Clarifying problems with modules

[Quertus]

Help me brainstorm categories of problems when trying to describe what's wrong with a module. (Yes, this is related to my "worst module you've ever seen” thread)

So, the most "definitely am error" errors are editing errors: typos, text doesn't match table, author can't count. Both of those are annoying, and unprofessional, but usually pretty easy to catch and fix.

Then there's issues where things are just missing, where it looks like the author just didn't finish their work. And that's

And then there's the Railroading. Now, you can try to claim that it's just a matter of style; however, in this thread, I'm requiring it to be counted as an error.

One special form of Railroading I really want to address is the artificial time constraints / "speed of plot" timelines. But that'll have to be a future post.

So, what other categories of errors have you seen in modules? What do you think is the clearest label for these or other types of errors? How do you think that they should they be grouped to optimize a reader's ability to assess the compatibility of the module with their group's playstyle?

[Khedrac]

One annoying one I came across was "author not understanding the rules". We were playing 3.0 but the adventure was probably 3.5 and I think it was from Dungeon.
The BBG cast timestop and then used a number of spells during the timestop, one of which caused the target to switch sides if the save was failed (which I did). Whilst it was quite fun to play, it is also explicitly not permitted for this to happen during a timestop.

Other annoying modules are ones where the players are expected to figure something out from either insufficent clues, or clues which require the players to be on the same mental wavelength as the author and are easy not to get.

[Pleh]

Point of order.

Doesn't the premise of running a module ratger imply some level of railroading?

If you improvise and leave the preplanned content, you are no longer running the module, right? From there, you are running homebrew inspired by a module. That's not wrong, and is probably the optimal outcome. But we can't say we're still running the same module.

[KillianHawkeye]

Then there's issues where things are just missing, where it looks like the author just didn't finish their work. And that's

You had me here for a second.



So regarding railroading, I think my least favorite form is when the writer of the adventure has a specific method of resolution in mind and doesn't put any thought into "what if the PCs don't do that".

My story of how I started writing all my own adventures instead of pre-written modules comes from the free adventure series for Star Wars Saga Edition that was on the Wizards website. I forget what the series was called, but it was released in several parts such that it sort of echoed the feel of the original movie trilogy.

One of the middle parts involved joining a high stakes card tournament, Casino Royale style, on Cloud City in order to intercept a shipment of Imperial weapons or some such thing. The players were supposed to infiltrate the casino, get into the finals of the tournament by whatever means necessary, and observe the coded information from the Imperial agent that would let them hijack the ship or whatever. But instead, MY players decided to kidnap the agent, thus preventing him from attending the final rounds of the tournament entirely, and I just didn't know what to do at that point.

It was so many years ago that I don't remember all the details, but I think the person they kidnapped was the one who was supposed to be receiving the coded information, so there was really nothing to be gained from their actions (not that they knew that). And the module just gave me nothing to go on. The whole campaign ended there, unresolved, because I was still a relatively new DM and I just couldn't make something up.

[Quertus]

One annoying one I came across was "author not understanding the rules". We were playing 3.0 but the adventure was probably 3.5 and I think it was from Dungeon.
The BBG cast timestop and then used a number of spells during the timestop, one of which caused the target to switch sides if the save was failed (which I did). Whilst it was quite fun to play, it is also explicitly not permitted for this to happen during a timestop.

Other annoying modules are ones where the players are expected to figure something out from either insufficent clues, or clues which require the players to be on the same mental wavelength as the author and are easy not to get.

Hmmm… "author not understanding the rules" a) is shockingly common; b) sounds pretty descriptive to me.

The clues one is tough. Because there's "vague clues", "only one path / failed rule of 3", and straight-up "insufficient clues" / "insufficient/no foreshadowing".

Point of order.

Doesn't the premise of running a module ratger imply some level of railroading?

If you improvise and leave the preplanned content, you are no longer running the module, right? From there, you are running homebrew inspired by a module. That's not wrong, and is probably the optimal outcome. But we can't say we're still running the same module.

"The bad guy gets away".

"No matter how long it takes you to get there / how quickly you get there, you have X time left / are X time units too late".

