No, but a group could need to go somewhere (via agreement to play an adventure), and a player could insist on burning down the docks or sinking the ship. Then, when they get to the stables, the player slaughters or poisons all the horses.
Both are ridiculous. And personally, I have only seen it done by very young players. But both are railroading.
Again, I do not see this as "railroading," because players simply, emphatically do not (in general) have the power to
declare that such events happen. DMs do have that power, consistently.
Like...are you honestly trying to claim that a player can simply say they burn down the docks? That they can just say "I poison all the horses" and suddenly that's definitely absolutely 100% true?
I'm not saying player choice can't be disruptive. It absolutely can. I've dealt with a very small amount of that in my own game (mostly from the two players, one of whom has left the game for unrelated reasons, who had never played TTRPGs before). I see this as either a failure of understanding (which is what it was in my group) or a failure to truly get everyone on the same page for what the game is about (which happened in a friend's game where I'm a player).
But--again--there are key differences between "player who doggedly
pursues disruptive actions" and "DM that does not
permit disruptive actions." The former, yes, can as a consequence change the state of play so that some events are no longer realistically possible, but--
again--that is true if LITERALLY ALL player behavior that has meaningful consequences. If "player burns down the docks" counts as "railroading" solely because it alters the future events of the game, then "player choosing to make a heroic sacrifice with the character" is "railroading" because it prevents all future interactions between the now-dead character and all other characters, PC or NPC. This,
again, means
literally anything that causes consequences with any degree of permanence is "railroading" your party, which is obviously nonsense.
Railroading has a valid meaning: It isn't just "bad DM is bad," it very specifically means "setting a fixed--not merely unchanging but
unchangeable, route through the narrative of the game." It isn't automatically disruptive, either; as I mentioned earlier, if the railroad happens to go unnoticed, then by definition it can't be disruptive (e.g. if it was designed in such a way that the players on it never
would attempt to go off the rails, such that they never realize it's happening.) Whereas the disruptive player behaviors people repeatedly give examples of are (a) not at all fixed and unchangeable, (b) literally overt so they can't possibly go unnoticed (meaning, they're automatically disruptive), and (c) gated through the action resolution systems that ALL character actions are gated through, while the DM retains the freedom to manipulate things without such resolution metrics. (I don't roll dice to determine whether a lying NPC successfully lies in DW, for example; players have the responsibility and tools, usually meaning one or more rolls, to find out whether and why the NPC is lying.)
When players can literally MAKE the adventure follow one, and ONLY one, fixed and unchangeable path, without having to resort to merely ATTEMPTING to do so, then I will grant that players can railroad. Until then, it seems absolutely clear to me that players cannot do this thing, while DMs can, because players have fewer powers over the narrative and progression of the story than DMs do.
Anytime the player makes an action declaration he’s forcing the GM and other players to respond to that particular change in fiction.
In some sense the whole game is about applying force.
...so everything is always railroading all the time?
Are you actually biting the bullet now? I want to be sure of what you're saying. Because now it starts to sound like you genuinely do believe that literally all player actions ever taken, in any game ever, have been "railroading," despite your earlier implied rejection of my assertion that your definition was over-broad.
Doesn't it depend why the attack roll missed? (And maybe you had that in mind in your post and didn't think it needed spelling out.)
For instance, if the attack roll misses because the GM makes up or alters the NPC's AC to make it higher than the player's roll, that could be an instance of railroading, by way of "fudging".
If the attack roll misses because the GM decides, at some point during the resolution process, that there is a friendly invisible helper in the scene who places a sudden buff on the NPC or who casts a "Turn aside the shafts" or similar spell on the PC's arrow, etc - then that, too, looks like railroading, by manipulating the fiction in the scene so as to control the outcomes of action resolution.
If the attack roll misses because the NPC has a very high AC, which (i) the GM doesn't "fudge" in the moment but which (ii) the GM has deliberately made much higher than it "should" be, given the fiction and the prior outcomes of the PCs' action as declared by their players, in order (iii) to try and ensure that a particular pre-conceived scene or event is part of the fiction - then that could also be railroading, by way of deployment of authority over backstory and resultant scene-framing.
As these examples show, my view is that railroading is about using authority over the backstory and/or scene-framing, or using authority over mechanical processes (ie "fudging"), to ensure that a pre-conceived scene or event is part of the game even though - if one was following the fiction and/or honouring the success/failure of action declarations - it shouldn't be.
Given my absolute anti-fudging stance (I 100% see fudging as cheating and neither do it in my own game nor accept it being done in games where I play), I was assuming DMs that don't do that. It would be arguing in bad faith. I also consider "X thing made unjustifiably and arbitrarily high" close kin, but sure, not formally fudging as I define the term (that is,
secretly changing either the statistics or the roll result in order to cause a specific outcome, despite pretending that the outcome was the natural result of what happened with zero manipulation). It's the unjustified nature that bothers me so, and what makes it such close kin to fudging and other such methods.
But yes. If the DM is playing in good faith, using the system consistently, not trying to pass off false or manipulated results as true and unaltered ones, and not secretly fixing the results in advance by making absolutely insurmountable obstacles without justification, then I see no reason why an attack roll that results in a miss would count as "railroading." I find it rather tedious that I must make so many caveats on what should be a simple, straightforward statement, but I guess that's unavoidable at this point.
So. Given the foregoing, as a lengthy and (God I
hope) sufficiently comprehensive list of all the caveats necessary to remove all the underhanded or unjustified exercises of power a DM might potentially employ (and which the player usually has no possible way to employ themselves), we have a missed attack roll. That miss is a pure, honest miss, just an unlucky throw against a perfectly hittable target, which hasn't been manipulated or dismissed or whatever else. Do we agree that "player missed an attack roll," in this context, does not coint as railroading? Because the definition on offer from FrogReaver (and possibly Scott Christian) appears to count such things as railroading, alongside basically all other actions ever taken by PCs in every game of D&D ever played. I am not using hyperbole when I say that; when FrogReaver said, "Anytime the player makes an action declaration..." that seemed pretty conclusive to me that the definition of "railroading" offered is so broad it literally, actually does describe every action of every participant player in every game ever played.
When a term meant to single out a specific kind of behavior instead refers to literally all behaviors within the set being considered, the term is no longer particularly useful.