Investigations in DCC

I mentioned before that I'm going to be taking over as the head writer for Discerning Dhole's CRAWL-thulhu zine. I haven't made many decisions about the future direction of the zine yet, but I know I want it to be set in a fictionalized Gilded Age (encompassing roughly the period from 1880-1945) and I know I want it to focus on mystery investigations.

Which has got me thinking more generally about the question, how do you conduct investigations in DCC? How do you handle skills? How do you handle clues?

Zine by John Potts, cover and interior art by Todd McGowan

CRAWL-thulhu issue 1 has a mystery investigation adventure, but basically everyone is willing to talk to you, and all the clues are laying out in the open. The core mysteries arise from the fact that a key witness is dead and a key source of danger is invisible. Essentially any character should be equally likely to solve the mystery - deciding where to go, what to look at, who to talk to are all tests of player skill instead.

But often in mystery investigation games, there's an element of character skill involved instead. The basic idea being that not every character should be able to find every clue. Even in Trail of Cthulhu (and related games) where any character with the right skill can find a clue automatically just by asking for it, they still have to have the skill, and they still have to ask if there's a clue. In the original Call of Cthulhu, you not only need to have the skill and ask to use it, you also have to roll the dice to see if you succeed at finding it. This introduces an element of ambiguity - was there really no clue there? or was there a clue but you failed to find it? (I'll talk about some possible solutions to the "what if they don't find ANY of the clues?" problem at the end. Dungeon Crawl Classics HAS a skill system that involves rolling dice already - so I'm NOT going to propose adopting Trail of Cthulhu's diceless skills.)

CRAWL-thulhu issue 2 introduces skills, and the list would look pretty familiar to players of both D&D and Call of Cthulhu.

There are a few other people who've written rules for conducting investigations in DCC, so I'm going to look at Brent Ault's Cyber Sprawl Classics, Stephen Bean's Bloody Hound character class for Julian Bernick's Nowhere City Nights, and Paul Wolfe's Dark Seas. Luckily for us, these are all freely available online, so they're very easy to look at.

In DCC, there are two types of skills - the named, formal skills practiced by Thieves, and the unnamed, informal skills that every character learns from their zero-level occupation. A Thief's formal skills can usually be substituted by an ability score check - although the Thief might roll against a lower DC, and always benefits from a bonus determined by her alignment and level. The informal occupational skills are considered either "trained" or "untrained" - and about a dozen occupations are likely to be considered "trained" for any particular task. Untrained characters roll a d10 to attempt the skill, while trained characters roll a d20. So using a trained skill in DCC is basically the same as making an ability score check.

CRAWL-thulhu's skills build on this framework. All skills start out untrained, and you can roll a d10 to attempt them. You get one trained skill from your occupation, and you can roll a d20 for that. As you gain levels, you earn "skill points" that you can either use to train in untrained skills, or to improve your training in a trained skill - becoming an expert who rolls a d24 or a master who rolls a d30.

There are two really basic ways to find clues in a mystery investigation - talking to people, and finding / analyzing objects.

There are also two really basic dangers to designing skills for a mystery investigation. The first is having too few skills - most people would agree that a single "Clue" skill is too few, and likewise that a "People" skill and an "Objects" skill is still not enough. The second danger is having too many skills. Consider the question of talking to people - if each PC occupation could only talk to NPCs in the same occupation, then surely having 100 different "Talk to Person of the Same Occupation" skills is too many. (A third basic danger is making the skill tests too difficult, which is related to the "what if they don't find ANY of the clues?" problem I'll discuss at the end.)

Zine by Brent Ault, Cover art by Korotitskiy Igor

In Cyber Sprawl Classics (CSC), player characters know Etiquettes that help them talk to NPCs. CSC treats Etiquettes a bit like foreign languages - everyone knows the common tongue, but you need a positive Intelligence modifier, a Lucky Sign, or a class feature in order to learn an Etiquette. If you are smart or lucky enough to know an Etiquette, you get to roll a d24 when speaking to the relevant NPCs, instead of the standard d20. So in this game, everyone is "trained" to talk to everyone else, but if you know the relevant "foreign language," then you become a bit of an expert.

