Consequences, Railroading, and Schrodinger's Dungeon

[Stonehead]

I've been thinking a lot recently about the consequences of player actions, and what counts as railroading. Before I get into that, I need to talk a little about Schrodinger's Dungeon.

Basically, the idea is that the DM prepares a few rooms, or encounters, or pieces of treasure, and uses them as the need arises, without deciding beforehand on the order or location. I first heard about the idea from a post about a guy who prepared a 4-dimmensional hypercube dungeon, where each room had a different exit in each direction. The post continued about how the players lved it, because it felt like they had total freedom to explore hundreds of rooms; the catch was that the DM only prepared 6 rooms, and regardless of which direction the party went, they found the next room the DM had prepared.

From what I can tell general opinion on this practice is pretty mixed, some people think it's a great way for a DM to cut down on prep time, others hate it, saying it robs the players of any agency, and others consider it acceptable, so long as the illusion is never broken, and the players never find out.

One other issue that arises from that specific example is that a blind choice isn't a choice. In the hypercube example, the PCs were deciding between arbitrary doors, with no possible way to make an informed decision. The most fundamental property of an rpg that I can think of is that the players make decisions as these fictional characters, rather than themselves. If you reach a fork in the road, in a sense you're faced with the choice between going left and going right, but if you have no way of knowing what you're choosing between, you aren't making a choice. And if you aren't making a choice, you can't make the choice your character would have.

With all that out of the way, I can get to my main point, which is I have no idea how to handle the Schrodinger's Dungeon problem, and the more I think about it, the more the entire genre as a whole starts to fall apart. Most people are pretty comfortable condemning railroading as bad practice, because it takes away the player's ability to influence the game world. Regardless of what the players do, the end result is the same. So then, is Schrodinger's Dungeon just another form of railroading? That kind of makes sense, is you prepare one ogre fight, and regardless of where the players go or what they try to do, they encounter the ogre, then regardless of what they do, the result is the same.

If that's all there is too it though, then a lot of other common practices could comfortable fit under the umbrella of "railroading". A lot of improv-focused DMs will just come to the table with a general idea of where they want the story to end up. Isn't that kind of the same thing though? If you come to the table knowing that there's an informant who can tell the party about the king's shady advisor, you're going to find a way to bring the party to him. So then regardless of what the players do, they end up talking to this one npc.

Even examples that strike me as pretty good DMing, and adaptability could be considered railroading. Let's say the DM wants to start his campaign with the PCs as gladiators, and are freed by the quest giver, under the condition that they fulfill his request. The rogue throws a wrench in this plan, however, when he successfully escapes his chains before the fighting begins. The DM rolls with the punches, and says the quest giver was impressed with his skills, and offers him a high paying job. My gut reaction, is that the hypothetical DM did a good job of thinking on his feet, he doesn't have to rewrite his plot hook, and the player didn't have to participate in the gladiator match, so he feels like his actions affected the story. If you think about it though, the DM did kind of ensure that the result he wanted would come about, regardless of what the players did. Why doesn't that feel like he's railroading? Same thing applies to getting the party together, a good DM usually lets the players go wherever they want, but pull the strings so that they all meet up with eachother in session 1.

Or what about PC death? A lot of DMs don't like killing characters, and a lot of players don't like having their characters killed, so instead, it's fairly common to instead punish characters with setbacks of some kind. Instead of being killed my the minotaur, they get knocked out and wake up after the battle. Instead of falling to their death from the airship, they grab hold of the anchor at the last minute, and failing that, catch a ride on a passing roc. Instead of being executed for their crimes, they're locked in a prison that they'll inevitably escape from. These examples all start to sound like railroading when you really think about it. But that can't be true, can it? Doesn't everyone have more fun when their characters grab hold of a tree branch when they fail their balance check than when they fall to their death?

And that leads into the much bigger issue, that this applies to a lot more consequences than just death. Permanently losing gear is a big feel-bad moment for a lot of players, so a lot of DMs will always give an opportunity to recover it. But then regardless of what happens in the story, they'll end up with that same familiar sword. Where's the player choice? I remember in this thread it was brought up that consequences that are only temporary set backs are just a waste of time. Doesn't that cover most consequences though? I know there are plenty of DMs who are totally cool with having the party fail and the BBEG win, but a lot aren't, and I'm not going to say that all of them that aren't are just railroading their players.

I honestly don't know what the answer is. Why does Schrodinger's dungeon feel like railroading to me, when none of the other examples do? I don't know. I do have a couple of theories though. Maybe railroading isn't inherently bad, and it's only bad when the metaphorical leash is too tight. If the players are allowed to find the informant by following his tracks, or by talking to his business partners, maybe that's enough for the players to have agency, even if the end result is the same. Maybe the details, not the end result, are what actually matter. They're going to save the world either way, but the point of the campaign is to find out who they befriend, and who they kill on their way there; as long as the players are free to talk to who they like, and kill who they don't, it isn't railroading. Or, maybe Schrodinger's dungeon isn't railroading at all. Maybe preparing one encounter, and using it regardless of where the players go is fine, as long as you let them choose where they go. Or, maybe the difference is when the push happens. Railroading is when you force the end of a story to go how you want it, and being flexible is when you force the beginning of a story.

I'm really curious to hear other peoples' opinions because, as I said before, I don't have a good answer.

[Telok]

Well, its going to depend on the system...

For the gear issue: most supers games certain pieces of PC gear can have various levels of plot immunity. Mostly because the original literature works that way and certain characters basically are their gear. Most supers systems also directly address this in their GMing sections, a PC can temporarialy lose access for a couple scenes but the genera and original fiction dictate that it comes back or gets replaced. In less equipment+stuff oriented systems like Risus or some of the PBtA ones, any gear is basically descriptive fluff. In certain systems like say, Lancer or D&D 4e, the "gear" is essentially part of the character and without it you don't have a filled in character sheet and don't play the game. Therefore the no-gear issue should simply never come up in those systems.

So the gear/no-gear thing is a system dependent issue. Any decent system that understands how characters work in the system will address the issue if it could cause problems. If you find yourself in a no-gear situation thats causing problems and the rule books don't provide any advice or rules you can go punch the author and steal something from a better system as a patch.

Schrodinger's dungeon will fall apart if there should be multiple paths and the PCs backtrack to follow another path. If the PCs just proceed through a series of rooms without loops or backtracking then any "dungeon" format is no different an experience for the players than a Schrodinger's dungeon. You can make a full multi-path interconnected web with a working ecology, factions, and dynamic responses. If the players take one path without detours or pauses, they can't tell the difference (double "can't tell" if they murderize their way through without talking to anything).

The 4d dungeon is actually a good place for the Schrodinger's dungeon method. Ideally all the transitions between rooms look similar and you can say that the rooms shift position through the 4th dimension in response to weight changes, thus making the doors PCs passed through look like they lead to new rooms. Yes, I've run a 4d dungeon and that was a trap room*. Thirty identical doors, the entrance vanished & replaced once more weight on the far side tilted the room in 4th D, the remaining doors were zero distance loops, solution was for everyone to stand far away from the exit re-tilting the room and send one person back out for some heavy furniture.

Railroading has always been a sort of sliding scale & personal preference thing. For me the step from "similar result by a different method" and "same result no matter what you do" has always been when the DM specifically frags some tactic, gear, or ability to force the party somewhere.

*ye gads, d&d 3e had been out less than six months by then. Feel old now.

[NichG]

It's easy to get caught up in what things are called and lose track of why those things mattered in the first place. Whether something works or not at a given table will depend on the players, as well as how that thing is used, and the nuanced context of its use. My basic rule of thumb would be to first and foremost listen to what the players are trying to achieve or pursue, and then don't do things which undermine the idea that those things can be meaningful (e.g. by no-selling them, arranging things to fail, using an abstraction that renders the thing irrelevant like using fixed sell prices when the players want to explore merchant gameplay, etc).

