So, I'm trying to look into developing an investigative RPG. It doesn't need to be as crunchy as 3.X, so I'm trying to consider my options and hopefully see some productive input that I can crib off of. On the one hand, crunchy systems provide a good foundation for predictability that can be a source of clues. On the other hand, they can serve as a distraction to narrative immersion, as many clues are based on context rather than "how many castings of flamestrike does it take to incinerate a body?" Similarly, relying on rules mastery to solve clues cleaves too close to ivory tower design for my taste, so going for something more rules-lite is probably the way to go. Not to say we should eschew rules or structure, but aiming for something closer to Fate or even an OSR is likely the rough goal here (unless you think After Sundown is well-suited?).
A possible line of inspiration is to crib from Outfoxed, which is a logic game of whodunit. It's got a built in progression, meaning that missing a clue just means you get to try again (and the monster kills again). The problem with cribbing a Guess Who logic puzzle is incorporating it seamlessly into the context of a tabletop RPG without it just being a board game with funny voices.
I guess I should consider the kinds of scenarios I want to see out of this hypothetical game.
[*] A ghost is terrorizing a hotel. You need to find out why they're restless & fix it. Mafia goons may disagree with you knowing.
[*] Whether it's an isolated town, a cruise ship, or a locked down mansion, someone is a murderer and you need to find them before they kill again.
[*] A beast is attacking every night, but you don't know if it's a werewolf, a vampire, a wendigo, or a gnoph-keh. The identity is important, as each requires a different method to locate & defeat (though technically all of them burn).
From these scenarios, social mechanics don't need to be any deeper than handling interrogation; though it absolutely needs to be able to handle a couple different types. Combat can be simple & quick since it's not a tactical simulator, but obviously needs to be there as a source of tension. Resource management can be very readily hand-waved; though not totally ignored since monster weaknesses are a facet and choosing to carry a silver blade as opposed to iron should mean something.
I'm going to sleep on it and mention ideas sometime tomorrow.[/url]
Investigative RPGs (No PhoneLobster)
Moderator: Moderators
Investigative RPGs (No PhoneLobster)
Come see Sprockets & Serials
How do you confuse a barbarian?
Put a greatsword a maul and a greataxe in a room and ask them to take their pick
How do you confuse a barbarian?
Put a greatsword a maul and a greataxe in a room and ask them to take their pick
EXPLOSIVE RUNES!
The Alexandrian has a few articles about mysteries that I found interesting to read. He has an article that has something to do with using nodes as well for this but I can't pin down the name of the article at the moment.
If you're going for this kind of time-sensitive investigation, then a push-your-luck mechanic could suit things well.
Say you start out with a good chance to find clues/get info from suspects/track people down/etc
Each time you follow a lead your success chance goes down, until you restart ("the monster killed again!")
Alternatively, each try could deplete some metagame currency, in a way that leaves you uncertain when you'll run out.
You could even combine the two, paying to convert failures to successes. That way, you deplete little or none at first, and more and more as you go (but always with the chance to continue).
Meanwhile, you're getting clues. You could leave it to the players to solve the mystery, or have a more formal system for identifying and eliminating suspects.
players could end up in a situation where they've run down their leads on night one and don't have enough clues to be certain, so they decide to gamble on an educated guess and pull out the garlic and wooden stakes. Maybe they're right, maybe it's on to "night two" for more investigation (each night could be a week instead, or whatever arbitrary unit)
Say you start out with a good chance to find clues/get info from suspects/track people down/etc
Each time you follow a lead your success chance goes down, until you restart ("the monster killed again!")
Alternatively, each try could deplete some metagame currency, in a way that leaves you uncertain when you'll run out.
You could even combine the two, paying to convert failures to successes. That way, you deplete little or none at first, and more and more as you go (but always with the chance to continue).
Meanwhile, you're getting clues. You could leave it to the players to solve the mystery, or have a more formal system for identifying and eliminating suspects.
players could end up in a situation where they've run down their leads on night one and don't have enough clues to be certain, so they decide to gamble on an educated guess and pull out the garlic and wooden stakes. Maybe they're right, maybe it's on to "night two" for more investigation (each night could be a week instead, or whatever arbitrary unit)
Chamomile did a nice writeup on mystery adventures in his blog. He suggests building a directed graph where nodes are scenes, edges are clues, and each node has at least three edges. The total size of the graph and the length of the shortest path to the conclusion give some sense of how hard it is. His "at least three edges" rule deals with rule 1 of mysteries - everything is always more opaque to the audience than it is to the author. (If this is too abstract, try Cham's blog version. He doesn't invoke graph theory, and it's full of good GM advice.)
