Judgement calls vs "railroading"

[Tony Vargas]

If we want to imagine some parallel thing that might be an RPG, we've got to ask questions like "Who decided to make the ring the focus of the game" (as per my post 87), "How was it determined that Shelob stung Frodo", "How was it determined that Frodo lived or died, and who knew that at what moment of play?", "What parameters governed Sam's picking up of the ring?", etc.

We can't tell anything about railroading or RPGing techniques just from being told a story of what happened.

I suspect we might. If the story is full of plot-inconvenient screw-ups, dead ends, anti-climaxes, and just ends abruptly with the death of everyone you thought might have been a protagonist, or maybe sort of trails off ambiguously, chances are it was a 'sandbox' run by a DM 'by the book' with dice in full view, world/NPCs/&c set before it started, and no tweaks or adjustments, illusionism nor DM-force.

If it was a cool story (or even a cliched story), with plot, pacing, a climax and denouement, chances are it was a total 'railroad' or 'illusion' (or maybe some sort of collective-storytelling exercise). Or, y'know, an actual story written by someone who remotely knew what they were doing.

'Truth is stranger than fiction, because fiction has to make sense,' and all that.

 

[Jester David]

This seems to rest on a distinciton between "adventure" and "scene" which I think is unstable. What individuates adventures?

I'm thinking in narrative terms. Very broadly, because you can't determine "railroad" or "sandbox" from a single encounter or moment. What happens in an encounter or scene is a single data point. There's no context. No frame or reference. You don't know if it's the norm or an anomaly. You have no scale.

For example, something like a villain escaping will occur more in railroads. They're a scripted scene. But it's not impossible that a villain is just in a position to flee and does. The DM is just taking advantage of positioning.

To me, it seems that the existence of something called []the DM's plot[/I], which has certain needs/requirements, entails that there is only one story.

The catch being there is *always* a plot. Even in a pure sandbox there is one story, albeit one scripted entirely by the players' actions. It's just a matter of perspective and how much it was planned.
Similarly, a DM can have a hard plot going on (a villain, a motivation, scripted events, etc) *and* the story can still not be a railroad provided the players have a choice whether or not to involve themselves in the plot, how to deal with the problems, etc.
Skyrim and Fallout 3 and 4 are very much sandbox games but they still have a "plot". D&D is similar but with a twist, as the plot can change and advance in the background.

What determines if a campaign is a railroad or not is if the players can deviate from the plot or ignore the presented plot's existence.

This is why I don't feel the force of your scene/adventure distinction. When you talk about the GM making a judgement call because that's what the plot needs, you seem to me to be talking about a railroad. There is a pre-authored outcome (or series of outcomes), and the GM is adjudicating in such a way as to bring that about.

"Pre-authored" and "improvised" are really irrelevant to this discussion. In either instance the Dungeon Master is making a decision. Whether or not they do so before the session begins, during the session but before that scene, during the scene, or retconning after the fact is all the same. Either way the Dungeon Master is making that decision.

If the DM is making the decisions because it's what the plot requires, that is generally a sign of a railroad. But, again, single datapoint.
DMs often make decisions on what does or does not exist based on plot and narrative. That's literally their job at the table. That doesn't mean the plot is a railroad, but could instead be what makes the most narrative or dramatic sense. If there's a good couple hours left in the session, the DM might make a call that extends the adventure a little longer. If they're nearing the end of the session, then the decision becomes one that wraps up that story so the next session can be focused on what happens next OR serves as a cliffhanger so the next session is anticipated. Similarly, if things have been going smoothly, the plot might require a complication to create tension. None of the above necessarily denote a railroad. They just denote good storytelling.

 

[robus]

This seems to rest on a distinciton between "adventure" and "scene" which I think is unstable. What individuates adventures?

