The Nature of Railroading

[Quertus]

That seems to me rather more reasonable than blanket statements such as the following:

That's because the blanket statement is being misunderstood. "Burns? They're caused by fire."

Which is the sort of thing that prompted my wonder over the definition of "wanting something," since there's nothing else there in that sentence, and "wanting a cool sandbox that my players will think is impressive" counts as "wanting something."

Yeah, I tried to explain it a few times, but it didn't seem to get traction.

But I still think your modified view expressed above has some problems.

First, it creates major problems for verisimilitude. If the GM has created a sandbox of even slight complexity, and thinks that time should pass and NPCs take actions while the party adventures, then constraints will emerge on player agency as a result. But perchance then some player will cry "Rails formed in your mind! You wanted that offscreen thing to happen, but I didn't, I don't like it, and we didn't have a chance to do anything about it!"

You slept 3 days, the sun came up and set 3 times, deal. Following game physics is not railroading; ignoring game physics is.

So, if the GM wanted you to witness the next sunset, and had the sun just hang there in the sky while you slept 3 days, yes, railroading.

Which means, second, that this sort of thing is incredibly difficult to police. Because it isn't hard for a GM to decide they want a couple of specific things to happen,

Sure it's not hard to be bad. I'm saying don't do that.

and just set up the world in a way that those are plausible outcomes, without ever even looking like it's railroading. And since it wouldn't be hard to do that, players would have to be constantly vigilant to see it didn't happen, resulting in a climate of suspicion that would make a good-natured game devilishly hard to play (assuming the players cared about that kind of railroading).

I'm not trying to teach players how to spot rails, I'm trying to teach GMs not to form rails in the first place.

And third, the constant suspicion of the GM overstepping that line would seem to me to make the game very little fun for the GM, who isn't a player with a PC but is still a player.

So don't be the suspicious unfun GM.

More broadly, we humans just tend to think narratively. So while it might be possible to create a number of characters and drop them in a sandbox as "toys," I can't see how it could be interesting or complex without incorporating the narrative biases of the creator. Because in the act of creating such a sandbox, in order to flesh out the characters you've got to be telling stories, consciously or not,

and in connecting them all in the same sandbox you've got to tell yet more stories. And those stories are also the heuristics we use to approximate a world that functions plausibly, since we humans don't have the intellectual capacity to forge coherent imaginative worlds from whole cloth. So the whole thing ends up imbued with the creator's narrative biases,

Yes, "this is the story of the toy at T minus 1".

And I even accept "this is what would have happened without the PCs intervention - 'and I would have gotten away with if it weren't for you darn kids' - so long as the GM doesn't care how or even whether the PCs change that timeline". This is the one place where I do form rails - I know how my dozen plots would have turned out, and I *want* the players' actions to change *something*.

which will inevitably, in some ways, act to limit the stories the players can tell once they get to work--though in other ways it helps to enhance and heighten those stories, by giving them a meaningful context.

And there you've lost me. In what meaningful way does "the world had a history, the toys in the sandbox have depth" inhibit the players from meaningfully engaging game physics?

Perhaps all that doesn't rise to the level of the GM "wanting something specific," and thus isn't railroading. But I have gotten the sense that you are generally hostile to GM-crafted narrative

Yes, I'm against the GM deciding where the game will go.

--and my point is that it is definitionally always with you, and if someone were somehow able to dispense with it, their game would be flat, lifeless, and boring.*

Wow, quite the opposite. The GM's rails are flat and lifeless, whereas the players' story is anything but boring.

Well, usually.

Yeah, that's a balance between educating and having fun at the current session. Know your players. Some people don't mind having their character being killed by an assassin without any opportunity to act, since they did piss of a noble 10 sessions ago, and it's only realistic. But people are different.

Know your group. Touché.

So, again, why do modern groups feel that they can tick off a noble without researching how he's likely to try to get revenge, and then get upset when they didn't prepare for his revenge? When they take no precautions, get passed out drunk & sleep with their doors unlocked, then throw a hissy when their actions have logical consequences?

Dunno, Sandy Petersen seems to be deliberetaly using telegraphing when he designs horror CoC scenarios, see link below. First a hint of something amiss, then evidence of something being there, before showing the monster. It's for horror though, so you kind of need telegraphing to build the tension.

https://m.youtube.com/watch?v=3o9P9fYrIfo

Well, sure, but that's not what I'm talking about people seeming to want. "Did the party know that the giant was going to throw them in the pit?" "Did the investigators know that the Star Spawn was weak to X?"

IME, they're called "investigators" for a reason.

[King of Nowhere]

Railroading is when the players don't want to do a thing and the GM forces them to do it anyway. Either overtly or through sneaky sleight-of-hand. It is inherently a bad thing.

this is a different definition than what I am using.

I was using "railroading is when the actions of the players do not impact the plot", and by that definition I stand by my argument that it is ok if done with moderation and for a good reason.

quertus is using "railroading is when the DM is pushing for certain things to happen", which is so vague and difficult to distinguish from regular npc actions that I still can't get his point after all this talking in circles.

others are using "railroading is when the players are forced to do certain actions", and others are using yet slightly different definitions.

in the end, everyone has a slightly different concept of railroading.

[Resileaf]

For me, railroading is what the DM in DM of the Rings does.

[Gallowglass]

So, again, why do modern groups feel that they can tick off a noble without researching how he's likely to try to get revenge, and then get upset when they didn't prepare for his revenge? When they take no precautions, get passed out drunk & sleep with their doors unlocked, then throw a hissy when their actions have logical consequences?

I think there is about 15 miles of space between your two extremes for people to live in. I have no problems with the DM saying. "As you drink the night away and collapse exhausted and drunk into your beds, you are very surprised to wake up the next morning smelling deep sea air and feeling the rock of a boat underneath you. You blink your eyes open and feel chains on your ankles and see yourself in a barred cell in the hold of a ship" because I know that its a hook to an adventure. But I would never do that as a DM even though I'm fine with it as a player. No I would give the players a chance to notice the sea-brigands casing the inn last night while they drank. I would give them the benefit of the doubt about locking their doors or not drinking to oblivion without them having to expressly tell me so and even if I wanted to run a press-gang to adventure-island scenario, I wouldn't hinge my entire plan on being able to press-gang them even if I thought it would probably more likely succeed than not. But to your point, what was fun for me 20 years ago isn't so fun for me anymore. So the idea of living in a campaign where I have to do this just to keep my character alive...

DM: "Okay, so as the sun starts to set do you guys want to set camp or keep going..."

Player1: "Oh we set camp. First I start digging the typical trench around the campsite area and pull out the punji sticks from my bag of holding and start setting them up... And yes I inspect each one to make sure it doesn't need patching or re-poisoning. Let's see, forty sticks, I'll start making rolls and hand you them all on a sheet..."

Player2: "Yeah and I start stringing the netting through the trees so we don't have another flying monkey incident."

Player3: "Ok, I'll start a fire after making sure the wood and surrounding material haven't been doused in oil or scentless fire power like that one time of course and start cooking. As always I only use food in still sealed containers where my arcane mark hasn't been breached so that we don't have another quickling poison incident."

Player4: "let's see... looking at the map, we are on grid 7 cross grid 14. let me check. Okay, the three sessions we spent in the library at town says that the only known incidents around here involve a small troglodyte band and one case of a wandering minstrel show raping people seven years ago. Better prepare for both. Everyone stick wax in your ears and your noses."

