Saturday, April 27, 2013

Narrative Structure vs. the Sandbox

I'm thinking about sandboxing lately. Mostly I'm trying to find out how deep I could bury those bones, but that's for another post or two. This is about something else.

Having a sandboxy game is not as easy as reading blogs makes one want to believe. In reality, without reading blogs and working on the game in one way or another, it is hard for a player to "get" the OSR and the ideas (or philosophies, if you like) that keep evolving in our little neck of the woods. Today I want to write about some things that occured to me during our last two sessions.

Railroading is appreciated... (?)

Yes, really, I've been told so by one of my players. He said something like: "Just point us in the right direction and we'll do what needs to be done!" How did that happen, one might ask. Well, I gave them a sandbox, enough hints about the dangers and some rumours to complete the package. They were free to do whatever they wanted and go wherever they wanted. I was content with the results. And I was eager to give this way of playing the game a shot. Their desperate need for quests (or better yet, a script) was the reason for this not to work. They were intimidated by the possibilities.

It reminded me of a story (urban legend? quote? could have been a joke...) I'd heard somewhere some time ago. It was about someone commenting the fall of communism with: "Do we have to do now what we want?!"

Anyway. This shouldn't be a surprise in our consumption oriented society, but I really believed they wanted to have all those possibilities. No dice. They wanted to be entertained. At least some of them want to know the direction and the possible outcome (that is: who's the enemy and how much xp could be gathered).

It's no ones fault, to be honest. All of them have jobs and lifes to take care of and if you only play once a month, it's hard to connect with a game. Railroading is an efficient way to enjoy the tropes and fun the game offers, without the need to invest too much. And as a DM it's sometimes hard to understand the effort needed to get involved in a game.

We'll try now to play for 2 to 3 hours a week via g+ or skype between games. Hopefully it will close the gap between sessions and help the game. The way we play right now, giving the game a very strict narrative structure (that is: a railroad) seems to be the only solution.

You have to improvise eventually

What is a sandbox able to accomplish? Ideally it produces a world engine that substitutes arbitrary DM decisions with a randomized resolution system and generates a story without, in theory, needing a narrative structure per se. But is that true? I believe it is to some extend. It makes sense to use reaction tables or  setting-specific random encounters. And having a complete and reactive setting at hand is one of the nicest tools a DM could have.

There is a point, though, where a DM has to be more precise than that. It is possible to create a totally random encounter, up to gender, race, age and adventure seed. But a DM connects the dots and the narrative structure becomes important at some point. Essentially, it is always the exact point when he starts talking to his players.

That's not railroading, it's just the need in every kind of communication to create a coherent narrative. A DM decides where he starts to tell the story, but sandbox or not, he needs to frame what is happening to make it a consistent experience for the players. I read somewhere that playing it sandbox-style results in a picaresque narrative, but I'd argue that this is true for playing roleplaying games anyway (just going by the definition of the term picaresque novel and the nature of roleplaying games in general).

Truth is always a matter of perspective

In the end I came to believe that sandbox is an artificial construct and not very helpful at that. The same goes for the infamous railroad. I'd have to call disbelieve on both, as far as playing the game is concerned. They are legit when discussing how a DM prepares his game/setting/etc. (because it needs terms to describe different methods and all that), but when playing the game, narrative structure is always imperative. How to get to the point where the people start talking to each other in the game is either unimportant (in the game) or, in a worst case scenario, distracting.

There is no true railroad, because every DM needs to improvise just by using a language and interpreting the rules. There is no true sandbox, because a narrative structure will evolve from every game. Funny thing is, in a players perception there might not even be a difference.


  1. "There is no true railroad, because every DM needs to improvise just by using a language and interpreting the rules. There is no true sandbox, because a narrative structure will evolve from every game."

    I have to strongly disagree. The WFRP Bogenhaffen adventure I've been playing in yesterday has had all the classic hallmarks of the "infamous railroad".

    We've gotten stuck several times in solving the mystery and since the DM came to the table with a foregone conclusion to the adventure, he was forced to just throw some clue in our faces. I hate that. Yes, we got to the climactic end-battle, but I don't feel like I earned it.

    Similarly, when I would have a good idea for how to solve the mystery, the DM would have to make up reasons why it doesn't work since that way of solving it isn't in the book.

    My problem with a Railroad is that the decisions my character makes has no bearing on the game-world, since either way the result will be the same.

  2. Thanks! And you're right, of course, in seeing it like that. My definition of the term "railroad" might have been a bit lax. I know the adventure you write about. Have been a player in it for 2 sessions or so with a DM that had a similar problem and the same frustration you describe as a result. But I thought that to be pixel bitching, not necessarily railroading. Mostly because leading a "railraod" into a dead end seems counter-intuitive to me. What I describe in the post (the player wanting (!) to be railroaded) led me to believe there is a more, well, positive way of seeing this. Sometimes the lack of free will is appreciated in a game just because it mimics what the game should be like, without the need to invest too much. That might be sad, but definitely not wrong. I really like your attitude towards players leaving a mark in a world because of their decisions and I'd like my players to see it the same way. But it's a process and if or if not limiting choices is a bad thing, depends strongly on how much involvement is wanted by the players.

    And "using language" as well as "interpreting the rules" both should produce flexibility in a DM worth his salt, right? Playing D&D RAW provides enough rules to allow derivation in scripted play.

    But now I have to wonder if I understood the term "railroading" to literally and if there is another term for what I'm trying to describe...

    1. Ok, I think I understand what you mean that we're referring to different things when we describe a "railroad".

      If I understand correctly, you are referring to how you start out the adventure. Do you:
      a) just present a scenario to be played(railroad)
      b) place the characters in a location with lots of options and let them decide(sandbox)

      I guess I would say that's Sandbox vs. Railroad at the Campaign level(picking which adventure to do), whereas my example was at the Adventure Level(what do you do in the actual adventure).