If you remove the ability for the players to influence something they've never expressed an interest in influencing, certainly that can semantically fit as 'railroading'. But it's going to matter a lot less than if you nullify their efforts or desires. If you designed a map of a ruin in a valley that the players are exploring and they miss it (not consciously avoiding it, just literally they never go to the place you put the thing) and you end up reintroducing that same content on a different continent later in the campaign, with stuff scaled up and the flags and cultural markers and denizens changed appropriately, in some theoretical sense you negated their choice to rush through the valley without exploring it, but you're not really interfering with any expressed player desire. Now if those same players found the ruin, looked at it and said 'this looks like a death trap, lets skip', then putting that same ruin map in later in the campaign would be thwarting something the players expressed a desire about - in that case they made an active choice 'we don't want this content'.

[OldTrees1]

1) Yes a Schrodinger's Dungeon is railroading. A Schrodinger's Dungeon is a room based example of the Quantum Ogre. The "Improv GM that wants the story to get to scene XYZ and thus guides it there" and the "GM that rewrites the plothook to catch a PC that derailed" are also examples of railroading.

2) Railroading (verb) references an agency spectrum/continuum rather than a binary. It is not inherently bad, but the noun "railroad" is usually used by players when they have a criticism of the current degree of agency.

Personally I like to run sandbox campaigns. I may make predictions about where the story will go, but ultimately I am enjoying the players pilot the ship wherever they want in the ocean I designed. Did you know there are players that don't want that much agency? It is important to listen to the group's preferences and finding what works best for the group.

3) Different players will care about different parts of player agency more or less than others.

1. The number of options per choice and the number of choices the PCs face could be too much (choice paralysis in a sandbox) or too little (too much railroading).
2. The types of choices might be meaningful to the player and PC, or they might be of little consideration.
3. The amount of information about the options of a choice might be excessively verbose or excessively sparse.
4. The predictive power / prediction accuracy of the information about the options of a choice might be excessively accurate or unreliable.

[Logalmier]

A lot of improv-focused DMs will just come to the table with a general idea of where they want the story to end up. Isn't that kind of the same thing though? If you come to the table knowing that there's an informant who can tell the party about the king's shady advisor, you're going to find a way to bring the party to him. So then regardless of what the players do, they end up talking to this one npc.

Seems like one way to address this from the DM's perspective is make it so that there's not just one potential outcome to any encounter/dungeon/campaign. The DM could come up with 3 or 4 plausible outcomes, and have the player's actions determine where the story ends up in relation to them. Now, this doesn't really fix the fundamental issue you seem to be describing here - you're not getting rid of railroading, you're just adding more rails. But if those rails actually go different places, and players can choose between them in a semi-informed way, then I don't think it can be described as Schrodinger's Dungeon.

Unfortunately, this approach (and any solution to Schrodinger's Dungeon probably) means more work for the DM. That's why railroading exists in the first place - creating a branching story is way more effort than creating a linear one. I feel that most player's aren't going to scratch too deep beneath the surface, so the illusion of choice is just as effective most of the time, and saves the DM a bunch of hassle.

[Stonehead]

I like the idea that it's intention that determines what matters. If the player just by chance doesn't go to the blacksmith where the important npc is, they've never made a choice to avoid them, and so you've never subverted their agency. That gets kind of weird when you start applying it to failure in general though. If there's a roll involved, you could argue that it's the dice exerting their influence on the story, not the DM so it doesn't count, but in most games I've ever been in, it's pretty common to fail without a roll. Sometimes it's clearly railroading, if you just auto fail any attempt to sneak into the castle's back door, or climb up to a window because you prepared a cool trap at the front door, that's pretty sketchy DMing. Failing to force open a door though, are less clear. The player's intention is clear, he wants the door to open, and the DM is overwriting that to instead force them to look for the key. Maybe it has something to do with believable results. There are a lot of doors that are just too tough for a human to knock down, that's kind of the point of a lock. So maybe forcing a result like this is only an issue if it doesn't make sense. If a villain was able to force open the door in a previous scene, then it does become an issue if the players can't even roll for it.

On the semantic side of things, it's interesting to think about why some things could be labeled as railroading, but I'm much more interested in why they start to become a problem. I was assuming people used "railroading" in a way that implied there was an issue of some kind.

[Telok]

On the semantic side of things, it's interesting to think about why some things could be labeled as railroading, but I'm much more interested in why they start to become a problem. I was assuming people used "railroading" in a way that implied there was an issue of some kind.

Sci-fi game:
Player 1: "We need to visit this reclusive weirdo outside the city to get info."
Player 2: "Everything outside the domed city is a toxic radioactive blasted wasteland with tentacle elephants and acid pools."
Player 3: "So? We rent an air-car. We rode one in from the spaceport."
-- later after having to hijack an air-taxi because there are none for rent, sale, or theft --
DM: "After you land and get out a giant radioactive tentacle elephant rises up from behind the hut and attacks."
Player 1: "Its how big? ... Thats bigger than the hut."
Player 2: "How come we couldn't see a huge purple monster against gray rocks without any cover?"
DM: "I guess it has a good hide check? The adventure dosen't say. It gets a surprise round and attacks."
Player 3: "We get back in the air-car and shoot it to death from 100 feet up."
-- later after murderizing the only person in the hut, who refused to talk and just attacked --
DM: "When you come back out the air-car is a burning wreck and there are soldiers who start shooting at you. You guys missed some encounters by not walking and the adventure says you need to do them. So this is one and you have to walk back to town."

I mean, adventures can be railroady and inexperienced DMs can have problems when the PCs aren't tame little plot-robots, but that was pretty bad. Actually I think that whole adventure path had bunches of that sort of stuff. Didn't matter how fast or slow the party was, you were always plot distance behind the bad guys. Didn't matter how stealthy or perceptive, enemies always saw & ambushed you. Transport didn't matter, you were supposed to walk everywhere and weren't allowed more stuff than you could carry. A good, experienced, DM can usually handle this stuff but newer DMs feel they have to choose between the adventure working or the players getting to make meaningful decisions.

[NichG]

I like the idea that it's intention that determines what matters. If the player just by chance doesn't go to the blacksmith where the important npc is, they've never made a choice to avoid them, and so you've never subverted their agency. That gets kind of weird when you start applying it to failure in general though. If there's a roll involved, you could argue that it's the dice exerting their influence on the story, not the DM so it doesn't count, but in most games I've ever been in, it's pretty common to fail without a roll. Sometimes it's clearly railroading, if you just auto fail any attempt to sneak into the castle's back door, or climb up to a window because you prepared a cool trap at the front door, that's pretty sketchy DMing. Failing to force open a door though, are less clear. The player's intention is clear, he wants the door to open, and the DM is overwriting that to instead force them to look for the key. Maybe it has something to do with believable results. There are a lot of doors that are just too tough for a human to knock down, that's kind of the point of a lock. So maybe forcing a result like this is only an issue if it doesn't make sense. If a villain was able to force open the door in a previous scene, then it does become an issue if the players can't even roll for it.

I think there's a difference between placing a surmountable obstacle in a player's path or providing opposing pressure, and actually negating a player's effort or intention. If you put an obstacle which is intended to force the player to change what they want, that's going to come off differently than an obstacle which makes the player work a bit or think a bit or have to weigh relative values but in the end allows the player to choose to keep their direction if they're willing to push through.

So e.g. the 'climb up to a window' thing:

If the player is really about 'my character is an acrobatic thief who does second-story hijinks, and is really good at them', and you say 'the wall is un-scalable', then you're no-selling that concept. If you say 'the wall has a DC of 35 to free-climb' but if the player tries a grappling hook+rope combo then it works just fine at DC 15, then you're probably not negating the player's effort or investment, even if in the end they have to use the tool rather than their skill check. And if the player really wants to push the raw skill free-climbing thing, well, in the future they can get a magic item to give a buff or have a party member buff them or try until they roll a 20 or whatever. Now, if the next smooth+rainslicked wall happens to have a DC of 50 because they managed to scare up +15 from sources between now and then, then yes, you're no-selling their effort.