Player2: "No wait... then we won't be able to smell or hear anything else. Um. whomever has the cross-section of the best perception check and the best fort save skip the wax in the nose and whomever has the cross-section of the best perception check and best will save skip the wax in the ears. Someone check the manual of venn diagrams. I'll make sure we have potions of break enchantment and cure poison ready with the poptabs and easy to access."

Player5: "Uh, you know guys, to save time, can't we just say 'We do the normal campsite preparations?"

*all the other players look at the DM*

Dm: *smiling innocently* "anything else you guys want to do?"

Player1: *sigh* "Ok, everyone check your notes under campsite, subsection deep forest and see what we are forgetting."

...is no longer fun for me. I would rather get to the story than spend my limited hours trying to out-think and out-plan the person who controls the world and has perfect knowledge of everything and, apparently, takes some measure of delight in punishing me for my lack of omniscience.

[kyoryu]

If it sounds like a bait and switch that is a problem in communication between me and the forum, not me and the players.

Maybe. But we only have your information to go on, and the fact that apparently at least one of your players felt like they were being railroaded. Your approach can either be "I will prove that they were not!" or "okay, let me figure out why they feel this way, and ensure that I'm either not railroading them if I was doing so implicitly, ensure they knew the reasonable things that they could have done to avoid the railroad, or be able to tell them that, yes, that is the type of game I run."

I highly recommend the latter. Telling people that what they're feeling is wrong almost never works out well.

This is a hex-crawly campaign. The players have known about this chasm for months, and made the decision to head there in the previous session. They knew that the adventure was about exploring a haunted gorge, and their goal while there was to kill a large monster (of unknown nature) and to find a large treasure (again of unknown nature), well before they made the decision to head out.

So they knew about the chasm but avoided it. "The adventure was about" doesn't seem to indicate a ton of choice.

Although, your post seems to hinge on the idea that a "twist" or "reversal" makes for a railroad, I don't think I much agree with that notion.

Nope. First off "you were fighting the giant and he threw you into the chasm!" isn't a twist or a reversal. Finding out that the king was actually the usurper is a twist. Saying "you're now doing this, like it or not!" isn't a twist.

Secondly, that kind of thing isn't necessarily railroading, but it's often a sign that railroading is occurring. It's, to use the parlance of another thread, a red flag for railroading.

Here's the ACTUAL crux of my post:

To me, a roleplaying game is basically the following conversation*:

GM: "This is the situation. What do you do?"
Me: "I do the thing."
GM: "Okay, the situation has changed becuase of that and is now this. What do you do?"

That's it.

The key here is that what I say has to change what happens. That, to me, is the game. If the new situation doesn't vary based on what I say, then what I say has no meaning and I'm effectively not playing the game.

Saying that "twists" are central to how I define railroading is a bad read at best, and a strawman at worst.

That's because the blanket statement is being misunderstood. "Burns? They're caused by fire."

The problem is that GMs should want things and that's perfectly okay. Saying "railroading is caused by GMs wanting things" implies that GMs wanting things is something to be avoided or cautious about.

It's not. Not at that blanket level. What needs to be avoided is certain categories of things the GM wants - mainly around sequences of events involving the players, and the choices and actions of players.[/quote]

Yes, I'm against the GM deciding where the game will go.

Yet in most (non-railroad) games, "where the game goes" is a combination of both the GM's and the players' contributions to events. This is good and healthy.

this is a different definition than what I am using.

I was using A"railroading is when the actions of the players do not impact the plot", and by that definition I stand by my argument that it is ok if done with moderation and for a good reason.

quertus is using B"railroading is when the DM is pushing for certain things to happen", which is so vague and difficult to distinguish from regular npc actions that I still can't get his point after all this talking in circles.

others are using C"railroading is when the players are forced to do certain actions", and others are using yet slightly different definitions.

in the end, everyone has a slightly different concept of railroading.

(Letters added)

I think A and C are, practically, equivalent. If your actions aren't impacting the plot (especially if "the plot" is defined as a sequence of things the players go through), then you're forced to do things. And if you're forced to do things, then your decisions aren't impacting the plot because you're not really making any.

[Talakeal]

For me, railroading is what the DM in DM of the Rings does.

Yeah, the "very specific level of tired" scene is what I would call classic railroading.

[NichG]

I think to get the 'DM wants' thing, its actually more informative to look at a case in which there's no question of agency involved: an author writing a piece of fiction. In that case, lets contrast an author who arranges setting elements and then follows their feeling of what would make sense to happen and where the interesting stories are within the interacting setting elements, versus an author who wants to tell a particular story and thereby arranges the setting elements in such a way that the particular story will happen.

If these seem like identical situations, then it won't yet be possible to understand the distinction between a DM whose NPCs want something and therefore act, versus a DM who wants something and therefore has NPCs act.

Also take into account that if you're trying to understand this, it should be with the goal of understanding a specific type of play experience rather than as a negotiation over the meaning of the term 'railroading'. The specific description is more description, and the conversation has gotten to the point where the terminology is likely hurting more than helping. It's also helpful to approach the situation from the perspective of trying to understand what others are reacting to rather than reading things as a persuasive interaction - if you let go of the idea that what's being discussed must reflect what you yourself should like or dislike about play, its easier to come to an understanding about what other people might like/dislike about play. That is to say, if you say 'I dislike railroading' and someone else says 'to me, railroading is fuzzy puppies', then its better to say 'words aside, lets discuss fuzzy puppies' than 'why should I dislike fuzzy puppies?'

[Talakeal]

So they knew about the chasm but avoided it. "The adventure was about" doesn't seem to indicate a ton of choice.

They didn't choose to avoid it, they just explored the bridge first as it was the most obvious landmark in the area.

We are playing a psuedo hex-crawl, The entire campaign is about exploration and killing monsters / searching for treasure / making politcal alliances along the way.

So when the previous week the players told me they wanted to explore the haunted gorge, I prepped for them exploring the haunted gorge.

Maybe. But we only have your information to go on, and the fact that apparently at least one of your players felt like they were being railroaded. Your approach can either be "I will prove that they were not!" or "okay, let me figure out why they feel this way, and ensure that I'm either not railroading them if I was doing so implicitly, ensure they knew the reasonable things that they could have done to avoid the railroad, or be able to tell them that, yes, that is the type of game I run."

I highly recommend the latter. Telling people that what they're feeling is wrong almost never works out well

The player in question is sullen and grouchy, he gets mad often frequently and seldom explains why.

But when he accuses me of something it does make me self conscious, and I am seriously doubitng if I have been using the term railroading wrong all of these years.

For example, last month when he said he hates how my NPCs monolgue all the time, and I couldn't figure out what he meant as I typically go years between what I would consider a villain monologue. It turns out he didn't know what monologue meant, and he was using it to mean "reveal information through conversation", and as he was a hack and slash player he didnt like having important informatiom come ojt during talky scenes as he prefered to just tune out dialogue and wait for the next fight.

Nope. First off "you were fighting the giant and he threw you into the chasm!" isn't a twist or a reversal. Finding out that the king was actually the usurper is a twist. Saying "you're now doing this, like it or not!" isn't a twist.

Secondly, that kind of thing isn't necessarily railroading, but it's often a sign that railroading is occurring. It's, to use the parlance of another thread, a red flag for railroading.