Did see the AngryGM's break down of the major components in RPGs? Here's a relevant sample:

Scenes and Encounters

Actions happen within the context of scenes and encounters. Scenes and encounters begin with a dramatic question, a statement of the character’s goals in the form of a question. The GM then presents the scene to the players, usually in some way hinting at the dramatic question or goal. If the answer to the dramatic question is uncertain because of a source of conflict within the scene, the scene is an encounter. The scene or encounter plays out as a series of actions until the dramatic question is resolved. Then, the GM provides a transition to the next scene or encounter.

Adventures

An adventure is a series of scenes and encounters that comprise a single, complete story. Adventures begin with a motivation or goal and a scene that presents that motivation or goal to the players. Adventures then play out as a series of scenes or encounters as the players pursue the goal. Once the players have accomplished the goal or definitively failed to accomplish the goal, the adventure ends.

 

[Lanefan]

If it's unavoidable, it's a railroad. They're trucking along on rails with none of their choices able to take them off of that track. You are obstructing choices by the way. Each and every choice that would allow them to save the Marquis is obstructed, as are all choices that could bring him back.

If the PCs have no way of knowing an assassination is coming until it happens, they can't exactly be expected to do anything about it. The only difference between the Marquis getting killed now as opposed to last night is the dramatic effect of having the PCs see it happen.

For my game, I'd allow the possibility of saving him if they could come up with an idea that would work. The plot would still move forward with an attempt on his life. It would just move forward a bit differently. Instead of trying to solve a murder, the PCs would be trying to solve an attempted murder. Perhaps with the resources of the Marquis if he and the PCs could get along. The group making the attempt might or might not plan another go at it. Maybe they go get more powerful help.

[MENTION=6801286]Imaculata[/MENTION] can correct me if I'm wrong, but my guess is that your plot would immediately become different from his. We don't yet know why the Marquis had to die (maybe he's supposed to reappear as an undead somewhere down the road); but your plot says he doesn't have to die. Different plot.

It isn't that the Marquis had to die to move the plot forward. It's that he had to die so that the plot could move forward the way YOU wanted it to go. That's sticking the PCs on rails.

No, it's sticking the world's backstory on rails...which is right where it should be. As I said above, the only difference is that this particular bit of backstory takes place in full view of the PCs. It's a more up-close-and-personal equivalent to my idea of the meteorite crashing into the mountain...the PCs can't prevent it, they can only react to it after the fact.

And that's the crux: your plot makes the PCs proactive in a situation where they're supposed to be reactive.

Lan-"if one train leaves Waterdeep at 3:35 and another leaves Neverwinter at 4:10, at what time does the Marquis die?"-efan

 

[Imaculata]

@Imaculata can correct me if I'm wrong, but my guess is that your plot would immediately become different from his. We don't yet know why the Marquis had to die (maybe he's supposed to reappear as an undead somewhere down the road); but your plot says he doesn't have to die. Different plot.

Exactly. Its the equivalent of letting the players prevent the red wedding. Sometimes a plot needs to go in a specific direction, and that has more to do with storytelling, than with railroading.

The players are in a reactive position, rather than one where they can prevent the event from happening completely.

I don't feel that a DM is taking choice away from the players, if he decides that a specific thing happens in the campaign. It's only railroading in my opinion, if the DM is obstructing what the players are trying to do, in a disruptive way, because he is unwilling to alter the course of his plot.

DMs often make decisions on what does or does not exist based on plot and narrative. That's literally their job at the table. That doesn't mean the plot is a railroad, but could instead be what makes the most narrative or dramatic sense. If there's a good couple hours left in the session, the DM might make a call that extends the adventure a little longer. If they're nearing the end of the session, then the decision becomes one that wraps up that story so the next session can be focused on what happens next OR serves as a cliffhanger so the next session is anticipated. Similarly, if things have been going smoothly, the plot might require a complication to create tension. None of the above necessarily denote a railroad. They just denote good storytelling.

I agree with this as well. Often a DM will decide that something dramatic happens, to end the session on a cliffhanger.

