I've recently been tinkering around with an idea for a game of investigative cosmic horror, in which the characteristics of the creature are created during (and through) the investigation process.  The idea is partially based on my initial misunderstanding of what the Gumshoe system was going to be like -- I heard the line about the important part of an investigative game not being the finding of clues but the interpretation of them, and I extrapolated my dream Cthulhu game from that.  Inevitably, the game I had in my mind and the game they had in their book were two different things; which is fine.  Esoterrorists looks way cool anyway, and I'll probably be getting Trail of Cthulhu eventually.  But I like my game idea, too.

Basically, what I took that whole bit about Interpretations to mean was that at the start of the game the GM wouldn't have any better idea than the players exactly what they were up against or how to defeat it.  The GMs job would be to create juicy hooks in the form of somewhat ambiguous clues, and then put them where the characters would find them.  The characters' skills in investigative areas would determine how accurately they could interpret the clues.  (Or at least, the investigative skills PLUS the character's knowledge of the Cthulhu mythos -- someone with no experience battling horrors from beyond isn't likely to guess that they're up against non-Euclidean Horrors from Beyond.)  The skill is rolled in a sort of Hypothosis -> Test -> Result form, where the Hypothosis is the character's interpretation, the Test is the roll, and the Result is whether or not the interpretation becomes fact. 

So, as an example, let's say that the PCs are performing an autopsy on a victim with several deep wounds on his abdomen.  One of the characters, Hubert Cumberdale, decides to see if he can figure out what sort of weapon caused the wounds. So he uses his Forensics skill.  "I think that the wounds were caused by a wickedly shaped dagger," says the player.  The die is cast.  If Hubert succeeds, he is correct in his hypothosis, and the wounds were indeed caused by a ubiquitous dagger of blasphemous aspect.

If he fails, however, the interpretation is left to the GM, who can come up with anything she wants.  Perhaps the wounds were caused, instead, by the horrid teeth of some unnamable beast.  Perhaps they were self inflicted somehow.  Whatever she comes up with is kept unknown to the players for now.  I envision a GM's sheet of facts about the monster, where she keeps a list of the characteristics the players know and don't know about the monster.

In this way, the shape of the monster or perpetrator or cult or conspiracy gradually begins to take shape.  It's my assumption that once a picture has solidified to a certain point in the GM's mind, the clues become more substantial, the enemy becomes more concrete, and eventually -- at some point -- there is a big Reveal, where the Enemy is Unveiled, and madness and chaos generally reign.  Aha!  But by that point, along the way, there needs to have been established some means, some hope however remote, for the players to stave off the horror until another day. 

Obviously this is all really just a kernel of an idea at this point, but I did have a few questions.

First, I'm wondering about Player knowledge vs. Character knowledge.  Should the players be in on the actual data about the enemy, or should the GM's antagonist sheet be kept secret from them as it develops?  And if the players should be kept in the dark, would be better if they didn't know it when they failed a roll somehow?  Ehhh.  Probably not; though I do think part of the fun will be when the Big Reveal happens and the characters are all panicking about the bits they got wrong.

Second, even if the players know when they've failed a roll, should the characters be aware of it?  That is, should a failed roll mean they decide they don't know what the clue means?  Or should they stick to the hypothosis, thinking it's correct?
Perhaps when a character fails a roll, he still gets a clue about the GM's interpretation.  "No, it isn't human blood.  It's something synthetic."  Or perhaps all three are a possibility, contingent on how badly he fails? 

A critical failure -- The character thinks that his hypothosis was correct.
Failure - The character realizes that his hypothosis was incorrect, but the results are inconclusive.
Partial failure - The character realizes that his hypothosis was incorrect, and the results point him in the right direction.

Does that seem reasonable?  Any thoughts?

Ooh. I like.
No, I think you've got it right on the money there.
One suggestion I have for you is maybe giving characters some kind of 'arrogance' or 'jump to conclusions' stat or trait that makes the character more likely to think their hypothesis was correct.

There have been a couple of suggestions over the past few months regarding similar concepts to this.

I definietly think it's worth pursuing.

Perhaps a checklist of sorts where every successful skill test eliminates some of the possible options.

"We know it uses claws, so if can't be Creature A, B, G, H or L, because they don't have claws. We know that it is capable of striking during the day which rules out creatures C, F, I, J and M. That's limits it to Creatures D, E, K and N. D and K are hydrophobic and two of the attacks occured in a seaside town...it doesn't necessarily rule them out, but it makes them unlikely. How can we differential between creatures E and N?"