When the module dictates "this *must* happen this way", rather than, "this is what is there", that's railroading.

If the module does not cover the details well enough to handle when the PCs do something unexpected, that's… something else. And shortsightedness if it's something pretty obvious.

If the module cannot handle if something else happens (say, the PCs choose *not* to talk to stinky beggar #3), then the module is Fragile.

You had me here for a second.

Yeah, after autocorrect messed the title up as "clarifying" instead of "classifying", I just had to.

So regarding railroading, I think my least favorite form is when the writer of the adventure has a specific method of resolution in mind and doesn't put any thought into "what if the PCs don't do that".

Yeah, that needs a name.

My personal least favorite is probably the class of "the PCs cannot X / must X". No underlying reason. You're just required to railroad and change the facts to force that to be true.

My story of how I started writing all my own adventures instead of pre-written modules comes from the free adventure series for Star Wars Saga Edition that was on the Wizards website. I forget what the series was called, but it was released in several parts such that it sort of echoed the feel of the original movie trilogy.

One of the middle parts involved joining a high stakes card tournament, Casino Royale style, on Cloud City in order to intercept a shipment of Imperial weapons or some such thing. The players were supposed to infiltrate the casino, get into the finals of the tournament by whatever means necessary, and observe the coded information from the Imperial agent that would let them hijack the ship or whatever. But instead, MY players decided to kidnap the agent, thus preventing him from attending the final rounds of the tournament entirely, and I just didn't know what to do at that point.

It was so many years ago that I don't remember all the details, but I think the person they kidnapped was the one who was supposed to be receiving the coded information, so there was really nothing to be gained from their actions (not that they knew that). And the module just gave me nothing to go on. The whole campaign ended there, unresolved, because I was still a relatively new DM and I just couldn't make something up.

"The PCs fail" is a valid answer, if you ask me. Although the kidnapped agent sounds like something that the GM -or- the players could use to "fail forward".

[Unavenger]

So, the most "definitely am error" errors are editing errors: typos, text doesn't match table, author can't count. Both of those are annoying, and unprofessional, but usually pretty easy to catch and fix.

Some are easy to catch but not simple to fix - if something tries to force you to make an EL 1 encounter out of CR 4 creatures, it's difficult to know what you're meant to do. Was that meant to be more than EL 1? Friv thinks (I'm referencing To Hell and Back here, for those keeping count) it's possibly a placeholder that they never changed) or did the author just goof entirely? If there's a Dark Knight and an Evil Warrior statted up, and you're told that you come across an Evil Knight, which of the two monsters that actually have stats did they mean (I can't remember which two monsters they actually mashed up like that, but they made an identical error). What if you try to make an encounter out of hundreds of monsters in a space that won't fit them all in; how do you fix that? There are multiple possible answers and it's not clear which one was intended.

So, what other categories of errors have you seen in modules? What do you think is the clearest label for these or other types of errors? How do you think that they should they be grouped to optimize a reader's ability to assess the compatibility of the module with their group's playstyle?

Here's a couple of error categories that might be included, as well as some examples primarily from the infamous Tomb of Horrors:

Difficulty error

We don't expect even a truenamer to hit DC 45 checks at level 9, and we definitely don't expect a 9th-level party to take on multiple encounters of almost double their APL in CR, unless you're the Tomb of Horrors in which case apparently rogues need to make multiple DC 45 checks and everything is trying to kill you, or To Hell and Back where apparently people who are crippled by classes with no access to standard optimisation tricks find themselves taking 5d6 constitution damage out of the blue. Conversely, the Tomb contains laughable encounters made of individual CR 3 monsters or a few snakes who'll die to a burning hands spell, while To Hell will let you ambush an EL 9 encounter at level 25 and decide that's a meaningful combat. Generally, there's an acceptable band for difficulty, but the Tomb and To Hell will often make things too easy or too hard by any reasonable stretch.

Psychic Puzzle Error

To Hell has no puzzles and gets a free ride here (notably, Cain will tell you how to use the plot items if you get stuck), but the Tomb contains a black orb in the wall that kills you and another black orb that you need to go through to beat the module, as well as requiring that you guess which keys to use when and generally also killing you when you get it wrong, among other nonsense, oh and you have to fall into at least one trap deliberately, two if you want all the treasure and three if you want all the experience too. In general, anything where the correct answer is completely impossible to derive from the information you're given.