Here's the ACTUAL crux of my post:



Saying that "twists" are central to how I define railroading is a bad read at best, and a strawman at worst.

I am trying to figure out why the scenario comes across as railroading and am asking for clarification to try and put my finger on it.

It seemed to me that you were implying it was railroading because they thiught the adventure was killing a giant, while it was actually exploring the chasm, which seems to be a problem with a "swerve" causing a midpoint change in the nature of the adventure.

I think that the player is using railroad because he doesn't like losing, and (correctly) assumed that I set up a very difficult fight because having the players suffer a setback, recover, and then triumph is more fun from both a narrative and game mechanical perspective than a straightforward slug-fest where victory in inevitable.

Now, in my oppinion railroading is when the GM ignores the logic of the setting or the game rules (or just browbeats the players OOC) into following a script. It is not setting up a scenario where the players have limited options, althiugh that may well be a bad game.

For example, one time I heard a guy on the white wolf forums gitching about the iron will merit, because he couldn't motivate the players to accept going on missions without mind control. In my oppinion having an NPC mind control the PCs isnt railroading (unless you are ignoring the iron will merit or otherwise cheating / deforming versimilitude) but it sure as heck makes for a completely bad and unfun game.

[kyoryu]

They didn't choose to avoid it, they just explored the bridge first as it was the most obvious landmark in the area.

We are playing a psuedo hex-crawl, The entire campaign is about exploration and killing monsters / searching for treasure / making politcal alliances along the way.

So when the previous week the players told me they wanted to explore the haunted gorge, I prepped for them exploring the haunted gorge.

Well that's a fairly key bit of info. Somewhat hard to argue that giving the players what they asked for is railroading (though it could be, like if the chasm was a one-directional sequence of rooms with no decisions).

The player in question is sullen and grouchy, he gets mad often frequently and seldom explains why.

But when he accuses me of something it does make me self conscious

Well, I mean, we all do things, and people have preferences, so getting feedback is good, and it also doesn't mean you're wrong or bad. It often just is a matter of clarifying preferences, or ironing out miscommunication. I know it's hard, but it's generally best to view feedback as an opportunity for improvement rather than an attack.

and I am seriously doubitng if I have been using the term railroading wrong all of these years.

If the players choose to go to the chasm, and tehy end up exploring the chasm, that is not, in and of itself, railroading. That doesn't mean that railroading didn't happen or couldn't happen within that scenario, but that information, alone, is not enough for it to be railroading.

While nailing down a formal definition can be difficult, I think the general gist is "if there's a preplanned series of events that the players have to go through, and that's the design of the game." Some people might call that a "linear adventure" and say that only if the players push against it is it railroading, some people might call the whole thing railroading and further divide into participationism (you willingly get on the train) and illusionism (you're told there's no tracks, but there totally are, and the GM tries to conceal that).

But overall, the crux is really "you'll do A, then, B, then C, then D. And your choices don't change that."

For example, last month when he said he hates how my NPCs monolgue all the time, and I couldn't figure out what he meant as I typically go years between what I would consider a villain monologue. It turns out he didn't know what monologue meant, and he was using it to mean "reveal information through conversation", and as he was a hack and slash player he didnt like having important informatiom come ojt during talky scenes as he prefered to just tune out dialogue and wait for the next fight.

That's.... odd usage.

It's also indicative of the difficulty of using terms - you start getting possibilities for getting confusion about the definition. When possible, especially if you're confused, ask for specifics rather than terms. Instead of "railroading", ask what specifically happened that the player didn't like. And try, as best you can, to do so from a place of curiosity and genuine interest, not a place of defensiveness.

I am trying to figure out why the scenario comes across as railroading and am asking for clarification to try and put my finger on it.

Unfortunately, a brief description is not going to be super helpful for us to figure out what happened over hours of gaming.

It seemed to me that you were implying it was railroading because they thiught the adventure was killing a giant, while it was actually exploring the chasm, which seems to be a problem with a "swerve" causing a midpoint change in the nature of the adventure.

No, swerves can happen. Again, it's often a *red flag* of railroading (the GM wants x to occur, so they do something the players want, then force what they actually wanted to occur). But it's not inherently railroading.

I think that the player is using railroad because he doesn't like losing, and (correctly) assumed that I set up a very difficult fight because having the players suffer a setback, recover, and then triumph is more fun from both a narrative and game mechanical perspective than a straightforward slug-fest where victory in inevitable.

I think setting up an outcome of a fight is bordering on railroading, depending on specifics. My personal preference is that the GM not know how a given encounter/fight will go. And, no, a slim chance of it not going that way doesn't really change that.

Now, in my oppinion railroading is when the GM ignores the logic of the setting or the game rules (or just browbeats the players OOC) into following a script.

If you know the players will do A->B->C->D->E, it's a railroad. It's the script that makes it so, not the means for getting players on script.

And scripts are totally okay, if the group is into that.

It is not setting up a scenario where the players have limited options, althiugh that may well be a bad game.

Limited options? No, of course not.

One option? Yes.

For example, one time I heard a guy on the white wolf forums gitching about the iron will merit, because he couldn't motivate the players to accept going on missions without mind control. In my oppinion having an NPC mind control the PCs isnt railroading (unless you are ignoring the iron will merit or otherwise cheating / deforming versimilitude) but it sure as heck makes for a completely bad and unfun game.

If the players have no choice as to what they will do, yes, it is railroading. You seem to be defining railroading as "bad illusionism". Bad illusionism makes the railroading apparent, but the railroad was there already. Arguing that the railroad would be perfectly fine if players just didn't notice is not something that anti-railroad people really find acceptable.

If you're gonna railroad me, do so. Do it loud and proud. And I'll either get on the train or not play your game. And that's what that GM should have done. "Guys, this game is going to follow a story. You're going to get missions and directives, and for things to work you just kinda need to go along with that. Trust me."

Either that or figure out what actually was interesting to the players, and use that to create situations that the players found compelling, rather than just trying to force the players to do what the GM found compelling.

But, in this case, based on your description, I'm not entirely sure that the player doesn't just hate losing and find Internet Arguments to explain that, regardless of how appropriate they are. But, again, I'm not at your table. But your monologuing example is kinda weird, unless the NPCs in question were completely info-dumping for a long time.

Unfortunately, a lot of people in RPG circles are also bad at just saying "I don't like this" and feel the need to "prove" it in some way, often by pointing at some kind of guiding principle. "I'm not happy about losing and I felt the fight was too tough, but I don't think that just saying that is enough, so it's because... railroading! Yeah, that's it. We couldn't win so that's railroading!" But again, the key here is what the player was actually unhappy about, not the word that they used.

[Gallowglass]

Talakeal,

Wait. Is this the same ******* as the penny-pincher from the other topic you posted?

Cause, if so, man you gotta start tuning that turd out. The same way he tunes out your "monlogues."

I mean, that guy is never going to be happy, is always going to complain and is completely bat-**** insane.

So move on and ignore him. Stop letting him trigger your self-conscious need to be always liked.

The answer to every question you ever ask about him will be "he's a ****."

[Talakeal]

Talakeal,

Wait. Is this the same ******* as the penny-pincher from the other topic you posted?

Cause, if so, man you gotta start tuning that turd out. The same way he tunes out your "monlogues."

