Judgement calls vs "railroading"

[pemerton]

From this and other things you've posted here it seems your games are a good way down the spectrum towards co-operative storytelling. I'm coming at this from a long history of quasi-Gygaxian dungeoneering where a plot or world backstory sticks its head up now and then, takes one look at the party, and likely runs screaming back into its cave. So, small wonder we're talking in circles.

But I don't think we're talking in circles! You asked, "How would a game have backstory and a 'plot' if the GM doesn't provide it. I answered.

Yes, the answer includes doing things differently from Gygaxian dungeon crawl style. But we've known that, in the context of this thread, at least since [MENTION=16586]Campbell[/MENTION] posted outlining three different approaches to player-driven RPGing (Gygaxian; what he called "scene-framing"; and what he called "principle GMing" - I tend to blur those last two together as "modern" or "indie"-style, but that taxonomic issue shouldn't matter to you because it still makes the contrast with Gygaxian sandbox-style clear.)

And as I replied to [MENTION=996]Tony Vargas[/MENTION] upthread, there is nothing about 5e that stops it being run in a player-driven fashion of an "indie" style, provided that the GM adopts some appropriate a techniques in calling for non-combat checks and setting the DCs for them, and provided that the inspiration mechanic is engaged to a suitable degree.

it seems you're trying to take an overall foundational basis of those systems (i.e. a strong lean toward co-operative storytelling) and shoehorn it into D&D, which traditionally has had a more DM-driven style.

But D&D does not "traditionally" have a more DM-driven style, does it? It does assume GM control over backstory, but not GM control over the events of play. In Gygax's PHB and DMG, he assumes that the players will be the ones who choose what part of the dungeon to target, whether to negotiate, fight or flee from encountered creatures, what equipment to take with them, etc.

Plot, backstory, and game-world history are really all the same thing, with the only differences being time (whether current or past) and level of effect on or interaction with the PCs.

I think it's pretty helpful to distinguish backstory from plot. Designing a dungeon, mapping it, placing all the creature and treasure - that's backstory, which will include some game-world history.

But what the players choose to do when they encounter the dungeon - eg does Robilar free the trapped gods; or Erca's Cousin free Fraz-Urb'luu from imprisonment? - is not something the GM is at lbierty to make up. Those events are initiated and driven by the players, and they are what establish the plot of the campaign.

Of course from the perspective of the ingame inhabitants they are all just events, but that perspective is not useful for analysing the play of the game.

If you hadn't decided ahead of time whether there's a diamond there to be found or not, my statement that I'm looking for one and subsequent success on the roll to find it means *pop* I've just generated a diamond. Do this often enough and hey, who needs to adventure?

<snip>

It's significant to my character's wealth, if nothing else.

I think there is a narrow, more technical answer to this, and then a broader one as well.

The narrow answer: if you declare that you are looking for a diamond in the room, and the check is framed and you fail, then you are going to have to deal with the resulting consequence of failure. (For elaboration, see my reply to [MENTION=6785785]hawkeyefan[/MENTION] not far upthread of this.) And "let it ride" also applies. So the game is not just going to hang, static, waiting for you as a player to roll up diamonds for your PC.

Because no PC in this particular game has ever been on the hunt for diamonds, this particular issue has never come up. But a similar one has: a PC mage had returned with the party to the ruined tower where he once lived studying under his brother's tutelage before they fled when the tower was assaulted by orcs (this was part of the PC's backstory; the brother is the mage who was decapitated in the tower by the assassin, having been possessed by a balrog when the attempt to cast a spell to repel the orcs failed). Another part of the PC's backstory stated that, in the tower, he had left behind The Falcon's Claw, a nickel-silver mace which he was preparing to receive enchantments. Having now returned to the tower after lo those many years, a search was made for the mace. This was framed as a check in the same way as looking for a vessel in the bedroom. The check failed; and so the PC did not find any mace. Instead, he found a collection of cursed arrows in the ruins of what had been his brother's private workroom - this discovery therefore (i) implicating his brother in the manufacture of the arrow that had killed another PC's (the elven ronin's) master; and (ii) implying that his brother was evil before being possessed by a balrog, rather than as a result of it.

So for all sorts of reasons, of which the threat of adverse consequences on failure is just one, it makes no sense for a player to just go around declaring attempts to find stuff that his/her PC wants.