Goddamn Boring Error

One of the monsters in the Tomb does literally the same thing every round - save or die. It doesn't even move. No tactics required, and it's immune to everything that might make the fight at all interesting. I imagine bits of To Hell can get like this after slogging through the literal tens of thousands of enemies that you can face in a single act.

Yikes Error

Okay so this one needs some explaining: I consider anything that promotes an adverse relationship between people to be a kind of error of its own, whether that's the DM-vs-players relationship that the Tomb loves to play with - openly inviting the DM to mock the players and telling the DM that it's fun to make people's characters lose all their items - or hostility towards people in real life, whether that's using transness as a curse (that's the same goddamn door as the one that steals the character's items and tells you this is quite fun for the DM) or the exceptionally racist portrayal of the already slightly unfortunate flayers which, while only mildly uncomfortable in Diablo II, were suddenly made far worse when the game was translated over to D&D in To Hell.

Nonsense Error

Sometimes the module will call for something that just doesn't make sense, whether you're caught by a trap on your way to room 12 in the Tomb, which contains no room 12, or a tree is dropped on you while you're underground in To Hell. Some of these are easy to fix; others aren't.

No Creativity Allowed Error

Whenever you have a clever solution, the enemy or even the entire dungeon is immune. This one's all the Tomb; anyone who goes ethereal is eaten by demons and everyfrikkingthing is immune to spells. The Tomb does, however, have no walls and exactly one door which will put up any sort of resistance to trying to bash them down.

Unintended Consequences Error

This one's all you, To Hell. The ability to break plot weapons is honestly frickin' hilarious. In general, anything where the writers clearly didn't think through how something would work and it leads to problems later.

[Name_Here]

I think a pretty egregious one was from the climax of the crimson throne. Where you could pick up bits of information and do things to turn the BBEGs minions into your allies if the value reached a certain level but some of the minions didn't have enough avaliable points to turn as written.

[Pauly]

Existential error
Why are we here? Why does it matter? What’s the point of this module?
The opposite of railroading. Giving the players freedom to do anything without having an understandable goal to achieve. The opposite of railroading error.

Grinding error.
Have the party perform repetitive tasks throughout the module. Fetch and carry tasks are the usual offenders, but also needing to defeat X adversaries to progress.

[NigelWalmsley]

The BBG cast timestop and then used a number of spells during the timestop, one of which caused the target to switch sides if the save was failed (which I did). Whilst it was quite fun to play, it is also explicitly not permitted for this to happen during a timestop.

Technically that works if you're using Delay Spell. I would not be surprised to hear that the BBEG was not, but there are ways to use Time Stop offensively.

Doesn't the premise of running a module ratger imply some level of railroading?

A lot of modules actually don't require much railroading. Sunless Citadel is just a dungeon. There's not any real plot the PCs can wreck by sequence-breaking, they just go through the dungeon until they find the boss. Even something more involved like Expedition to Castle Ravenloft doesn't really require railroading. If the players figure out how to kill Strahd early, that's fine, because the point of the adventure is just "kill Strahd". Railroading only really comes into play with longer multi-module adventure paths, where there's a whole sequence of plot points everyone needs to arrive at in the same way. And even there, things should be written to provide a minimal level of resilience (ideally, while still respecting player agency, but that's a tall order).

[Telok]

Since When Has This Been A Thing?

A space fantasy module, a piece set on a asteroid with an unbreathable toxic 1/5th pressure "atmosphere" at level 6. That's ok becaise everyone has armor with 5 or 6 days of air, and two PC classes get 1st level spells that provide forcefields with 6 days of complete environmental protection and air supply for 6 people with every cast.

Make fortitude saves against a poison gas trap and later against a bad smell. Because the smell of rotten food goes right through fully sealed space suits and forcefields. Yes, even if you jammed the door open on the way to let all the breathable air out of the base.