Almost like a game of "Guess Who?"

V
 

The GMs job would be to create juicy hooks in the form of somewhat ambiguous clues, and then put them where the characters would find them.

Thanks for the comments everyone!  I was thinking last night about the other side of the roll, where the investigator succeeds, and came up with this:

Critical Success - The investigator is correct, and gets to add a detail to the Enemy's weakness.
Success - The investigator is correct.
Partial Success - The GM gets to say "Yes, but..." and add some contradictory detail.

chronoplasm wrote:
One suggestion I have for you is maybe giving characters some kind of 'arrogance' or 'jump to conclusions' stat or trait that
makes the character more likely to think their hypothesis was correct.

Callan wrote:
Could it also be that the players don't need to find them? That whatever they start focusing on, becomes a clue? Like as GM you just described the bellhop on a whim, but now he's becoming like the center of a massive conspiracy? Perhaps have some small bonus, like a 5% bonus, on clues the GM made, so players do look for them. But when they make a mistake and go after something that isn't a clue, it doesn't matter much system efficiency wise - their mistake will ultimately make a richer, more unexpected game. Just an idea that seems exciting to me :)

Vulpinoid wrote:
Perhaps a checklist of sorts where every successful skill test eliminates some of the possible options.  "We know it uses claws, so if can't be Creature A, B, G, H or L, because they don't have claws. We know that it is capable of striking during the day which rules out creatures C, F, I, J and M. That's limits it to Creatures D, E, K and N. D and K are hydrophobic and two of the attacks occured in a seaside town...it doesn't necessarily rule them out, but it makes them unlikely. How can we differential between creatures E and N?"

Almost like a game of "Guess Who?"

Hmmm... Maybe instead of ruling out mythos creatures you rule out normal creatures?
"A human being couldn't have possibly done this."
"Whatever did this, it wasn't a gun."
"How does a person get attacked by a shark in the middle of a cornfield?"
Perhaps as players rule out the normal possibilities, you can decrease the 'sanity level' of the game or whatever.
On the other hand, maybe it does turn out that there is a perfectly plausable situation for something? Maybe the wound really was just caused by a knife and not some kind of claw or fang. Then again, maybe the attacker was a Cthulhu cultist and the game can shift toward the paranormal from there?
It'd be like... players decide how wierd they want the game to get, and how quickly.

So, is the player making this deduction based on what the GM describes, or is it whatever the player makes up?  If it is the latter, you'd need a mechanism for keeping the player from making up something wacky or that would take the players far from where they ought to be headed.  The GM might have meant for it be the result of a demon impaling someone with a wavy horn.  If the player decides it was a wavy dagger and is "right", the GM has to either get them back to looking for a demon, or make up a bunch of new material about guys with wavy daggers.

apeiron wrote:
So, is the player making this deduction based on what the GM describes, or is it whatever the player makes up?  If it is the latter, you'd need a mechanism for keeping the player from making up something wacky or that would take the players far from where they ought to be headed.  The GM might have meant for it be the result of a demon impaling someone with a wavy horn.  If the player decides it was a wavy dagger and is "right", the GM has to either get them back to looking for a demon, or make up a bunch of new material about guys with wavy daggers.

What if there is a "Reputation" (there must be a better term!) resource that describes the trouble you are in in the mundane world.  Run out of Reputation and you get arrested and thrown in prison or an asylum.

Failed investigations will lead to a new scene, with a chance for success, but at the cost of Reputation.

Maybe high Reputation helps with certain kinds of conflicts (at the dinner party, trying to get Captain Briggs to reveal exactly what he saw on the ill-fated expedtion up the Congo), low Rep with others (stealing that car so you can get to the Summoning in time to prevent it)

Nick

Mr. wrote:
Basically, what I took that whole bit about Interpretations to mean was that at the start of the game the GM wouldn't have any better idea than the players exactly what they were up against or how to defeat it.  The GMs job would be to create juicy hooks in the form of somewhat ambiguous clues, and then put them where the characters would find them.  The characters' skills in investigative areas would determine how accurately they could interpret the clues.  (Or at least, the investigative skills PLUS the character's knowledge of the Cthulhu mythos -- someone with no experience battling horrors from beyond isn't likely to guess that they're up against non-Euclidean Horrors from Beyond.)

If he fails, however, the interpretation is left to the GM, who can come up with anything she wants.