I mean, that guy is never going to be happy, is always going to complain and is completely bat-**** insane.

So move on and ignore him. Stop letting him trigger your self-conscious need to be always liked.

The answer to every question you ever ask about him will be "he's a ****."

Yesh, but I lack a thick enough skin to ignore him entirely, and I keep thinking that even a broken clock is right twice a day.

While nailing down a formal definition can be difficult, I think the general gist is "if there's a preplanned series of events that the players have to go through, and that's the design of the game." Some people might call that a "linear adventure" and say that only if the players push against it is it railroading, some people might call the whole thing railroading and further divide into participationism (you willingly get on the train) and illusionism (you're told there's no tracks, but there totally are, and the GM tries to conceal that).

But overall, the crux is really "you'll do A, then, B, then C, then D. And your choices don't change that."



No, swerves can happen. Again, it's often a *red flag* of railroading (the GM wants x to occur, so they do something the players want, then force what they actually wanted to occur). But it's not inherently railroading.



I think setting up an outcome of a fight is bordering on railroading, depending on specifics. My personal preference is that the GM not know how a given encounter/fight will go. And, no, a slim chance of it not going that way doesn't really change that.



If you know the players will do A->B->C->D->E, it's a railroad. It's the script that makes it so, not the means for getting players on script.

And scripts are totally okay, if the group is into that.



Limited options? No, of course not.

One option? Yes.



If the players have no choice as to what they will do, yes, it is railroading. You seem to be defining railroading as "bad illusionism". Bad illusionism makes the railroading apparent, but the railroad was there already. Arguing that the railroad would be perfectly fine if players just didn't notice is not something that anti-railroad people really find acceptable.

If you're gonna railroad me, do so. Do it loud and proud. And I'll either get on the train or not play your game. And that's what that GM should have done. "Guys, this game is going to follow a story. You're going to get missions and directives, and for things to work you just kinda need to go along with that. Trust me."

Either that or figure out what actually was interesting to the players, and use that to create situations that the players found compelling, rather than just trying to force the players to do what the GM found compelling.

But, in this case, based on your description, I'm not entirely sure that the player doesn't just hate losing and find Internet Arguments to explain that, regardless of how appropriate they are. But, again, I'm not at your table. But your monologuing example is kinda weird, unless the NPCs in question were completely info-dumping for a long time.

Unfortunately, a lot of people in RPG circles are also bad at just saying "I don't like this" and feel the need to "prove" it in some way, often by pointing at some kind of guiding principle. "I'm not happy about losing and I felt the fight was too tough, but I don't think that just saying that is enough, so it's because... railroading! Yeah, that's it. We couldn't win so that's railroading!" But again, the key here is what the player was actually unhappy about, not the word that they used.

I personally think there is a fundamental difference between a linear adventure and a railroad, and I don't think illusionism plays into it.

To me a linear adventure is fine, and for many groups its actually preferable. My PCs, for example, actually like adventures a good deal more linear than I do, because they suffer from severe analysis paralysis (and memory issues) and frequently end up bored and squabling if they arent presented with one correct right answer.

Railroading, in my mind, occurs when the PCs are trying to escape from the confines of the adventure for whatever reason, and rather than going with the flow and adapting their adventure to its new instrad fight tooth and nail to actively push the players back onto script.

[DMThac0]

Railroads have two directions, forward and backward, no other choices. Once you are on the rails you cannot turn if the rails do not turn, you cannot switch to a different rail unless there is a specific point that allows you to switch. If there is something on the rails that is heading backward relative to your forward there is no way to avoid a collision. If you stop moving on the rails, nothing happens, you can only then choose to continue heading forward or you can head back the way you came.

It is possible to make everything about an TTRPG game seem like a railroad.

A "DM wanting things" as a definition of railroading is a difficult stance to keep, as we have seen. From the moment a DM chooses a module or crafts a homebrew, we are inherently seeing railroads form by this definition. They both "want" the players to participate in the adventure that is written, you are "railroaded" into following that adventure otherwise there is no forward movement.

A "sandbox" game where the DM doesn't do anything unless the players choose to is a recipe for chaos, and forces the DM to create "rails" if there is nothing happening. The DM cannot give the players a setting because that is a "want", cannot create a plot arc because that is a "want", cannot make any decisions prior to the players' input because that is a "want". If the players do not make any decisions, then the DM must create something for the players to do, which is "railroading" since you're forcing the players' hands.

The fact that you are using the rules/guidelines presented by the books is, inherently, "railroading" since you are forced to follow them. The flow of combat "railroads" your players into only being able to perform specific actions. The classes "railroad" your players into only being able to accomplish certain acts as described by those classes. The races "railroad" your players into certain abilities, you're "railroaded" in how weapons and spells are handled.

----

There is a need for rails in a game, it helps, it's what makes the games work. The problem is that some people are unwilling to relax and allow for the players to explore more than what they, the DM, are willing to work with. Whether it be from inexperience, an unhealthy need for control, or a misunderstanding of the social contract, it's a problem. On the PC side it's more often used as a weapon, and a misunderstanding of what is actually going on in the story, or displeasure in not being able to succeed in whatever they were trying to do. The resolution is as simple as talking it out and explaining both sides of the situation. If neither side is willing to budge, then it creates conflict and may indicate a need to remove one of the offended people. If both parties are willing to listen and come to, at the very least, a compromise, then the game can grow into something everyone can enjoy.

[kyoryu]

I personally think there is a fundamental difference between a linear adventure and a railroad, and I don't think illusionism plays into it.

To me a linear adventure is fine, and for many groups its actually preferable. My PCs, for example, actually like adventures a good deal more linear than I do, because they suffer from severe analysis paralysis (and memory issues) and frequently end up bored and squabling if they arent presented with one correct right answer.

Railroading, in my mind, occurs when the PCs are trying to escape from the confines of the adventure for whatever reason, and rather than going with the flow and adapting their adventure to its new instrad fight tooth and nail to actively push the players back onto script.

And this is why I hate these terminology things.

At the end of the day, I think we both agree on the following things:

1) Some games are arranged as a prewritten list of things the players will do: A->B->C->D and so on and so forth.
2) Some people like this and that is cool.
3) Some people do not like this and that is also cool.
4) When people don't like this type of game, and they play in one, they usually push back against the path. If the GM pushes them back onto it in whatever way, that leads to very uncool situations.

(I mean, I think that the issue actually occurs when people say "I don't like that" and the GM does it anyway, but that point has a baffling to me amount of contention).

A "sandbox" game where the DM doesn't do anything unless the players choose to is a recipe for chaos, and forces the DM to create "rails" if there is nothing happening. The DM cannot give the players a setting because that is a "want", cannot create a plot arc because that is a "want", cannot make any decisions prior to the players' input because that is a "want". If the players do not make any decisions, then the DM must create something for the players to do, which is "railroading" since you're forcing the players' hands.

The fact that you are using the rules/guidelines presented by the books is, inherently, "railroading" since you are forced to follow them. The flow of combat "railroads" your players into only being able to perform specific actions. The classes "railroad" your players into only being able to accomplish certain acts as described by those classes. The races "railroad" your players into certain abilities, you're "railroaded" in how weapons and spells are handled.

There are ways to play that completely avoid being either railroads or "empty" sandboxes.

[DMThac0]

There are ways to play that completely avoid being either railroads or "empty" sandboxes.