There are seven Etiquettes - Academic (for talking to scientists and doctors), Corporate (for talking to CEOs and white-collar workers), Gang (for talking to criminals), Security (for talking to police and military), Runner (for talking to hackers), Socialite (for talking to "industrialists" and "the elite"), and Street (for talking to blue-collar workers and people who provide services to criminals.)

If that list sounds familiar to you, it's probably because it's so similar to the list of backgrounds available in 5e and the GLOG. Before looking at what other DCC writers were doing, I made a list of the way I would divide up Gilded Age society, and CSC's list is very similar to what I came up with. It's probably very similar to the list you would come up with, if you were thinking about how to divide virtually any Western society.

If I were to alter CSC's list, I think I would combine the Corporate and Socialite Etiquettes. In the Gilded Age, "society" was basically synonymous with the corporate elite and their families. That might be different in a cyberpunk game - indeed, in such a game, it might even make sense to have two skills for talking to the same person in two different environments, at work and at leisure. I might also do away with the Runner Etiquette, or combine it with Gang, since there isn't really any group analogous to hackers in a Gilded Age setting, and since the motives of any analogous individuals would be essentially criminal.

I like "etiquette" as the name for this kind of skill though. I'd thought of calling them "interaction skills," but I think "etiquette skills" might sound better.

There's also a question of how common these skills should be among characters. In 5e and the GLOG, essentially every character starts with one Etiquette due to their background. In CSC, only a fraction of characters know any Etiquettes. In the heroic fantasy of 5e, character backgrounds are mostly relevant for receiving material support from NPCs, and the support most NPCs provide is food and shelter, and perhaps friendship with a specific faction. In CSC, Etiquettes might have many uses, but they're optional, a bonus. You get along fine without them, you just get along better if you have them. But "etiquette skills" could be treated as a skill like any other, a skill that you could either be "untrained" or "trained" in - but doing that changes something else fundamental about how social skills work though.

If everyone has an "etiquette skill" (or, what amounts to the same thing, if not everyone has one, but nobody needs one) then it's possible to have other social skills as well - separate skills for persuading people, for tricking them, or for intimidating them. Those are the kinds of social skills we're pretty used to seeing. But, if not everyone starts the game with an "etiquette skill" and every NPC needs you to have one, then I don't think you can have separate "traditional" social skills as well. If the party wants to blackmail a robber baron, I think it's too much to ask for them to have both a "corporate etiquette" and a "blackmail skill."

So the question becomes, which is more interesting for a mystery investigation game? Is it more interesting if you have a skill to interact with corporate types in whatever way you please? Or is it more interesting if you have a skill to blackmail any NPC you come across? Which leads to more interesting dilemmas if you don't have the skill? Is it more interesting if you have "academic etiquette" and you have to try to find a scientist who can talk to the robber baron for you? Or is it more interesting if you have "intimidation skill" and you have to find someone you can bully into setting up the blackmail?

Roleplaying games, including D&D, including Call of Cthulhu, have traditionally answered the latter - that it's more interesting to use character skills to define a particular approach and then let the PC use that approach on any kind of NPC they want. But part of me wonders if it might not be interesting to try the former. Perhaps it's more interesting to use skills to define a kind of NPC and then let the players use whatever approach is situationally appropriate - but only on the correct kind of NPC. At least for a mystery investigation game, where (paradoxically) the whole point of skills is to not let every character find every clue. To misquote Maslow, if all you have is a Seduction skill, every NPC looks like a nail. But if the only kind of NPCs you can talk to are workers, then perhaps it forces you to get creative to figure out what happened inside that share-holders meeting.

Nowhere City Nights by Julian Bernick, Bloody Hound by Steven Bean

The "Bloody Hound" character class (BH) is an investigator character that Steven Bean wrote for Nowhere City Nights and published in the 2017 Gongfarmer's Almanac, volume 7. BH includes six skills for mystery investigations. The Bloody Hound character class gets all six, every other character gets a single skill based on their background.