Whatever she comes up with is kept unknown to the players for now.  I envision a GM's sheet of facts about the monster, where she keeps a list of the characteristics the players know and don't know about the monster.

It's my assumption that once a picture has solidified to a certain point in the GM's mind, the clues become more substantial, the enemy becomes more concrete, and eventually -- at some point -- there is a big Reveal, where the Enemy is Unveiled, and madness and chaos generally reign.  Aha!  But by that point, along the way, there needs to have been established some means, some hope however remote, for the players to stave off the horror until another day. 

First, I'm wondering about Player knowledge vs. Character knowledge.  Should the players be in on the actual data about the enemy, or should the GM's antagonist sheet be kept secret from them as it develops?  And if the players should be kept in the dark, would be better if they didn't know it when they failed a roll somehow?  Ehhh.  Probably not; though I do think part of the fun will be when the Big Reveal happens and the characters are all panicking about the bits they got wrong.

OK, i think i got it.

So, the GM should describe events in a way that give clues (and maybe a red herring) to the players.  Then the player's assumptions or role-playing interpret that (which is very realistic, there is no reality, only perception).  The GM can then either steer them or riff off the player. 

Lovely. 

Metagaming is always a possibility, to that i offer "Challenge!"

Dave, there is no way your character would think that!

Um well, he's got X ranks in Whatsitmancy.  Remember last month when event X lead in direction Y with outcome Z?

Oh, yeah, OK, go on then.  You clever bastard.

OR

Bull-oney! I'm going to give you a penalty on your roll for that. 

Sort of like an opposite to the Cthulhu Mythos skill, perhaps?

Oh, definitely!  That's totally the way to go -- the PCs can think pretty much anything is a clue, and roll for it.  If they fail, of course, the GM can decide whether or not it was legitimately "mythos" or not, and if they succeed, it can be whatever they want it to be.  And even more than that, maybe the players should be able to come up with clues themselves -- "How about, on the ground in the middle of the ruins, we find a broken bit of bone sticking up from the ground?"    "Okay, roll for it."

I like that idea, but I almost don't want to have pre-created creatures out there.  I think what I'm going for is more the *feel* of the Lovecraft stories, created dynamically, rather than either taking the entities from the established mythos itself, or creating my own mythos for the game.

Sorry, got busy with work so I haven't had a chance to think on this much lately.

Nathaniel makes some good points.

Nathaniel wrote:
Isn't this a good place to integrate player knowledge rather than have it solely rely on character skill + character knowledge?  If the players are in the game and are buying into the idea of creating the details of the yog-sothotian threat, then perhaps the players can have input that's sort of a "it's more than you know" moment to their own characters.

The direction dice that let you know how to narrate could change based on how the question is asked.

Who did a specific act:  Monster, Willing human, Unwilling human (is there any other answer that is still interesting to the story?)
Physical characteristic about the monster: Tangible, Intangible, Behavior
Motives, goal or plan of the monster (and it's minions): Cosmic, Physical, Communal

Monster -- the monster itself did the act in question
Willing human -- a willing follower of the monster (despite how ignorant they themselves are of the nature of the monster) did the act in question
Unwilling human -- someone does it either out of insanity or mind control or even blackmail or something

Tangible -- some physical characteristic about the monster "oooh.... those 6 eyes... like fire... like fire..."
Intangible -- some consistent feeling experienced by those who get too close "my blood boiled.  I remember getting so angry with Doris.  She was supposed to pick up the kids yesterday and I had to do it.  Go out of my way.  I don't know quite what happened but in that moment I wanted nothing more than to strangle her.  And... sob... little Jesse too... tell me I didn't... tell me I didn't..."
Behavior -- something the monster does.  "After analyzing the bones, you're quite convinced the substance found on them was digestive fluid.  Something ate these bones.

Alfryd wrote:

Sort of like an opposite to the Cthulhu Mythos skill, perhaps?

Yeah, sort of.  Increases maximum sanity through confidence and daring, but makes erroneous conclusions more likely.

Oh, definitely!  That's totally the way to go -- the PCs can think pretty much anything is a clue, and roll for it.  If they fail, of course, the GM can decide whether or not it was legitimately "mythos" or not, and if they succeed, it can be whatever they want it to be.  And even more than that, maybe the players should be able to come up with clues themselves -- "How about, on the ground in the middle of the ruins, we find a broken bit of bone sticking up from the ground?"    "Okay, roll for it."