But don't worry, in the second half of the module these things totally protect against a poisonous radioactive atmosphere for the entire week+ you spend on a different planet.

[Vahnavoi]

Doesn't the premise of running a module ratger imply some level of railroading?

No.

Let me demonstrate:

There's a piece of terrain contested by three kingdoms. Each has an encampment in the area. The player characters can attack these in any order, they can lose and either be killed or surrender to any encampment, they can ally themselves with any one encampment, they can negotiate peace betweem any two encampments to gang up on the third or can they can win over and loot all the encampments.

This is a very simple sketch of a gamespace for a module. Using just this information, how many different ways are there to go through the module?

[Yora]

Possibly the worst is: Make a successful skill check to find the adventure.

Like when there is only one hidden entrance into the dungeon and you can only find it by making a search check. If the die falls with a number not high enough, the adventure enda there. This comes up surprisingly often in tenfootpole reviews.

Existential error is probably the one that makes adventures fail for me most frequently. When I run campaigns, killing all the monsters for XP and looting their treasures just isn't the style of the game. But particularly with older adventures, that is commonly all there is.

Earlier this month I finally ran my dream adventure Dwellers of the Forbidden City, which I wanted to do for ages. The players knew that somewhere in the ruined city is the stronghold of the humanoid villains that have been causing trouble in the region, very much like the original setup for the module.
Once the players got inside and scouted out the ruins, they found a fortified hill where bullywugs made their lair. But they decided to ignore it because they were not looking for bullywugs. Then they came to the ruins where the bugbears made their camp, but decided to ignore it because they were not looking for bullywugs. They discovered the trees with the tasloi village, but decided to ignore it because they were not looking for tasloi. Then they found the mansion of the evil wizard commanding the raiders, cleared it out, and left the still half unexplored city.
Admittedly, this was a module for AD&D first edition, and it was originally written as a one-shot tournament module. But I always thought this one looked like one of the most interesting and complex ones, with lots of different factions to possibly get involved with and a hook other than collecting loot. (Ending the bandit raids instigated by the wizard.) But most of the content doesn't really relate to that goal and is placed in ways that it can be completely ignored.

[KillianHawkeye]

"The PCs fail" is a valid answer, if you ask me. Although the kidnapped agent sounds like something that the GM -or- the players could use to "fail forward".

Yeah, I'm sure I'd handle it better now, but I was still a novice DM back then, and I just threw in the towel. It was, I think, the second adventure/campaign I'd ever run.

[Silly Name]

It's Actually a Novel! or, the Mimic Module

As I mentioned in the thread about worst modules, for me stuff like the Avatar Trilogy is an enormous offender because not only it is extremely railroady ("Midnight/Elminster must [do X]" is the name of the game here), it further spits in the face of player agency by making the PCs simply watch the plot happen.

In a "normal" railroad, the PCs are still part of the plot - sure, they're hamstrung along, but at least the module will insist that the PCs must do things.

No such things in the Actually a Novel modules. The PCs are entirely superfluous, and exist only to witness other characters get to be cool and move things along. The GM is asked to simply read lengthy excerpts to the table a lot of the time, with no space for players to intervene or interject. In short, you were duped into buying a novel and reading it aloud.

[Eldan]

Point of order.

Doesn't the premise of running a module ratger imply some level of railroading?

If you improvise and leave the preplanned content, you are no longer running the module, right? From there, you are running homebrew inspired by a module. That's not wrong, and is probably the optimal outcome. But we can't say we're still running the same module.

Depends. D&D modules ,yes, but I've seen modules from other game systems that basically give you a small sandbox, an incident that starts the story, and a few ways the PCs might tie into it. Then fills the rest of the module with conditionals, like "if the PCs do this, this might happen".

[Yora]

It's Actually a Novel! or, the Mimic Module

I was immediately thinking about Avatar Trilogy. And from what I heard, basically everything for Shadowrun and Vampire.
But there's also the hilariously named "Freedom" for Dark Sun.

[Pauly]

Deus ex machina error.
No matter how badly and repeatedly the party screw up there is always a kindly NPC to put them right. Having DeM happen once is good, twice is pushing it, but any more than that just rewards the party for proving Charles Darwin wrong.