I believe I said that in the last paragraph:

TL;DR: Rails are necessary for a game to work, as is being able to go off rails when necessary. It is a poor DM or Player that makes railroading bad.

[Resileaf]

See, this is why I prefered my analogy of roads, with highways and side roads (and the occasional off-road venture). It's way more versatile a description when you can both have a main road to take and side roads to use to do things differently than expected.

[King of Nowhere]

(Letters added)

I think A and C are, practically, equivalent. If your actions aren't impacting the plot (especially if "the plot" is defined as a sequence of things the players go through), then you're forced to do things. And if you're forced to do things, then your decisions aren't impacting the plot because you're not really making any.

There's still much difference if the railroading is not taken to the extreme. if your actions aren't impacting the plot, you are still free to choose what your character does. this can happen because of in-world reasons, so it can be done in moderation. It borders on not-railroading.
On the other hand, if the DM tells you which actions you are taking, the DM is taking away your agency. While politely asking is ok ("uhm, guys, that's the only adventure I prepared, you can refuse it but know i will have to improvise otherwise") forcing the players is not ok.

As an example, I will use my own railroading:
- the high cleric of vecnawas trying to ascend to godhood by draining the soul of everyone in a city. the party was trying to stop him.
now, I knew that, regardless of the fight outcome, the ritual would have fizzed, and it would have blasted everyone in the room with godlike essence. All those dead or destroied would have been resurrected on the spot, and everyone would have developed some minor divine ability. I had yet to decide which was for the players, but I knew the vecna guy would get an ability to ignore antimagic fields. I also knew that, with the ritual failed, the guy (regardless of whether he won or lost the first time) would have no interest in taking risks to fight the party again, so he would escape. He had plenty of quickened teleports and the capacity to cast quickened disjunction. Also, he got a sort of demilich form, without most of the demilich powers (would be way too strong for the power level of the campaign) but with the fast fly. and his freedom of movement now worked in antimagic.
So, I basically railroaded the fight into being a phyrric victory for the good guys. The villain was not going to complete his ritual, and he was going to escape; the only difference was that by winning the battle, the party got better rewards. I admitted that with the players afterwards.

However, that outcome stemmed from the way I built the magic ritual. It made sense in the situation. It served to bring in the major conflict of the campaign, a conflict between forces that vecna had been rallying for centuries and everyone else.
The players were very happy with the session.

Now, suppose that instead I had said "ok, and just as you are about to defeat the villain, he escapes"
"But i stop him with [action X]"
"You don't do that. He escapes"
Now, that would have sucked.

So, maybe in the end contriving circumstances so that some things happen regardless of the players is just railroading with a better camouflage. Still, there is a world of difference.

[Gallowglass]

There's still much difference if the railroading is not taken to the extreme. if your actions aren't impacting the plot, you are still free to choose what your character does. this can happen because of in-world reasons, so it can be done in moderation. It borders on not-railroading.
On the other hand, if the DM tells you which actions you are taking, the DM is taking away your agency. While politely asking is ok ("uhm, guys, that's the only adventure I prepared, you can refuse it but know i will have to improvise otherwise") forcing the players is not ok.

As an example, I will use my own railroading:
- the high cleric of vecnawas trying to ascend to godhood by draining the soul of everyone in a city. the party was trying to stop him.
now, I knew that, regardless of the fight outcome, the ritual would have fizzed, and it would have blasted everyone in the room with godlike essence. All those dead or destroied would have been resurrected on the spot, and everyone would have developed some minor divine ability. I had yet to decide which was for the players, but I knew the vecna guy would get an ability to ignore antimagic fields. I also knew that, with the ritual failed, the guy (regardless of whether he won or lost the first time) would have no interest in taking risks to fight the party again, so he would escape. He had plenty of quickened teleports and the capacity to cast quickened disjunction. Also, he got a sort of demilich form, without most of the demilich powers (would be way too strong for the power level of the campaign) but with the fast fly. and his freedom of movement now worked in antimagic.
So, I basically railroaded the fight into being a phyrric victory for the good guys. The villain was not going to complete his ritual, and he was going to escape; the only difference was that by winning the battle, the party got better rewards. I admitted that with the players afterwards.

However, that outcome stemmed from the way I built the magic ritual. It made sense in the situation. It served to bring in the major conflict of the campaign, a conflict between forces that vecna had been rallying for centuries and everyone else.
The players were very happy with the session.

Now, suppose that instead I had said "ok, and just as you are about to defeat the villain, he escapes"
"But i stop him with [action X]"
"You don't do that. He escapes"
Now, that would have sucked.

So, maybe in the end contriving circumstances so that some things happen regardless of the players is just railroading with a better camouflage. Still, there is a world of difference.

So, let me ask you a question.

Let's say during the fight, you had a player who just absolutely HATED the guy you had pinned to escape after turning into a demi-demilich. And concentrated all his devious concentration on going after that guy instead of concentrating on disrupting the ritual. And you weren't expecting that.

Let's say that guy had a plan you didn't expect.

DM: Okay Brad, your turn...

Brad: Okay, I'm going to use my quickened dimension door ability (*) to teleport here, right behind Lucius Badguyov then backstab him. *rolls* hey nat 20! *rolls* oh wow, that's really good. 112 damage!

DM: *checks bad guy sheet. Has no defense against that particular attack, no reasonable way to survive it. has 98 hp left*

*please, for the nature of this argument, don't reply with how Lucius Badguyov would've escaped that particular scenario. Or explain that dimension door doesn't work that way. I know that. Please fill in the blank with any particular scenario that could/would have worked if tried. (and, as an aside, if your response is "well there's nothing the PCs could have done that would have worked because of how I built the scenario and Lucius' defenses" then you have a problem and we won't ever get anywhere by discussing it)

In this case would you have had Lucius die and moved forward without him turning into a demi-demilich and returned for future games or would you have found a convoluted way to keep him alive or bring him back for the role you intended for him?

Because, I would've reward the player who OBVIOUSLY wanted that win and took the effort to play outside the box with his well-earned victory. Lucius is dead. Go Brad. Then figured out some other, new, unrelated badguy "Mucious Worseguystan" to fill in the future role I need Lucius for.

But if you would've "fudged" to keep Lucius alive, then that would -feel- like railroading to the player (because it was)

And if you would've found a way to bring Lucius back before he was needed next time (Bad Guy resurrects him), well that still would have -felt- like railroading to the player (even though it wasn't. It was the legitimate thing the bad guy would have done.)

Ultimately, like many people has said. Its a matter of knowing your table. Some players would be super pissed to have Lucius come back after working so hard to kill him because it feels like it takes their agency away. Some would be happy to have a recurring villain return so they can kill him again. That lucius. What an asswipe.

[King of Nowhere]

So, let me ask you a question.

Let's say during the fight, you had a player who just absolutely HATED the guy you had pinned to escape after turning into a demi-demilich. And concentrated all his devious concentration on going after that guy instead of concentrating on disrupting the ritual. And you weren't expecting that.

Let's say that guy had a plan you didn't expect.

If the plan is a good one, I let it work.
vecna dude has a few lieutnants that could become bosses of the coalition if he was dispatched. none of them is powerful enough to challenge the party on his own, though, so they'd have definitely gotten rid of the major boss earlier.
And I would have strenghtened the vecna coalition a bit to keep the competition going. It's that, or tell the party "ok, you won the campaign. there is literally nothing else that can give you real trouble in the world". that, or I would have had some of the other groups of potential villains I know of join them.