We've been having a prolongued dicussion on the merits and pitfalls of the approach in this thread.  There is a general danger with any system where the players establish the world as they go- either subtle inconsistencies may creep in that you can't easily retcon, or the players options become more and more constrained by prior events.  In your case, the list of previous clues will wind up limiting the conclusions that players can draw without contradicting themselves.

I like that idea, but I almost don't want to have pre-created creatures out there.  I think what I'm going for is more the *feel* of the Lovecraft stories, created dynamically, rather than either taking the entities from the established mythos itself, or creating my own mythos for the game.

Okay.  But how do you intend to terminate the process of investigation, if there's no 'finished product' you can identify by a process of elimination?

Something ate these bones whole."

Cosmic -- Summoning an elder god, waking the Great Old One Cthulhu.  Something big that will not be good for humanity.  So a single clue might be that there's something ritualistic going on.  Further investigation might reveal another clue about the exact nature of what the purpose of the ritual is.  If previous roles were physical or communal, the insight would then be how those physical or communal goals are fitting with the overall cosmic goal.
Physical  -- the monster might be pursuing alien goals, but it's accomplished through physical means.  Perhaps it needs to feed on 100 bodies so it can go into another hibernation cycle.  Perhaps the "monster" is a human sorcerer who is trying to accomplish some personal physical aim, like getting lots of money to fund further research.  Or something physical that must be accomplished to get the overall plan moving forward.  Getting a certain ancient idol, a certain book, etc.,.
Communal -- the setting up of a following, dynasty, family.  Like Deep Ones wanting to capture human men & women to mate with humankind.  A monster developing a cult to do it's bidding.  Or perhaps a cult growing to gain power (be it to get something physical or something cosmic as revealed by other rolls).  Or the stark absence of such goals.  Whatever the statement is, it's about how the monster/minions are planning regarding other people.  And a communal result, perhaps the characters learn that the cultists don't get it and are not part of the monster's plan and are just dupes.

I find that when the dice just determine who gets to narrate but with no guidance as to what that narration should contain or how it should slant, things can fall a bit flat.

I don't know.  Just brainstorming.

Alfryd wrote:
In your case, the list of previous clues will wind up limiting the conclusions that players can draw without contradicting themselves.

But how do you intend to terminate the process of investigation, if there's no 'finished product' you can identify by a process of elimination?

May i introduce you to The Burning Wheel?

It works just like what you are describing.

Apeiron,

Do you mean mechanically, the dice pool thing, or investigatively, the antagonist sheet thing, or something else entirely?

In terms of the players writing the story with the GM.  i was in a game of this last w/e and it was great fun.  You take a story like 'Rescue the Princess' and then the players answer questions about it.  "The Princess has the means to escape, why hasn't she done so?"  Our answer was that she was in on it.  "Why was the Princess kidnapped?", i suggested it was an arms dealer/mercenary trying to start a war.  Maybe it's not the same, but it has some elements that might inspire you.  The difference might be that in your game the players can suggest things, that may or may not be true.

Interesting....I'll have to check it out.  Burning Wheel's complexity has intimidated me up to now, although I've heard many good things about it.  Perhaps it's time I got someone to run it for me.  Thanks for the tip!

So, just to recap, then:

1. There's a GM, and there are players.
2. On the table, where everyone can see, there is a sheet called the Antagonist sheet, which has places to fill in elements of your Antagonist.
3. The players can take anything in the game and make it into a clue, simply by investigating it.
4. Investigation involves a contested roll between any players with different ideas about the clue, and the GM.
4a. Just to clarify, if more than one player has a theory, more than one player will be rolling dice.
5. The roll is a dice-pool, determined by the PC's investigative skills.  Only each player's top two dice will count.
6. If the top two dice of any player exceeds the GM's top two dice, that player's character succeeds in their interpretation of the clue.
7. The "truth" of the interpretation, however, is determined by whoever has the highest single die.  That person gets to write the truth on the Antagonist sheet.

To continue, it seems like there are quite a few levels that lovecraftian horror could work on, and they could all lead from one to the next.  You start out with something small, and only as you begin to solve that issue do you find that no, it's actually pointing to something else, something larger....

(With apologies to Dogs in the Vineyard)

1. Something bad happens.  Someone turns up dead, or some weird shit happens around town, or someone just plumb goes crazy.

That's because...

2. There are Local problems.  There's something or someone doing the killings, or causing the weird creepy shit, or making people go crazy.

If those local problems are bitey enough, then take care of them and break out the champagne.  Otherwise, the local problems are there because...

