The Time before Sandboxes and Railroads

Today’s entry is about the terms “sandbox” and “railroading” in role-playing games. When I stopped playing games about the mid-1990s, these terms were not used among role-playing gamers, unless they referred to a box of sand where children played and the tracks along which trains and locomotives traveled.

All the RPG bloggers today seem quite familiar with these terms in new, derived senses. The sandbox is a virtue and railroading is a fault.

A “sandbox” or a “sandbox campaign” is one in which the world is wide open and players can go in any direction to explore it.

“Railroading” is the sin committed by a Referee (DM, GM) when he or she denies players a choice of action. The implicit ideal, by contrast with this sin, is that players should have a sandbox at their disposal. The plot, if there is one, should be generated by the players, and their understanding of the setting should be “emergent” as they explore. Player choice should dictate direction, not Referee plans.

Sandbox and railroading are therefore effectively complementary and, as ideals, mutually exclusive.

The virtue of “railroading”

Those who rail against railroading can become dogmatic. One fault with their complaint about railroading is that it can overlook character motivation. Skillfully designed characters will frequently have inherent motivations that are conducive to scenarios where player choice is not unlimited. The quest for a specific in-character goal that has one end and the sandbox ideal sit awkwardly together.

Character motivation may be simply the background declared for a character by the player, but some game systems practically demand motivating characteristics on the character sheet, as each character will have traits with game mechanics attached that help players to guide character behaviors. (Think of GURPS and Pendragon or even those newfangled games that are “Powered by the Apocalypse.”)

In a game in which character advancement results from loot and murder (XP for GP and for defeating monsters), a sandbox may work well. The goal is built in: kill monsters and take their shiny stuff. The players sally forth and search for those two things, usually finding the one behind the other. But in a game in which character advancement is related to accomplishing goals or playing a role, a sandbox may be less suitable because the motivation or the goal is mismatched to the mode of exploring at random. When characters have goals besides wandering and plundering, plots may become more directed not by Referee engineering but by player design.

Another fault with the complaint about railroading is rightly pointed out by the seasoned and thoughtful DM David, who usefully distinguishes linear adventures from railroading. (More on linear adventures below.) Not all players enjoy having unlimited of possibilities. They want choices but they also want a purpose, a goal, a direction. I have seen the “paradox of choice” (see this and this) alive in the games I run for my kids. Too many options can lead to paralysis, as one sees among young or inexperienced gamers.

Some scenarios that players may desire, such as a Quest, should entail a sequence of specific steps in order. When characters are motivated to undertake those specific steps, the negative accusation of railroading has no place because the course of action, duly planned by the Referee, accords with character motivation. In these cases, story structure is motivated by elements within a character’s design. Railroading is not then a sin, but I can imagine players, once taught to look for railroading, decrying the reasonable structure of a quest.

As DM David further notes, a dungeon environment works well to reduce choices available for players to a visibly finite number. Do you go left, right, or straight? Given clues and descriptions of sensory cues by the Referee for each of these three directions, players make meaningful choices about which way to go without facing the mind-boggling possibilities of going just anywhere and doing anything at all.

Where the term sandbox adventure comes from

As the terms were not used when I was a young gamer, I had to spend about five minutes poking about the internet to determine how the two terms entered the jargon of table-top role-playing games.

The “sandbox” is an import from video games. Some MMORPGs of the late ’90s were described as “sandbox worlds.” That is, they were settings in which the player’s avatar could wander and explore at will. Of course, the expression “the world is your sandbox” is the background of this, conveying a sense of freedom and play.

Perhaps the fact that wargamers at the dawn of role-playing played with miniatures on “sand tables” made the metaphor especially appropriate for import to the table-top.

The sandbox is a metaphor for rebellion against the dictates of someone who would restrict your creativity and force you down an unvarying track. Bad Dungeon Master!

But, as I have argued above, there is no reason to think that a linear adventure design restricts all kinds of player creativity. If role-playing games were solely about choices right or left, take or leave, and not about strategy and tactical decisions, or roles or drama, or storytelling, then linear adventure design would truly be a disaster. But there is a lot of room for creativity and fun in a linear scenario. If player skill means defeating threats, and that is all, then you had better have a sandbox because your players are not doing much more of interest than wandering in your world.

