Friday, March 30, 2012

Another Inevitable Post on Railroading

I talked a little last week about the idea of sandboxes and finales. In reading through some of the excellent comments to the post, I began to wonder about the terms I was using and the values associated with them. For example, I think different people may have a different sense of what’s meant by a sandbox game. Some might see it as completely without form to start- with the players supplying drive and direction and the GM responding entirely to that. On the other hand, a sandbox game might be defined as open choice in a setting- where the players have the freedom to explore the setting outside of a central plot or story.

An interesting parallel experience comes from one of the designers for the PS3 game, Flower, essentially an exploration game. Jenova Chen in an interview on Gamasutra says,

“The game was designed to be an open world at the beginning. Then, people said, "What am I supposed to do after the first 20 minutes of straight awesomeness of enjoying the nature?" And then they all end up saying, "Oh, there's no purpose. I don't know where to go.”… So, we kind of have to figure out a solution there. And we iterated it to more like a semi-open world and semi-linear structure….You have to kind of trigger these flowers in sort of an order, and then we see players just wander off into the distance to the very end of the level, and there's nothing there. They were just all confused. So we actually had to kind of block them before they could move on to the next area. It's almost like we wanted to throw away the traditional game design, but we end up picking up all the pieces we threw away and putting them back because we know those are actually needed to deliver a good guided experience.”

I’d consider most of my games open- I try to supply a setting and incidents for the players to respond to. I set up situations, but how the players decide to deal with those ought to remain open. No suggestion should be discounted outside of the game or setting logic. Keith Baker talks about elements of sandbox games here. The blog Reality Refracted breaks down some of the types of campaigns, and I’d say I fall somewhere usually between Plot and Campaign. I really want my players to have Agency, and I think the Rhetorical Gamer really lays out the concerns of agency in two great posts: here and here. One problem in doing any assessment or analysis of these kinds of gaming techniques is that I can’t really be objective: my sources are anecdotal- based on my experiences running and playing. Given that gamers likely play with the same groups most often, certain patterns or beliefs may arise in there. My group wants to discover the story the GM has planned out, and wants to help get that to the table. But they don’t want to play a module. My gaming groups generally hold to those values, though individuals may have different interpretations.  

Related to that, one of the splits I see is that between Improvisation and Railroading. There’s some suggestion that improv-approach games reduce the problems of rails or offer a better power balance between players & GM. I’ll admit to being a pretty freeform GM- I write up notes and sketch ideas, but I’m also pretty loose about how important that ends up being at the table. I rarely stat encounters or enemies, unless I really want a big scene or I’m pretty sure the players have it in mind to take on the opposition in question. I’m a fan of Play Unsafe and his mentor Impro, but I’m less comfortable with fully improvised sessions. Beedo, at Dreams in the Lich House, considers the levels of improvisation players might be comfortable with, in the context of the excellent Armitage Files. Some people have suggested that in practice we end up with a continuum of practices running from Railroad (Control) to Improvisation (Freedom).

What do I think the far poles are? On the Railroading side, I think it takes two forms: active and passive. Active Railroading has the GM making declarations about what the players can and cannot do. Doors cannot be opened, obstacles cannot be passed, monsters may not be defeated- regardless of rules, character abilities and rolls. The Rhetorical Gamer’s story about the GM berating the group for taking actions undercut the GM’s set-up is a classic example. My favorite incident occurred in an Exalted game, when the GM declared to a player, “The giant tree reaches out and grabs you.” The player naturally said she wanted to dodge. The GM repeated that she’d been grabbed. The player pointed out the various skills and charms she had available to avoid that. Angrily, the GM told her to roll- at which point she obtained 10 successes, an EPIC and ABSURD number. “OK, he grabs you and knocks you out,” replied the GM. At which point I think the rest of the table burst out laughing at him.