The broader answer is this: the reason for playing a RPG, as I take it, isn't so that one's PC (who is purely imaginary) experiences wealth and pleasure in the fiction (which is all purely imaginary). It's so that you, the player, actually experience, in the real world, the satisfaction of playing a game. If that satisfaction mostly comes from having a really long equipment list full of diamonds, then probably the sort of game I run is not the best for you. But that's not the only way to get satisfaction out of a RPG.

 

[pemerton]

How did you determine that the giants were or were not open to negotiations? That seems like a pretty big stake and something that would be uncertain but it feels like (and I could be wrong here), that you did in fact create an outcome around what you wanted to happen in the narrative. If that isn't the case what differentiates this from the vessel example earlier? Where saying yes or no to the player would be considered a railroad??

Okay I'm unfamiliar with Cortex Fantasy (so I might need some basic explanation of the mechanics), but in the post I quoted you stated you narrated this part (the PC's arriving and being allowed to enter the giant's village) without any dice rolls...right? So how was the d6 Invitation to Enter asset established if there were no dice rolls?

The first actual action of the game, after the initial setting up, was when one of the players declared "I go to the gate of the setting, knock, and call out a greeting." I narrated an outcome without any dice being rolled - a deep voice (correctly presumed by the players to be the voice of a giant) responded gruffly. That is not railroading (as characterised by me in my OP), because it is not a case of me shaping an outcome to fit a preconceived narrative. It is simply "saying 'yes'" to the player - the player wants his PC to open negotiations, and I am letting that happen.

The negotiations themselves then unfolded, and were resolved via a check (or perhaps a series of checks - I don't recall the details now) - the upshot being that the player successfully established that he was invited to enter the steading, and hence the gates were opened.

the player declared actions and was successful. In mechanical terms, he established a d6 Invitation to Enter asset, which he was subsequently able to leverage in his rolls to impose a Persuaded to Help complication on the giant chieftain.

I've reposted the relevant bit of the first post that you quoted from, and my first reply.

As it states: when the player wanted to try and negotiate an invitation to enter, I "said 'yes'" to the opening of negotiations for that purpose ie when the player said "I got to the gate of the steading, knock, and call out a greeting" I didn't respond "A giant spear comes hurtling towards you - roll your reaction pool!" Rather, a gruff voice responded. (I can't remember the response, but it would have been either "What do you want" or "Please go away".)

If the player had failed in the attempt to receive an invitation, then the result of that failure could have been the hurling of spears. But it wasn't.

If the question is, why is it not railroading to allow the player to have a chance at successfully wrangling for his PC to enter the steading?, I think the answer is straightforward: letting a player have a chance to do something s/he wants to do with his/her PC is not a railroad.

it looks like it was created as an action during the Social Conflict to get in. It looks like it was probably part of the dice pool that stressed out the opposition and won the Social Scene for the PCs.

From memory, I think the opening of the gates was, in the fiction, the correlate of establishing the asset mechanically. The action scene didn't actually come to an end until the giant chieftain Loge was "stressed out" by stepping the Persuaded to Help complication up above d12. And yes, the Invitation to Enter asset was part of the pool that achieved that final result - together with a d6 for the giant shaman's advice being given to the chieftain, which had been established by the same player as a Social asset by spending a point to activate an opportunity that I rolled.

 

[hawkeyefan]

Nor do I. Obviously setting a DC affects player success, but I do it all the time, and in my OP distinguished that sort of judgement call from railroading.

I don't know if you made a strong distinction between judgment calls and railroading. You gave an example of a judgment call involving setting a DC, and then you explained why this example was not a railroad because you were not basing your decisions with the intention of maintaining your predetermined course of action.

You then kind of implied that leaving the presence or not of a receptacle for the mage's blood up to the GM's judgment was a railroad. Which it could be...a GM could make his decision based on maintaining what he expected to happen. But that need not be the case. The GM could simply answer yes or no about the presence of a receptacle based on the prevailing conditions established by the fiction without knowing the player's intent....his desire to seethe railroad maintained need not enter into it.

Hence why I would say that in this instance, the DM's judgment to decide yes or no is about the same as it would be to set a DC for the player.

I suppose that we could say that one approach opens up the possibility of railroading more than the other. But then we need to get into value statements about railroading and what the term means and so forth.

The check is to find out whether or not the PC, looking for a vessel in the room, is able to notice one.

Okay, thanks for clarifying. I misunderstood and I thought that the skill check was being used to determine some of the elements of the world rather than simply the PC's ability to interact with the world.