If you say 'you scale the wall just fine, but through the window you see that 90% of the estate's guards happen to be collected in this bedroom for some reason, oh, and there's a powerful trap on the window (that was going to be on the front door)', then yeah, you're no-selling the implied intention of 'sneak in to avoid a fight' vs 'bust in'. If you put adamantine bars on the window, maybe that's fine and maybe it's not, but if you complain when they use an adamantine dagger to carve a second window out of the stone nearby then that's probably not.

Dice don't absolve you from that kind of reasoning, they just fuzz things a bit so it might be less clear that the DM has some intent to block a course of action. It's a classic DM manipulation to ask for more rolls when you want something to fail and not ask for rolls at all when you want something to succeed.

[kyoryu]

Ultimately "railroading" (I prefer the term illusionism) is about pretending that players have a choice, while they don't.

As in, it's totally fine to run a completely linear game where the players have no real choice but to follow the tracks - just be honest about it

(Giving choices with no information is basically the same thing).

As long as you're not doing that, you're fine. There's a lot more stuff that can be done in terms of how to present consequences of actions and stuff like that to get away from railroading, but that's a bigger discussion.

[BRC]

I honestly don't know what the answer is. Why does Schrodinger's dungeon feel like railroading to me, when none of the other examples do? I don't know. I do have a couple of theories though. Maybe railroading isn't inherently bad, and it's only bad when the metaphorical leash is too tight. If the players are allowed to find the informant by following his tracks, or by talking to his business partners, maybe that's enough for the players to have agency, even if the end result is the same. Maybe the details, not the end result, are what actually matter. They're going to save the world either way, but the point of the campaign is to find out who they befriend, and who they kill on their way there; as long as the players are free to talk to who they like, and kill who they don't, it isn't railroading. Or, maybe Schrodinger's dungeon isn't railroading at all. Maybe preparing one encounter, and using it regardless of where the players go is fine, as long as you let them choose where they go. Or, maybe the difference is when the push happens. Railroading is when you force the end of a story to go how you want it, and being flexible is when you force the beginning of a story.

I'm really curious to hear other peoples' opinions because, as I said before, I don't have a good answer.

Schrodinger's Dungeon is railroading because it doesn't really leave room for the Players to actually influence the story.

Railroading usually manifests in one of two ways: Lazy or Overbuilt

Lazy Railroading is where stuff springs up every time the PC's try to step off the intended path to put them back on track, because the GM is invested in them taking a specific path.


Overbuilt railroading is where the scenario works nicely to keep the PC's on the intended path, BUT it's a bit contrived and the players notice that they're being herded through artificial barriers, rather than working in a naturalistic scenario.


The escaped gladiator scenario isn't railroading because 1) Everything in that scenario flows perfectly naturally. "You escaped, this impressed me and I'm going to offer you a job", and 2) The Rogue could say no (Doing so would effectively mean not engaging in the campaign, but the SCENARIO isn't forcing them to say Yes).

Getting players to follow your plot hooks is part of GMing.


Schrodinger's Dungeon is a form of railroading (I've heard it called "The quantum ogre") because it deprives the Players of any actual agency in the scenario. This is different from Scenarios where there is an optimal choice (Like the Employer offering the gladiators freedom), or scenarios where there is implicit buy-in because saying "No" means just not engaging in the campaign (Like how a lot of session 1's go). It may seem fine because it's clever and the PC's don't notice, but they're not making any real decisions either:

This kind of ties into another rule which is that if you don't give the players sufficient information to make reasonable decisions, you're also depriving them of Agency, even without actually railroading.

Let's say you make a "Real" version of Schrodinger's Dungeon, a bunch of rooms with identical doors, and no information about what is behind each door. Even if you have everything mapped out, and stick to that strictly, you're still not giving the players any actual agency. Sure, you're not picking what they do, but neither are they.

I like the idea that it's intention that determines what matters. If the player just by chance doesn't go to the blacksmith where the important npc is, they've never made a choice to avoid them, and so you've never subverted their agency. That gets kind of weird when you start applying it to failure in general though. If there's a roll involved, you could argue that it's the dice exerting their influence on the story, not the DM so it doesn't count, but in most games I've ever been in, it's pretty common to fail without a roll. Sometimes it's clearly railroading, if you just auto fail any attempt to sneak into the castle's back door, or climb up to a window because you prepared a cool trap at the front door, that's pretty sketchy DMing. Failing to force open a door though, are less clear. The player's intention is clear, he wants the door to open, and the DM is overwriting that to instead force them to look for the key. Maybe it has something to do with believable results. There are a lot of doors that are just too tough for a human to knock down, that's kind of the point of a lock. So maybe forcing a result like this is only an issue if it doesn't make sense. If a villain was able to force open the door in a previous scene, then it does become an issue if the players can't even roll for it.

On the semantic side of things, it's interesting to think about why some things could be labeled as railroading, but I'm much more interested in why they start to become a problem. I was assuming people used "railroading" in a way that implied there was an issue of some kind.

The crime of Railroading is that you deprive your players of any chance to control the story. There are a lot of ways to do this.

It's a Problem when your players don't feel like they have any control over the outcome of the game, like they're just witnessing a story rather than actually playing.


A lot of Railroading as a problem is because the GM is effectively declaring their intent to not let the Players make any choices.


Like, "You need to break into the castle. The front gate is heavily guarded, but there is a back door"

"Can I climb the walls?" "No, the walls are magically unclimbable and constantly patrolled" "Can I get in through a window?" "No, all the windows are unbreakable and super-reinforced"
"Can I bribe a servant to let me in?" "No, all the servants and guards are fanatically loyal to the lord of the castle" "Can I use my Raven familiar to scout the castle?" "No, all birds that fly overhead are shot out of the sky by the guards".


The lack of realism is the problem because it's a declaration of intention. It shows that the GM cares more about keeping the players on-track than the scenario making any narrative sense.


Like, it's perfectly fine to build a scenario with only one reasonable approach in a way that makes perfect sense. If you do it over and over again your players might notice the pattern, especially if they want more narrative control over the game.

Railroading is usually when the players notice that none of their choices will actually matter.

Schordinger's Dungeon the players may not think that they're "On Rails", but the scenario depends on them being denied the knowledge they would need to make meaningful decisions, so it comes out the same way.

[icefractal]

Dungeons are interesting, because they basically are linear unless you have some actual information behind the choice of path you make. For example:
* Visiting the same dungeon multiple times
* Having a map (not necessarily complete, but enough to make the choice informed)
* Having (and using) scouting and/or divination abilities

But for instance - a dungeon with a lot of branching where you choose from thousands of possible routes, and all of those are fully statted ... but all the choices are blind and you never backtrack or visit it again? That's effectively a linear series of encounters, and it might as well be fake branching because the players are getting no benefit from the real branching.

[kyoryu]

Dungeons are interesting, because they basically are linear unless you have some actual information behind the choice of path you make. For example:
* Visiting the same dungeon multiple times
* Having a map (not necessarily complete, but enough to make the choice informed)
* Having (and using) scouting and/or divination abilities

But for instance - a dungeon with a lot of branching where you choose from thousands of possible routes, and all of those are fully statted ... but all the choices are blind and you never backtrack or visit it again? That's effectively a linear series of encounters, and it might as well be fake branching because the players are getting no benefit from the real branching.

That's why scouting is critical, as well as making some info about the dungeon available (usually within the dungeon).