BH's skills are Search Scene (for finding clues within a crime scene), Analyze Physical Evidence (for learning information from objects), Analyze Medical Evidence (for learning information from dead bodies, primarily), Interrogate - Charm (for making people want to talk to you), Interrogate - Intimidate (for making people talk to you even though they don't want to), and Conduct Surveillance (for staking out a person or location to see what happens.)

In terms of the effects of skills, BH distinguishes between finding a clue (with a "clue" here meaning an fact from an interrogation or an object discovered at a crime scene), making a deduction (which means analyzing the fact/object to learn what it tells you), and discovering an answer (which refers to piecing together several deductions to solve the mystery, or at least an important part of it.) So for example, finding a shell casing next to a murder victim would be "finding a clue," figuring out what kind of gun fired that bullet would be "making a deduction," and realizing who the shooter is would be "discovering an answer." Note that to discover the answer, you would need another strand of the investigation that tells you what type of gun a specific person has, so that you could later discover that that person is the shooter. BH also awards XP for each of these activities.

So BH makes a few key distinctions. First, it distinguishes between finding a clue and learning something from the clue. Those are two separate steps, and it's important for anyone adopting this approach to keep in mind that adding a step increases the chance of failure, especially if adding a step means adding a dice roll. Difficulty Classes that look intuitively too low individually can easily become too high collectively if you make ultimate success contingent on succeeding each roll in sequence.

Second, BH distinguishes between clues from objects and clues from talking to people. It does this in two ways. First the obvious - you use one set of skills to find and analyze objects, and a second set to learn information from NPCs. But second, and less obviously, you only have deductive skills related to objects. You make one roll to find an object at a crime scene, and a second roll to learn something from it. But when conducting an interrogation, you make one roll to learn a fact, and then ... It's possible that you make a second roll on the same interrogation skill to get the person to tell you what you deduce from the clue. It's also possible that making deductions from verbal clues is a player skill, and not a character skill.

I agree that "discovering an answer" - that is, finally solving the mystery - should be a player skill that doesn't rely on rolling the dice. I'm not sure if I agree that "making a deduction" should be a player skill, or at least, not always. Some information NPCs give you is going to be clearly useful. It will either already be a deduction, or it will clearly point to a deduction that the players can make. But if an NPC tells the players something, and they just have no idea what to do with that information, it seems like it might be nice to have some mechanism in place to let them ask the judge for help. The danger of that is players relying on that mechanism instead of their own thinking, or judges insisting on that mechanism even when the players are able to deduce on their own. If you don't create such a mechanism, then no one can abuse it. But also, no one can use it in a real emergency. I guess it's the same problem you run into with traps in D&D, where it's inherently ambiguous whether you should find them with player skill or character skill, and where any GM hoping to rely on player skill is at the mercy of the adventure writer to provide enough detail to make that possible. (Although we're veering into "what happens if they don't find ANY of the clues?" territory here, so let's come back to this.)

What I find especially useful in the "Bloody Hound" class description is the idea that learning from clues in a mystery investigation is a two-step process, and that it might be profitable to separate those steps.

Dark Seas by Paul Wolfe

Dark Seas (DS) is a mini-setting with it's own fairly complete set of rules modifications that Paul Wolfe wrote and published in the 2017 Gongfarmer's Almanac, volume 4. 2017 was a good year for DCC mysteries! DS doesn't have any specialized skills for investigation, but what it does have is a really excellent interpretation of clues and how to use them.

Let me start with what I consider to be the key takeaway, and then back up. Every clue is an object. You might find some clues by talking to people and other clues by looking around the environment, but what you GET when you find a clue, what you KEEP once you have it, is a physical object. Like any other object, it goes in your character inventory.