For convention one shot modules
There’s a mole in the party error.
Often used by lazy module writers to pad out an under written adventure. Unless you are specifically in a genre where in-party backstabbery is the norm (eg Paranoia) it sucks the fun out of the game. It can also lead to Meta gaming of knowing particular authors like to use moles so players go into the module on a mole hunt.

[Psyren]

So, the most "definitely am error" errors are editing errors: typos, text doesn't match table, author can't count. Both of those are annoying, and unprofessional, but usually pretty easy to catch and fix.

Then there's issues where things are just missing, where it looks like the author just didn't finish their work. And that's

And then there's the Railroading. Now, you can try to claim that it's just a matter of style; however, in this thread, I'm requiring it to be counted as an error.

One special form of Railroading I really want to address is the artificial time constraints / "speed of plot" timelines. But that'll have to be a future post.

I see what you did there

I'd agree with any module containing required PvP or adversarial party member elements to be a mistake - it's not impossible to do well, but it's very difficult.

[Pleh]

No.

Let me demonstrate:

There's a piece of terrain contested by three kingdoms. Each has an encampment in the area. The player characters can attack these in any order, they can lose and either be killed or surrender to any encampment, they can ally themselves with any one encampment, they can negotiate peace betweem any two encampments to gang up on the third or can they can win over and loot all the encampments.

This is a very simple sketch of a gamespace for a module. Using just this information, how many different ways are there to go through the module?

My question then being, exactly how are we defining a module?

This example sounds to me more like a premise for a game and less like a written out module.

Point being, when you write out a plot, it's gonna feel more railroading, because any deviation from the plot will take you out of what the module has prepared, no matter how many different ways you can pursue the plot.

[NigelWalmsley]

My question then being, exactly how are we defining a module?

I don't think going into the weeds of definitions is super helpful or necessary here. Again, look at something like Sunless Citadel. Seems hard to argue that it isn't a module, but it doesn't require any real railroading. Railroading only comes into play when you have an extended sequence of events that all are supposed to happen in a particular way, which is more a problem of APs, or longer modules like RHOD.

[gijoemike]

Point of order.

Doesn't the premise of running a module ratger imply some level of railroading?

If you improvise and leave the preplanned content, you are no longer running the module, right? From there, you are running homebrew inspired by a module. That's not wrong, and is probably the optimal outcome. But we can't say we're still running the same module.

There are 3 levels of module presentation found in printed material. Some people would only consider option 1 to be true railroading. I have played many an adventure where it was world building and connected NPCs.

My Way or Bust / This is a novel

The author has envisioned a very specific story. This is practically a novel. Adventures are presented a task or puzzle in which they must go to location a, solve a problem using social skills, that will lead them to location b where a fight will break out and they are captured, the party/prisoners are taken to lactation c where the rogue will use disable device and open lock to free them avoiding the guards, and then on the escape they find x and solve the mystery.

It is impossible to get to location b, or c, x without going to a first. And the puzzle must be solved exactly how the DM envisioned. Very little player agency.

Summons to the Mount

Author presents up front a set of locations/problems (A,B,C). The party then solves those problems in the way they see fit. Multiple solutions and order applied lead to the same set of solving A or B leads to location/problem D and E. This is a magicians choice. No matter what, you get to location X but the path varies based on your initial choices. Players have 100s of choices but true player agency is mostly an illusion. It isn't possible to do A and then go to X. But one could go to B, then C, then D to arrive at X. But it is viable to do A, then E, then C, then B to arrive at X.

Diverse World Building

The author gives little direction many choices are given up front. The PCs may have a difficult time even finding the adventure. A to X works, A C D W Y, alpha to X also works. Players decided everything. There modules are left vague because it is not possible to plan for everything but everything is acceptable. The players may never know they could have visited a B, E, or S location. These modules allow for maximum replay value.

[gijoemike]

I find it irksome when an author clearly knows the rules but doesn't' understand them worth a pile of dung. This is closely related to doesn't know the rules.

