Creating Satisfying Mysteries When Nobody Knows the Answers

I’m working on a PbtA game where a core part of the tone/genre revolves around characters solving mysteries.

Now, in your standard D&D and whatnot, the theoretical approach is that the GM has planned something out beforehand and the players have to work to figure out what it is. I don’t want to do that. I want this game to be more collaborative and emergent, with the GM and the players working together to create the story.

But how can I design for that and also design for satisfying mystery arcs, with clues leading up to a twist or revelation that makes sense in the game world? What tools can I give players and the GM to make that happen? And are there any systems out there that do this well that I should look at?


  • Have a look at this thread: Myth Arcs in GMless Games (
  • edited August 2018
    The basic theory for this sort of thing is that you need to offer the "clues" (facts that establish the puzzle) in an appealing yet clear way, and then you need to make space for the players to fit those clues together. Whatever they come up with is the answer to the mystery. It may help doing the design if you think of this as providing the players with a bunch of jigsaw puzzle pieces that they're asked to put together, except you just chose those puzzle pieces randomly, and not from the same puzzle even. It's just that instead of physical pieces that fit together in various ways you have clues to an event they are trying to put together.

    To get more specific, you could also "test" the solution the players choose. Maybe the first solution they think up proves false when the characters try to act according to it, in which case this fact alone becomes another puzzle piece; the true solution has to account for it as well as the rest of the pieces. This could easily occur when the characters try to solve the puzzle before finding all the pieces.

    Because you're providing puzzle pieces without knowing what the full picture looks like, it's possible that what you provide has no good solutions - the pieces instead of closing off possibilities until a clear picture emerges close of all possibilities, becoming a paradoxical impossibility. For those situations your game process also needs a way for the characters to discard puzzle pieces: further investigation proves that one or more of the clues are red herrings, planted lies, misunderstandings. The dropped piece doesn't need to be the most conflicting one; often it's more interesting if it is a less controversial piece that gets eliminated.

    You could implement all this by simply saying that the GM is an endless reservoir of new puzzle pieces: the players can pump new clues by doing more investigation, and they can continue doing this until they have enough for a solution - which necessarily occurs sooner or later as the GM provides new information. Or, you could say that there are a specific number of clues, and that's all that can be discovered. They may not be enough for a detailed solution, but perhaps they will point a way for getting the rest of the way on guess-work? Or, maybe the mystery could be left unresolved despite the effort, that's something the genre is familiar with as well.

    There is a game in doing this sort of thing, no doubt. As far as I know there is no true master implementation out there yet, though. Dirty Secrets comes closest that I can think of, but it's got other stuff going in it, too.
  • In essence, Eero is describing abductive reasoning. You needn't even create the 'clues' that the players fit together to invent the solution. Rather, you can simply provide the starting state--the 'scene of the crime--and then let them come up with the clues as they ask themselves (you) questions. For example:
    * There's a body on the floor and a window is obviously broken from the outside and left open.
    * A player supposes that there could be clues outside the window, under it (say, in the shrubbery?).
    * You ask them what they think they might find.
    * They suppose footprints or physical evidence snagged on the shrubbery. Good Idea! Yes, there's footprints.

    ...and so forth. Almost like a story jam where you Say Yes a lot. Alternately, you could hook into system using (hidden) 'Investigation' rolls to determine if their 'found' clues are relevant or red herrings... but be warned about too many red herrings, unless plot twisting is you jam ("No, the GARDENER didn't do it! You've been chansing a grroundling when, clearly, what left by the window NEVER TOUCHED THE GROUND!"). Just know that more red herrings means a MUCH longer resolution. Further, you could also hook into 'Deduction' rolls to verify or falsify their solution (as opposed to the Say Yes/Good Idea automatic validity in the bullets above). But again: More delay, more twists, longer resolution time.

    But either way, when they have an 'airtight case' after accumulating 'found clues' and propose a solution (in finest confrontation scene tropes), then if/when it is correct... roll initiative. ;^)

  • edited August 2018
    Check out InSpectres as a seminal entry in this field, though it has its problems.
  • Great advice and suggestions for games & other threads to look at – thanks, all!

    The idea of abductive reasoning and the puzzle piece metaphor make a lot of sense. And playing with the idea of red herrings is really smart. I’ve actually stopped including pre-planned red herrings when I GM mystery games, because the PCs always latch onto them and invest them with all kinds of meaning. And once they’ve done that, it seems mean to say, "No, that means nothing”, when I could instead reward them for innovative thinking, so I always end up making their red herring ideas “true” in the end. So really, it seems like all I have to do design-wise is formalize that process as part of the actual rules. And having the red herring as an out for clues that don’t fit is a very smart solution for loose ends that the players can’t logic into some kind of story pattern.