3. There are Bigger problems at work here.  The local problems are symptomatic of other forces at work; for example, perhaps the whole town is corrupt.  Perhaps the creature killing people is the scout preceding an invasion.  Perhaps the murderer belongs to a cult.

Now, if those Bigger issues seem dramatic enough, great.  Once they're dealt with the characters can call it a night.  But if the players want something more...

4. There are World-Wide issues at stake.  Even those bigger issues are just the effects of an even greater cause.  That cult?  It's a world-wide conspiracy.  That alien infestation?  A precursor to planet-wide "terraforming" and global conquest.

Now, world-wide is pretty dang big.  But if that's not quite enough for your players, maybe the problems are really...

5. Cosmic Horror.  The world-wide cult is trying to summon their Dread Lord into our world, effectively ending all life as we know it.  The alien invaders?  They were just trying to escape the real danger that's been chewing up galaxies for millenia -- and now it's reached us.  And all those people going crazy?  That was our reality crashing into another dimension -- the dimension of Darkness and Pain.

Only, instead of the GM making this up before the game, it happens in-game, with everyone contributing.  I'm not sure how this fits in with the Antagonist sheet -- perhaps we need a different sheet for each of these levels?

Ok, I don't know if this is helpful in any way, but while reading your last post, the image in my head was of a large sheet with a series of concentric circles. The middle one is "local", the outermost one "cosmic" with all the layers in between. Then there are smaller circles within, linked like one of those cloud/bubble chart thingies on ManyEyes, with words in them. The word-bubbles start in the middle "Local" circle, and link to other bubbles, most of which will be in the Local circle, but eventually may start appearing in the next ring out.

Am I making any sense?

I don't know how useful that is, but that's what it looked like in my mind's eye.

Concentric circles on a single page!  Great idea.  Much better than having multiple sheets for different types of antagonists, depending on how much information ends up on the sheet at the end.

I've been thinking lately of something very central to most Lovecraftian horror games, and that's Madness.  I want to connect sanity loss with the differences between your character's perception of the antagonist, and the Truth about the antagonist determined by whoever gets the authority on a roll.  Essentially, I want Sanity loss to occur (or at least the possibility of sanity loss to occur) whenever the difference between the character's belief of as aspect of the antagonist is revealed to be different from the truth of that aspect.

Additionally, I've been thinking about how the GM can introduce scenes where the characters actually interact with the antagonist in some way before the antagonist sheet is complete.  I think the GM should be free to have the characters encounter the antagonist (in some form or another) in such a way that he or she only reveals whatever is already on the antagonist sheet, plus one element (possibly just a clue element?).  I think of the scene in Se7en, where Brad Pitt's character finds the killer's apartment, and then the killer comes home; we see him in the distance, pulling out a gun and shooting at Pitt, and then (if I remember correctly -- it's been a couple years since I saw the film), there's a chase scene.  It's all the character interacting directly with the antagonist, but we don't learn, you know, all of the details about the killer -- maybe just one or two more than we already did.  This could scale to the broader types of antagonists, of course, where the characters come across a cult but don't learn the name of the god, etc, etc.

But this ties back into Sanity and Madness in that, in any scene, the GM can use any of the details already on the sheet; and at any time these can conflict with the ideas in the characters' mind, which might lead to madness, a loss of reason or sanity, that sort of thing.

I'm also thinking that the characters will have an "Area of Expertise" stat, which is their main avenue of investigation, and then perhaps a couple lesser ones, and then a few contact NPCs with expertise in other areas (one each) that they can bring in to help.

I'm a bit stuck on combat at the moment: not sure how detailed I want it to be, or how abstract.  I'd like to fit it into the "roll dice, some players win conflict, highest single die narrates" scheme, as above, but I'm not sure how.

I guess I need a strategy for how the player character stats are implemented.  Any suggestions?

The group I'm in currently is playing a game of Prime Time Adventures over Skype (actual play over at rpg.net.), and since cards are a bit tough to use over voice or text chat, we've adapted the rules for dice instead; me with the thought that it might come in handy for my game here, since the dice mechanics are going to be similar.

1. Even numbers are "successes"
2. High evens trump low evens on ties.
3. The person with the most successes wins the conflict
4. The person whose dice, added all together, total the highest, wins Authority.

I don't know about the "evens are successes" for my cosmic horror investigation game (I really need a better title than Cthulhu Now, don't you think?); it might be something like "3s and above on a d6 are successes"; which might give me the ability to play around with different modifiers to what numbers are successes or not (like in, say, Arkham Horror?).