In 3.X custom npc stat blocks would have Wizards with combat reflexes, or dex 10 fighters/barb with combat reflexes. The author knows a given npc should have X feats but then chooses terrible feats, or feats that don't benefit the character in any way. Spell focus on a school that a caster has few to no spells prepared. Things like that.


But I will also second author has no idea what the rules are as a very common issue.

I once played a module where the end boss was a hag. That according to the tactics and author was to etheral, pick up the unconscious npc hostage. Wave them around in the air like they were possessed while casting spells at the party to battle them.

The hag had no way of going etheral or even blink, once on the etheral plane they couldn't have picked up a body, the hag didn't have a strength score to wave the npc around, every single spell the hag did have couldn't be cast across planer boundaries (No force effects). Our GM had to invent a whole new battle instead. The author had a total disregard of the rules but had a cool sounding fight scene.

I have also played in mods that use npcs and spells way outside the tier/challenge rating. I have seen a maximized burning hands spell used in a module that was only for 1st level characters. The caster was 5th level using a metamagic reducer to use a 3rd level slot. They party was 1st level. 5d4(20) damage to multiple characters DC 15 ref for half.

[jjordan]

I think you need to clarify what modules are designed to do before you start listing errors. Some modules ("Tomb of Horrors") aren't designed for standard gameplay. Some modules are meant to immerse players in a setting/story they really enjoy. Some modules are meant to be plug and play. Some modules have no idea what they're trying to accomplish.

Things I dislike in modules:
Lack of articulation of motives/plans. The villains are trying to do something for reasons that make sense to them and they're doing it on a schedule that should at least make sense to them. The DM should be able to make things work at the speed of plot, but should also have guidelines on what the NPCs are doing and when so they can improvise if players take things in a different direction.

The one solution fallacy. Put the characters into a reasonable and logically consistent situation and let them come up with a solution. They didn't solve the riddle but smashed in a wall to escape the trap? Good. And the module should be sufficiently robust to allow the DM to handle that.

[Vahnavoi]

My question then being, exactly how are we defining a module?

...

Point being, when you write out a plot, it's gonna feel more railroading, because any deviation from the plot will take you out of what the module has prepared, no matter how many different ways you can pursue the plot

A module is a piece of premade and prepackaced content you can add on to a pre-existing game or game system. Nothing, anywhere, demands that a module needs to have a "plot" in the sense of a script the players must follow. As should be plainly obvious from this and the "worst modules" thread, a lot of people consider a script the players must follow to be a design flaw in a module; it's a deliberate design schema that a module doesn't have to follow.

This example sounds to me more like a premise for a game and less like a written out module.

That's only because I'm not going to write out an entire module into a forum post.

[kyoryu]

Railroading is not required for a singular module - but I'd argue it is required for an "Adventure Path".

Also.... the problem isn't that it's a linear adventure. The problem is not getting your players on board with it, but running it anyway. Like, if you're gonna run a linear adventure, make sure the players know that and are on board with it.

Constraining players onto a path causes problems. Having the players willingly follow a path doesn't.

(Note: I am not a fan of linear adventures)

[Pleh]

Railroading is not required for a singular module - but I'd argue it is required for an "Adventure Path".

This is what I'm talking about.

What's the difference?

Presumably, an AP is a module, but a module isn't necessarily an AP. But why?

Is it only because APs require railroading while modules only might us railroading?

I think if we're codifying module errors, knowing this distinction is rather imperative.

[kyoryu]

This is what I'm talking about.

What's the difference?

Presumably, an AP is a module, but a module isn't necessarily an AP. But why?

Is it only because APs require railroading while modules only might us railroading?

I think if we're codifying module errors, knowing this distinction is rather imperative.

An AP is, generally, a series of modules.

The longer prewritten content is (outside of settings), the more it must constrain action.

In order for module #2 to begin, the players have to have ended module #1 at a particular place. And if there's ten modules (or the equivalent amount of content), then that starts looking more and more linear.

The alternative is that the content stops looking like "an adventure" or "a story" and starts looking like a setting. Which is cool, but it doesn't seem to be what most folks want.

I also disagree that railroading is inherently a problem. I was playing an Avernus game, and am currently going through a Dragonlance rework for 5e. They're linear as hell, but everyone knows that and so happily stays within the path. It's not my favorite type of game, but it works.