    So it seems like that’s my solution – present clues that fit with the world, let players piece them together, and allow their solution to be narratively true. And it should fit nicely into the PbtA idea of “play to find out what happens.” My follow-up is: Is there any way to bake that into the game mechanically, beyond laying it out procedurally as GM advice? Does it even need to be done mechanically?
  • I started designing a game utilizing the approach you seem to have settled on here, but I ended up with another kind of who-dunnit game that we’ve had some real fun with. So the original mechanical idea I had hasn’t been further developed. Here it is: dixit... kinda. The voting mechanic of the card game dixit really.
    My idea was intended to make mystery stories a possibility for gmless games. I think it might work better w a gm though. Here’s the Just: each clue revealing scene would contain 3 or so things connecting to something in a previous scene. These could be made up on the fly. At the end of the scene these clues would be voted on in a secret ballot going to gm. Vote for two. The gm will secretly choose a clue that got some but not all votes and will make it the useful clue. The group won’t know which one is selected though. Any clue that got votes from everyone will become a red herring. The gm will need to think of a good story/rationale to justify the red herring and will need to weave this into the story. The next scene should offer opportunities to recognize/deduce that the red herring is just that. The useful clue is the one you use to build the next clue in the next scene. This means that the game will involve following up multiple leads w some getting pruned off the possibility tree.

    Anyway, that was the basic idea. Make any use of it that you see fit.

  • We Used to Be Friends likewise addresses this.
  • @davey – Cool system! I especially like that the most popular clue becomes the red herring.

    @Lisa Padol – Will definitely have to check out We Used to Be Friends. It seems very similar to what I want to do (though the genre specifics are different).
  • Davey,

    That's slick, I agree!
  • There's a Swedish game of Film Noir mysteries that does this called Nerver av stål (Nerves of steel):äs-nerver-av-stål-du-borde-läsa-mitt-spel

    I don't know how accessible it is through google translate. @Simon_Pettersson wrote it, so maybe he has some interesting perspective here?
  • @DBB I suggest you keep an eye on the mystery's role in this design. I think there are many potential roles, but two really clear ones:
    • The players try to solve the mystery.
    • Solving the mystery is not attempted; rather, the role of the mystery is purely narrative, to entertain the participants and provide some inspiration for fictional situations in play.
    It seems clear to me that you aren't going for the former. I think you're going for the latter, but I'm not positive, as many people who attempt what you're attempting want some blend of both.

    My thoughts on the matter are these:
    • Blending both is very hard, but if you can keep the players correctly oriented at all times, so that solving efforts are rewarded and fruitless solving efforts aren't attempted, then it can be a lot of fun.
    • With absolutely zero ability to pursue the "real truth" of a mystery, "mystery" loses its value. Players aren't invested in pursuing answers they just get to make up later; and randomly-generated content doesn't produce good stories; and as plot frameworks go, "it's a mystery!" doesn't inherently have any of the rising action we usually expect and prefer.
    Not trying to discourage your endeavor, just throwing out some cautions that I hope might help you zone in on what exactly you want the "mystery" element to provide. Examples:
    • In Animal Crime, the point of the mystery is to riff of mystery fiction tropes, hilariously contrasting their seriousness with the cartoony element of sentient animals.
    • In The Eye in the Pyramid, the point of the mysteries is for everyone to spout clever or funny conspiracy theories about what's really going on.
  • Check out City of Mist for PbtA rules about clues. I dont think that its terrific, but could give you some idea.

    Also I have this from an old thread somewhere

    When you dust for prints, you collect samples of any prints in the vicinity, and roll +focused. On a hit, you get to ask questions.
    10+, ask 3
    7-9, ask 1
    On a miss, the Mystery Master can introduce a complication, a red herring, or else you just can't find anything.

    How many different prints are here?
    How old are the prints?
    Was this thing touched by someone?
    How was this thing held or used?

    Oooh! And:

    When you send a sample (DNA, fingerprints, trace, camera footage) to the lab geeks, roll +hard.
    On a 10+, hold 3. 7-9, hold 1.
    Spend 1 hold to get one of the following (MM's choice):
    Introduce a new suspect
    Introduce a secondary crime scene
    Locate a suspect
    Identify a murder weapon

    Oh! And of course:

    Once per session, when you make a terrible pun over a dead body, take +1 forward. YEEAAHHH!

    When you break a suspect, hold 3. Spend hold 1 for 1 to ask questions:
    What did you see?
    Who was with you?
    Where were you at the time?
    What do you know about X?

    When you rough up someone in interrogation, roll +hard
    10+ you break them.
    7-9 you break them, and the MM chooses one.
    On a miss, the MM can introduce a complication, advance a countdown clock, or give a red herring, and the MM chooses two.
    - They're hurt pretty badly
    - You get caught
    - Someone else sees you

    When you talk a suspect around to cooperating, offer some genuine insight into their motives, and roll +sharp.
    10+ you break them
    7-9: they give you a titbit (MM's choice)
    On a miss, the MM can introduce a complication, advance a countdown clock, or give a red herring, and the MM chooses two.

    When you leave a suspect to stew, the MM can advance a countown clock, but hold 1 (max 3). Spend hold 1 for 1 to take +1 on any roll to break the suspect.

    My favourite from a generalised move view: Whenever you take significant time away from the case, roll +intuition. 10+ something happens that makes clear something you didn't know before. 7-9 you couldn't stop thinking about some piece of evidence. If there's nothing more to Know, the MC will tell you.