The way I plan the campaign could be likened to a valley surrounded by mountains. the players are encouraged to follow the river, and if they do nothing they are transported by the flow of the current. They can also choose to move around the valley, deviate from the river (which, in this metaphor, is the planned course) easily as long as they don't deviate too much. However, the more they deviate, the more they have to climb. If they sit, gravity will move them towards the river gradually.
If they really put effort into climbing the mountains, though, there is no hard barrier to it.

But besides being reasonably certain that I had covered all angles, one of my major safeties here is that I put effort into making him a cool and interesting villain. I'm sure the party preferred to have him as an arch enemy, rather than an obscure underling that hasn't even gotten a name yet. If that was not the case, then I failed at creating a compelling plot.
And besides sprining the occasional surprise (that ritual came completely out of the blue, the church of vecna had been pretending to be reformed until that point, and it even helped the players a couple of times), I generally discuss with my players the direction they prefer to take.

Which makes me think of another layer in this discussion: that of players wanting, or not wanting, to follow the DM.
If the DM creates a compelling story that interests the players, the pplayers want to follow it. They will not mind being railroaded a bit, because they actually want to experience the cool stuff that the DM has planned.
If the players are not interested in the story that the DM is telling, then they will try to go off the rails. A bad DM will try to keep them boxed. A good DM will try to figure out the problem. Better players, however, may simply tell the DM "look, we don't like this plot, can we move to something else?"

This whole conflict on railroading is born on the premise that the players cannot trust the DM, nor the DM can trust the players. that the DM wants to have the players as spectators, while the players care nothing for the world or the plot and only want to mess up with the DM.
If there is open discussion and reciprocal trust, DM and players will move in tandem like a group of dancers to push the campaign in the direction everyone agreed, and there is no conflict.
I think I finally found out what cheeses me in those extremistic stances on railroading.

[Tajerio]

You slept 3 days, the sun came up and set 3 times, deal. Following game physics is not railroading; ignoring game physics is.

So, if the GM wanted you to witness the next sunset, and had the sun just hang there in the sky while you slept 3 days, yes, railroading.

And there you've lost me. In what meaningful way does "the world had a history, the toys in the sandbox have depth" inhibit the players from meaningfully engaging game physics?

Yes, I'm against the GM deciding where the game will go.

Wow, quite the opposite. The GM's rails are flat and lifeless, whereas the players' story is anything but boring.

Well, usually.

Most of what I might have said in response has already been well put by DMThac0.

The part I find bizarre is that you agree that the world should have a history, and that its elements should have depth. But then you don't concede the logically necessary consequence, which is that the existence of those things constrains the players' actions. Not all things are possible, some developments are more likely than others, and NPCs will do things that change conditions for the players, in part because of the narrative elements that a world with verisimilitude and the ability to engage the players requires.

Unless you don't care about verisimilitude at all/aren't human, there will never be a game that satisfies the conditions of being an interesting sandbox, being run impartially, and valorizing player agency to the maximum.

[NorthernPhoenix]

I railroad all the time, there's really no reason for me not to. However, I never "force" players to do anything or constrict their choices or actions. It's Quantum Ogre every time. Why does this matter? Because "traditional" railroading involves moving the players to the plot (or any "thing"). With Quantum Ogre, it's always the plot that moves to the players. The difference in perception is what matters.

[Cygnia]

I think a big part of it is if you're using the carrot or the stick approach. As I like to say: "If I have to railroad, it's gonna be with a well-stocked dining car".

[Pippa the Pixie]

If these seem like identical situations, then it won't yet be possible to understand the distinction between a DM whose NPCs want something and therefore act, versus a DM who wants something and therefore has NPCs act.

Look identical to me.

1.The NPC bandits want to rob the PCs.
2.The DM wants to rob the PCs....with the bandits

So how can a player...ever....know if a bandit is just a bandit ''being a bandit" or if the bandit is some sort of set up the DM is using to get what they want?

Railroading, in my mind, occurs when the PCs are trying to escape from the confines of the adventure for whatever reason, and rather than going with the flow and adapting their adventure to its new instrad fight tooth and nail to actively push the players back onto script.

TO my mind, only a crazy player would even want to ''escape" from an adventure. For a player to say they don't want to go on an adventure makes no sense. And if the player really wanted to ''escape" the adventure so baddlly....well, they could just sty home and watch TV or something.

The ''adapting" part really does just show that Railroading is a thing for Clumsy DMs. Like say the DM wants ''X", and the players go ''escape crazy" to not do "X". Well, it really does not matter. At all. If the DM wants ''X" to happen: then it will. There is nothing the players can do....other then ''really" escape from the game and leave.

Let's say during the fight, you had a player who just absolutely HATED the guy you had pinned to escape after turning into a demi-demilich. And concentrated all his devious concentration on going after that guy instead of concentrating on disrupting the ritual. And you weren't expecting that.

Let's say that guy had a plan you didn't expect.

That last part is very important. And it's something that a good, expereinced DM does not let happen. For such a DM there is not unexpected stuff. In general, the players are not going to think of crazy unknowable things.....they are much more just going to go for Obvious Thing One, or Two, or Three. Yes, the DM might not know the details, or the ''how" the players will do something: but they can see the general idea.

Now the plan part. So the DM had the plan that NPC A would be a reaccoring lich villain. So the players do Obvious Thing Three...and kill NPC A.

Ok? So what. The DM still has...oh, a trillion OTHER NPCs in the world that they can make in to a reoccoring lich villain. So, the DM can just do that.

And even if the DM really wanted NPC A to come back...well, they could do that too. In just about any game, there are at least a couple fantasy fiction ways to do it. A clone...a ghost...a robot...a construct...a time traveler...and so on. So a DM can ALWAYS do one of them.

[NichG]

Look identical to me.

1.The NPC bandits want to rob the PCs.
2.The DM wants to rob the PCs....with the bandits

So how can a player...ever....know if a bandit is just a bandit ''being a bandit" or if the bandit is some sort of set up the DM is using to get what they want?

Well, let's go back to the case of the novelist. Can you see a difference between a story in which the protagonist visits the poor and under-policed outskirts of a crumbling kingdom and has to deal with bandit attacks and a story in which the protagonist comes into super powers and the next day gets to show them off by handily defeating a bunch of bandits that attack them?

For a more obvious example, murder mysteries. Everywhere Hercule Poirot goes, someone is getting killed.

[The Insanity]

I railroad my players... into having fun.

[Pippa the Pixie]

Well, let's go back to the case of the novelist. Can you see a difference between a story in which the protagonist visits the poor and under-policed outskirts of a crumbling kingdom and has to deal with bandit attacks and a story in which the protagonist comes into super powers and the next day gets to show them off by handily defeating a bunch of bandits that attack them?

For a more obvious example, murder mysteries. Everywhere Hercule Poirot goes, someone is getting killed.

Sure....thoes two are diffrent, Somehow?

Are you talking about the player permission thing? Like:

A.Ok, DM, I am officaly stating that my character crosses the offical border into the officaly marked Bandit Lands. As I have had my character enter that lands freely of my own will, I hearby give you, the DM, limited premission to occasianly have my character be attacked by bandits for as long as they are in the Offical Bandit Lands.