As a DM, I try to think of what is most interesting narratively, and what makes the most sense given the situation. I may decide that an unexpected character suddenly enters the inn, because it adds more drama to the scene. I may decide that a specific npc loses his life, because it makes sense, and because it adds the drama and stakes that the campaign needs at that particular moment. Sometimes I want my players to just react to what is happening, like when my players arrived at a harbor town, and saw smoke rising from it. Something had happened in their absence, as a consequence of choices they had made earlier in the campaign. Something bad had happened while they were away, and now they could only react to it... or ignore it.

Scripting a scene, is not the same as railroading. Obstructing your players is railroading. Forcing your players to go fight the dragon, is railroading. But having the dragon suddenly attack a nearby village, that is not railroading, -even if you had been planning it for weeks.

 

[Campbell]

The medium impacts storytelling techniques. With the Marquis example, it depends on a couple of factors for me. Under the principles I operate under the most critical consideration is how the players see the NPC. Is he our guy or your guy? If the players have invested significant effort to make him an ally, and their conception of their characters is rolled up in the relationship killing him off summarily without a chance to do anything about it feels like a cheap shot. If they are not emotionally invested, screw him. He dies to prove a point. If they are you want to make whether he lives or dies the result of player action or inaction. It will feel more impactful if their decisions had an impact. Being a fan of the players' characters means you don't just take away the things they have earned through play lightly.

Apocalypse World has this concept of hard and soft moves that I think applies to this situation. A hard move has finality of resolution. Something happens that changes things forever. A soft move serves to threaten, change things momentarily, or reveal possible consequences. Generally it is a good idea to use soft moves first. It's a way to apply foreshadowing in a game where we cannot know what will happen. It gives players a chance to think about their actions, and emotionally prepare for those hard moves. Soft moves make the hard moves feel right. It is also important that the hard move actually comes to pass. Threaten, then deliver.

For me, this comes down to bleed, the idea that player and character emotions influence each other. You are playing with people's emotions here and should have enough respect for the other players to treat that responsibility very seriously.

 

[pemerton]

The medium impacts storytelling techniques. With the Marquis example, it depends on a couple of factors for me. Under the principles I operate under the most critical consideration is how the players see the NPC. Is he our guy or your guy? If the players have invested significant effort to make him an ally, and their conception of their characters is rolled up in the relationship killing him off summarily without a chance to do anything about it feels like a cheap shot. If they are not emotionally invested, screw him. He dies to prove a point. If they are you want to make whether he lives or dies the result of player action or inaction.

Absolutely agreed.

Not all RPGing involves story (eg playing through White Plume Mountain in the classic style). But when it does involve story, there can be no consideration of how an event of significance to the story might come about without thinking about that in the context of the investment and orientation of the participants. This is the basic idea of "no failure offscreen" - and what counts as a failure for the PCs is obviously dependent primarily on the players'/I] conception(s) of the situation, not the GM.

 

[pemerton]

in any given moment all of the players (including the GM if there is one) are both participants and audience. That is why playing to find out is so important to me. I have a genuine interest in not knowing how things will shake out.

Definitely this.

The session I talked about in the OP opened with the PCs and the NPC mage whose tower they were in bursting into the bedroom just as the assassin decapited the unconscious, recuperating mage. This was established from the close of the previous session.

So we all knew that the session would, in some fashion, involve this confrontation in the tower. But the stuff that actually happened: the collecting of the blood; the escape of the two PCs from the tower, carrying the decapitated body and also the body of the unconscious assassin (who collapsed from the strain of trying to cast a spell to escape); the apprehension of the PCs by the night watch as they fled across town with their bodies and blood; the details of their (unsuccessful) attempts to talk their way out of that apprehension ("It's OK! I've got the head!"); none of these was known or even knowable before we actually played the game.

Next session I expect we will learn some things, like what happens to the bodies, the head and the blood. But no doubt new things will happen which take the game in a direction that was not known, and not knowable. (By definition, as it were, I can't usefully speculate about them in advavnce.)