[Talakeal]

I have been thinking about railroading for a while now, as it seems my players (or sometimes the forumites) accuse me of it anytime the game doesn't go in the direction they would like, and I have come to realize that it isn't a single thing, rather it is a combination of multiple spectrums.

The first is campaign premise. This should probably be worked out before the game even begins. People who sign onto the premise should agree to be railroaded; if I sign up to play a game about gangsters in 1920s Chicago, and my character decides to get on a train to become a farmer in Arizona, well I don't think I am entitled to have the game change setting and genre on my account. And if all of the players decide that they want to do this, I don't think the GM should be forced to run a completely different game than he signed up for.


Next we have classical railroading. This is, in short, warping either the rules or the setting to tell a player no. But, this is a spectrum, and on the other side of the spectrum is not playing the game by the book (imo that's the center point) but rather warping the rules or the setting to tell players yes, i.e. the rule of cool.

And often times players will accuse you of railroading when you are playing by the book, for example the long litany of perfectly RAW reasons why I told the players that they couldn't ride a griffon they found one time. And of course, if you bend the rules some of the time, players will often consider it a form of railroading or sense inconsistency / unfairness.


Then we get into sandbox games vs. linear plots. At first glance the sandbox seems to be free of railroads, but on the other hand linear plots often give players more agency. In a fully fleshed out sandbox game, stuff is happening in the background all the time, and nothing guarantees players will be in the right place at the right time to affect the world in major ways; this rarely happens in linear games which make sure to put the PCs and their decisions front and center of most of the important things that are going on.


Then we have the improv vs. prepped GM. This thread has already discussed improv and railroading, but on the other hand, there are limits to preparation. And by definition, if you go outside of the material that has been prepped for you, the GM either has to resort to improv or, more likely, find some excuse to keep you from going out of bounds.


And the last trio is objectives, obstacles, and solutions.

These are things the GM puts in the world to make it interesting and give the players something to do. Without them, you risk players getting bored or frustrated.

An example would be "Some dawarves want you to get their mountain back (objective), but there is a large dragon guarding it (obstacle) who has a weak spot on his left breast (solution)."

Now, each of these are also spectrums. For example, an obstacle to killing a monster could range from "it has a high AC" or "it is immune to fire" to "it can only be killed by tapping the spoon of St. Cuthbert three times on its right wrist". Too high on the spectrum and they start to look like puzzles; which are often railroads and not much fun for anyone.

A common problem with all of these it someone (either the players of the DM) looking at these as straight-jackets and assuming that anything which was not explicitly spelled out is off the table. If the DM refuses to allow solutions (or objectives or even obstacles) not explicitly laid out by them, that can easily be railroading. But at the same time, if the players assume that the DM will shoot down anything they haven't explicitly put in front of the players, that can easily be perceived as railroading.

[Stonehead]

I think there's a difference between placing a surmountable obstacle in a player's path or providing opposing pressure, and actually negating a player's effort or intention. If you put an obstacle which is intended to force the player to change what they want, that's going to come off differently than an obstacle which makes the player work a bit or think a bit or have to weigh relative values but in the end allows the player to choose to keep their direction if they're willing to push through.

So e.g. the 'climb up to a window' thing:

If the player is really about 'my character is an acrobatic thief who does second-story hijinks, and is really good at them', and you say 'the wall is un-scalable', then you're no-selling that concept. If you say 'the wall has a DC of 35 to free-climb' but if the player tries a grappling hook+rope combo then it works just fine at DC 15, then you're probably not negating the player's effort or investment, even if in the end they have to use the tool rather than their skill check. And if the player really wants to push the raw skill free-climbing thing, well, in the future they can get a magic item to give a buff or have a party member buff them or try until they roll a 20 or whatever. Now, if the next smooth+rainslicked wall happens to have a DC of 50 because they managed to scare up +15 from sources between now and then, then yes, you're no-selling their effort.

If you say 'you scale the wall just fine, but through the window you see that 90% of the estate's guards happen to be collected in this bedroom for some reason, oh, and there's a powerful trap on the window (that was going to be on the front door)', then yeah, you're no-selling the implied intention of 'sneak in to avoid a fight' vs 'bust in'. If you put adamantine bars on the window, maybe that's fine and maybe it's not, but if you complain when they use an adamantine dagger to carve a second window out of the stone nearby then that's probably not.

Dice don't absolve you from that kind of reasoning, they just fuzz things a bit so it might be less clear that the DM has some intent to block a course of action. It's a classic DM manipulation to ask for more rolls when you want something to fail and not ask for rolls at all when you want something to succeed.

The rain example raises some pretty interesting questions. To me, rain is a pretty believable, realistic occurrence. But here, the rain raising the climb DC to an impossible level seems like a form of railroading. Does it matter if it was raining before the players approach the castle? If the player attempts to climb the wall, and then the DM declares that it's raining, making the climb impossible, it feels like he's shutting down player choice. But if the DM anticipates that this PC is great at climbing, and establishes the rain beforehand, is that still railroading?

Or, what if the DM decides it's raining, for purely aesthetic reasons, but it still shuts down the climb plan. To me, the first two options at the very least feel like railroading, and the third one does not. Not super confident on why, just my gut reaction.

The escaped gladiator scenario isn't railroading because 1) Everything in that scenario flows perfectly naturally. "You escaped, this impressed me and I'm going to offer you a job", and 2) The Rogue could say no (Doing so would effectively mean not engaging in the campaign, but the SCENARIO isn't forcing them to say Yes).

Getting players to follow your plot hooks is part of GMing.


Schrodinger's Dungeon is a form of railroading (I've heard it called "The quantum ogre") because it deprives the Players of any actual agency in the scenario. This is different from Scenarios where there is an optimal choice (Like the Employer offering the gladiators freedom), or scenarios where there is implicit buy-in because saying "No" means just not engaging in the campaign (Like how a lot of session 1's go). It may seem fine because it's clever and the PC's don't notice, but they're not making any real decisions either:
Like, it's perfectly fine to build a scenario with only one reasonable approach in a way that makes perfect sense. If you do it over and over again your players might notice the pattern, especially if they want more narrative control over the game.

It seems to me like the "quantum ogre" can only exist behind blind choices. The original example was a combat encounter, but the same idea should apply to other forms of content, like talking to plot-important npcs. So if you have some traveler you want the players to talk to, and you have him sitting at the magic shop expecting the players to find him, but they go straight to the castle, missing him completely, I think it's fair to say he's a quantum ogre if you behind the scenes change your plans to have him wait at the castle. I think the big question is whether or not this is railroading (or whether or not it's bad railroading, if you think that could be a neutral term).

The way I see it there are a few options. Maybe it's ok, because the players had no intention to ignore him, so you haven't subverted their choices. Maybe it's not ok, because the players' actions had no impact on the next step in the story. Or maybe it depends on the circumstances, and if it would be believable for the traveller to still meet the party, then it's ok, and if not, then it's not.

Dungeons are interesting, because they basically are linear unless you have some actual information behind the choice of path you make. For example:
* Visiting the same dungeon multiple times
* Having a map (not necessarily complete, but enough to make the choice informed)
* Having (and using) scouting and/or divination abilities

But for instance - a dungeon with a lot of branching where you choose from thousands of possible routes, and all of those are fully statted ... but all the choices are blind and you never backtrack or visit it again? That's effectively a linear series of encounters, and it might as well be fake branching because the players are getting no benefit from the real branching.

Kind of a tangent, but is that how most people run dungeons? Whenever there's a fork in the road, most DMs I've played with will give you some kind of clue as to what's down each path. Some times it's obvious, like signs on doors, but usually it's just some kind of sensation, ie, you hear footsteps down one path, and nothing from the other; or torchlight is faintly glowing down one path, and the other is dark. Trying to navigate the dungeon with limited, but extant information is part of what makes it fun.