If the check fails, then (as a GM) I have to narrate failure. Obviously if the check is failed, then the character has failed to spot a vessel in the room. But as a general rule there has to be some other consequence that drives the action onward (as the BW website put is, "the consequences for failure lead to the next conflict. There are no dead-ends"). This is the basic premise of so-called "fail forward" or "no whiffing" adjudication.

Sometimes I indicate in advance what these consequence of failure will be; sometimes I leave it implicit in the situation; sometimes I just make something up. The night watch apprehending the PCs after the failed check to lug the bodies through town is an example of the second (ie it is implicit in the situation that if you tire and slow down lugging bodies through a town at night, you might encounter someone, including the watch); but in the case of looking for the vessel, it would really have had to be in category three - I would have made something up. (Handling communication of consequences of failure, and expectations around that, is another important domain of GM judgement calls.)

I don't know what I would have done at the time - I can't remember if I had anything in mind, and it's a bit hard to put myself back into that situation and recapture the feel. But possibilities I can think of now would include (1) "Yep, there's a jug on the table - but as Jabal [the mage whose tower it is] backs away from Halika [the assassin, who having killed the unconscious mage was now trying to escape by cutting down Jabal], he knocks the table and the jar falls to the ground and breaks", or (2) "You can't see any vessel, but in a disgusting display you can see Jabal's familiar, a bony raven-like creature, licking up the blood as it flows under the divan."

(1) would create a new context for decision - eg maybe the PC can summon a spirit that can put the jar back together again. (2) would increase the pressure on the PC to save the blood - not only is there no vessel, but the blood is getting eaten!

Either way, the player has to make a new decision about how (if at all) the PC is going to make sure that the naga gets the blood.

So do you think that one cannot achieve the same results with simply relying on DM judgment? Or is it only less likely? Does a failed check really open so many alternative paths that a DM saying "there is no chamber pot" does not? I think it would vary depending on many factors...but I'm not sure if I would say that it would tend to do so more often.

Do you feel that from the player's point of view there is a significant difference between a failed check compared to the DM simply saying no?

 

[pemerton]

I misunderstood and I thought that the skill check was being used to determine some of the elements of the world rather than simply the PC's ability to interact with the world.

Maybe there is still some misunderstanding?

No one at the table knows whether or not there is a vessel in the room. I haven't described the room in great detail; I certainly haven't drawn a picture of it. There has been no reference to anyone pouring a drink, or using the chamber pot.

The player says (speaking as his character; and subject to the fact that I'm paraphrasing based on recollection), "There must be a lot of blood spilling out! I look around for a vessel - maybe a jug, or a chamber pot - that I can try and catch it in."

I set a low DC, because spotting a vessel of some sort in a bedroom is not a hard thing to do; the roll was made, and succeeded, and so the PC did indeed notice the vessel he was hoping to.

As I said, had the check failed (unlikely, but not impossible - a bit above 1%, from memory) then I would have had to narrate some consequence for failure, perhaps along the lines I described. As you've seen in my examples, (1) involves spotting a vessel but not being able to use it to catch blood, because it becomes broken in the course of the spotting (in the terminology of Burning Wheel, this is described as "successful task, failed intent" - a pretty standard approach to "fail forward" narration of consequences), while (2) leaves the question of a vessel open but tends to imply there is not one there (there is no reason to suppose that someone has hidden a vessel in the room, and nothing has been said to suggest that the room is so cluttered that a vessel might have been missed, and indeed the action that has unfolded so far tends to suggest that the room is fairly sparsely furnished).

So do you think that one cannot achieve the same results with simply relying on DM judgment? Or is it only less likely? Does a failed check really open so many alternative paths that a DM saying "there is no chamber pot" does not?

I'm not sure what you have in mind by the same result.

No doubt, at some RPGing tables, the GM could just narrate one of the things I suggested, without there being a check.

Can that sort of GM narration - what is sometimes called "GM fiat" - yield drama? Yes. As I posted upthread, I've played in CoC railroads which are exciting and engaging, mostly because of the oratorical/theatrical skill of the GM. It's essentially storytelling, and being told a story by a good storyteller can be an engaging experience.