    When you gather evidence, roll +Savvy
    On a 10+, you gain 3-Clues. The MC will tell you what you find.
    On a 7-9, you gain 1-Clue.
    On a miss, you have gathered evidence, but you can't make heads or tails of it.

    This leads into players pursuing clues-- the ability to generate new scenes.

    When you follow a lead, roll +Cool
    On a 10+, a new piece of evidence crops up, and take a +1 forward to gathering evidence.
    On a 7-9, new evidence crops up.
    On a miss, you have alerted your quarry of your investigation, and they may hinder you.

    Lastly, the players may finger the culprit. It works similarly to the Savvyhead's Oftener Right ability. The player comes up with a theory based on the evidence related to the case.

    When you finger the culprit spend up to 3 Clues and roll +(clues spent)
    On a 10+, ...

    When you gather evidence, tell the MC what sort of thing you're looking for and roll+savvy. If you are investigating a PC, rol l+Hx. On a 10+, you get 2-Clue +1. On a 7-9, you get 1-clue. On a miss, you have evidence, but you can't make heads or tails of it. If your Clue is +1, you get a +1 to following any leads from that clue.

    When you follow a lead, you roll +cool.
    On a 10+, the MC gives you a scene where new evidence crops up, and take a +1 forward to gathering evidence. On a 7-9, the MC gives you a scene where new evidence crops up. On a miss, you have a confrontation, expose your position, or alert the quarry.

    When you put the pieces together, spend your Clue Holds (up to 3) and roll +(Clues spent). On a 10+, you get 3 of the following, and take 1-Dread.
    On a 7-9, you get 1 and take 2-Dread
    On a miss, take 3-Dread
    you learn their motives
    you learn their weakness or Dread Secret
    you learn where they’ll be in the near future
    you learn the extent of the conspiracy or power

    When you cover your tracks, roll+Dark. On a 10+, you get all 3. On a 7-9, you get 1.
    you misdirect their investigation
    you slow their investigation down
    you make them waste precious resources

    If used on a PC, on a 10+ you remove 2 clues from their pool, on a 7-9, you remove 1 clue. On a miss you have a confrontation with the PC.

    When you try to find out more about the current mystery, by research or interviewing witnesses or whatever, roll +Sharp. On a 10+ pick two (or the same one twice). On a 7-9 pick one.
    Tell me something that has already happened.
    Tell me something about what we are hunting.
    Tell me one of the monster's weaknesses.
    Tell me something that is likely to happen.
    Tell me where the monster is headed.
    Tell me who the next target is likely to be.
    Tell me if anyone is hiding something, or something isn't quite right.

    The Keeper may ask “how do you find that out?” If you don't have a good answer, choose one of the other questions instead.

    On a miss, you're going to reveal some information to the monster or whoever you are talking to. The Keeper might ask you some questions, which you have to answer truthfully.
  • My theory is, something(s) have happened and probably continue to happen.
    They are generally bad things - murders, stealing, missing people, strange sightings, etc, etc...

    So there are two issues as far as the meddling kids are concerned - First and most pressing is How do we stop the bad things from continuing to happen? This is the immediate issue. This is mandatory.

    The second, optional issue is: why have these things been happening? This is part of the larger picture that may be played out over multiple sessions.

    I say mandatory and optional, because the problem with mysteries is: how do things progress? failure to look for, find and correctly interpret 'clues' are story blockers. Clues kind of suck in that way. Too many opportunities for 'nothing happens'.

    So the How clues are all given to the players - it doesnt matter what they do, you will force feed them breadcrumbs to advance the story. THEY come up with the places they look, the people they talk to, etc... And you just roll with it - keep feeding them unambiguous information about how to progress the 'how'. NPCs are your best bet for this. Eventually, they will be able to put a stop to the bad things happening (at least in the short term). The uncertainty exists in two areas - what does it cost them to stop the bad things, and ...

    But Why? They killed/captured the murder/kidnapper/monster/whatever - but why was he kidnapping left handed blonde children or stealing ancient african artifacts? wtf knows? This is optional as far as progressing the story from an initial situation thru to an end of session conclusion. This is where you can lay ambiguous, uncertain to find clues that you dont need to know the meaning of yourself (yet) Let the players find them if they look for them and listen as they discuss what this all might mean. Eventually, over a series of related sessions, you yourself can come to a conclusion about the greater story of why.

    A really cool part of mysteries is after you as a group come a conclusion about why all this stuff is happening, you can point back towards earlier scenarios and go 'ah, thats why they needed to be left handed!'

    Its like the TV series LOST, yeah? Every week some crazy shit happened, and things advanced in the immediate term during an episode, but nobody, including the writers, knew why any of it was happening for the first several seasons I bet.
  • edited August 2018
    This Lost way of putting the meaning in the beginning, without certainty that there will be one in the end, can go wrong. A simple open frame (who how why when where for whom) can be useful for managing this risk, setting expectations, boundaries.
  • The gamemasterless Okult does this, doesn´t it? Today I´ll play it with three friends. Maybe I´m able to tell something substantial afterwards.
    edited August 2018
    @David_Berg brings up some good words of caution, and @stefoid starts to get at some of the answers, I think.