I am not saying you should force players down a linear path to help them develop role-playing skills. I always design scenarios in which player choice is meaningful. You can’t please every player, but you can minimize the range of player preferences you disappoint.

Anyway, to get back to the point: sandbox is a term from computer adventure gaming (which derived its premises in turn from table-top role-playing games).

Where the term “railroading” comes from

Why are linear adventures bad? Because Ron Edwards and his friends in the Forge forum seemed to say so.

For those who missed it, as I did, the Forge was an internet forum (floruit 2001-2012) presided over by a games theorist and game designer named Ron Edwards. He and his friends and followers thought hard for many years about how role-playing games work and proposed how they should work. They developed hundreds of technical terms serving as in-group jargon that had the effect of excluding pedestrian gamers from their illuminated ranks as they determined the Perfect Way to design and play role-playing games. They believed that there were just a few kinds of kicks that gamers got from games and that each game should deliver just one of those. Games that tried to do more were deemed incoherent. This, at least, is my take on it after reading a few too many of their essays.

One of the climaxes of the Forge was Ron Edwards’ argument that bad game design can literally lead to actual brain damage in players. You can read about it here or you can read it for yourself here. There are many traces of discussion about the Forge because of this kind of controversy. Also, the people who attacked it elevated its importance far beyond its worth.

Speaking of railroading and brain damage, I am reminded of the Alexandrian’s blog entry that says that railroading is “player abuse.” (The link he gives leads right to Ron Edwards.) The Alexandrian wrote a “manifesto” about railroading, too. Apparently gamemasters can really traumatize their friends by depriving them of choice, like Ron Edwards’ brain damaging role-playing games.

One of the things that the Forge did most effectively was to tell players off behind a screen of intellectualism, insisting that others besides them were having fun in incoherent and wrong ways. Much of their jargon is nothing but terms for mistakes gamers are likely to commit: how not to play role-playing games and how not to design role-playing games. A whole catalogue of perceived problems was the result. Take the now-widespread term “fantasy heartbreaker.” Maybe it’s useful, but it’s ultimately just another one of the Forge’s terms for things you should not do. In this case, it means “don’t design your own house rules version of D&D because other people did it a lot before you.” In effect, this sounds to me a lot like Gygax at his most imperial (also here).

Wittingly or not, the Forge turned into awful scolds and, in my point of view, they have left lasting damage on the hobby, inciting people to take sides in debates charged with intellectual terms. At the same time, even self-appointed OSR pundits who attacked them adopted their terminology, like “railroading.”

Anyway, “railroading” was one of the bits of jargon from the Forge for one of the innumerable cardinal sins of gamemastering. Specifically, the earliest use of “railroading” in this sense that I could find, without spending much time on it, is in a Forge essay of January 2003, where it appears to be an established term already. I would guess they cooked it up in 2002 or slightly earlier.

All this took place during the period when I was not paying any attention to role-playing games. I am glad I missed it all while I was not gaming!

The upshot is this. From the start, the term “railroading” was negative. If you say a Referee is railroading, it’s a criticism. Railroading is an outrage. The sandbox is the cure.

Before sandbox and railroading

One of the side-effects of the internet is that we who rely on it today tend to forget whatever was discussed in print before it developed.

As it happens, early game designers and players did think a lot about the distinctions between different kinds of story structure. We often knew what we were doing. We had ideas corresponding to railroad and sandbox already.

I think of a short essay by the smart game designer Ken Rolston from way back in 1982. This appeared in Different Worlds, one of the pro magazines about role-playing games that featured far more intelligent game commentary, in its early years at least, than Dragon magazine ever did.

Rolston wrote a page about the how to pick and choose among published role-playing game scenarios. To this end, he distinguished three types of scenarios.

• Linear
• Multiple Option
• Open-Ended

What made these three different was illustrated by the graphic here.

 

As you can see, these are clearly “railroading” versus “sandbox” designs, with one intermediate design between them. This was long before gamers started to use the terms I have been discussing in the senses I have described. Rolston’s advice is devoid of any Forge-style judgment about which is best and which is worst. These are just three descriptive types, and each is good for something.