[BRC]

It seems to me like the "quantum ogre" can only exist behind blind choices. The original example was a combat encounter, but the same idea should apply to other forms of content, like talking to plot-important npcs. So if you have some traveler you want the players to talk to, and you have him sitting at the magic shop expecting the players to find him, but they go straight to the castle, missing him completely, I think it's fair to say he's a quantum ogre if you behind the scenes change your plans to have him wait at the castle. I think the big question is whether or not this is railroading (or whether or not it's bad railroading, if you think that could be a neutral term).

The way I see it there are a few options. Maybe it's ok, because the players had no intention to ignore him, so you haven't subverted their choices. Maybe it's not ok, because the players' actions had no impact on the next step in the story. Or maybe it depends on the circumstances, and if it would be believable for the traveller to still meet the party, then it's ok, and if not, then it's not.

So, the real question is "What is the Next Step in the Story".

Generally speaking, the Players would have some REASON to go to the magic shop, or the Castle, or the Inn, or what have you. The real question is, does meeting this traveler prevent or invalidate any choices.

Railroading is bad when it replaces what SHOULD be meaningful choices.

If they're just going about their business, and you arrange things so they happen to run into this guy no matter where they go, that's fine, since 'Run into this guy" or "don't run into this guy" isn't really a meaningful choice. Their meaningful choice (Go to the Castle vs the Magic Shop) Still happens. Presumably they had a reason to go to the castle, and they still get to do castle stuff there, they just also run into this guy.

If running into this guy locks the story into a certain sequence, then you have other problems, but forcing the initial encounter isn't really one of them.


An example of where it might be a bad thing would be

The PC's arrive in town, they don't intend to stay very long, and only really have time to do one thing before they leave.

The GM expects them to go to the Magic Shop to restock their potions and scrolls. Instead, they decide the most important thing to do is meet with the Baron in-person to tell him what they found (Rather than leaving a letter).


They go to the castle, only to be told that the Baron is out today, but if they leave a note it will get to him. In the meantime, this random traveler guy is willing to talk to them.

In this case, the PC's chose a thing they wanted to do (Meet with the Baron), but the GM blocked that, and instead forced them to do the planned encounter (meet with this traveler). It's not super unreasonable that the Baron is out and unable to meet with them right now, nor is it unreasonable that this traveler happens to be at the Castle, but the PC's were given a decision (Pick 1 thing to do while in town), and when they chose "Wrong", their choice was invalidated and replaced.


Edit: The Quantum Ogre is different than, say, ogres on a random encounter table. Consider the following scenarios, which all end up the same way.

Scenario 1: There are three roads to take, the GM says "All 3 roads are menaced by Ogres". The PC's pick Road A, and fight an ogre.
This isn't, like, great GMing, but there's nothing wrong with it. Not every choice needs to be super meaningful.

Scenario 2: There are three roads to take, The GM says "Road B and C have ogres on them", the PC's pick Road A, still fight an ogre.
This is the Quantum Ogre, the PC's took a choice that they were told was not Ogre, but they still met the Ogre.

Scenario 3: There are three roads, The GM decides ahead of time "The PC's will fight an ogre as they travel along the road". The PC's Pick Road A for reasons unrelated to Ogres, and fight an Ogre.

This is fine. The reason the PC's picked Road A is still intact, they fact that they happened to fight an ogre there doesn't invalidate or eliminate any meaningful choices.

[NichG]

The rain example raises some pretty interesting questions. To me, rain is a pretty believable, realistic occurrence. But here, the rain raising the climb DC to an impossible level seems like a form of railroading. Does it matter if it was raining before the players approach the castle? If the player attempts to climb the wall, and then the DM declares that it's raining, making the climb impossible, it feels like he's shutting down player choice. But if the DM anticipates that this PC is great at climbing, and establishes the rain beforehand, is that still railroading?

Or, what if the DM decides it's raining, for purely aesthetic reasons, but it still shuts down the climb plan. To me, the first two options at the very least feel like railroading, and the third one does not. Not super confident on why, just my gut reaction.

Well, it's because motive matters, and it matters even beyond a specific instance. Trust is built over time, passing through a number of situations where one party leaves themselves vulnerable to the other and isn't betrayed by that. So if the DM says before going to the estate that it's raining, and as a result the climbing fails, what matters for the ongoing relationship with that person is 'why was it raining?', not what actually happened in-game. If you later find out it was raining because the DM thought a few steps ahead 'I need climbing to not work and one of the PCs is an uber-climber, but I don't want to be caught making stuff up, so I'll establish beforehand that it's a rainy region, and I'll make a weather table where 8 out of 10 results are 'it's raining' and I'll roll for weather in the open in front of the players so I have plausible deniability', well, that DM did a whole bunch of stuff but their reasons were still problematic. They held an intent to thwart the players. How that intent was executed, whether it was hidden well or really obvious, doesn't really change what it means to be playing a game where the person you're giving the power to decide what happens doesn't want you to do get what you're looking for out of the game.

Similarly, it doesn't matter if the word railroading can be technically applied to a situation or not, what matters is: is this person going to work with you to help you enjoy yourself in the way you want to, or are they going to work against you or e.g. try to force you to enjoy what they like but you don't? How much can you trust them that the game will have been worth playing?

[Stonehead]

Generally speaking, the Players would have some REASON to go to the magic shop, or the Castle, or the Inn, or what have you. The real question is, does meeting this traveler prevent or invalidate any choices.

Railroading is bad when it replaces what SHOULD be meaningful choices.

If they're just going about their business, and you arrange things so they happen to run into this guy no matter where they go, that's fine, since 'Run into this guy" or "don't run into this guy" isn't really a meaningful choice. Their meaningful choice (Go to the Castle vs the Magic Shop) Still happens. Presumably they had a reason to go to the castle, and they still get to do castle stuff there, they just also run into this guy.

If running into this guy locks the story into a certain sequence, then you have other problems, but forcing the initial encounter isn't really one of them.

Yeah, my dumb examples are really being pushed to their limits, but I think you have a point. The "intentions matter, not just any consequence" argument looks pretty strong.

Well, it's because motive matters, and it matters even beyond a specific instance. Trust is built over time, passing through a number of situations where one party leaves themselves vulnerable to the other and isn't betrayed by that. So if the DM says before going to the estate that it's raining, and as a result the climbing fails, what matters for the ongoing relationship with that person is 'why was it raining?', not what actually happened in-game. If you later find out it was raining because the DM thought a few steps ahead 'I need climbing to not work and one of the PCs is an uber-climber, but I don't want to be caught making stuff up, so I'll establish beforehand that it's a rainy region, and I'll make a weather table where 8 out of 10 results are 'it's raining' and I'll roll for weather in the open in front of the players so I have plausible deniability', well, that DM did a whole bunch of stuff but their reasons were still problematic. They held an intent to thwart the players. How that intent was executed, whether it was hidden well or really obvious, doesn't really change what it means to be playing a game where the person you're giving the power to decide what happens doesn't want you to do get what you're looking for out of the game.

Similarly, it doesn't matter if the word railroading can be technically applied to a situation or not, what matters is: is this person going to work with you to help you enjoy yourself in the way you want to, or are they going to work against you or e.g. try to force you to enjoy what they like but you don't? How much can you trust them that the game will have been worth playing?

I don't think all instances of planning on shutting down player abilities is railroading, or if it is, it isn't always bad. If you give the BBEG a mind-control shield because one of your players has been steamrolling over combat with his mind control, I think that's just making sure everyone can have fun in combat (and fairly realistic if the boss has any communication channels at all). The relationship idea is interesting though, because that would kind of hint at the frequency being an important factor too. Maybe if it's raining once, because the DM just can't figure out any other way to get the players through the front door that's acceptable, and the problem arises when random occurrences like this shut down their plans more often than not.