If the story is full of plot-inconvenient screw-ups, dead ends, anti-climaxes, and just ends abruptly with the death of everyone you thought might have been a protagonist, or maybe sort of trails off ambiguously, chances are it was a 'sandbox' run by a DM 'by the book' with dice in full view, world/NPCs/&c set before it started, and no tweaks or adjustments, illusionism nor DM-force.

If it was a cool story (or even a cliched story), with plot, pacing, a climax and denouement, chances are it was a total 'railroad' or 'illusion' (or maybe some sort of collective-storytelling exercise). Or, y'know, an actual story written by someone who remotely knew what they were doing.

Well this is pretty much the crux of it, isn't it?

If the only two ways to do RPGing were sandbox or railroad, then what you say would be correct. But they're not. We know they're not because we have actual, concrete examples of RPGs that are intended to be played in neither of those ways. I have further additional evidence, in that I've actually GMed RPGs (both the "modern" ones I just alluded to, and more "trad" ones drifted) in ways that are neither sandbox nor illusion/force/railroading, and the result has been cool (if cliched) story.

(The possible exception being denouement. Like other forms of serial fiction, I tend to find episodic RPGs a bit light in the denouement department. I think you have to depart more radically from traditional models than I have ever done to get that. I have done denouement via campaign-wrap-up-narration, having been inspired by the Nicotine Girls idea of the PC getting her dream, but that's not really RPGing - it's free storytelling between GM and players.)

To elaborate: if we unpack your examples, we see an implicit assumption that "by the book" with dice in full view means poor (or no) dramatic pacing, dead ends and random protagonist death. But none of thes assumptions need hold true. Consider the dramatic pacing inherent in a game like MHRP or 4e, with their scene-based resolution and in the first case the Doom Pool, in the second the "comeback" narrative ensured by the combat mechanics. Consider "fail forward" resolution in the context of "say 'yes' or roll the dice"-type approaches to calling for checks; consider various devices, from fate points to stake-setting practices, for modulating the risk of PC death and correlating it with dramatic stakes.

'Truth is stranger than fiction, because fiction has to make sense,' and all that.

Again, this seems to assume that all systems are, more-or-less, process sim, modelling the ingame causal processes ("truth") and hence that to get story (ie "fiction") we need the GM to use force at various points to suspend or override the mechanics.

But that apparent premise about system is just false. It's now 25 years since the first publication of Over the Edge, and nearly 35 years since the James Bond game was published.

Which takes me back to . . .

As a participant I have some stuff to say and expect what I say to really matter

A lot of discussion about railroading and the like focuses on what the GM will or will not permit in terms of action declarations. But that's to focus on only one small part of the overall picture, which I tried to get at upthread with my reference to the negation of player choices.

If the stuff a player has to say is going to matter, then it helps if there are devices that support that: eg mechanics that allows players to try harder when the stuff that really matters is at stake; and principles around the finality of results, such that if a player succeeds on some check, the GM is not just at liberty to undo that outcome by free narration, or to introduce into the shared fiction some other story element that defeats the player's success.

(On the issue of trying harder/I]: one reason wizards are so popular among experienced players of classic D&D is not just their power, but the fact that they have a built-in mechanic that allows the player to try harder - by using spells, which are handily scaled in a way that roughly correlates difficulty of using to significance of impact on the ingame situation. If not much is at stake, a 1st level magic-user doesn't cast at all, and a high level one uses a low level spell or a charge from a wand. If the stakes are high, however, then the 1st level MU unleashes the Sleep spell, or Charm Person, or similar; and the high level one lets fly with 10 dice fireballs, Death Spells and the like. A big part of what distinguishes 4e from other versions of D&D is that all PCs have mechanics that enable the player to try harder when the stakes are higher.)

 

[Imaculata]

The medium impacts storytelling techniques. With the Marquis example, it depends on a couple of factors for me. Under the principles I operate under the most critical consideration is how the players see the NPC. Is he our guy or your guy? If the players have invested significant effort to make him an ally, and their conception of their characters is rolled up in the relationship killing him off summarily without a chance to do anything about it feels like a cheap shot.