So if there's an error, the error is with the GM not getting buy-in from the players to actually follow the path.

[Silly Name]

I agree that a proper definition of a railroad isn't "having a plot". A railroad happens when the plot keeps going in a certain way even after the players tried to steer it another way.

If you always stay within the tracks of the planned plot by your own decision, it is not a railroad. An example of a railroad is a phrase like "the party must head to City A. They cannot be allowed to go to City B", because then if the party tries to go to City B the GM is urged to force them towards City A.

Of course no module can account for every possible decision a group of players may make, it's simply impossibile. But a good module can offer enough information and examples of what decisions the players may take that it is possible to react efficiently to unplanned actions.

A bad module will tell you that there exists limited number of permitted actions and paths and to bend over backwards to ensure those are the only actions and paths that will be taken. Especially bad modules have a singular path that is indicated as "correct".

There are ways to mitigate this - I think crafting illusion of choice isn't that bad. If a module wants there to be a big boss fight at the end which should always happen unless the players actively flee the scenario... That's not bad, as long as the players have the option to arrive at the fight in a number of different ways, and the door is kept open to other resolutions. Sometimes a fixed plot event isn't bad.

[Vahnavoi]

This is what I'm talking about.

What's the difference?

Presumably, an AP is a module, but a module isn't necessarily an AP. But why?

Is it only because APs require railroading while modules only might us railroading?

An "Adventure Path" is a string of modules meant to be played in sequence, typically governed by an overarching story. If an "Adventure Path" requires railroading at all, it's because the designer couldn't think of another method to make one part of the path segue into the next.

Fundamentally, this happens because modules designed with actual player choices branch out. In the example I posted, there are four distinct total victories and dozens of distinct partial victories and ways to lose; trying to write continuation for each individually would become infeasible fast. But let's imagine each module on a path only has two outcomes: the players win or they lose. So after the first module, there's two branches (win and lose). After the second, there are four (win-win, win-lose, lose-win and lose-lose). So on and so forth; if the branches aren't made to converge in a heavy-handed manner, their number keeps growing exponentially.

[PhoenixPhyre]

An AP is, generally, a series of modules.

The longer prewritten content is (outside of settings), the more it must constrain action.

In order for module #2 to begin, the players have to have ended module #1 at a particular place. And if there's ten modules (or the equivalent amount of content), then that starts looking more and more linear.

The alternative is that the content stops looking like "an adventure" or "a story" and starts looking like a setting. Which is cool, but it doesn't seem to be what most folks want.

I also disagree that railroading is inherently a problem. I was playing an Avernus game, and am currently going through a Dragonlance rework for 5e. They're linear as hell, but everyone knows that and so happily stays within the path. It's not my favorite type of game, but it works.

So if there's an error, the error is with the GM not getting buy-in from the players to actually follow the path.

This. I'm a big fan of JRPGs. Which are linear and completely railroaded from beginning to end, even in the "open world" portions--those open world portions don't advance the main story-line at all. But since I know that going in, I can enjoy the ride.

As for "railroading", a distinction from statistical physics might be useful here. There are microstates, which represent the actual state of all the particles in the system. In a TTRPG, this might be the exact history of all the actions of the PCs and their effects on the world. Then there are macrostates, which represent the bulk-level behavior of the system as a whole. Many different microstates can result in the same macrostate--in fact this is essential for things like entropy and temperature. But that's a tangent. In a TTRPG, the macrostates are the few distinct "end states" of that portion of the proceedings.

Using this, you don't have to plan for every possible action. Instead, at the end of each module in an AP, you have to sort the actual microstate into one of a few general buckets/macrostates that you carry forward to the next module and which determines the initial conditions for that next module. Sure, you lose some resolution/detail doing so, but you keep the branching down to a manageable amount. Maybe 3 macrostates: good success, qualified success, qualified failure. Outright failure means that you didn't make it to the next module at all (due to TPK, destroying the world, whatever), so it's not a valid start state for the next AP. It's still tons of work, which is why they just assume that the previous module either ended with complete failure (and thus you're not using the next one at all) or complete success.