    I should clarify that I think there are two kinds of mysteries (in writing and in game design). You've got what I would call a Whodunit, where you're presented with an array of clues and possibilities, some of which are valid and some of which aren't, and the goal is to figure out which of the possibilities is the correct answer. You might also call this a branching mystery. An Agatha Christie novel is a classic example of this. Then there's what I would call a Breadcrumb Mystery, where you don't initially know what's going on, but you gradually uncover a trail of clues that leads you to the answer. You might also call this a linear mystery. Doctor Who episodes are usually structured like this.

    I'm very much going for Breadcrumb Mystery in my design intent. I mostly agree with @David_Berg that players won't be invested in answers they make up later – if they're playing a Whodunit. If the game is, you have an array of clues, some are valid and some aren't, and we're generating both what the clues are and which ones are valid on the fly, then solving which set of clues – which path – is the correct one is meaningless because the game treats every path as equally valid. But a Breadcrumb Mystery doesn't have this problem, because, narratively speaking, all possibilities always collapse into a single valid path. This gets at what you by "fruitless solving efforts aren't attempted," yes?

    So I'm definitely trying to get players to "solve the mystery", but they should be doing so by following one clue to the next to the next. I'm just also trying to square this with the PbtA "play to find out what happens" ethos. I want players to pull the string and follow the trail back to the solution to the mystery, but I don't want the GM to have that solution planned beforehand, I want it to arise through play.

    Lost – specifically the Lost writer's room – seems like a good touchstone here. I've never watched the show, but I've talked to a number of friends who have. They've specifically told me about interviews where the writers said, "Yeah, we didn't know any of the answers, we just wrote something cool in the moment and relied on our own future ingenuity to make it work later." I want to build a game that supports doing that. Though I'd definitely be interested in @DeReel's thoughts about how front-loading meaning without direction can go wrong.

    Right now, the best solution I've arrived at, based on this thread and my experience GM-ing pre-planned Breadcrumb Mystery plots, is to lean into PbtA's "ask questions and build on the answers" principle. So in a scene, the GM should say to the players something like, "You notice something is out of place here – what is it?" And then their answer becomes either the seed of a new mystery, or a clue that represents the next step in an existing mystery. Then whatever way the players are able to fit that new piece of information into the overall picture becomes canon – or, if the GM has thought of something else (or thinks of something else later), it was a misdirect or belonged to a different mystery.

    Does that make sense? Is there a better or more formalized way to do that?
  • edited August 2018
    Throwing out dramatic events and characters with no idea what will happen in the end can go wrong in that it will lead to a disappointing ending, Even Lynch gets this sometimes, and my is he skillful.
    Maybe I don't need to explain this : just recall the zillions of bad stories who's ending did disappoint you. I can bet that throwing dramatic events, witty dialogues and kwel characters together never made a good story.
    A satisfying mystery means Closure, and more.
  • I can bet that throwing dramatic events, witty dialogues and kwel characters together never made a good story.
    A satisfying mystery means Closure, not just a fast knot.
    My experience with the dramatic arts suggest a more nuanced picture. Without going into deeper theory nitpickery, a succinct overview of my understanding:

    It is possible to create entertaining stories essentially piece by piece by ensuring that the modular components individually possess artistic virtue. Roleplaying games do this a lot, but it's done in other formats as well. For example, if you have a good character, that character can work essentially by itself: you put that character in any situation, anything at all, and the story's going to have some heat thanks to the character's virtues. Imagine just throwing say Sherlock Holmes into anything at all - you have something there just because Sherlock is a nice character; a farcical parody if nothing else. The same is possible with things like milieus, plot structure, expressive style... they can be treated as independent and arbitrary modules while still retaining some artistic merit.

    Meanwhile, there are truly mature art forms out there where the quality culture has grown to be truly impeccable in regards to the overall shape of the piece. The novel is a good example due to how popular it's been for the last hundred years, how many have been made, how effective the critical establishment has been, and how much money has been riding on that quality. The consequence has essentially been that what we require of a novel is a rounded, all-around excellence: having good characters is a minimum requirement, and there is not a single aspect to the work where you can afford to be anything but professional-grade if you wish to create something that matches to the expectations of the audience. What's more, the elements of the work absolutely have to harmonize aesthetically when you're writing a novel: harmonious interaction gives an unified artistic work what amounts to a boost in quality compared to a haphazard work, which means that even if you have some truly excellent stuff to contribute (a really good story idea, for example), it is impossible for your work to match up next to a professional-quality one unless you also master the emergent value of creating a harmonized overall presentation where absolutely everything in the work has been calculated to fit. Sherlock Holmes won't make a random appearance in a good novel; if he's there, there's a deep intent to it.