In my campaign, befriending him did not take any effort by the players. He was pretty much their ally right from the start of the campaign. And he was very much my guy. His daughter on the other hand is a love interest and a close ally. His death was basically the push to make his daughter a part of their crew. A reason for her to say goodbye to her old life, and join them on their future adventures. It also added an emotional side to the story, and to her character.

If they are not emotionally invested, screw him. He dies to prove a point. If they are you want to make whether he lives or dies the result of player action or inaction.

If you kill him off and they are not emotionally invested, then what's the point? Sometimes a death has to hit home, and actually feel sad. You can't do it often, but if you do it at least once, it gives the story a bit of emotional weight. It's like killing off Gandalf at the end of Fellowship of the ring, but then not bringing him back.

Being a fan of the players' characters means you don't just take away the things they have earned through play lightly.

Of course, but in this case, this didn't feel like something they had actually earned, and it had made them feel a little bit too safe. For the story to carry some weight, it felt necessary to me to kill off a character that had been their trusted ally since the start of the campaign. This paved the way for the players to actually get politically involved in who would be his successor, with some possible bad consequences if it was the wrong person.

Apocalypse World has this concept of hard and soft moves that I think applies to this situation. A hard move has finality of resolution. Something happens that changes things forever. A soft move serves to threaten, change things momentarily, or reveal possible consequences. Generally it is a good idea to use soft moves first. It's a way to apply foreshadowing in a game where we cannot know what will happen. It gives players a chance to think about their actions, and emotionally prepare for those hard moves. Soft moves make the hard moves feel right. It is also important that the hard move actually comes to pass. Threaten, then deliver.

This I sort of agree with, but I don't feel that every hard move needs to be foreshadowed. Sometimes you want to hit your players with a punch in the guts; -with a shocking twist that came totally unexpected... and yet made sense given the story. The red wedding is a good example. You can sort of see how events were building up to it, but it still came very unexpected when it happened.

For me, this comes down to bleed, the idea that player and character emotions influence each other. You are playing with people's emotions here and should have enough respect for the other players to treat that responsibility very seriously.

Of course, but this is also why I feel that the players should care when you kill off an important npc like that. It should be a catalyst for greater things that push the story forward. The villains in my campaign had already affected the lives of other characters in the campaign, but no one they were very close to. By killing off a character who was both a close ally, and the father of a character they really cared about, this was a very powerful moment. And while the players were triumphant in defeating the villains during their attack, they also suffered a great loss, that would instigate a whole new chapter in the story.

 

[pemerton]

you can't determine "railroad" or "sandbox" from a single encounter or moment. What happens in an encounter or scene is a single data point. There's no context. No frame or reference. You don't know if it's the norm or an anomaly. You have no scale.

One can tell, from a single event of play, whether that event of play was driven by the GM or the players (assuming one has all the relevant information). And given that that is what I'm interested in when it comes to the content of a railroad, that's enough to answer my question.

(There might be other questions - is the occurence of GM force atypical? or (perhaps more likely) is allowing an episode of play to unfold in a player-driven way atypical? But these are questions that would be relevant, say, to deciding whether to join a group. They don't seem to help me analyse the play that is taking place, or how various approaches and techniques are informing that play.)

something like a villain escaping will occur more in railroads. They're a scripted scene. But it's not impossible that a villain is just in a position to flee and does. The DM is just taking advantage of positioning.

This goes back to [MENTION=16586]Campbell[/MENTION]'s comments about the assassination of the marquis. Without telling me more about what is at stake, I can't make the call. To relate it to my OP, is the villain's escape an outcome in the relevant sense? Or mere colour. If it's an outcome, then I would never just declare it (ie the GM can't "say 'yes'" to him-/herself!) The dice would have to be rolled.

Did see the AngryGM's break down of the major components in RPGs?