Rolston advised that the choice between these types of adventures depended on how much work you, as GM, were willing to do before running your game. He wrote,

If you want an adventure you can use with a minimum of gamemaster prep time, you will want a package with a linear narrative design. If you are willing to do some work, but want to add the detail and imagination of professional writers to your own production, and in the process, save yourself some time and labor, you will be looking for multiple option adventures. If you intend to do most of the adventure design yourself, and are willing to put in a lot of time and effort into your work of art, you will want open-ended adventures.

The article assumes that GMs can identify which was which by reading published adventures. He says that, although linear adventures “are well-organized for gaming, they often read poorly, while background materials for more open-ended adventures can follow familiar non-fiction and fiction prose styles, making them more pleasurable reading.”

Rolston was also interested in a major element of scenario design in this discussion: where do you put the climax? The first two options have clear climaxes, but the open-ended scenario does not.

For Rolston in 1982, the variation between the linear and open ended was a trade-off. Scenarios that are more “defined by the designer” require less improvisation and end in a defined climax but may be less interesting. This goes along with the current idea that railroading destroys interest. By contrast, open-ended scenarios burden the Gamemaster and require improvisation and “extensive and detailed” knowledge of the game world setting to facilitate that. In today’s terms, sandboxes are more work.

In brief, this is all precisely the distinction between sandbox and railroad that gamers discuss today, but described in more neutral tones in 1982, two decades before the term railroading was developed by the Forge thinkers as one of the big sins now accepted by gamers all over the world.

In 1982, linear adventure design was a consciously known possibility conceived as possessing limited virtues complementary with those of an open-ended scenario.

If you read what Rolston wrote carefully, though, he makes a final point that reveals what he meant by a linear adventure. This is not what I expected.

The natural development of tastes runs from the linear adventures to more open-ended, campaign games, but I recommend a dose of dungeon-crawling when the burden of game-master prep causes waning enthusiasm in a large campaign. Note that dungeons lend themselves well to linear adventures, while cities almost demand an open-ended style, with the attendant gamemaster preparation responsibilities.

For Rolston, a dungeon-crawl was a linear adventure. Today’s gamers who advocate for sandbox-style play tend to see dungeon crawls as the basis for a sandbox. For Rolston, the restriction of choices entailed in a dungeon crawl--where you can turn left or right and seek to go deeper in search of greater treasure or retreat--effectively rendered it a linear kind of scenario.

He seems to have viewed the dungeon crawl as a way for a GM to recharge, a scenario requiring less preparation.

From that point of view, today’s sandbox dungeons are a contradiction in types. Today, megadungeons may seem like sandboxes because you have many of avenues to explore, especially if you follow the admonitions of bloggers who want you to “jacquay” your dungeon. (This just means that there should be multiple points of entrance and egress in different sections and floors of your dungeon, i.e. non-linear dungeon layouts.) Moreover, the fad of using random tables for every adjective and noun in your encounters (if anybody actually uses them that way) may relieve the burden of GM prep, at least in theory, enabling a more sandbox-like dungeon.

Conclusions

It’s interesting to me that linear adventures got a bad name basically because of the Forge. I wonder whether the good old-fashioned Quest has suffered on the same account. Stories in which characters have a mission with a goal at which they succeed or fail and few avenues by which to arrive at the solution may be much less common now. Was anything lost?

It is apparent that those who decry railroading apparently have a lot in common, philosophically speaking, with “Story Game” designers.

The “sandbox” ideal, by contrast, is a term popularized by video games. Video games leave nowhere nearly as much choice to players as table-top role-playing does, so they had to advertise themselves as sandbox worlds. I would guess that even a linear dungeon adventure in a table-top role-playing game is much more of a sandbox than any video game. Then again, I do not play video games, so I would not really know.

It’s also interesting to me how little these current ideas are really new. Gamers have been thinking seriously about adventure design since the earliest days of gaming. Having fun is engrossing. Gamers tend to come up with similar solutions to the same problems over and over, but internet groups of gamers become polarized over key terms even when they agree on pretty much everything deep down inside.

I’ll have more on the theme of early role-playing game theory in future entries.