    The interplay of these two phenomena - aspects of a work having merit while it also being possible to create truly unified works with emergent value - means that different artistic cultures and mediums put weight on different aspects of the creative process depending on what the art is like in each case. This is important because the roleplaying art, when interpret as a narrative artform, is very different to how novels or movies or comic books are put together. The rpg activity is immediate, improvisational, interactive, and it should be obvious how these qualities impact the way storytelling works in roleplaying games: a well-crafted unity of form, so prized in writing and movies, is essentially an impossible ideal for an art form like this. Roleplaying is more like musical improvisation or a dance hall experience; authentic, personal, ever-shifting. Harmonizing in the sense I mean here can occur, and it's truly a great surprise when it does, but you can't count on it. Much of the art of game design is about harmonizing the players in a way that still retains the immanent virtues of the form, so it doesn't turn into simple co-writing of a novel.

    For storytelling in roleplaying games I would argue that it is very much realistic and reasonable to adopt a relatively loose and haphazard story design, much more so than would ever be possible for a novel. There are many roleplaying games that do precisely this, very intentionally: their preparations focus very hard in creating compelling modular content that will contribute value to the final product, but the connections between the pieces are kept loose, enabling the players/performers to combine and recombine their material flexibly as the game experience takes shape. This type of storytelling definitely loses out on the emergent values of matching this character to this story with these clever foreshadowing elements and this well-prepared climatic scene; however, on balance we gain the very heart of what makes rpgs interesting and relevant: extreme interactivity, personalization and creative audience participation. The virtues are on a level that a novel or movie can absolutely never reach, but the price for that should be clear: we're going to miss out on opportunities to carefully match elements so each complements the other. The storytelling's going to be rough and spontaneous compared to a novel.

    I'll mention a few games that work exactly like this, "throwing events and characters together to make a kewl story": The Shadow of Yesterday, Apocalypse World, The Mountain Witch, Dust Devils and Fiasco are all games that seem to occasionally produce what the participants have considered "good stories" despite explicitly including instructions for play that literally consist of little except an exhortation for the GM (or the group) to prep some compelling fictional ideas, for the players to create some interesting characters, and for the group to put all that stuff into a pot and see what comes out, metaphorically speaking. These games definitely fail to produce a well-coordinated novel or movie story, with regular dramatic arc, unambiguous protagonists and consistently developed thematic arguments. I would, nevertheless, disagree if somebody said that they do not create good stories. The stories are good, they just have to be gauged in their own art form, not compared to what a story is like in a novel. (Or if you insist on comparing, at least do the rpg replay the favour of editing it to shape. It's less than a first draft, after all - it's just a story jam! Have that novelist do one of those and see how he does before insisting on a rpg session to harmonize well.)
  • edited August 2018
    Of course you can find good stories without the closure I describe (Don Quixote, The castle, Twin Peaks s1, Le fantôme de la liberté) and also story games with a strictly framed ending (The mountain witch, Fiasco, Call of Cthulluh). I meant you don't get a good story with dramatic events and characters *alone*. (Harmony, frame, both mean a story is lead. by an intent / a meaning obviously)

    But we're discussing Mystery with a satisfying closure. Feel free to propose a better definition than mine (a reveal that comes as a surprise yet totally makes sense). In the meantime this definition is specific enough that it can be implemented (how do you manage a surprise, a foreshadow, an ending).
  • edited August 2018
    I find it pretty funny that Lost is the particular example being used here:

    That show sprawled a million "mysteries" with no clear explanation, going for shock value and an appeal to the curiosity of the viewer every time.

    To the same extent that it was endlessly creative, the show was a failure. The viewers were frustrated when the end of the show made clear how unfocused the "mystery" had been: there was nothing there, as everyone could clearly see.

    They had been duped, and they could all see it now. Many felt that they had wasted their time watching the show at all.

    That's the thing with mysteries: coming up with mysterious events to intrigue an audience is EASY. It's drawing it all together in a clever and satisfying way which has a compelling sense of logic and inevitability... that's where the art lies.

    Lost is, in fact, a perfect example of this.
  • For sure, I lost in interest in Lost after one season or so for exactly that reaon, but that show went on FOREVER, endless disappearing up its own clacker.

    To summarise my long rambling post, keep two threads going - one of which you are in control of - the immediate situation from moment to moment for the players - full of bangs and cliffhangers.

    And as for the long term mystery of it all, listen to the players. They will be constantly theorising with each other as to the meaning of it all. Its actually pretty easy to construct a story from disparate elements - seeing patterns and connecting dots in apparently random information is what human brains are really good at. Seeing animals in clouds.

    Maybe not david lynch stuff, tho...
  • The more pieces you find the harder it gets.
  • edited August 2018
    I think Eero has it right, this medium has to be considered in its own context. A "satisfying story" in an RPG will be much like a "satisfying mystery" I think. As long as the players feel like their actions lead to meaningful consequences, it works.

    I often watch terrible sci-fi and Fantasy series and think, "there are some good ideas here but they are implemented so poorly!" But, at the same time I think "this would make a great RPG!" That's preciously because an RPG story and a Novel story serve different purposes.

    If you want to present a singular vision, a world and story with clear artistic intent that makes its consumers feel a particular way, come to specific conclusions, write a novel .