Do I think the check yields more drama? At my table, absolutely. The "fiat" narration would be an instance of what, in the OP, I described as railroading - the GM is shaping an outcome to fit some pre-conception of how events should unfold. (To hark back to my recent reply to - I think - [MENTION=6801286]Imaculata[/MENTION], it counts as an outcome and not just colour or framing because it goes directly to the players' goals for his PC.) Whether that is good or bad is a matter of taste, though my use of the "railroad" label probably reveals what my taste is. And part of what shapes my taste is that I find it more dramatic to learn, with the rest of the table, what happens in the fiction; and I think for my players it is more dramatic to participate in setting the stakes and then invoke the mechanics of the game to find out what happens, rather than simply be told a story by me.

As for the possibility of narrating failure (whether following a failed check, or by fiat) simply as "No, you can't see an vessel" - I think that sort of "dead end" response tends to shut down creative and engaged play from the players, and instead encourages them to look for the "correct" solution that the GM has in mind. In other words, it tends to turn the game into puzzle-solving. I occasionally use puzzles in my game - eg in my main 4e campaign, which recently entered its 9th year, I think I've used one password puzzle and three riddles, where in each case there was a pre-determined answer and the players had to figure it out to get what they wanted for their PCs - but I regard it as very much in the "handle with care" bag of techniques, because of the risk of dead-ending, and certainly wouldn't adopt it as a general approach to adjudication.

Do you feel that from the player's point of view there is a significant difference between a failed check compared to the DM simply saying no?

Yes.

One paradigmatic illustration of the difference would be a TPK in a combat that was (given the ingame context and table expectations) fair, vs the GM just narrating "Rocks fall, everybody dies."

Cashing it out: if there's a check the player is able to bring his/her resources to bear. S/he is playing the game. The GM just narrating failure doesn't allow the player to actually play the game in that sense. (One reason why CoC lends itself well to GM fiat narration is because players have almost no resources, and so the game doesn't set up any expectation that the players will actually play the game in the sense of deploying their PCs in a mechanical fashion.)

 

[hawkeyefan]

Maybe there is still some misunderstanding?

No one at the table knows whether or not there is a vessel in the room. I haven't described the room in great detail; I certainly haven't drawn a picture of it. There has been no reference to anyone pouring a drink, or using the chamber pot.

So the player's roll actually does determine if the receptacle exists, not if the PC sees the receptacle. Or rather, not simply if the PC sees it...the roll determines both the presence and the viewing of it.

The player says (speaking as his character; and subject to the fact that I'm paraphrasing based on recollection), "There must be a lot of blood spilling out! I look around for a vessel - maybe a jug, or a chamber pot - that I can try and catch it in."

I set a low DC, because spotting a vessel of some sort in a bedroom is not a hard thing to do; the roll was made, and succeeded, and so the PC did indeed notice the vessel he was hoping to.

So setting this DC with about a 1% chance of failure is more meaningful than simply saying "yup, it's there"? I mean, clearly the player is up to something by asking....so Justin saying yes would seem to save some time and no the really impact the drama.

As I said, had the check failed (unlikely, but not impossible - a bit above 1%, from memory) then I would have had to narrate some consequence for failure, perhaps along the lines I described. As you've seen in my examples, (1) involves spotting a vessel but not being able to use it to catch blood, because it becomes broken in the course of the spotting (in the terminology of Burning Wheel, this is described as "successful task, failed intent" - a pretty standard approach to "fail forward" narration of consequences), while (2) leaves the question of a vessel open but tends to imply there is not one there (there is no reason to suppose that someone has hidden a vessel in the room, and nothing has been said to suggest that the room is so cluttered that a vessel might have been missed, and indeed the action that has unfolded so far tends to suggest that the room is fairly sparsely furnished).

I'm not sure what you have in mind by the same result.

In your failed check/fail forward example, I don't see how it's the check that really does all that much. In this example, you've said the the result of the roll is pretty much a given. And I don't think that you mentioned anything about the degrees of failure or success based on how far the roll was from the DC...so all the alternatives (the broken chamber pot, etc.) are all still the product of DM judgment.

No doubt, at some RPGing tables, the GM could just narrate one of the things I suggested, without there being a check.

Can that sort of GM narration - what is sometimes called "GM fiat" - yield drama? Yes. As I posted upthread, I've played in CoC railroads which are exciting and engaging, mostly because of the oratorical/theatrical skill of the GM. It's essentially storytelling, and being told a story by a good storyteller can be an engaging experience.

This is where I question your choice to label it as a railroad. So far, I am not seeing a meaningful difference.