      I agree that having Sandboxyness at the Campaign Level is less important than at the Adventure level.

      When I ran the first session of New Tilia ( I had a similar problem to you. We started right out with rumour generation, but the players were a bit overwhelmed by the number of options. I told them that regardless of which rumour they followed, they were likely to have something interesting happen, which helped. They then debated and eventually decided which rumours to follow, but it took a full 2 sessions to arrive at an actual "adventure" locale(not the one they planned by the way--they got distracted by a random encounter).

      On the other hand, while they were wandering around they got to know the game world quite a bit. Oh and they probably caused a regional war(they won't find this out till they get back to town), so having PCs wander around the game world between adventures, killing things they shouldn't be can make things quite interesting :)

  3. Yes, I think it is a good idea to distinguish between campaign and adventure level and that's pretty much what I had in mind.

    Thanks for the link. It's a good read. And maybe it might help to illustrate my argument a bit further. I'm not that familiar with the WFRP-system, but at what point did the treant decide to pay attention to the adventurers? Was it a random occurrence or did it happen because they were in the area? At what point did you decide what happens next and how to describe it? And how did the manufacturing part of that have an impact on what happened at the table?

    The fact that you're able to tell a somewhat coherent story about what happened, was what I was aiming for. Narrative structure will arise from both ways of playing the game. And what a DM narrates is, to some degree, independent from how he gathered those informations. Railroad (for lack of a better term) or sandbox are both "just" ways to prepare and handle the game, but for playing the game a DM needs to translate those results into a more or less cohesive interactive story. Done right (or without them knowing), it's nearly impossible for the players to tell the difference.

    That's not saying the sandbox-approach is not superior to railroading the players. I think it is the way to go (as your post illustrates very well). But to get there and to get the players (with their different likes and dislikes) to accept it is a different issue altogether.

    And I really like the fact that destroying a big critter will have an impact on a regions power balance. Very nice touch :)

    1. "And I really like the fact that destroying a big critter will have an impact on a regions power balance."

      That's largely due to "Renegade Crowns". It encourages the DM to think about the relationships/politics between different groups while building the map. Why don't those Goblins up North just destroy the humans to the South. Well, they generally don't go South because of the Treant in the South. Once you create this "Political Web" between parties, then as soon as one disappears it follows naturally that you have a power vacuum and regional conflict.

  4. So I created the New Tilia setting with "Renegade Crowns", a WFRP product for creating your own principality.

    You randomly create settlement descriptions and you roll for "interesting features". So the party was near a homestead whose feature was this Treant that protects the area, mostly from the Greenskins to the North.

    I made a successful random encounter check and my next roll determined that it was a monster(my first roll for random encounters is 1-monsters/2-people/3-animals). So rather than rolling on the random monster table, I decided it was the Treant since the Treant keeps most other monsters away from this area.

  5. Thanks for replying in such detail! It really helped clarifying some arguments :)

    Renegade Crowns sounds like quite the product. I've read about it and only stellar reviews. Maybe I need to take a closer look...

    1. It's somwhat WFRP specific, but it's pretty easy to adapt to another system, and in general it's a good education about how to build a sandbox.

      It has you randomly generate the map, settlements, resources. But it also helps you create the politics of the region, which is just as important for creating a living, breathing game-world, rather than just a collection of loosely-connected adventure modules.

      The one thing missing from Renegade Crowns is turning the abstract descriptions of Adventure Locales into detailed adventure locales. So like for the Locked House(Sessions 3-5 I decided that the house would be based on "House on the Borderlands" and for the lower level I took a map from someone else's blog and re-keyed it.

      Great post/discussion--cheers!

  6. I tend to believe that railroad vs. sandbox has more to do with how the GM presents the game than how it's published. The most railroady written adventure can fail to be a railroad if the GM simply doesn't run it that way, and vice versa.

    Because gaming as an idiom bears some resemblances to storytelling as an idiom (not trying to get into the semantic wrangling associated with "storygames"--whatever the heck is meant by that term) to me it makes sense to borrow some conventions and practices from successful storytelling idioms, like television, movies and novels. Things like pacing, the utilization of the session as a dramatic chunk with it's own ebb and flow of drama, tension and suspense, etc. is just good GMing. Any GM that fails to utilize these techniques is not GMing a roleplaying game, he's just facilitating whatever the players are doing.

    But that doesn't invalidate a lot of the benefits of sandbox theory. Any game in which the players don't feel that they can go do something else, or do something in a different way than expected, is one where they can also legitimately claim to be stuck on a railroad, and nobody tends to like that.

    Where I think a lot of sandbox theorists fall flat is that they fail to recognize that players don't feel a sense of connection to the setting OR their characters immediately. In that situation, having a bunch of options open to them is meaningless--because they don't CARE what happens to either their characters or the setting. It's all too abstract to be meaningful.

    In my experience, the way to successfully run a "sandbox" game is to ease the players into it. Start the game off not by letting them "wander around wondering where the game is" but with some immediacy to things that need to be dealt. This kind of in media res "railroading" then gives way to a more open field where the players can start making most of their own decisions about priorities, and what they want to pursue and deal with--but after they've already had a chance to get a solid handle on both their characters, the setting, the interaction between the various players and characters in the group, etc.

    And unless you want the game to kind of trail away into a dissatisying fade-out, I also firmly believe that a firmer hand in gathering loose threads and pulling them together is the GM's responsibility at the end of a campaign as well.

  7. Thank you, very well put. And I agree wholeheartedly.


Recent developments made it necessary to moderate posts again. Sorry about that, folks.