Friday, September 4, 2009

Is investigative roleplaying railroading by default? A CoC case study

Some of you might have followed along in the discussion on Grognardia's retrospective on Shadows of Yog-Sothoth. The question arose if it has to be the case that investigative roleplaying with a "trail of breadcrumbs" leads the players along a railroad. Since I have just posted about my experiences of Beyond the Mountain of Madness, and accused that campaign of being a railroad, I felt it could be an interesting theme to continue and comment on. I decided to take out some of the CoC greats from my shelves and do some case studies.

First out is a golden oldie, Trail of Tsathogghua by Keith Herber. It was published in 1984, roughly the same era as Shadows of Yog-Sothoth. It's a collection of two loosely tied scenarios and one unrelated. The first one. The Trail of Tsathogghua is an expedition to icy lands, where an artifact has been found. It looks a lot like Mountains with the expedition leaders being NPCs and there is a timeline with things that will happen regardless of player actions. Also, at certain points interactions with authorities will happen, and they all happen "off stage". Roll some dice and the Keeper reads you the next clue your expedition member have uncovered. Should you fail, then there are NPCs around to tell you what you need. While the list of NPCs looks like it could be great fun to interact with some of them, it's still a good example of a very railroaded adventure. The second part The Curse of Tsathogghua is tied to some of the facts that became unearthed in the first part, and that one survivor from that expedition acted upon those facts and now the investigators follow his, incidentally, tracks. The idea of having a NPC with this special role who must survive, is just bad adventure writing. Having an adventure hinge upon something to happen is bad. The last and final part is totally detached from the first two. Interestingly enough, this last adventure The Haunted House is very different in style. It describes the background of a big mansion and why it is haunted. The the adventure is just a description of the adventure site, with appropriate hauntings described in different rooms. Go where you like, experience what you like. A good example of an investigative adventure that's wide open.

The second piece I wanted to look closer at is another adventure written by Keith Herber. This time it's his The Sanatorium in the Chaosium volume Mansions of Madness. It happen to be considered his best adventure, so it's extra interesting as a comparison. In this adventure there is a site (a closed environment), and a bunch of NPCs. There's also a menace. Compared to Trail this is very different. Something have happened, and you will have to get the information from the NPCs at the site before the menace kills you. Both adventures present a situation and a "story" that you will piece together. That's the common thread in both, as investigative roleplaying. But, while what happens in one is only at the players initiative, in the other it's always at the instigation of an NPC. There are no things that will happen, only factions with intentions and private goals. If the players wont interact with the NPCs, they wont find the clues to stop the menace, but they have all the power in the world to decide what happens. No NPC will push the story forward without you as a player pushing him or her first.

Worth noting is that one of these adventures is site based, and the other is based on a journey. The Haunted House is also site based, and is more open in style. The campaign that started the discussion Shadows of Yog-Sothoth, is also based on a lot of travelling. The other classic I have on my shelf, The Fungi from Yoggoth, is also a campaign where you travel a lot. I guess there's a strong temptation to hide one clue at each location and make them depend upon each other in a linear fashion. The sin here is, in my view, to make that the "story" that's supposed to happen. Do it have to be that way, though? Thinking on the format, you would imagine that it would be possible to avoid the worst sin by removing the linear dependencies to some degree and write the campaign so that visiting a site is not dependant on a string of former events. I know DGP worked on that kind of "plotting", with their nugget system for the Traveller material they published.

To make a parallel to T&T and D&D, where the dungeon is the backbone of the "story", we have many good pieces of analysis written in the OSR about multiple entrances, and interconnectivity between levels. I think this is a way of thinking that could be used while designing for investigative roleplaying as well.

So, in order to wrap this up I think the golden oldies tend to fall into a tendency of railroading when a lot of travel is involved. My hypothesis is that it makes it too easy to confuse all that travelling with "story" and thus needlessly limiting the players. Investigative roleplaying is not this by default, but it sure has happened more than once. Should I win the lottery I will take a closer look at the other classic as well, since it's supposed to do it just right.


  1. In an investigation, some level of railroading is necessary. In any instance where the PCs have to reach an end goal (solve the mystery, confront the villain), there will be evidence of tracks as the GM has to put the things they need to get there somewhere in the game world. I don't think a degree of railroading is a bad thing, its extreme railroading that is bad form.

    A dungeon connected to a story/purpose is by its nature a railroad, regardless of how many t-intersections it has. A truly site-based dungeon full of disconnected monsters with no central storyline is fairly open, even though the space and choice is still rigidly defined. The players are there purely to explore it for the loot potential and can come and go as they please (traps and the like excepted). But story = rails. The skil for the GM is in making those rails invisible.

  2. I've seen event-based adventures compared to site-based adventures in the sense that a room is like an encounter. In order to get to Room C you must first have gone through Room B, and previous to that through Room A. But it's quite obvious where to go next; effectively the "clues" to get to the next encounter are as simple as they get. A door.

    But what if the door isn't so obvious? The adventurers may enter Room A, not find your secret door, and leave. Adventure over. If they expect to find a secret door they may be persistent and clever in their search, but you can't guarantee anything without just tossing it out at them. This is the way those CoC railroad adventures fixed the problem: if you don't get the clue, we give it to you for free so we can move on.

    So one way to fix the problem of getting from one room to the next is to make the connections simple to find. The corollary to the event-based adventure is that the clues must not be too cleverly hidden. But they should be difficult enough to find that you're not insulting your players.

    Another way is to make the connections more numerous. A dungeon with 26 rooms should not require that you pass through all 26 in order to get to Room Z. You should be able to follow a path of, say, Rooms A, F, M, T, and Z. Or Rooms B, C, L, M, R, S, Z. And that of course is just the optimal paths - in a dungeon you can retrace your steps or go off on a tangent.

    So if you make a web-like map of all your event encounters, make a line for each type of clue that could lead you to them. You can include clues that allude strongly to the final encounter but don't make any that would let them bypass everything and go there straight away. They may make a mistake and do it anyway, but you shouldn't build an elevator from the entrance straight to the treasure room.

    So more clues, and easier to find clues.

    Now what type of clues should you place? Some players will want to talk with people, some will want to search the premises, some will think to hit the library and research. So include clues of every type.

    For example, let's say you have a murder. You should include info the neighbor will know (the victim was getting a lot of strange packages lately), the contents of his bookshelves (lots of vegan cookbooks, stuff on Buddhism), the contents of the victim's garbage (in the past couple days he started eating a LOT of red meat and nothing else), the crime scene report, coroner's report, friends in his address book, and the contents of his pockets (a dead glo-stik, a baggie of strange pills, and a matchbook with a woman's name and number on it from a club downtown).

    From that list of clues there are a lot of places the investigators can go. And you haven't left a Cthulhu statuette on his mantlepiece or mind-rending texts in his magazine rack.

  3. Thasmodious,

    I find the idea of story an interesting evolution in the hobby. Glad to see you mentioning it. As you say, there are always tracks of some kind, but the art lies in how they are (mis)used. I have ran my Dungeon of Voorand totally without story, but will now probably experiment with story, and what it might entail.

  4. 1d30,

    Interesting and well thought out addition. I guess you have knowledge of the Gumshoe system from Pelgrane Press? How to handle clues in investigative roleplaying is an interesting topic. Justin Alexander has written about that topic and his ideas are well worth thinking about.

    I also find your comment about the web of interlinked rooms of a dungeon. I have actually thought about dungeons as a string of bangs, which might be fairly un-orthodox, but might be a fruitful way of seeing it.


Copyright 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016 Andreas Davour. All Rights Reserved. Powered by Blogger.