Do I think the check yields more drama? At my table, absolutely. The "fiat" narration would be an instance of what, in the OP, I described as railroading - the GM is shaping an outcome to fit some pre-conception of how events should unfold. (To hark back to my recent reply to - I think - @Imaculata, it counts as an outcome and not just colour or framing because it goes directly to the players' goals for his PC.) Whether that is good or bad is a matter of taste, though my use of the "railroad" label probably reveals what my taste is. And part of what shapes my taste is that I find it more dramatic to learn, with the rest of the table, what happens in the fiction; and I think for my players it is more dramatic to participate in setting the stakes and then invoke the mechanics of the game to find out what happens, rather than simply be told a story by me.

As for the possibility of narrating failure (whether following a failed check, or by fiat) simply as "No, you can't see an vessel" - I think that sort of "dead end" response tends to shut down creative and engaged play from the players, and instead encourages them to look for the "correct" solution that the GM has in mind. In other words, it tends to turn the game into puzzle-solving. I occasionally use puzzles in my game - eg in my main 4e campaign, which recently entered its 9th year, I think I've used one password puzzle and three riddles, where in each case there was a pre-determined answer and the players had to figure it out to get what they wanted for their PCs - but I regard it as very much in the "handle with care" bag of techniques, because of the risk of dead-ending, and certainly wouldn't adopt it as a general approach to adjudication.

But the DM need not say no. He can say yes....the equivalent of setting such a low DC as to make the check virtually meaningless. So why bother with the check? I don't think it adds any real drama in this sense.

This is not to say that I always advocate for DM fiat or anything like that. Not even that this should be a general approach. I think there are absolutely times to call for a check in order to determine success or failure. And I like alternative paths being presented instead of a dead end...I just don't see the required connection between the dead end point and DM judgment.

 

[Hussar]

Railroading is something that has always been nebulous to define. It's defined most easily by it's center, rather than its edges - the DM decides that no matter what you do, X will always happen is a pretty easy thing to call railroading. Although, even there, there might be exceptions, but, I think you get the point.

For me personally, I define railroading slightly differently.

Railroading: When the DM defines the outcomes of player actions in such a way that it renders all player choices immaterial, AND, the players object (or would object if they knew the DM was doing it).

So, all sorts of pretty standard DMing techniques - starting a scene in medias res, cut scenes, fast forwarding scenes, etc. are only considering railroading IF the players are unhappy with it. If the players are cool with it, it's not railroading.

 

[pemerton]

The way I see things, there may be a main plot when you retrospectively look back on a campaign, particularly if you focus on one set of players and their PCs in a particular timeframe, but there are also supplementary plots and sequences of events occurring all the time that crisscross with the main plot.

When I talk about railroading I've got in mind mostly what I think you would call "the main plot" - ie the bulk of the action that the PCs were engaged in over the course of the timeframe in question.

in RPGs, plots are mutable, intersecting things, constantly being interfered with by both PCs and NPCs. Yet, since they all advance in time even if we do not interfere with them, they're all plots in the sense that they are all sequences of events causally linked together.

I don't really get the "interfered with".

PCs and NPCs do things in the fiction - but in the fiction there is no "plot", only life (the imagined life of those various imaginary people).

The plot is something that is created by the actions of real people - the players and the GM - and when I talk about railroading I'm focusing on who creates that plot, and how.

if the DM is trying to present a world in which things happen that aren't simply focused around the PCs (or one group of PCs)

<snip>

just because a DM has events planned out, if the PCs can choose to interfere with them or not interfere with them, I don't think it's a railroad. Plots operating in the background that affect the PCs in various ways, whether positively or negatively, aren't railroads if the choices the players make still matter.

If the GM is presenting something to the players, then - unless it is a sheer narrated vignette (a viable technique, but not a very traditional one) - presumably the PCs are somehow connected to it, if only to the extent that they learn of it.

I'm opening with the above paragraph not to be pedantic, but to link that point to a more general one - every RPG world will involve some stuff that is connected to the PCs only in that rather trivial sense, that they learn of it, or that it forms some background in a circumstance where the real focus of the action is something else.

But that sort of stuff - by dint of being background, or only trivially connected to the PCs - is not the plot of the game.

If the GM starts using that sort of stuff to actually determine outcomes of action resolution then its connection to the PCs becomes more than trivial - within the fiction, it is exercising some significant causal power in respect of them. In those circumstances, I do regard it as railroading.

But preparing events isn't the same as using stuff to determine outcomes. A GM might have a list of "Stuff that would be fun/interesting to happen", and then when some action declaration or some new framing context makes it appropriate, s/he takes something from the list. That's not railroading - because the GM isn't shaping outcomes towards something predetermined.