    But, if your goal is to ask lots of interesting open-ended questions and let other people come up with their own answers, then write an RPG.

    My point is that a "satisfying mystery" in an RPG just has to be one that everyone involved had a meaningful participation in uncovering, and one that doesn't feel so bonkers than it breaks anyone's suspension of disbelief.

    That doesn't put RPG story on a lower level of quality, it just means it has a different set of goals on it outset.

    ***Lost might be a great example for me here, it might have been a bad show in regards to a satisfying television mystery, but might have made a great RPG
  • @Kenny_J a great rpg, right up until the players, exasperated at the end of their 30th session ask the gm - ok, what actually IS going on here? and the GM replies - I have no idea!

    justifiable homicide ensues.
  • @Kenny_J a great rpg, right up until the players, exasperated at the end of their 30th session ask the gm - ok, what actually IS going on here? and the GM replies - I have no idea!

    justifiable homicide ensues.
    The desire for an absolutely true answer is completely understandable. However, life in its myriad ways does not always provide.

    But seriously, in real life how often do we get a completely verifiably true answer. Of all the great mysterious that are out there how many have non-fictional answers, fiction of one kind or another.

    I do agree though that a satisfying game should involve some sort of satisfying answer to the mystery. In this context a satisfying answer is one that fits the evidence right? I guess their's nothing to be done about the potential doubt, Is there ever really a way to know the truth for sure?
  • @Kenny_J a great rpg, right up until the players, exasperated at the end of their 30th session ask the gm - ok, what actually IS going on here? and the GM replies - I have no idea!

    justifiable homicide ensues.
    More specifically regarding your comment...

    -30 sessions!? If someone sat through that amount of time the must have been having fun on some level.

    --30 sessions!? If they weren't having fun they clearly lacked the ability to walk away from a bad gaming table, and therefore they had a bigger problem than an unsatisfying mystery...

    -Finally, isn't the OP about making a satisfying mystery where the GM doesn't know? I mean that's the basis for the entirety of discussion here?

    *P.S. After I'm thinking about your comment I'm wondering if it wasn't a dig at a plot like the show "Lost" more than the rest I what I was saying. If that is the case I owe you a head tilt and a nod sir.

    In using "Lost" as a basis for a game I only meant it potential rather than its execution, though I take your point there.
  • edited August 2018
    Sure you can settle for "a sense of mystery". What do you do with the reveal, then ? Do you add a confession scene to seal it up ?
  • edited August 2018
    Double post
  • I'm not sure why this has to be so complicated. The first thing I thought of when I read the original post was Ghost Echo. If you read any of the APs for the game, posted on this board. You'll find that most of them turned into action mysteries. So the solution seems to be creating an in media res start for your game where the players are somehow put in danger because of the mystery, with set of question the players must answer to find the source of the danger and deal with it somehow.
  • The other thing about a Ghost Echo style game is the list of suggestive names for people, place and things. They are suggestive but not defining which makes the a good source of clues, in a PtFO game.
  • It is the simplest solution to the design challenge posed in the OP, which is what I meant by saying, "I'm not sure why this has to be so complicated."
  • Yeah, I get what you're saying - I don't think it's particularly complicated either. As I said earlier I cannot think of a game that would've made this type of mystery structure into the centerpiece of the game, but the basic activity is practically omnipresent in many shared storytelling games anyway. If you don't have a GM, particularly, the story creation in many games necessarily becomes a matter of first producing interesting stuff and then discovering the connections between the elements, revealing the story that was there "all along". I used to call the moment when the story "resolves" in this sense the "genesis moment"; it's very distinctive in e.g. Zombie Cinema, because when the players come to a concord over what the story is "about", the creative direction of play changes to seek resolution; instead of introducing new stuff in an essentially random way, the players start developing the existing elements instead to resolve the conflicts established between them.
    My follow-up is: Is there any way to bake that into the game mechanically, beyond laying it out procedurally as GM advice? Does it even need to be done mechanically?
    You could mechanize - formalize - much of this, yes. For example, you could have the GM define clues as discrete, modular game components that are manipulated explicitly. The game could place requirements on scenario generation, say: "A medium-length scenario has 5 GM-prepped clues." This being PbtA, you could have GM moves like "reveal a clue", and player moves like "discover a clue", all explicit. "Discard a clue" would be a powerful and necessary move in case the mystery grows too tangled, of course, or as a GM twist that reveals a clue to be a false one.

    All this becomes much more concrete if the individual clues are written down on separate slips. This enables the group to add new ones visually, and remove false clues, and group the clues in various ways as the players attempt to fit them together. Most importantly, such props help the players be absolutely clear about what elements in the continuous narrative "count as clues"; this is often not completely obvious to less talented players, necessitating constant reiteration.