And

B. Bandits exist in the game world. Everywhere. At any time...in any place...it's possible the character might encounter bandits. And it does not matter what the player thinks or does.

[Malifice]

A certain amount of railroading is expected and indeed encouraged.

The DM has designed encounters during the week for his group. Simply purchasing an adventure and prepping it is 'railroading' seeing as the expectation is the players will participate in that adventure, and they will be nudged in that direction by the DM (via hooks and similar DM tactics).

If the players dont buy into those hooks and decide they dont want to participate in that adventure (or story) however, they shouldnt be 'forced' into it.

The DM should instead have a re-think about his DMing style, his understanding of what his players want in the game.

You dont put encounters or an adventure in front of your players (as DM) if you dont think they'll enjoy it, or participate in it. It's more than just a total waste of time for the DM, it's a massive Red flag he doesnt know his players.

If they refuse to engage with that adventure or series of encounters, it's a message to the DM that either he doesnt know his players, or he's doing something wrong.

Nothing is more frustrating as a player as when your actions and choices dont matter. In super railroady games, it literally doesnt matter what you do, or what actions or choices you make, the same thing happens. You can have your character even sit down and do nothing, and the next 'scene' just happens, and you just stay on those tracks.

A good DM knows how to make the rails invisible, even when running Adventure Paths and similar 'story' campaigns.

We've all played with bad DM's and good DM's. It's ultimately down to him or her as to how invisible railroading is, and how well they handle things when PCs get off those rails and do something unforseen.

[NichG]

Sure....thoes two are diffrent, Somehow?

Are you talking about the player permission thing?

Nope. The reason I'm using novels as the example is that there's no player, so agency and permission issues are set to zero.

In the former case, the author is extrapolating from premises that have been established and I can borrow the author's work in thinking through consequences to reach some new insight, understanding, etc. If instead of doing that, the author decides where they want the story to go first, then where the story ends up doesn't follow from established causes, so I can't borrow the author's brain in the same way anymore.

If I read a story about four orphans, three of whom steal bread to survive but then each have misfortune befall them (let's say they all get cancer), while the fourth suffers but is eventually adopted and becomes wealthy, should I conclude that stealing causes cancer and starvation causes wealth? No - the author presented things as natural consequences in order to convey their own ethical system, and the story has no bearing on medical fact. If I read realpolitik fiction about the difficulties faced by minor participants in the Cold War era, can I learn anything about politics, espionage, etc? Maybe, maybe not - it depends on the degree to which the story is shaped by real considerations, versus shaped by trying to include certain external considerations.

Up to this point, agency is a non-issue. But there's still a difference in the relationship between reader and author in those cases. The confusion in the railroading discussion is that things which feel to people like railroading don't necessarily have to center on questions of agency first and foremost. But at the same time, agency can be influenced by this difference.

If I'm playing in the game of the first of those two authors, then I can take seriously the proposition that the logical consequences of my actions (from the DM's point of view) will follow, and I can plan around that fact. I don't want to deal with bandit attacks, so I help shore up the crumbling central government, and consequently bandits don't attack me. The more consistent the DM, the further my planning horizon could in principle go.

For the second author's game, actions with planned consequences tend to fade out and be replaced with metagame considerations. I can't actually cure the 'random encounter' problem because the players get bored if there's no combat for 10 sessions straight, so something is going to attack the group. Or, if we take out the BBEG in session 3, there will be a secret bigger BBEG behind them.

That's the consequence to agency, but agency isn't the root difference.

I will note here, I'm not advocating for or against either of these styles, just trying to clarify what seems to be a point of confusion about the declared stances.

[Quertus]

NichG - again, clearer than my explanation. Kudos!

A "DM wanting things" as a definition of railroading

Is something noone has advocated. Either address the actual contention, or this is a willful misunderstanding / strawman.

is a difficult stance to keep, as we have seen. From the moment a DM chooses a module or crafts a homebrew, we are inherently seeing railroads form by this definition. They both "want" the players to participate in the adventure that is written, you are "railroaded" into following that adventure otherwise there is no forward movement.

A "sandbox" game where the DM doesn't do anything unless the players choose to is a recipe for chaos, and forces the DM to create "rails" if there is nothing happening. The DM cannot give the players a setting because that is a "want", cannot create a plot arc because that is a "want", cannot make any decisions prior to the players' input because that is a "want". If the players do not make any decisions, then the DM must create something for the players to do, which is "railroading" since you're forcing the players' hands.

Either you honestly believe this, in which case I doubt talking to you would be productive, or you don't believe this, in which case case I doubt talking to you would productive.

But I guess I'm not a Boov. I'll give it a shot.

A sandbox game where every actor follows their desires and game physics, modified by the players actions, is not a recipe for chaos, nor does it necessitate the GM to create rails for anything to happen.

Are you honestly incapable of comprehending this concept? Can you honestly not comprehend that the thousands of ingredients at the grocery store did not railroad me into making a PB&J? Can you honestly not comprehend that the bread would go bad much faster than the other ingredients because of "physics", not because of rails? Can you honestly not comprehend a sandbox with active toys which follow game physics which do not require the construction of rails?

The fact that you are using the rules/guidelines presented by the books is, inherently, "railroading" since you are forced to follow them.

Nope, completely wrong. Being forced to follow game physics prevents railroading, as railroading is defined as ignoring game physics.

I believe I said that in the last paragraph:

TL;DR: Rails are necessary for a game to work, as is being able to go off rails when necessary. It is a poor DM or Player that makes railroading bad.

Rails are not necessary for a game to work, as evidenced by working game without rails.

Railroading is pretty much inherently bad, like murder or slavery is inherently bad.

Conclusion: you are using your words differently than a) I am, and b) I suspect most Playgrounders are (although I could be wrong on b).

If the plan is a good one, I let it work.

That's your good. Well, conditionally - if "is good" is defined by game physics, not by your subjective values, unless a) we're in "rule 0" territory, or b) the players have bought in to a narrative game.

If they really put effort into climbing the mountains, though, there is no hard barrier to it.

This is also your good. Just to check, though - if a player had a character who can metaphorically fly, would you invent new obstacles? That is, is it the effort, or the logical ability to get there, that lets one bypass the mountains in your game structure?

But besides being reasonably certain that I had covered all angles, one of my major safeties here is that I put effort into making him a cool and interesting villain. I'm sure the party preferred to have him as an arch enemy, rather than an obscure underling that hasn't even gotten a name yet. If that was not the case, then I failed at creating a compelling plot.
And besides sprining the occasional surprise (that ritual came completely out of the blue, the church of vecna had been pretending to be reformed until that point, and it even helped the players a couple of times), I generally discuss with my players the direction they prefer to take.

Which makes me think of another layer in this discussion: that of players wanting, or not wanting, to follow the DM.
If the DM creates a compelling story that interests the players, the pplayers want to follow it. They will not mind being railroaded a bit, because they actually want to experience the cool stuff that the DM has planned.
If the players are not interested in the story that the DM is telling, then they will try to go off the rails. A bad DM will try to keep them boxed. A good DM will try to figure out the problem. Better players, however, may simply tell the DM "look, we don't like this plot, can we move to something else?"