[Tanarii]

Going straight to opinion: I find quantum ogres, Shrodinger's dungeon, sure thing 'plot' events (as part of an improvised game or not), saving PCs from death, and saving PCs equipment all distasteful. Both as a GM and as a player.

Btw on picking a dungeon door not being a choice ... different doors should be able to be examined. If they're all identical, block noise, air, heat, have the same amount of undisturbed dust around them, and directions are unknown (relative to the outside world and relative to the rest of the dungeon) ... yeah, then there's no choices to make. Passages usually carry even more hints to help make a choice, since sound and air travel more freely along them.

[Pex]

I brought up something similar before calling it "trolley tracks", a light rail railroad. Playing a module is this, but it can also be a homebrew Campaign Plot. Some people call it a linear campaign. One-shot game sessions are also this. Players buy into it. They will do the Campaign Plot, whether it's a module or homebrew. No question, they're doing it, it's the point of coming to the game and play whatever it is the DM prepared. The player agency comes in by the players Solving The Plot by however they choose to do so. They decide where to go, what NPCs to talk to, how they approach an encounter, etc. The DM does not enforce anything to go a certain way, but there are consequences good and bad as appropriate to what the players do.

[NichG]

I don't think all instances of planning on shutting down player abilities is railroading, or if it is, it isn't always bad. If you give the BBEG a mind-control shield because one of your players has been steamrolling over combat with his mind control, I think that's just making sure everyone can have fun in combat (and fairly realistic if the boss has any communication channels at all). The relationship idea is interesting though, because that would kind of hint at the frequency being an important factor too. Maybe if it's raining once, because the DM just can't figure out any other way to get the players through the front door that's acceptable, and the problem arises when random occurrences like this shut down their plans more often than not.

Well it comes down to this - is the GM going to use their power to work with the players to help them experience what they're looking for, or are they going to use their position to exchange player satisfaction for their own fun/ease of running/etc? If the GM gives the BBEG a mind-control shield when one player has been building up to 'I'm here to play a mind control fantasy', then that's working against what that player wants to explore. Now the GM does have to consider everyone, but maybe that means that instead of a mind-control shield that just no-sells the ability, the BBEG has a spell up providing that effect which can be dispelled. Or maybe the BBEG has multiple subminds where each successful control starts to co-opt the BBEG's action economy a bit at a time, letting the character target movement, attack, knowledge, etc. Or maybe the BBEG has a one-off (like Iron Heart Surge) which lets them throw off all negative effects once during the fight. Or maybe the BBEG just has a bunch of minions/summons/etc on the field, so even if they're mind-controlled during a surprise round there's still a fight as the minions try to kill the party or dispel the control.

The thing that matters is that the GM acknowledges how the player decides they want to interact with the world, and works with the player rather than against them in order to make that work. The semantics of whether something is technically railroading, whether it can be justified from realism or randomness, etc all just distracts from that point.

[Tanarii]

I brought up something similar before calling it "trolley tracks", a light rail railroad. Playing a module is this, but it can also be a homebrew Campaign Plot. Some people call it a linear campaign. One-shot game sessions are also this. Players buy into it. They will do the Campaign Plot, whether it's a module or homebrew. No question, they're doing it, it's the point of coming to the game and play whatever it is the DM prepared. The player agency comes in by the players Solving The Plot by however they choose to do so. They decide where to go, what NPCs to talk to, how they approach an encounter, etc. The DM does not enforce anything to go a certain way, but there are consequences good and bad as appropriate to what the players do.

The examples given in the OP were a lot more than buy-in. Solid steel rails & medium engine at least.

Except for the gladiator / quest giver one. That one might not be. In that case it could be that NPC still needs the job done, and offering it to someone that has proven their mettle in an alternate way from winning gladiator fights is exactly what a smart NPC would do in that situation. Yay for the DM they still get to use their content ... if the PCs accept. Of course, if the NPC was the type to be more pissed off that a slave escaped, that might not happen not. Point is it's not automatically quantum or 'plot' driven. It's character and NPC driven.

Or it's not, because the DM is just scrambling around trying to get things back on the (trolly) rails.

[Morgaln]

If that's all there is too it though, then a lot of other common practices could comfortable fit under the umbrella of "railroading". A lot of improv-focused DMs will just come to the table with a general idea of where they want the story to end up. Isn't that kind of the same thing though? If you come to the table knowing that there's an informant who can tell the party about the king's shady advisor, you're going to find a way to bring the party to him. So then regardless of what the players do, they end up talking to this one npc.

As someone who has an improv-heavy style of GMing: that's not how it works. I don't decide beforehand where my stories end up, because I'm not the only one affecting the story. The players have a big part in that too. Usually, I have a main problem that the story is about. I also tend to have 2-3 scenes in mind that I would like to incorporate into the game. Note, this is not per session, it is per chapter of a longer game, so over 5-10 sessions.
But often, I don't even know how and when these scenes will fit into the story. I will incorporate them if an opportunity arises. If I can't find a good time to have that scene, well, tough luck, it's not happening. Other than that, I provide the players with situations that I don't know how to solve. What that means is, I certainly could come up with ideas on how the situation could go and how to solve it, but I don't. I wait for how the players decide to tackle the situation, what solution they try for solving whatever problem(s) come up and have the world react to that.
So yes, I might have that informant show up at some place and tell the players about the shady advisor. But then, I will let players choose what to do with that information. It's completely up to them to try to expose the advisor, or blackmail him, or offer up the informant to him to get into his good graces or completely ignore that matter or anything else they can think of. I don't consider that railroading at all.

Admittedly, I have the advantage of playing with people I've played with for up to twenty years. I know how they think and how they will likely react.
For example, in a recent Werewolf: the Apocalypse game, the players ran across a human that was about to become a Fomor (=possessed by an evil spirit). This process is usually irreversible, and killing the human the proposed solution. However, I complicated matters by making said human a 15 year old girl. As expected, the players didn't really want to go and just slit her throat, so they set out to find a cure after all; and lo and behold, I had a game.
Note that this worked, because the characters where people who wouldn't just kill a girl. I have another game (same players, different characters) where this would have been over in five minutes, as those characters would have killed the girl without a second thought and consider it a mercy. So if you know your audience, you'll have an easier time guessing where the story might go.

Which is not to say that I haven't had games that went into completely different territory than I expected, and I'd roll with it. To steal Talakeal's example, if my players decide their 1920s gangsters want to go and become Arizona farmers, then that is the game we are playing. Have fun talking about the prices of seed and dealing with bad weather destroying your crops or coyotes killing your livestock :P Of course that doesn't mean the enemies they made in their former lives won't be able to track them down eventually...

[oxybe]

I think it bares to mention that we should look at one of the longest standing railroads in TTRPGs, and one that doubles as a prime example of metagaming too.

GM: sits down at table, orders his notes and peers above the screen at the awaiting players
"Alright... You all meet at a tavern"

This trope right here is proof that presentation and intent matters.

Everyone at the table has (hopefully, due to a successful session 0) agreed to make characters that would be in this tavern and willing to take a job handed to them by some mysterious, cloaked old man while also working with 3-4 other strangers.

Experienced gamers may roll their eyes a bit at the trope being used, but it's still a beloved classic that exists just gets the ball rolling on that first adventure to baptize the party.

Everyone agrees to be metagame'd and railroaded into that first adventure.

[Vahnavoi]

"Railroading" is not an exact technical term, it's a metaphor built on analogy: a train moves along set tracks and a passenger, once on board, has no control over where the train will go, only whether to get out or keep going at the next stop.

"Railroading" outside of games refers to a feeling of being socially pressured, of having your opinions and options negated in a negotiation (etc.) to reach a set outcome.

These are applicable to roleplaying games, but if a lot of distinct things start to sound like railroading, beyond of what the term is normally used for, then you're likely better off eliminating the analogy from your vocabulary alltogether. Just talk about the things that bug you directly and forget about whether they count as "railroading" or not.