    The actual process of play would also have a very clear formal arc: the scenario begins with investigation scenes, because you cannot solve a mystery without discovering what the mystery is about. Once the players have a minimal number of clues they can start attempting to solve, but trying it too soon risks having the GM reveal a contradictory clue that proves their solution wrong. Either way, through investigation and failed solving attempts, the players gather enough clues to discover a true solution - defined as a solution that accounts for all the clues on the table. Once this occurs, there might be an action sequence of some sort, depending on the genre and the particulars of the mystery; it's not uncommon for mystery stories that the revelation of the secret is immediate cause for some sort of action.
  • You can check out

    It's based on my two page game The Murder of Mr. Crow, that doesn't come up with an all satisfying story. But I think this "boardgame" takes it even further because it comes with questions that should be answered along the way, which will steer the story in unexpected directions.

    Another thing to think about is to create a game with several game masters and only one player, just like The Coyotes of Chicago, where all the game masters try to puzzle together and adapt their plans to each other.

    A third way would be to use an oracle that randomize a starting situation, play out what happens in a scene and then – after the scene – ask questions about where the story is going. The already mentioned Nerver of stål (Nerves of Steel) uses this method.

  • A third way would be to use an oracle that randomize a starting situation, play out what happens in a scene and then – after the scene – ask questions about where the story is going. The already mentioned Nerver of stål (Nerves of Steel) uses this method.
    I'm all for this too. Here's an -incomplete- idea for the starting random table:

    Crime (here would be a list of possible crimes, I've got one somewhere but can't find it right now)

    Roll 1d6
    1-2 It's revealed at the start of the game
    3-4 Serial crime, players must piece out a profile of the criminal and his/her modus operandi to stop it
    5-6 There are but a few clues that a crime has/it's being/will be commited.

    Roll 1d6-1 to find out how many suspects. The criminal may even not be among them.

    You would need further tables to create 3 details surrounding the suspects, crime scene, and primary clues, but otherwise you can totally rely on the players questions to come up with further clues, scenes and twists. Play to find what happens.

    Amazingly useful tool here: Tenra Bansho's Emotional Matrix. Basically, PCs and NPCs will have an emotional response to their presence, actions, questions, etc. Emotional reactions are clues too, so they will help to tie things together.

    Otherwise all I see missing to make this work is a mechanic and a clear way to define the endgame. The way I see it:

    -The more the PCs mess up with the mistery, the more and more nervous the culprit will be. In PbtA's tradition that would be a Clock, right? First of culprit reactions would be "pretend you don't know a thing", then "try to hide or mess with the evidence", next "do something against whoever seems to be getting close" and then "get the hell out of there". The last part of the clock will mark that the culprit got away.

    Of course, this doesn't mean the case isn't solved. Players will get the full picture at the end, but not closure if they take too much time.

    -All players and the GM need to keep a list of questions to answer: What, who, why, how? When the GM decides one of the answers on his list is definitive, the clock of the criminal moves one step. The idea here is that the GM should use all the material provided by the players as inspiration to get the answers right, to have everything make sense. Rolls to determine if a piece of information is right or not should happen only when the GM isn't sure about one of the answers, but as soon as something is written down the GM should give a straight answer, no roll needed. It would be the same with the emotional matrix.

    The general idea is that random input should only happen when there's uncertainty and stop when things are defined to help everything come to a rational conclusion.

    -To challenge the players a bit more, the setting may demand that they get proof of whatever they suspect. This could give the culprit enough time to try to escape.
    "You notice something is out of place here – what is it?"
    Excellent provocative question, you need more of these! I'll help a bit:

    -Something important is missing here. What is it?
    -This is too obvious, could it be made like this on purpose?

    Also, there are the classic questions the players should always keep in mind
    -Who may benefit from this?
    -Follow the money
    -Somebody commits a crime. What does that person do to hide it?

    Hope any of this helps. Certainly this thread totally inspired me, I'll probably do something with this anyway.
  • Another thing to think about is to create a game with several game masters and only one player, just like The Coyotes of Chicago, where all the game masters try to puzzle together and adapt their plans to each other.
    That looks great! Will play. :)

    All this brings to mind a setup that may work... gmless, let everyone play a character, but also everyone has authority over some secret aspect of the situation, and if 2+ players claim to have authority over a single adjudication then they must roll-off/draw/etc. Then maybe previous codified pieces of evidence were written on index cards and the loser(s) may mark some as red herrings to retain their personal secret aspect's coherence.
  • Lots more really great resources and suggestions! Thanks all! Really love seeing everyone’s varying thought process on this and digging into some of the suggested mechanics and existing mystery games.

    Reading through has really clarified my thinking about what I even mean by “mystery” and the ways in which I’d like that to interface with the design. I’m definitely trying to bake in “mystery as a driver of narrative”, rather than “mystery as narrative”. By which I mean, I’m much more interested in getting players invested in questions that they solve by taking action, rather than questions they solve by examining available evidence. Less Poirot or Sherlock, more Sam Spade or Blade Runner (not that I’m trying to design noir, mind you). So there should be clues and the players should be following them back to a resolution that reconciles those clues, but the story is about what they do to get the clues and what the clues prompt them to do subsequently rather than the clues themselves.

