Thursday 11 July 2013

The "Masks" School of Mystery Design

1. The On-The-Fly Approach

I've mentioned before how I like to run a mystery:

  1. DM develops a general idea of "what is going on"
  2. Clues/leads randomly generated or made-up on the fly for the PC's as their investigation advances
  3. These lead to further people/places with more clues/leads

2. The Masks Approach

But there's another approach: the Masks of Nyarlathotep Approach.  And it's all about Master Craftsmanship.  Here, the clues/leads are painstakingly prepared in advance, with actual pictures/text to cut-out and give to the players.  To give you an idea of the amazing design-work that went into this, here's a graph of clues and the people/places/things they lead to constructed by one dedicated fan.  Keep in mind that the adventure has 5 such locations:

My On-The-Fly approach is clearly easier to prepare/run.  But Masks' approach means that you can actually give players a physical clue to hang-on to, that they can analyse in detail and will actually remember between sessions.  Thus the game becomes less about killing stuff and taking their treasure, and more about putting together all the clues and figuring out the mystery.

3. The Player Skill Problem

As I mentioned recently, one of the difficulties in running a mystery is that of Player Skill.

Player Skill- solving mysteries can be difficult, and additionally, what seems obvious to the DM may not be to the players. So what does the DM do when the party is stuck, besides throw a tasteless Deus Ex Machina at them?

Both approaches above deal with this problem in different ways:

In the On-The-Fly Approach, I randomly generate clues for each location, so that even if the players miss a critical clue at some location, they may still find another one elsewhere.

In the Masks Approach, there are so many clues that even if the players don't find them all, they still have a good chance of finding some clue to advance them in their investigation.


  1. You also have the Trail of Cthulhu solution, which ensures that player's automatically finds the core clues which drives the plot forward as long as the players look for them and have the relevant skill.

    1. Thanks--I'll have to look that one up.

  2. I've certainly seen some schools of thought regarding clues that either say that if a clue is critical, it should be found automatically, or that you should have at least 3 opportunities to find each clue.

    Personally, I work it more like you, although I don't roll clues randomly. If a PC is looking for something, and rolls successfully, I make up a relevant clue on the spot. Seems to mostly work.

    1. I don't like running things so linearly, but that makes sense assuming you want a simple linear structure.

  3. "If a PC is looking for something, and rolls successfully, I make up a relevant clue on the spot. Seems to mostly work."

    As in, there are clues everywhere (potentially)? I like this! I'd been leaning towards giving out clues and letting the players decide what to DO. But in a game such as CoC (or RuneQuest, or WFRP) this would mean that a great number of skills and specialisms would be pointless. Why be skilled in Library Use if any old score would turn up the vital clue. Far better to reward use of these investigative skills even more (after all, use physical and combat skills is pretty central to most games) and allow people to turn up context dependent clues from all kinds of inventive 'searching'.


    1. Yeah I guess it's sort of the "Quantum Clue".

      There is a significant limitation though which is "It must make sense in the context of the gameworld."

      So for instance, in the Polish Resistance game I ran, if the party really wanted to understand how the zombies worked, they would need to do some heavy research. But there were a number of directions they could take this: libraries, museums, finding occult experts, breaking into government laboratories.

      So on one hand, there were many options, on the other hand, academic skills would still be a big help.

  4. Well, the problem with that is that if you make the investigative skills a requirement for turning up clues, then you run into the structural problem of "nobody has the skills to find this clue". Which means it isn't a clue. It's just some wasted time on a piece of paper.

    You can also use clues as general pointers. Using CoC as an example, let's say the PC is examining the cultists, who you know are Cultists of Nyarlathotep, and want to steal the artifact in the museum.

    The PC says, well, I know a lot about religion, so I want to see if anything religious pops out at me. You hadn't really planned any clues, but they roll successfully, so you go, "The cultist has a tattoo that looks a lot like the old Sumerian representation of Y-tho-tet, God of Opening Pathways. You know that there is a big display of Sumerian material at the Museum, and the Silver Order library also has a big section on ancient Sumerian myth."

    You can use the little context clues as pointers towards the places or larger clues, but always slanted towards the skill they used.

    Furthermore, you'd normally only do this for investigative skills. That gives the players a solid incentive to take them, because they are often useful, rather than just situationally useful.

    1. "nobody has the skills to find this clue" isn't usually a problem, at least how I tend to run games. The party can always seek out a helpful NPC who has the skill. So having the skill yourself just saves you time/effort on your investigation, as opposed to opening doors that would otherwise be shut tight.

      I like your cultist tattoo example--as game designer, you aren't going to think of every possible clue that might be there in advance, and even if you do, it will be hard to run a game with that many potential clues written out in advance. So using the "On-the-Fly" approach, limited by "makes sense in the gameworld" has a lot going for it.

  5. Hi

    I only just found your blog and it's very interesting, as a solo player who likes mystery style games (at least some of the time) it's always good to hear to hear of others 'make it up on the spot' ideas.

    I'd be interested in knowing if you've looked at TechNoir, there are some free Transmissions (adventures) on the game's website that give a general feel if you haven't got the full rulebook.

    The other game that seems might be a good mine of ideas is all of those Fiasco playsets (also completely free), I need some time to sit down and look at both of these in depth to work out which give the most useful adventure construction on the fly approach.

    Just a thought you and some of your readers might find interesting.

  6. On-the-fly can also be difficult to run. How relevant are the randomly created clues/leads?

    I think Masks of Nyarlathotep does it best. An abundance of clues and leads prepared in advance. BTW, I think Trail makes it too easy to "drive the plot forward". Sometimes, the PCs should get stuck or have the chance of failure.


  7. I didn't find it difficult, but that's just me. I have a post with an example table:

    I'm not knocking masks at all--it's an amazing work. But at the same time, not many have managed to duplicate it, so if you want to run some other mystery, there is a lighter-weight alternative than creating your own extensive graph of clues in advance.