I'm familiar with the concepts. My point is that I don't think they can bear the explanatory load that is being put on them.

For instance, we're told that "An adventure is a series of scenes and encounters that comprise a single, complete story." But what marks a story as complete? Who gets to decide that nothing more is at stake? If it's the GM, then we're right back in the realm of GM force - so the idea of an adventure as a meaningful unit of play brings railroading with it per se; if it's the players, then when they decide that there's nothing more at stake for their PCs presumably the campaign is done.

The catch being there is *always* a plot.

<snip>

What determines if a campaign is a railroad or not is if the players can deviate from the plot or ignore the presented plot's existence.

This seems to assume that "the plot", if it is to exist at all, must be authored by the GM - and hence that there is no difference between "the plot" and "the GM's plot" - the latter being the phrase you used in your earlier post, to which I responded.

But that assumption is simply false. And not just false in an abstract or theoretical "it might be otherwise" fashion, but in the very concrete "I am currently GMing 5 campaigns - two 4e, one MHRP, one Cortex Fantasy hack and one Burning Wheel - in which there is no such thing as "the plot" in your sense.

For instance, as I wrote up here, on the weekend I GMed a session of Cortex Plus fantasy. The session involved the (fantasy viking) PCs travelling into the northern hills and mountains to learn about a curse/doom befalling their land.

How did we establish that that was why they were travelling north? Because the players determined it: one explained how there were perturbations in the northern light, a sign of trouble among the gods; another told how his PC had heard cries of anguish from the great spirits of the spirit world, and for that reason had come to the village to seek aid, despite normally being a solitary traveller; a third talked about the need to investigate the Dragon's Curse. So the core elements of the plot were settled by the players, not the GM.

The first encounter the PCs had was with the steading of the giant Loge? Is Loge a friend or foe - an ally in the quest, or another force bringing blight to the land? That wasn't known at the outset. There was no "plot" in this respect. We learned that Loge had a shaman in his steading who thought that the PCs were right to be concerned about dire portents, and who therefore urged Loge to align with them rather than eat them, because one of the PCs spent a plot point to establish that shaman as a resource. And we learned that Loge was able to be persuaded by his shaman because in the resolution of the social conflict the PCs who achieved the final victory had the d6 representing the shaman as one die in his pool.

There is no "presented plot" here, which the players might deviate from or ignore. There's a set of tropes - eg everyone knows that the land of vikings is full of giants in their halls, who sometimes can be allies but ultimately are not to be trusted; there is an overall orientation chosen by the players - a curse or blight or doom that needs to be lifted/prevented; there is a situation that engages the players - will they treat with Loge, fight him, or (perhaps, but in my experience less likely) ignore his steading altogether? But that is a series of premises or thematic elements. It's not a plot.

"Pre-authored" and "improvised" are really irrelevant to this discussion. In either instance the Dungeon Master is making a decision.

Well, the whole point of the OP is to put this claim under scrutiny, by asking "Which GMing decisions in the adjudication of the game tend towards railroading, and which don't?"

(In passing: "improvised" is ambiguous between "no prep" and "no pre-authored plot". The latter doesn't entail the former. For instance, I din't make up stats for the giant on the spot - I used a stat block from the Cortex Plus Hacker's Guide, which is to say I relied on (someone else's) prep. But the story wasn't pre-authored.)

If the GM has already decided how things will turn out, that is a railroad.

DMs often make decisions on what does or does not exist based on plot and narrative. That's literally their job at the table. That doesn't mean the plot is a railroad, but could instead be what makes the most narrative or dramatic sense. If there's a good couple hours left in the session, the DM might make a call that extends the adventure a little longer. If they're nearing the end of the session, then the decision becomes one that wraps up that story

What you describe here might be one approach to the GM's job at the table. It is not the only one. For instance, it depends upon there being "a plot", "a story" which the GM is (quite literally, in your examples) curating and parcelling out to the players. I personally don't see how that doesn't count as railroading. It's certainly not a case of the players determining outcomes in the fiction.