Passive Railroading, on the other hand, is more common. It can take many forms. A GM might excise all other possible options, not bothering with any attention or details aside from the plot and story they want to play. Any questions or exploration loop back to that. More sinisterly, the GM may simply ignore player inputs and “voting” in favor of the story they have in mind. They don’t actively negate, but ignore results or shift them in a way antithetical to the intent. We were playing in a game where we weren’t quite sure of the direction or plot- but we’d at least identified some of the bad guys. We ended up meeting a lieutenant of the BBEG, a follower out of family loyalty and nationalism. The group decided that interacting with and convincing this person of the plots afoot would be the best approach. So we spent a good part of the session interacting with the NPC, establishing ties, building up connections. One of the PCs moved to open a potential romantic interest with the NPC. We liked the character and could see how this would move us towards our goals and provide new opportunities. At the end of the session, the GM essentially walked the NPC off-stage, never to return. Our attempts to contact and follow up with the NPC were stymied. Instead, many sessions later at the climax, the character showed up- in the service of the bad guy but now “mind controlled” to explain his siding with them.


Of course one of the problems with defining a rails vs. improv continuum is that these positions can shift depending on what’s happening at the table. A GM might be improvising a scene, but still be driving players in one direction. Or a GM who has their plot worked out down to the beats might freeform a fight.

I have some questions and they arise from interesting discussions like that of Casting Shadows here and the negativity towards some kinds of story presented by Monsters and Manuals here.

At what point does GM input become railroading? At what point is a GM imposing story? Is it in limiting choices? In placing obstacles? In forcing conflicts? In setting up dilemmas and tough choices? In setting up incidents? And do different players see the “imposition” of these details differently?

On the flip side, as suggested in the opening quote- shouldn’t a GM be obligated to provide some structure to the players? How much is too much? What elements of drama and story-telling right fit in-- rising action, act structure, reversals, inner vs. outer conflict, repetition for emphasis, controlling ideas, symbols-- and which deliver “railroaded narrative”?


I’m thinking about these questions because I want to try to find the right balance between control and freedom. I see a value in a heavily improvised approach- with the idea of an emergent story. But I also see value in active story-crafting by the GM. I have stories and ideas I think the players might enjoy playing around with. If I didn’t, I wouldn’t be a GM. But those stories aren’t complete. Instead they’re fragments that can get filled in at the table or ignored if the players choose another, better path. But I think story isn’t a dirty word. What bothers me about the “Hickman” model mentioned above is that in its extreme forms of play you end up with one of two static approaches. On the one hand, you get a story where your characters don’t matter- any could be slotted in. The PC natures and backgrounds don’t impact it. On the other, you have stories where the players don’t matter. The characters have been created and fulfill roles, but the actual choices made by them have little or no impact on events (i.e. Dragonlance).

I’ve also played in games without signposts. Some of them have been interesting in the level of freedom allowed the players. But more often, the group’s been lost. Perhaps traditional narrative has served as a crutch for us too long, so that when we find that space we can’t cope. But sometimes it has felt like the GM’s given up or wants to show off a mastery of the form. As in the Flower example, we walked to the end of the level, but still remained confused.

Which means either extreme fails for our group. I believe a great game requires a blending of control and freedom. Despite the dangers, I believe that story offers benefits in play. Story-telling techniques- requiring GM control- offer benefits at the table. But those benefits only arise when tempered with free and meaningful choices for the players. What I really want to figure out, what I try to figure out in every game I run, is how to do that. 

Incidentally, since I mentioned it in my last post, this weekend I got to hear an angry rant by one of our group who loves Mass Effect, but hated the ending. We talked about it for some time and his major complaint was SPOILERS that all of the endings were the same. With the curtain pulled back, he hated the sense of powerlessness that created for him (the player) vs. him (the character). The Hopeless Gamer has an ending idea, though I'm not sure it would exactly solve that problem, but it is worth reading if you're interested in the game.


  1. It seems to be a balancing act between the two. I try to keep all avenues open to the players/characters to do whatever they want, yet as you mention there needs to be some structure in place, too. Something specific for the PCs to do, a direction. Otherwise there could be a lack of focus or "analysis paralysis".