This whole conflict on railroading is born on the premise that the players cannot trust the DM, nor the DM can trust the players. that the DM wants to have the players as spectators, while the players care nothing for the world or the plot and only want to mess up with the DM.
If there is open discussion and reciprocal trust, DM and players will move in tandem like a group of dancers to push the campaign in the direction everyone agreed, and there is no conflict.
I think I finally found out what cheeses me in those extremistic stances on railroading.

See, this touches on what I've learned since joining the Playground: that some people like to be railroaded. Or, perhaps more accurately, like to be guided along a more linear story? I never used to believe that was possible until I started posting here.

Anyway, I've not had a group in quite some time that didn't seem allergic to communication. And, before that, I was mostly looking at gaming with either like-minded individuals, or ones willing to accept, Participationism style, my fanaticism against rails. So I can't really relate from experience.

I can only say that I'm not at all interested in the GM's story - I'm here to see what story these particular characters will tell with the GM's content.

Most of what I might have said in response has already been well put by DMThac0.

The part I find bizarre is that you agree that the world should have a history, and that its elements should have depth. But then you don't concede the logically necessary consequence, which is that the existence of those things constrains the players' actions. Not all things are possible, some developments are more likely than others, and NPCs will do things that change conditions for the players, in part because of the narrative elements that a world with verisimilitude and the ability to engage the players requires.

Unless you don't care about verisimilitude at all/aren't human, there will never be a game that satisfies the conditions of being an interesting sandbox, being run impartially, and valorizing player agency to the maximum.

... I honestly cannot decide if you're being serious or not. Let's pretend you are serious.

The GM saying, "D&D, 3e, 15th level" greatly constrains my character options. But I don't scream "railroad". Playing chess constrains me from landing on Park Place, but I don't scream "railroad". Following game physics is not railroading. Constraining player options by Park Place not existing on Toril is not railroading.

Player Agency may also need defining (is everyone secretly DU? I jest - I have trouble with subtle (or not so subtle) points of definitions myself), but it's not being violated by things having history and logic and physics - in fact, I'd argue that it's being enabled and validated by the existence of such things.

The maximum player agency occurs in a sandbox being run impartially. I'm lost how you could try to write that equation otherwise.

[King of Nowhere]

Wow, quite the opposite. The GM's rails are flat and lifeless, whereas the players' story is anything but boring.

Well, usually.

This makes me think you had bad experiences in the past with some DM, and you became allergic to a DM-driven story as a result.
Because I don't see any reason the players driving a story would be intrinsically better than the DM driving the story. In fact, if several players have different ideas, they may spend half of each session arguing over it.
Myself, I had passive playes, so I became convinced that you can't go anywhere if you don't nudge them here and there.

This is also your good. Just to check, though - if a player had a character who can metaphorically fly, would you invent new obstacles? That is, is it the effort, or the logical ability to get there, that lets one bypass the mountains in your game structure?

That's... complicated to answer. I'll try my best.

I would say that you have the relation obstacle-capacity backwards here. I build the obstacles around what the characters can't do.
One of the first, and most important things I determine for a world is "what can't a high level party do?". I follow that line of thinking because I find limitations, and ways to circumvent that, are interesting. A character walking in a straight line without any obstacle that he can't overcome is boring. A character finding an impassable obstacle and having to circumvent it somehow is better. For some deeper musings on this argument, I invite everyone interested to read sanderson's second law of magic. That, and the other two laws, are very sensible principles that have a deep influence on my worldbuilding.

So, I establish those things that cannot be done, at least not directly.
Once I do that, I automatically set the power level, and possibly I do a bit of houserulings, to make sure of it. Once those limitations are established, I preemptively ban or nerf anything that would break them. Those limitations are part of how the campaign world works. And if I were to describe my houserules to new potential players, I would not make a list of rules; I would make a list of limitations, and then possibly explain what they entails.

You could also say that when the fluff and the number crunch don't align, I change one or I change the other to make them align. And if I established the fluff earlier, I change the crunch.

All this means that if some player finds out a way to bypass the mountains, I only allow it if the mountain was not one of the things established as impossible. If it was established impossible, then any attempt to bypass must have some counter to it, least the whole premise of the world collapses. I do not think this is related to railroading, simply to establishing some principles.

For example (I got into an argument with you a few weeks back, which got me quite angry; but now I am contextualizing your point better, and I am also figuring better answers), I established that in my world the greater powers would have some extremely fortified magical strongholds that would be capable of repelling even a high level party. They may be taken with great numbers and effort, but a single adventuring party cannot, period.
That's before I even considered how they were made.
So after I figured out a bit how they were made, if a player were to bring a new obscure spell and say "with this spell I can break in easily", I have to say that either the spell does not exist, or it was specifically countered, because that's a basic premise of the world. If they bring me a long ritual involving hundreds of casters? that could fly, because it keeps with the "may be taken with numbers and effort". (the ritual of vecna dude weakened the defences as a collateral effect, and I managed to show that; that's how I created the opportunity for the party to defeat him)
Or, I established that a high level party cannot just take over a powerful nation*. A powerful nation has one of the aforementioned strongholds to protect its most valuable possessions and personnel, and it can call allies and hire people and eventually it vastly outpowers a single party. So, if any players says "I walk into the throne room and dominate the king", the attempt to take over the nation is doomed to fail. And I make that clear with the players well in advance. On the other hand, undermining a nation through selective strikes and political manuevers can be done; the party brought a great nation to heels as early as level 12, but they had to pull off a clever scheme.
Of course those obstacles are there for everyone in the world. The players allied themselves with a great nation and use the nation's stronghold as a base, if some enemy managed to teleport in there while they cannot they'd have every right to complain.

* My worldbuilding stemmed from me wondering "how would the world react to the existance of people with superpowers? either it would fall to anarchy, with everyone high level taking what they want, or they would find a way to contain them. Let's assume this second case" and from there I started figuring out what it would take. Impregnable strongholds, inescapable prisons, ability to rally high level help of your own. How could it be done, what principles would it be founded upon, how it would impact the world at large and the goals of adventuring groups specifically.

[Tajerio]

... I honestly cannot decide if you're being serious or not. Let's pretend you are serious.

The GM saying, "D&D, 3e, 15th level" greatly constrains my character options. But I don't scream "railroad". Playing chess constrains me from landing on Park Place, but I don't scream "railroad". Following game physics is not railroading. Constraining player options by Park Place not existing on Toril is not railroading.

Player Agency may also need defining (is everyone secretly DU? I jest - I have trouble with subtle (or not so subtle) points of definitions myself), but it's not being violated by things having history and logic and physics - in fact, I'd argue that it's being enabled and validated by the existence of such things.

The maximum player agency occurs in a sandbox being run impartially. I'm lost how you could try to write that equation otherwise.

You are again picking really obvious and mechanical examples that don't have much to do with the more subtle nature of the issue I was trying to express, but there's at least one point of agreement there--player agency is enhanced by the world having depth.

My point at the end is that you will never have a sandbox run impartially, because anyone who runs it will bring existing biases about realistic paths for the development of institutions and actions of NPCs in the game. If that's covered by your usage of "game physics," then that's more than a little odd, but carry on.

However,

Railroading is pretty much inherently bad, like murder or slavery is inherently bad.

this is the point at which it becomes clear there can be no further profit in engaging you on this point. Either you're arguing in really bad faith, or you've drifted so far from reality that you honestly support that comparison you've made. And in either case, you ought to sit down and have a word with yourself about it.