Same goes Schrödinger's dungeon and Quantum Ogre. If you ask me, these are horrible misnomers created by appropriating terminology from unrelated field of science.

The "Quantum ogre" isn't quantum. It's inevitable. The game master has prepared an encounter with an ogre and by God there will be an encounter with an ogre. There is no quantum superposition anywhere, the game master has decided their next move will be placing the ogre no matter what move their players make.

The hypercube dungeon you describe is nothing like Schrödinger's cat experiment. The point of that thought experiment was to show how states reliant on a random element would be trapped in superposition, being both true and false until the outcome of the random element is resolved through observation. A true example of a "Schrödinger's dungeon" would just be a randomly generated dungeon, where the exact state of the dungeon isn't known until dice are rolled. Your example of a hypercube dungeon is not random at all, it is a linear dungeon that uses special geography to disguise the fact. A proper hypercube wouldn't even be linear, since it would have multiple ways of moving through the rooms.

Having imperfect information in a game rarely requires terminology stolen from theory of quantum mechanics to describe, so stop doing it.

---

So, now that I've established the special terminology isn't useful, what seems to be the issue?

It seems to me, you want games to have choices. Okay. Well, a game master constantly making choices that counteract their players' choices will obviously reduce number of effective choices their players can make. And any game rule which relies on a random function for its outcome won't count as a player's choice. But these are relevantly different. A game master acting that way is typically eliminative, causing game paths to converge, while the latter is typically generative, causing game paths to diverge. Both have their uses for game design, so their mere presence somewhere in a game tree aren't worthy of concern. As a rule of thumb, don't fret about game master intervention if after a round of elimination there are still at least two distinct options per decision prompt, don't fret about randomness if for each random decision there is a decision reliant on the player. Only start to worry when either element becomes so dominant players are reduced to spectators for large fractions of real playtime.

As many people have already pointed out in slightly different ways, it's hard to tell a linear game from a non-linear game on one playthrough. The reality is even worse: it's possible to value having a lot of options, but it's impossible to experience all those options. The larger the move space of your game is, the harder it is to see more than a fraction of it. In a stochastically branching universe, there may be many possible futures, but only one will be actualized for you. You can't travel back in time to see if things could've gone differently than they did, no-one can. What you can do is check the movespace of your game to see if it allows for multiple paths through it in the first place and then put different playgroups through it to see if those different paths emerge. That's the information you want to prove choices exist and matter in a game.

[Quertus]

Sci-fi game:
Player 1: "We need to visit this reclusive weirdo outside the city to get info."
Player 2: "Everything outside the domed city is a toxic radioactive blasted wasteland with tentacle elephants and acid pools."
Player 3: "So? We rent an air-car. We rode one in from the spaceport."
-- later after having to hijack an air-taxi because there are none for rent, sale, or theft --
DM: "After you land and get out a giant radioactive tentacle elephant rises up from behind the hut and attacks."
Player 1: "Its how big? ... Thats bigger than the hut."
Player 2: "How come we couldn't see a huge purple monster against gray rocks without any cover?"
DM: "I guess it has a good hide check? The adventure dosen't say. It gets a surprise round and attacks."
Player 3: "We get back in the air-car and shoot it to death from 100 feet up."
-- later after murderizing the only person in the hut, who refused to talk and just attacked --
DM: "When you come back out the air-car is a burning wreck and there are soldiers who start shooting at you. You guys missed some encounters by not walking and the adventure says you need to do them. So this is one and you have to walk back to town."

I mean, adventures can be railroady and inexperienced DMs can have problems when the PCs aren't tame little plot-robots, but that was pretty bad. Actually I think that whole adventure path had bunches of that sort of stuff. Didn't matter how fast or slow the party was, you were always plot distance behind the bad guys. Didn't matter how stealthy or perceptive, enemies always saw & ambushed you. Transport didn't matter, you were supposed to walk everywhere and weren't allowed more stuff than you could carry. A good, experienced, DM can usually handle this stuff but newer DMs feel they have to choose between the adventure working or the players getting to make meaningful decisions.

This sounds like it could have been a contender in my "worst module" thread.

Edit: The Quantum Ogre is different than, say, ogres on a random encounter table. Consider the following scenarios, which all end up the same way.

Scenario 1: There are three roads to take, the GM says "All 3 roads are menaced by Ogres". The PC's pick Road A, and fight an ogre.
This isn't, like, great GMing, but there's nothing wrong with it. Not every choice needs to be super meaningful.

Scenario 2: There are three roads to take, The GM says "Road B and C have ogres on them", the PC's pick Road A, still fight an ogre.
This is the Quantum Ogre, the PC's took a choice that they were told was not Ogre, but they still met the Ogre.

Scenario 3: There are three roads, The GM decides ahead of time "The PC's will fight an ogre as they travel along the road". The PC's Pick Road A for reasons unrelated to Ogres, and fight an Ogre.

This is fine. The reason the PC's picked Road A is still intact, they fact that they happened to fight an ogre there doesn't invalidate or eliminate any meaningful choices.

As much as I like this, there's the issue of The PC's Pick Road A for no stated reason, or The PC's Pick Road A for many reasons, an unstated one being related to Ogres.

[BRC]

As much as I like this, there's the issue of The PC's Pick Road A for no stated reason, or The PC's Pick Road A for many reasons, an unstated one being related to Ogres.

What the PC's say is less relevant than what they've been told.

If they've been told that Road A does not contain Ogres, then if they go to road A, they shouldn't meet any Ogres. It doesn't matter if they said anything about Ogres when they picked Road A.

It kind of breaks down to 2 rules

1) PC's should be given meaningful choices

2) If the PC's make a choice, that choice shouldn't be invalidated.

If there are 3 roads, the PC's should have enough information about those roads such that choosing which one to take is a meaningful choice. It's fine for all 3 roads to have ogres, but if part of the information their given is "Road A is the road to take to avoid Ogres", then Road A better not have Ogres.


Thinking about it, Rule 2 is just a subset of Rule 1. If the PC's make a choice, and that choice is invalidated, then the choice wasn't meaningful.

[Easy e]

Controversial post:

Railroading is good, IF the illusion of choice is never removed. The only thing bad about Railroading is clumsy railroading that breaks the illusion of choice.

Playing a game of D&D is just one big magic trick, there is the set-up, there is the distraction, and then there is the big-reveal (This is often called the 3-act method). The trick is getting players to feel like they have total freedom. However, the truth is that they do not have total freedom. There is no such thing.

Quantum Ogres and Schrodinger's Dungeon are just types of the magic trick, but the end result is that the illusion of choice is always present.

A corollary- The illusion of choice is only important if that is what your players need to have fun. If they do not need the illusion of choice to have fun, then there is no need to provide the illusion. Most Video games railroad the heck out of you as a player, but they are infinitely more popular that RPGs. Therefore, Railroading is not a bad game design; it is simply a game design style.

[BRC]

Controversial post:

Railroading is good, IF the illusion of choice is never removed. The only thing bad about Railroading is clumsy railroading that breaks the illusion of choice.

Playing a game of D&D is just one big magic trick, there is the set-up, there is the distraction, and then there is the big-reveal (This is often called the 3-act method). The trick is getting players to feel like they have total freedom.

Quantum Ogres and Schrodinger's Dungeon are just types of the magic trick, but the end result is that the illusion of choice is always present.

A corollary- The illusion of choice is only important if that is what your players need to have fun. If they do not need the illusion of choice to have fun, then there is no need to provide the illusion. Most Video games railroad the heck out of you as a player, but they are infinitely more popular that RPGs. Therefore, Railroading is not a bad game design; it is simply a game design style.

The above is true, and also terrible advise that I would never give to any GM.

Yes, pre-planning everything while maintaining the Illusion of Choice theoretically opens the door to the best game.