    Something else to take into account is that the GM is a player as well. She should have the chance to use the GM tools that are available to make things happen in the world. Rather than just reacting to the PCs, the GM can and should be able to shake things up, or at the very least return the game to some sort of focus.

  2. Awhile back, I was struggling with my word choice when it came to this topic. One of my players insisted I knew what was going to happen next, while I insisted that I was just as much in the dark about where they were headed.

    It came down to semantics. You mention "story" not being a dirty word, but similar to the word "plot" it makes me wince when used in the context of a role-playing game. Yes, after events unfold, there is a story to tell. I even enjoy retelling them after the fact.

    During play, yes, I have an idea how a NPC will react if the players do such and such. Yes, things happen in the world around them. I know if that door is locked and where the key is located. Yet, the world doesn't revolve around the players. They aren't special. The odds are against them.

    A generic example: a town needs saving.

    You ask, at what point does DM input become railroading, impose a story, limit choice, etc?

    In the generic example, if the town is not saved, the town is destroyed and the people are taken as slaves or some other variant.

    Actions (or lack of action) should have consequences, right?

    Those consequences don't necessarily punish the players. It really wasn't their problem in the first place.

    A minor tangent would be the players burning down the town, but that's another matter.

    It's a slippery slope when you make it their problem. If the players at the table don't care, why try and convince them that they do?

    "Oh, by the way, your parents live in the town."

    Sticking with the example . . .

    So you spent all night writing up these awesome encounters for a town being attacked by goblins. You've even taken the time to map out the goblin cave and link them to possible other factions that are involved. The players show up and say, "Screw that. Let's go buy a ship and sail to the west. I want to visit the city of Greyhawk."

    It becomes railroading when the guy selling the ships says the ships were all burned by a recent goblin raid and his men taken as slaves. You know, the same group that attacked a nearby village where their parents were murdered.

    The players want to go to Greyhawk. Why stop them?

    As I mentioned to Runeslinger on his latest Shadowcast, let the players play. They don't care about your goblins.

    Personally, I would find it extremely boring as a Dungeon Master if I knew what was going to happen next. I'd rather be excited with the players about this voyage across the sea, because a lot of fun stuff could happen on that trip. Why not go on the journey with them and find out?

    You mention that using story telling techniques provides benefits at the table. Who is it benefiting? Without any examples, I am hesitant to comment much more.
    However, telling a story to the players versus creating a story with the players are two different things.

    Does the group want to play a game or tell a story?

  3. Well I've been a player and GM in both sides, railroading and total improve, and had fun both times. I think a lot of 'how much is to much' also has a lot to do with both the group you have and the game you are running. Group wise is pretty self explanatory, some folks just like a certain style. But game wise it can be different. For example I could set down and run a completely off the cuff game of Vampire:TM or Little Fears at the drop of a hat. But others like Sla Industries and Cyberpunk 2020 I fair better with more ahead of time planning (but then again those two are more 'mission oriented' type of games).

    I've also done railroading up to a point. When I used to run a COPS game regularly this was pretty much a given. The character where police officers on patrol and it was rather simple to get them to the point of action. They get told to go do something or investigate a distrubance. Once they get there... well then it goes off into the other direction of off the cuff as they deal with it how they want to. So I believe you can have both in a game but at some point it switches over between the two.

    Just my two cents.

  4. @Peter I would not consider that railroading, assuming the players wanted to play cops. They aren't being forced to do something they did not want. It follows they will get calls and need to investigate. Sure, they could ignore the chief and go GTA on the town, but why waste everyone's time rolling up cops if you're going to do that?

    And I agree in regards to group dynamics. As with anything, it matters what your group enjoys.

  5. @Peter:agree with burnedfx on that point. Providing players agree to the framework of a game, you aren't unduly limiting choice.

    Nice post! I particularly like the section on railroading passively.