Tuesday, February 21, 2012

Build Your Own Investigation

This week's blogthulhu is looking at different approaches to constructing investigations.  The goal of the series has been to outline a step-by-step approach to building a Cthulhu campaign that can be run as an investigative sandbox; previous columns are here:  Blogthulhu.  Let's look at the advice the core books have provided on structuring investigations.

Call of Cthulhu
The core book presents the classic overview of how a scenario unfolds:
  1. A mystery or crisis is posed
  2. The investigators become linked to the problem.
  3. The investigators attempt to define the mystery.
  4. The investigators use the clues and evidence to confront the danger.
  5. They mystery or problem is solved.
The key step in the classic view is step 3, defining the mystery.  Clues are gathered, NPC's are questioned, and the problem solving happens.  However, there's not a whole lot of method in the old COC book, just some sample investigations.  The other thing I've always remembered from COC is its use of an onion metaphor; each phase of the adventure is like the layer of an onion - once the players penetrate the first layer, they realize there's another layer underneath, and so on.

Trail of Cthulhu
Trail challenges the Keeper to develop three pieces of information - what is the plot hook that engages the characters, what is the horrible truth behind the scenario, and what is the trail of clues that leads from the initial plot hook to the horrible truth?

It seems like common sense to start with the end of the scenario (the horrible truth) and work backwards; start with the center of the onion and cover it with layers, working outwards to the last peel, so to speak.  One thing to be wary of with the Trail approach is this reliance on a breadcrumb trail of clues leading from point A to point B to point C; they call it "the spine", and if the idea is taken too literally, the investigation will seem linear.  In practice, though, it's fairly easy to make sure the path meanders, and to ensure there are events that ratchet the tension or sidetrack the action as the players probe the mystery - Trail calls them "confrontational scenes", and they're a way to introduce reactions by the other side.  Finally, there's this idea of "floating clues", which I'm not terribly fond of, unless I'm using a lot of improvisation in the scenario.  Floating clues are a tool to get an investigation back on track.

Last week's post mapped the flowchart of "The Haunting" as a dungeon, and it can be analyzed using the Trail approach fairly easily.  The horrible truth is that a previous owner of the house, now interred as an undead monster in a secret chamber in the cellar, exerts a baleful influence throughout the house.  There are two trails of clues to follow, a research path that goes through various library, newspaper, and court records, or a physical path that involves careful searching of the basement.  Finally, the plot hook involves being hired by the landlord, after the last set of renters ended up in the asylum.  Simple!  Phrased thusly, you could write an investigation like that, couldn't you?

There are many scenarios that don't use a linear breadcrumb trail of clues to meander along a path, but rather a "cloud" of clues surrounding the subject matter.  I rather think of them more like Call of Cthulhu's "onion peel".  There's a layer or two of obfuscation that conceals the truth, not single path.  I was taking a look at "Edge of Darkness", a popular introductory scenario for COC, and recently read another popular scenario, "Mister Corbett" (from Mansions of Madness), and both scenarios follow a similar structure:  they start with a simple, awful situation.  The players learn enough up front to head right to the site of the scenario, guns blazing (metaphorically, at least), or they can choose to take a circumspect approach and do non-linear research.   For that matter, both "The Haunting" and "The Kingsbury Horror" from last week are really close to this structure as well; while a few of the clues in both scenarios have prerequisites, creating a slight breadcrumb trail, the group is otherwise free to choose between following a research path or kicking down doors.

There are some high profile Cthulhu campaigns that involve lots of directive action by NPC's; the group of players is a bit like Rosencrantz and Guildenstern, popping in to observe a key scene when they're sent for, or sent to, a certain locale, and the whole thing involves a lot of puppetry by the game master.  Blech.  A few bad apples have created this reputation that "horror game means rail road".  This is false.  As I review more Cthulhu scenarios, I'll call out the ones that require Keeper puppetry to move the investigation along; there are plenty of well-done alternatives.

One final note; the past two weeks have focused on structure, and identifying structures that support the kind of agency we enjoy in old school games, like the dungeon crawl.  Structure does not equal content.  I'm considering what is to be said about making a good mystery, how much information is too much, and how to create an elegant puzzle that challenges the players.  Another thing to look at is the other elements I like to put in a campaign.  I've got a week to decide which one comes next...


  1. I wonder if anyone has ever played with the idea of treating investigations like they are in the real world. That is, they aren't set up as self-contained "adventures", but instead consist of a series of clues which may or may not be connected with the case at hand. Over time, the investigators might put together a series of clues gained in various sessions to puzzle through a layer of an onion.

    This would be a very slow-burning campaign, in which the players would have to keep files of information and work hard to puzzle out a variety of cases (some mundane, others more cosmic in nature). Not something that I would want to run, probably, but I might be interested in playing in it.

    1. I have, sort of. I don't run a ton of mysteries personally, but I have some plotting tools that I use when I do.

      I tend to use a relationship map, a timeline, and a progress tracking tool I don't have a name for. The tracking tool basically divides the mystery into "acts" with each act beginning and ending when a certain threshold is irreversibly crossed. Between each threshold are a list of important facts, rather than specific clues. Some facts are enough to catapult the PCs to next act immediately, while others are red herrings. The threshold of each act represents some critical moment, either an irreversible choice by the PCs or a reaction to their search by their enemies.

      It's a bit hard to explain, but it's pretty obvious in play. It looks like a sheet with a bunch of lines drawn down it and some sentences written in between the lines, with a few arrows and numbers.

  2. Right on. I find GUMSHOE's core assumption of "core clues" to be a bunch of poppycock if you don't build a railroad. The best CoC scenarios (and I'd say the standard, these days) are "clouds" of clues. No single piece of information lays out the truth of the matter, but each provides a bit that hints at something awful, and through comparison of each they can triangulate on what is actually happening.

    1. +1 to this, I think you've nailed the heart of it here succinctly.

    2. If you actually read GUMSHOE adventures, their actual structure tends to be a series of mini-clouds with choke-points between them. 'Shanghai Bullets,' for example, has one initial cloud (the less-dangerous investigative alleys) and then a fuzzily-defined transition to the second cloud (the chase); 'Death Laughs Last' is two clouds of different scales (the murder scene and a social scene) leading to a final location; 'Dimension Y' is a trigger event leading to navigating a cloud while the clock ticks on the elements which make up the cloud.

      The advantage of GUMSHOE's information presentation is that you actual can have clues which are relevant only in context, allowing the clouds to reinforce each other. You don't need to have infodumps in letters in desks, etc.

  3. @faoladh: There's a Trail campaign called The Armitage files that dumps a lot of information on the players at a time, creating that slow-burning campaign that requires choices on which to investigate. It's a bit like the infodump approach I advocated back in Target Rich Horror Gaming.

    RP: I'll map out some more scenarios as work flows, I'm curious if this "cloud of clues" is a phenomenon. There are scenarios I like, others I don't, and more often than not, when I go back and analyze the structure, I like the ones that are fairly free form, just present a weird situation, now "go".

  4. We live in an age of universal investigation, and of exploration of the sources of all movements.