If you start adding skill-based or other mechanical roadblocks that may prevent finding clues, you'll need more clues to compensate. Maybe you make sure there's at least three clues found per scene on average. Or maybe you make sure there's always three paths to the solution with pure narrative-driven clues and no failable mechanics, and the mechanical versions are just shortcuts.
(Also I really like Whatever's suggestion of push-your-luck mechanics in a mystery game. That's a very good genre fit.)
If you start adding skill-based or other mechanical roadblocks that may prevent finding clues, you'll need more clues to compensate. Maybe you make sure there's at least three clues found per scene on average. Or maybe you make sure there's always three paths to the solution with pure narrative-driven clues and no failable mechanics, and the mechanical versions are just shortcuts.
(Also I really like Whatever's suggestion of push-your-luck mechanics in a mystery game. That's a very good genre fit.)
-
OgreBattle
- King
- Posts: 6820
- Joined: Sat Sep 03, 2011 9:33 am
Poker is investigating the mystery of what's in other people's hands and who's lying, you get more clues as the rounds go on but also increasing risk to stay in the game.
the "Total Party Kill" of an investigation is everyone not knowing what to do or literally getting killed by the unsolved conspiracy or returning god.
the "Total Party Kill" of an investigation is everyone not knowing what to do or literally getting killed by the unsolved conspiracy or returning god.
At first glance, Chamomile's post seems pretty derivative of the Alexandrian articles MGuy mentioned. That's totally okay of course, but I would have expected to at least see a citation or reference.jt wrote:Chamomile did a nice writeup on mystery adventures in his blog. He suggests building a directed graph where nodes are scenes, edges are clues, and each node has at least three edges.
Out beyond the hull, mucoid strings of non-baryonic matter streamed past like Christ's blood in the firmament.
I feel it's fundamentally flawed to look to mystery Novels as a guide for how to write Mystery Games. They're two very different mediums. And even mystery video games tend to lean heavily on the players object spamming and/or save scumming to brute force things when they can't figure out the puzzle. I can present every item in my inventory to a videogame NPC until one of them triggers the next dialogue scene, I can't really do that in a tabletop.
I think it's important to identify the function you want the mystery to serve in the game before you attempt to tie mechanics to it.
- Is the Mystery here to facilitate exploration or an adventure?
- Is the Mystery there to be a solveable puzzle?
- Is the Mystery there to provide suspense? A ticking time bomb?
After all there's a different feel to trying to solve a cold case vs an active murder investigation. Or a treasure hunt vs a conspiracy.
I think it's important to identify the function you want the mystery to serve in the game before you attempt to tie mechanics to it.
- Is the Mystery here to facilitate exploration or an adventure?
- Is the Mystery there to be a solveable puzzle?
- Is the Mystery there to provide suspense? A ticking time bomb?
After all there's a different feel to trying to solve a cold case vs an active murder investigation. Or a treasure hunt vs a conspiracy.
Good RPG systems have both narrative rules and mechanical ones and for something as broad as investigating mysteries you're going to need have both sides locked down. If the players make someone reveal more than they intended you should be able to translate that narrative event (“alibi said more than intended”) into a mechanical output (+1 Clue). Ideally you could do it the other way too by printing lots of suggestions for what players finding +1 Clue could mean exactly.
Lets say our game has a mechanical value to clues: Clues give +X to Investigate checks per clue. After you gather your clues it's a D20+Mod roll VS Mystery DC whenever players decide they want to try to solve it.
Now you create a set of definitions. You define that there’s three kinds of clues, small, normal, and big.
Small Clue +1
Clue is +3
Big Clue +5
Define those three things narratively. Describe the differences in enough detail that 4 people around a table will probably agree on what “the alibi saying too much” would count as if they all heard it.
Now you have both mechanical and narrative definitions. You can write in your session plan that the PC’s will get 1 Big Clue if they go to the Warehouse, and create the narrative of what that clue is in the moment. In that situation you have the mechanics planned (PC’s find +5 to investigate token in Warehouse) but not the specific bits of dialogue and narrative output that will create it. Inversely if the PC’s are talking to someone you didn’t plan, or you improvise a line that would lead the PC’s in the right direction then you can define that narrative event as “a clue”. Where you put a mechanical value on the narrative input to the game.