BUT, doing so relies on 2 things

1) The GM can control the story while retaining the Illusion of Choice, which is to say, they have a 100% success rate in leading the PC's down the path they want them to take without the players noticing

2) The GM can perfectly predict how the players want the story to go.


Because there's more to "Meaningful Choices" than just tactical decisions. It's also a way for the Players to control how the story goes, and make sure the story they are playing through is one that they enjoy.


Every GM who railroads thinks that they can perfectly predict and control their players without them noticing the rails. They're almost always wrong.

[kyoryu]

Ultimately, it's a linear game if the GM knows, in advance, what's going to happen and (for the most part*) in what order they're going to happen, and the GM will not allow anything to veer from that plan.

It's railroading/illusionism if the GM pretends that's not the case.

How you accomplish it is irrelevant, and things that can be used to railroad can also be used in non-railroad situations. I feel like a lot of times these discussions ignore the intent of the GM, and it kinda feels like that's done to come up with a set of "blessed" techniques to use for railroading.

I also highly disagree that well-executed illusionism is the best game. A) some people really do play games to make choices and see the results of them and B) I think it's always a bad idea to be dishonest about the nature of the game. If you want to run a linear game, do so - just be up front about it.

* there's a certain amount of optional encounters/side quests, and reordering of things that can be done while still being, effectively, a railroad. Every BioWare game is basically a railroad. Bethesda games are a little murkier (they're usually a railroad main story with lots of subquests and other systems that can be played through - the main/subquests are railroady but the fact that you can in many cases just play around in the sandbox with teh systems can be argued to move them away from that designation, but I'd still personally argue that they're railroads, just with a ton of emphasis on optional stuff)

Then we get into sandbox games vs. linear plots. At first glance the sandbox seems to be free of railroads, but on the other hand linear plots often give players more agency. In a fully fleshed out sandbox game, stuff is happening in the background all the time, and nothing guarantees players will be in the right place at the right time to affect the world in major ways; this rarely happens in linear games which make sure to put the PCs and their decisions front and center of most of the important things that are going on.

I'm not sure how that follows. A game with no real choices gives you more choice than a game with choices? Like, it might be a more interesting ride, but it's not one with more choices.

There's also lots of ways to keep players "in the plot" without locking them on rails - there are other models behind the classic "wander around and look for stuff" sandbox. Engaging players with the plot isn't railroading/linear if you as a GM don't know what will happen.

It seems to me like the "quantum ogre" can only exist behind blind choices. The original example was a combat encounter, but the same idea should apply to other forms of content, like talking to plot-important npcs. So if you have some traveler you want the players to talk to, and you have him sitting at the magic shop expecting the players to find him, but they go straight to the castle, missing him completely, I think it's fair to say he's a quantum ogre if you behind the scenes change your plans to have him wait at the castle. I think the big question is whether or not this is railroading (or whether or not it's bad railroading, if you think that could be a neutral term).

I mean, it probably is railroading. Is the NPC being thrown in their path to ensure that the players get back on the sequence of events the GM has planned? If so, railroad.

The way I see it there are a few options. Maybe it's ok, because the players had no intention to ignore him, so you haven't subverted their choices. Maybe it's not ok, because the players' actions had no impact on the next step in the story. Or maybe it depends on the circumstances, and if it would be believable for the traveller to still meet the party, then it's ok, and if not, then it's not.

Is it being done to ensure that the players stay on the GM's planned story, and going through the encounters the GM planned? If so, railroad. If not, no.

Kind of a tangent, but is that how most people run dungeons? Whenever there's a fork in the road, most DMs I've played with will give you some kind of clue as to what's down each path. Some times it's obvious, like signs on doors, but usually it's just some kind of sensation, ie, you hear footsteps down one path, and nothing from the other; or torchlight is faintly glowing down one path, and the other is dark. Trying to navigate the dungeon with limited, but extant information is part of what makes it fun.

Yeah, blind choices are just bad design for any number of reasons.

I don't think all instances of planning on shutting down player abilities is railroading.

Of course not. It's railroading if it.... forces players down a particular path the GM has planned.

I brought up something similar before calling it "trolley tracks", a light rail railroad. Playing a module is this, but it can also be a homebrew Campaign Plot. Some people call it a linear campaign. One-shot game sessions are also this. Players buy into it. They will do the Campaign Plot, whether it's a module or homebrew. No question, they're doing it, it's the point of coming to the game and play whatever it is the DM prepared. The player agency comes in by the players Solving The Plot by however they choose to do so. They decide where to go, what NPCs to talk to, how they approach an encounter, etc. The DM does not enforce anything to go a certain way, but there are consequences good and bad as appropriate to what the players do.

This is close to what I do, maybe. There's a "plot", but I don't plan what hte players will do. Buying into the plot is part of joining the game, but you can approach it how you want, and I don't know how it will resolve itself. I don't see this as railroading.... you agree to the premise, we sit down and find out what happens.

As someone who has an improv-heavy style of GMing: that's not how it works. I don't decide beforehand where my stories end up, because I'm not the only one affecting the story.

This. Exactly this.

The players have a big part in that too. Usually, I have a main problem that the story is about. I also tend to have 2-3 scenes in mind that I would like to incorporate into the game. Note, this is not per session, it is per chapter of a longer game, so over 5-10 sessions.

Pretty much. The "plot" is what happens when the players intersect with the problem.

Other than that, I provide the players with situations that I don't know how to solve. What that means is, I certainly could come up with ideas on how the situation could go and how to solve it, but I don't. I wait for how the players decide to tackle the situation, what solution they try for solving whatever problem(s) come up and have the world react to that.

Yes. 100%. If you provide problems and let the players come up with the solutions you pretty much automatically create agency, and have to deal with the results of their actions.

(Again, not the only way to run a game, but it's in the "not a railroad but not a wander about blindly" space that I personally prefer)

So yes, I might have that informant show up at some place and tell the players about the shady advisor. But then, I will let players choose what to do with that information. It's completely up to them to try to expose the advisor, or blackmail him, or offer up the informant to him to get into his good graces or completely ignore that matter or anything else they can think of. I don't consider that railroading at all.

Sure. The GM controls the world, including the NPCs and when they show up. As long as that's not being done to subert player choices, not railroading

And any game rule which relies on a random function for its outcome won't count as a player's choice.

Meh, not entirely sure I agree with that.

If I offer players two paths, one through the goblin woods and one through the ogre mountains, what they encounter might still be the result of a random roll - but it should be on different tables. So in that case the result is random, but agency is preserved.

What the PC's say is less relevant than what they've been told.

If they've been told that Road A does not contain Ogres, then if they go to road A, they shouldn't meet any Ogres. It doesn't matter if they said anything about Ogres when they picked Road A.

It kind of breaks down to 2 rules

1) PC's should be given meaningful choices

2) If the PC's make a choice, that choice shouldn't be invalidated.

If there are 3 roads, the PC's should have enough information about those roads such that choosing which one to take is a meaningful choice. It's fine for all 3 roads to have ogres, but if part of the information their given is "Road A is the road to take to avoid Ogres", then Road A better not have Ogres.


Thinking about it, Rule 2 is just a subset of Rule 1. If the PC's make a choice, and that choice is invalidated, then the choice wasn't meaningful.

Eh, mostly, though this isn't really railroading, per se. If you say "absolutely no Ogres" that's one thing, but I think it's fine to say "take this road to avoid Ogres", and have them be much lower on the random encounter chart (1 in 100 vs. 1 in 5).

If the Ogre is there because that's the encounter that you as the GM have decided on? Railroading.

[Telok]

This sounds like it could have been a contender in my "worst module" thread.

Starfinder "Dead Suns" ap. Also known for several instances of "didn't do any math" face palming and a couple times the entire group stopped game to critique bad writing.