    Given that, the more I think about overly mechanical solutions, the less I like them (for what I’m doing – I love some of these suggestions for a detective game). It’s starting to seem to me as if baking the clues into the game moves (at least on the player side) will put the players in the wrong mindset. I’d rather give the GM tools to establish mysteries and let the players solve them by taking action in ways that are themed to the setting (rather than ways that are themed to the mystery genre). To that end, I think the scene framing questions are the most useful tool.

    I actually did an early draft playtest yesterday, and it went exactly as some of you predicted – I asked the players what was out of place in the starting scene, they threw out suggestions, I built on them, and by the end we had a nice little mystery on our hands, with multiple clues that all hung together in a cohesive structure. They really didn’t need much prodding to do it, and I’ve already figured out a solution that explains the whole mystery we invented. (And if they come up with a different solution in a future session, we can just pivot to that, or I can use my “actual” solution as a twist reveal.)

    The other major tool that I’ve really neglected to consider here is what happens in between play. I don’t think it’s contrary to the spirit of my endeavor (create cohesive mysteries from emergent storytelling) for the GM to do planning between sessions or during a break in play. Generating the puzzle pieces on the fly and then stepping back to see how they might fit together is totally in line with “play to see what happens,” as long as the GM remains open to the other good solutions that will inevitably arise at the table. So that should get baked into the GM rules & advice. I think explicitly including some advice to the effect that players are smart and they’ll figure out how to make things work at the table, and the GM is smart and they’ll figure out how to make things work between games would go a long way.

    As a philosophical side note, I can’t help but wonder if audience disappointment at media where mysteries don’t end in satisfying conclusions is a case of internal vs external processing. By which I mean, when you’re external to the story (an audience watching a tv show) you’re trying to solve the mystery but you don’t have any say in how it unfolds. So if it goes a way you dislike, you’re bummed. But if you’re participating in the story (a player in an RPG), then you’re an active creator. Even if the mystery unfolds in a way that wouldn’t make sense to an external observer, you’re satisfied, because you created the thing. It doesn’t seem fragile or weird or nonsensical, because you built up the whole thing from the inside and you did it the way you wanted.
  • There's a Swedish game of Film Noir mysteries that does this called Nerver av stål (Nerves of steel):äs-nerver-av-stål-du-borde-läsa-mitt-spel

    I don't know how accessible it is through google translate. @Simon_Pettersson wrote it, so maybe he has some interesting perspective here?
    Thanks for mentioning this today at lunch, Jonatan! I'd have missed it otherwise.

    Yes, NAS absolutely does this, and it does it in a beautiful way that flabbergasts me every time. As in, it's so much better than pre-planned mysteries. The mysteries are better. Doing it in this way, where nobody knows the answer, guarantees that the mystery works the way a mystery should work. The red herrings are always believed until they are proven false. The surprising twists are always surprising. Chechov's gun is always fired. All loose ends are always tied up (except when you deliberately leave one piece unsolved, for an open ending). I think it's pretty close to impossible to have such a satisfying mystery story when running pre-planned, at least without some serious illusionism. NAS is the most elegant and solid game I've ever written (sorry for tooting my own horn). Pretty much every game I've played ended up with the feeling that "this could be a movie, scene by scene".

    There are two keys to making this happen. One is documentation. After every scene, you write down any questions the audience is thinking about. "Who killed Lawrence?", "Why was he killed?", "What's in that mysterious box?", "Who has the nazi gold now?". You also cross out all the questions that have been answered in the scene. Then, when setting a new scene, you look at the paper. Can you answer one or two of the open questions? Or, if there are few open questions and a lot of time left, you introduce new questions.

    The other key is rigorous adherence to logic and continuation. Any time something happens which doesn't seem to fit with what's already been established, the scene is paused and we talk it out. "You just said that Jones ran out the door, but previously we established that she's blind." Perhaps I just forgot she's blind. Or perhaps it was intentional. "She must have been faking it!" Either way, you absolutely must not let any plot holes slip by.

    There is of course more things to know about how to do this, what questions to write down, when to cross them out, how to think, etc, but that's the basics. If people are interested, I'd be happy to do one of those "Simon explains how his game works" threads, which I've done before with "Svart av kval, vit av lust" and "Utpost". The game is not likely to come out in English translation, however, unless someone else does it. Perhaps French, though! :)

    Oh, it's GM-less, though. Might be harder to do it with a GM. Dunno.
  • I, for one, would love to see a Story Games treatment/discussion of this game after hearing about it for so many years!
  • @DBB well? what happened?!??!?!
  • @DBB, are you familiar with Technoir and how the transmissions and emergent conspiracies work there? If not, give it a look!
  • @Simon_Pettersson – This sounds like an excellent system and I’d love to know more about it. Going to jump over to your explainer thread and give it a read. The writing down and crossing out of questions and the way logical inconsistencies and red herrings are handled sounds perfect and very satisfying.

    @stefoid – I’ve got a bit more work to do before I’m ready to reveal the specifics of the project, but I’ll include a recap when I post about it for forum feedback.

    @hamnacb – I have not and I’ll take a look! Thanks!
Sign In or Register to comment.