If you create narrative definitions for a word you can translate any input into a mechanical output, and vice versa. The only task then is to make the mechanics of the game feel like they mirror the pace and feeling of a show about detectives solving a mystery. If the roll and result happen too quickly it will feel bad cause that’s not the image people have, if you don’t have to pursue leads and talk to dockworkers it will feel bad because that’s not the image people have. Toughest of all you have to consider outside systems, cause it may turn out that the best way to solve mysteries is to cast Sadism on yourself and kill 50 rats for +100 to your Investigate check. Because that's definitely not mirroring people's mental image.
But once you’ve got your narrative and mechanical definitions for the pieces you think are important to appear in your mystery it’s just playtesting some bullshit game with those elements and seeing which bits players want more of and less of. Maybe they want clues to expand into bigger clues if they spend time following them, so you do that. Maybe they get stuck a lot so you make "Hint" metacurrency which lets people just have a hunch where to go next. Who knows. Create your definitions, narrative and mechanical, put them into a game that does your best at mirroring the genre you're aping, then run players through it and modify modify modify.
Lets say our game has a mechanical value to clues: Clues give +X to Investigate checks per clue. After you gather your clues it's a D20+Mod roll VS Mystery DC whenever players decide they want to try to solve it.
Now you create a set of definitions. You define that there’s three kinds of clues, small, normal, and big.
Small Clue +1
Clue is +3
Big Clue +5
Define those three things narratively. Describe the differences in enough detail that 4 people around a table will probably agree on what “the alibi saying too much” would count as if they all heard it.
Now you have both mechanical and narrative definitions. You can write in your session plan that the PC’s will get 1 Big Clue if they go to the Warehouse, and create the narrative of what that clue is in the moment. In that situation you have the mechanics planned (PC’s find +5 to investigate token in Warehouse) but not the specific bits of dialogue and narrative output that will create it. Inversely if the PC’s are talking to someone you didn’t plan, or you improvise a line that would lead the PC’s in the right direction then you can define that narrative event as “a clue”. Where you put a mechanical value on the narrative input to the game.
If you create narrative definitions for a word you can translate any input into a mechanical output, and vice versa. The only task then is to make the mechanics of the game feel like they mirror the pace and feeling of a show about detectives solving a mystery. If the roll and result happen too quickly it will feel bad cause that’s not the image people have, if you don’t have to pursue leads and talk to dockworkers it will feel bad because that’s not the image people have. Toughest of all you have to consider outside systems, cause it may turn out that the best way to solve mysteries is to cast Sadism on yourself and kill 50 rats for +100 to your Investigate check. Because that's definitely not mirroring people's mental image.
But once you’ve got your narrative and mechanical definitions for the pieces you think are important to appear in your mystery it’s just playtesting some bullshit game with those elements and seeing which bits players want more of and less of. Maybe they want clues to expand into bigger clues if they spend time following them, so you do that. Maybe they get stuck a lot so you make "Hint" metacurrency which lets people just have a hunch where to go next. Who knows. Create your definitions, narrative and mechanical, put them into a game that does your best at mirroring the genre you're aping, then run players through it and modify modify modify.
Last edited by Dean on Sat Sep 26, 2020 6:38 pm, edited 1 time in total.
DSMatticus wrote:Fuck you, fuck you, fuck you, fuck you. I am filled with an unfathomable hatred.
-
JigokuBosatsu
- Prince
- Posts: 2549
- Joined: Tue Aug 10, 2010 10:36 pm
- Location: The Portlands, OR
- Contact:
I had had high hopes for the Sphynx RPG, and now that I have gotten to read the English translation I am quite disappointed. The people who had played it years ago spoke highly of it having a unique research/investigation mechanic but it pretty much has no mechanics at all. It's a mother-may-I story game that uses counters on a couple things, mainly to track the progress of exploring the ruins. Otherwise, meh. It barely even helps you create your researcher and the ruins they explore. Dean's last post was a more complete game in this vein.
So... there continues to be plenty of space for investigation-focused RPGs.
So... there continues to be plenty of space for investigation-focused RPGs.
Omegonthesane wrote:a glass armonica which causes a target city to have horrific nightmares that prevent sleep
JigokuBosatsu wrote:so a regular glass armonica?
My main problem with puzzles (and most mystery minigames are just a variant of a puzzle, in my experience) is that only one or maybe two people are engaged with it while the rest of the players are just sitting around. At least in an escape room you can split into teams of one or two and solve multiple puzzles in parallel.
