I must create a system, or be enslaved by another man's. I will not reason and compare: my business is to create.

- William Blake

Monday, May 09, 2022

Railroading in space and time (and the "future timeline" method)

Here is another post about a frequent subject in this blog. Tackling it from yet another angle...

In the last post, we talked about the quantum ogre; a form of railroading in space ("wherever you go, you'll find the ogre").

This type of railroading is almost universally condemned in the OSR (in modern D&D, it can go either way).

However, one could theoretically do railroading in time ("whenever you arrive, you'll find three goblins playing dice", or, even more extreme, "whenever you arrive, you'll see the bandits approaching a princess as she sleeps...").

This form of railroading (that reminds of Schrödinger's cat, maybe even more than the quantum ogre) is usually accepted without much questioning (or, at least, it isn't discussed as much).


My guess it is that it is almost inevitable. Even if you have a very detailed mega dungeon, it would be hard to add another dimension (time) to every place. You might make some concessions ("there are 1d6 goblins in the room at any time, with 1-in-6 chances they are sleeping..."), but it can never be perfect: what if the PCs take a month or a year before visiting the dungeon?

It might take hundreds or thousands of pages to realistically portray how a small community may change over time with enough detail (unless we are talking about undead, automatons, etc.).

It is usually assumed that the goblins are there in a state of suspended animation until the PCs enter (or listen to the door, or hear rumors, etc. - Schrödinger's cat indeed!). Adventurers are written in the present tense ("there are goblins..."). But the present is indeterminate (and fleeting); events happen in determinate times.

Let's illustrate this.

Here is one example from The Wretched Hive
Day versus night
The hive is active both day and night, but night makes it significantly more dangerous, because many creatures go out during the day to hunt for pollen, captives, etc. In addition, unless the PCs can see in the dark, their sources of light will make them easier to notice at night.
During the night, whenever you roll randomly to determine the number of creatures, re-roll all natural 1s. For example, if one room has “2d6 bee soldiers” and you roll a 1 and a 3, this would mean 4 bee soldiers during the day, but re-roll the 1 if the PCs enter at night (if you roll 1 again, there are four soldiers; if you roll 4, instead, there are seven soldiers total, and so on).
This rule applies to all encounters within the hive, random or not. On the upside, there is a 50% chance for every demon to be sleeping during the night.
Here is another example, this time discussing the time dimension:
[E] The Barracks
2d6+3 imps are here. During the day, half are asleep, but they do not seem to care with people coming and going. During the night, only a third of them is wake, but they'll be suspicious of outsiders.
The rest of the imps are playing disgusting games with small vermin and malformed dice.
A counter example would be the entrance to the Hive (room A). It is guarded by seven imps and a demon-ogre. No reference to time. Do they guard the entrance 24 hours per day? Obviously not. It is assumed that the PCs can wait for them to go to sleep, for example, but it isn't spelled out.

Finally, here is part of the aftermath (spoiler alert!):
If both the Queen and Malavor are still alive, the hive expands. In 2d4 weeks, the number of demons and bee-soldiers is doubled, and the hive’s defenses are reinforced. In another 1d6 weeks, Malavor manages to mutate himself into a bee-demon, half-insane, but with full control of the bee-people. The bloated and sick avatar dies after a while, but this no longer affects the bee-soldiers, that can now be cloned in the underground. Three months after the characters left, Malavor unleashes his army against the nearest village.
A list of pre-planned events? A pre-written ending? There are people that would think this is what railroading is all about! It is usually considered bad form if a module describes what happens next. You're supposed to find that out during play!

My point is: this does exactly the opposite.

A list of preplanned events might be the perfect solution to avoid railroading in space and time. I've made this point before, but I think it is easier to grasp with examples.

Here is a timeline of 12th century England:
  • 1135 Death of Henry I, accession of King Stephen to English throne
  • 1137 Beginning of a civil war between King Stephen and the Empress Matilda over the succession to the English throne; accession of Owain Gwynedd, the first Welsh ruler to style himself prince of Wales
  • 1154 Death of King Stephen, accession of Henry II to English throne
  • 1164 Constitutions of Clarendon, a set of laws which governs the trial of members of the Church in England
  • 1170 Assassination of Thomas Becket; death of Owain Gwynedd, prince of Wales
  • 1189 Death of Henry II, Richard I accedes to the English throne.
  • 1192 Richard is captured by Duke Leopold of Austria whilst returning from the Crusades
  • 1194 Richard is ransomed and returns to England; accession of Llywelyn ab Iorwerth to the throne of Gwynedd
  • 1199 Death of Richard I, King John accedes to the English throne
Since these are real events, most of them happen in an exact place and time. 

Now, let's say you're creating a campaign where the PCs are loyal to King Stephen, and the campaign lasts from 1135 to his death in 1154.

If you want your PCs to meet Owain Gwynedd before he ascends to the throne for some reason, but they decide to travel to Scotland instead, you just forced yourself to avoid railroading your players: they cannot meet Owain because he is fighting in Wales!

