in all it made for a very uneven level of enjoyment on my part, though as a whole the group seemed to be having fun.

 

[howandwhy99]

I don't think there is spectrum between a "sandbox" and a "railroad". One is a game and the other is a story. A sandbox is played and succeeded or failed at due to the players' actions and abilities. The latter is where the "player" reads or listens to a story.

What I think confuses contemporary gamers is many computer games try and remove gaming from their design and predetermine a linear story. Future events are almost entirely not a consequence of players playing a game, but really pushing the buttons to feed themselves more of the linear story. This is sometimes called a gauntlet design. The player has to proceed down the gauntlet and overcome every challenge along the way as a game, but if they fail they simply resurrect over and over and over again. "Bad" gauntlet games then become ones where the gaming is actually considered a hindrance for receiving the "pellets" of story the designer is feeding them.

Compare that to a game that is actually designed to support many of the assets found in games rather than stories, Chess for instance. The number of pieces and types of moves for every piece are highly limited, but the quantity of potential future situations is staggering. That's what actual D&D is: a massive quantity of potential futures shifting as you take each action in the game. Every action matters for the entire length of the campaign, but resurrection is highly limited and if you die you have to start over at the beginning - not the starting location, but back at level 1. You need to self select to face easier challenges for your role.

 

[Lanefan]

One thing I've noticed too is that the player's perspective on the level of freedom available to them can change if a player determines that a character is "bought in" to whatever stakes are at hand. When this happens, then the player and character are "invested," and plot linearity (if any) fades to the background. The character is doing what the GM "wants" because the character's fictional positioning would indicate that they would act in that manner.

I call this the players railroading themselves - the GM, knowing the characters and what motivates them, can put down fifty adventure hooks and know exactly which one will be chosen.

I've done it to myself, in the past: played a character that in effect railroaded himself into things that wiser characters might have avoided, just because of his personality and motivations.

The problem comes in when a player has a character idea that doesn't quite mesh with the fiction as authored. I had this problem in the last campaign where I was a player. My character, as envisioned, had VERY specific backstory and active plot "hooks" that were germane to the concept of the character, but in play were never addressed in anything but token fashion by the GM.

This is one issue. Thoe other happens when-if two or more of these self-railroading characters are on tracks that go to different places. The simplest example:

Character 1 will always altruistically do the noble and-or heroic thing regardless of reward, and actively seeks out such deeds to do no matter how dangerous they may be.
Character 2 will always take the most potentially-lucrative option without regard for ethics or any other considerations except maybe the lower the risk the better.

Drop a half-dozen adventure hooks in front of these two, stand back, and watch the fireworks.

Lanefan