What I'm currently thinking is for the investigative rules to instead be essentially a rubric for developing a module. One of my examples was identifying the monster so you can confront it more readily, so that is what I will focus on. I'm hoping the structure I develop here can be kit-bashed for the other two examples.
What I'm currently thinking for this is for the setting to have well-developed bestiary to avoid the Steve problem. Each monster will have a set of traits that, if you figure out enough, you can identify the monster and use that knowledge to know its weaknesses. As a benefit, the investigative traits can be used for procedural generation. The specific traits I'm thinking of are those that would be discerned from investigation; the tentative list being...Well, it's got six legs, which means this jasmine will either kill it, make it even more violent, or do nothing.
- Tracks (literal, collateral damage, droppings, etc)
- Time of Attack
- State of Victim's Corpse
- Rough Appearance (eyewitness accounts shouldn't/can't be detailed)
- Location of Attack
- Preferred Target
- Reaction to Investigation
Come see Sprockets & Serials
How do you confuse a barbarian?
Put a greatsword a maul and a greataxe in a room and ask them to take their pick
How do you confuse a barbarian?
Put a greatsword a maul and a greataxe in a room and ask them to take their pick
EXPLOSIVE RUNES!
In this list, Tracks is a pretty extensive category and Time of Attack is a pretty narrow one compared to the other three. You might want to split "collateral damage" out of Tracks (for which Spoor is probably a better name) and put it in an Environmental Impact category, which would include incidental damage like snapped branches, major damage like burning buildings, and things like "it looks like the victim was dragged along the ground, so it's possibly an X because they like to play with their food" or "the door is locked and no one broke in, so it's probably a Y because they can go through walls."virgil wrote:The specific traits I'm thinking of are those that would be discerned from investigation; the tentative list being...
- Tracks (literal, collateral damage, droppings, etc)
- Time of Attack
- State of Victim's Corpse
- Rough Appearance (eyewitness accounts shouldn't/can't be detailed)
- Location of Attack
Similarly, you could broaden Time of Attack into Mystical Circumstances, which would include time of day, significant dates (e.g. solstices and holy days), astrological/astronomical phenomena (e.g. eclipses and conjunctions), and so on. Though in that case it might make more sense as one of the more subtle clue categories, as that can get into fairly arbitrary territory (e.g. "the Dark Sidhe only appears at midnight during a lunar eclipse").
Good options here would probably depend on the cosmology of the setting. Are all the supernatural beasties powered by parallel worlds like in After Sundown, such that e.g. noticing that there was a large reflective puddle nearby that could have served as a portal to the Mirror Realm will help you narrow down your target? Are monsters powered by (or do they imprint their surroundings with) certain flavors/colors of magic like in D&D or Magic, such that e.g. identifying the lingering magic in the air as the taint of Chaos rather than Necromancy helps you eliminate some suspects?What this means is that the PCs must either wait for the creature to strike again to generate more clues, or for more subtle clues to exist that aren't the first things investigators will aim for. Examples in this includeThese two are hopefully are strong clues simply because they require more time to pass, but I'd like more options here. I would, however, like to have more options here.
- Preferred Target
- Reaction to Investigation
For outputs you seem to want suspects.
If 0 clues gives you no suspects, 1 gives you 100? 10?
If you can convert clues to suspects, you could also introduce 'false clues' that might introduce false suspects ('faked tracks', purposely killing during the 'wrong time'), or have clues be misinterpreted. Suspects can also provide alibi's, real or fake to be removed from the suspects list right, or wrongly.
Having a list of known suspects, the investigators could try gathering supporting clues, to rule out or confirm a suspect, so activities like tailing a suspect, visiting known haunts of the suspect, interview suspect, investigation of suspects where abouts etc can provide plenty of follow up clue generation.
If 0 clues gives you no suspects, 1 gives you 100? 10?
If you can convert clues to suspects, you could also introduce 'false clues' that might introduce false suspects ('faked tracks', purposely killing during the 'wrong time'), or have clues be misinterpreted. Suspects can also provide alibi's, real or fake to be removed from the suspects list right, or wrongly.
Having a list of known suspects, the investigators could try gathering supporting clues, to rule out or confirm a suspect, so activities like tailing a suspect, visiting known haunts of the suspect, interview suspect, investigation of suspects where abouts etc can provide plenty of follow up clue generation.