(Of coursing, using real events gives you much more information to play with; but even a brief biography of Owain could accomplish a similar result).

One important caveat: the PCs should not be included in the timeline. It is up to them to decide where to go, what to do, etc. Including them in the timeline will encourage railroading.

Of course, if you make Owain immortal before 1170, than you are back to railroading. too So you must plan the events... and let the PCs affect them, for better or worse.

(This is just an example; I know nothing about this period. Maybe The Great Pendragon Campaign works something like this; I haven't played it. I did run a Game of Thrones adventure 15 years before the books with a similar method, although I never had to decide "what happens if the PCs kill Eddard Stark", but I'd allow it. I found this easy to run, since I could turn to some GoT wiki for lots of information about what could happen.).

In short: the best way to avoid "railroading" might be building an explicit "railroad" of events and letting your PCs free to derail it, IF they are willing and able.

What are some of the practical implications?

Mainly, that it would be immensely useful if existing adventures and campaigns included future timelines. "This is what happens if the PCs do nothing". Having a note saying:
1137 Beginning of a civil war between King Stephen and the Empress Matilda over the succession to the English throne
Will give you lots of ideas about what might be happening in 1136. Rumors, tensions, enmities, alliances, discussions, scheming, skirmishes...

Knowing this is helpful even when deciding what happens if Matilda dies before 1137! Are there other forces that would oppose Stephen (maybe Robert of Gloucester?). Or would the whole war be avoided? If the PCs want to avoid a war, what should they do? 

These questions will inspire GM to find their own answers. But the book itself might add some "ifs" and "buts" in order to make things even easier ("If both the Queen and Malavor are still alive, the hive expands...").

This is the implicit assumption of most campaigns, BTW. "The cult wants to summon Tiamat, and the PCs have to stop them...". But since there is no set time for the coming of Tiamat, it can never actually be summoned. The final showdown will wait for the PCs, and even if the PCs die before that, a new group of PCs will rise to face the "final scene". Some campaigns will be explicit about the deadline; they make it easier to know if you're railroading or not.


One method I haven't used is planning weather and encounters/events in advance. Those two are clearly distinct; no one would object the GM having detailed weather for the next 30 days, since the PCs usually cannot affect it. Having a pre-planned encounter in 15 days sounds icky, as if you're forcing the players down a railroad - but it is not so different from rolling it in advance.

When I ran Tomb of Annihilation, such script would be very useful - much better than rolling every day.

Let's see if I can add some of that to my next campaign.


  1. To answer the initial question (why it is not addressed as often): adventures and modules don't describe locales but experiences. If that experience includes six goblins playing cards, it must be so. That's a trivial example, of course, but imagine Stradh being on holiday when the characters arrive in Ravenloft ... or, even more crass, imagine him already defeated! As a DM, you offer stories. Those stories always come with assumptions. King Stephen might not die, but the heroes help him make it look that way or Matilda dies and the heroes help making it appear that she's still around and kicking! What I'm saying is, we play with expectations, even with things like "recorded history". If that ends up being a "script", I'd say that's a railroad ... but that's the extent of it: the DM threatening or taking away the players' illusion of agency. If you manage to keep that illusion up, it does not matter what you are using to make the magic happen. In a sense, the tools you describe for the Wretched Hive are more about reminding the DM that the place is alive than it is for the players. That might be another important distinction to make here: modules are not written for players, they offer material to DMs and those DMs integrate them in their campaigns for the players. The materials modules offer are not the story, they're additional tools for the DM (maybe to help manifest the style of a brand or the vision of an author), and once you see it that way, you can't help but notice that it HAS to follow different criteria than, say, a novel. My 2 cents. Your advice is valid, of course. The more a DM knows about the potential future of a scenario, the more he'll be able to broadcast the implications. Weather is the perfect example for that!

    1. Yup, makes sense. "Experiences" is a good term; I'd have said "events" or "situations", but experiences sound broader. Strahd is indeed an extreme example, but maybe one of his minions might be found dead if PCs wait for too long.
      I dislike illusions (of agency), but I agree with your point about additional tools for the DM. Part of the job of the DM is finding out "what happens next" (or "what happens if..."). Having it spelled out for them is very useful - always making clear that this are tools, and not exact precroptions.

  2. I think your key comment here was 'PC's must never be included in the timeline' - the timeline is your sketch of what will happen if the players decide to go fishing. Without interference, the dominoes will fall as the timeline dictates, but as soon as the players get involved I find it more helpful to have things like NPC or factions goals that they will continue to drive towards, even after the players interference.

    I think the timeline is very helpful to let you map out what is likely to happen - and things may well play out that way even once the players get involved if their interactions do not disrupt the key elements of the timeline. When the players, say, kill a key NPC or cause an event to play out a different way, factions and goals are better for smoothly adapting on the fly at the table.

    1. Yes, that is the idea! When the players interfere, they can cause a "butterfly effect" in the timeline, disrupting everything.