That said, you didn't use the word "preparing" - which is about the GM's "homework". You used the word "planning", which is a bit ambiguous as between preparing material so its ready to use, and sketching out a causal path within the gameworld. The latter needn't be railroading - the GM plans to narrate some event as occurring under such-and-such conditions, but the question of whether the conditions come about is ultimately subject to the outcomes of player action declarations for their PCs. But in practice I think it easily bleeds into railroading, because there can be a strong "pull" on the GM to use the background causal considerations that s/he envisages will bring this planned event to pass as factors in adjudicating player action declarations. And, as I said, that's something that I regard as railroading.

My own preferred approach, therefore, is that when these events, being causally driven by ingame forces, are planned by me, they become known to the players also. Eg in my main 4e game the players (and PCs) know the Dusk War is on the horizon; in my Burning Wheel game, the players (and PCs) know that the mage Jabal is engaged to marry the Gynarch of Hardby. That way these pending events become part of the shared context of action declaration, framing of checks, etc. And it becomes clear to the players how they might declare actions to delay or prevent them (as has happened in both cases, deliberately in respect of the Dusk War and a bit more inadvertently in respect of the wedding).

in my experience "plot" in a literary context has a much broader definition than the dictionary provides, and is no way limited to "the main events". As evidence, consider the usage of the terms side plot and plot weaving, both of which require the presence of multiple plots in a single work, a prospect that the dictionary definition does not allow for.

In an RPG context I think it makes sense to use the broader literary sense. Accordingly, I don't think it's right to use the dictionary definition to reject the concepts of flexible plots, breadcrumbs, or (optional) GM-provided plots within a sandbox.

The multiple plots in a single work are, nevertheless, sequences of events that are interrelated, and presented as such, and in some sense "main" or otherwise salient.

But some bit of backstory known only to the GM is not salient in the same way, in my view. if the audience (ie the players) don't even know of it, I don't see how it counts as plot.

The breadcrumbs, "flexible plots" etc are also not plots. They're not sequences of fictional events. The GM planning, or preparing, to make a certain event part of the shared fiction under certain circumstances, does not actually make it part of the shared fiction. And if it's not part of the shared fiction, then it's certainly not part of the plot of the game.

 

[Campbell]

From my vantage point, the difference between a failed check and a GM simply deciding if my character is successful absolutely matters to me. I am assuming that check was made in good faith and will have an impact on the ultimate resolution. In some cases having a check may be preferable - in others it may not. Generally, I am not really a fan of perception and knowledge checks. If something is in plain view in the fiction or if my character would know some relevant detail I would prefer to simply be told so. This is part of always saying what honesty demands - an important precept of the sort of role playing I prefer. Only by providing meaningful information to players can decisions truly matter.

I also care very deeply why the GM is making the decisions they are making. It carries a lot of water to me if something is being introduced to offer players meaningful decisions or to prod players down a particular path. In general what I am looking for when I play a roleplaying game is to be offered the chance to make informed decisions, have those decisions actually matter, sometimes experience bleed, and give the game its say.

 

[pemerton]

Some posts in this thread seem to be running together GM preparation - which might consist of notes, ideas, plans, etc around what sort of events would be fun or clever or appropriate to introduce into the shared fiction - with the plot of a game - which is an actual series of events that occur in the shared fiction and constitute (subject to qualifications around interwoven plots, sub- or side-plots, etc) the "main action" of the campaign.

But these things must, at least in principle, be distinct - because a plan to make something part of the shared fiction is not the same thing as actually doing that.

In the OP I described railroading as "the GM shaping outcomes to fit a pre-conceived narrative." If the GM is shaping outcomes to ensure that the plot of the game (ie that series of actual events in the shared fiction that amount to the main action) corresponds with his/her preparation/planning in respect of the game (eg his/her scripting of a sequence of events to unfold in the campaign), then in my view that absolutely is railroading. (I put adjudication by reference to backstory known only to the GM - "secret backstory" - in this category.)

(It follows that, if you don't want to railroad in this sense, then some sorts of preparing/planning might be not all that useful.)

 

[Lanefan]

Sandbox 101. Players piss off evil warlord who sends 50 soldiers at them. Players die.

Characters die. Players, I hope, survive long enough to roll up new ones.

Lan-"pet peeve: characters are not players"-efan