But what that means for a mystery game, is that when you want to take stock of your investigation so far, you don't have to wrack your brain trying to remember every detail, you just look through your inventory and see which clue-objects are there. If you need help remembering what a particular clue told you, you just ask the GM to describe the object again. All this is probably easier than tracking ephemeral bits of information that are untethered from any specific reminder. I think this is brilliant, and I definitely plan to take Paul's advice.

So technically, in DS, Paul doesn't talk about "clues" but rather about Secrets. As mentioned, each secret takes the form of a physical object. Players collect Fragments like treasures as they explore - and 10 fragments combine to form one secret. In the example adventure, characters can collect fragments by doing things like searching a dead body, gathering rumors in a bar, inspecting magic items, questioning NPCs, they can be acquired like treasure from defeated monsters, and they're a reward for finding islands. The number of fragments acquired at one time is generally random, and is usually somewhere on the order of 1d10 fragments per investigative activity (although sometimes you get a full secret at one go).

I don't know if I would use this approach, but it encourages players to search as many places as possible, and it means that you don't need to know the meaning of every fragment, only the meaning of the final secret (clue) once it's assembled. And, you get to pick which secret you give them, which could maybe avoid the problem of finding a lot of clues hinting at one thing, while missing all the clues hinting at something else. Some examples of secrets in DS are port reports and charts of the sea, but also ghost stories and chess moves. Each character begins the game with a "starting secret" that grants them one boon, so for example you can have a political pamphlet that gives you an NPC contact, a last will and testament that gives you money, or a racy novel that gives you a bonus on certain saving throws.

I'm not completely convinced the experience system in DS would really work in practice the way Paul seems to want it to. When characters find fragments, they divvy them up, each character gets their own secret at 10 fragments. Characters earn XP for secrets - although not for finding them, but rather for divulging them to an NPC confessor. I think you're supposed to need a new NPC for each secret, although that could add up quickly. Raising 4 PCs from 0th level to 1st level would take 40 secrets and 40 NPCs ... which feels like kind of a lot. Starting secrets also need to be divulged in order to earn their benefit, which seems more appropriate. I'm quibbling over details at this point though - the big takeaway that every clue is an object is still absolutely brill.
 

Zine by John Potts, cover and interior art by Todd McGowan

 
Finally, as promised, let's address the question "what happens if they don't find ANY of the clues?" How are you supposed to run a mystery if there's a chance that the players won't find, or won't be able to interpret, ANY of the clues that are left for them?

1) First, and most obvious, give lots of of clues. The Alexandrian famously recommends including a minimum of 3 clues for any conclusion you want your players to draw.

The point is that in order for there to be ENOUGH clues for the players, there need to be what feels like TOO MANY clues from the perspective of the judge. The judge can see everything, the players will only ever experience a fraction of it. The judge also knows all the answers from the outset, and so can instantly see how each clue points to each conclusion. The players are assembling a mental image piece-by-piece, and it's not always immediately clear where each piece goes.

2) Second, provide multiple sites of investigation. Give the players several distinct places to go look for clues. Following the Alexandrian's advice again, at every site, leave clues pointing to the final solution AND clues pointing to the other investigative sites.

Realizing that there's another place to go look can feel like a discovery in itself, and leaving one site to go to another can feel like forward progress is being accomplished. Movement between sites also passes some time that gives the players a chance to think, and creates opportunities for new information to become available.

3) Third, use the random encounter table to provide breaks in the case. Mysteries don't necessarily need wandering monsters the way other D&D adventures do, but random encounters are still useful for pacing and for marking the passage of in-game time.

Each day that passes with no solution to the mystery, allow events to be in motion. Maybe the criminal keeps committing similar crimes. Maybe the criminal gets spooked and engages in some kind of cover-up. Maybe new witnesses come forward. Maybe new sites for investigation are revealed. Maybe an NPC investigator got killed but left a diary behind. These random events provide verisimilitude, they can be a way to just GIVE the players a clue they might need, and they should almost always open up some new avenue for investigation that wasn't available before.

4) Fourth, speaking of just giving the players clues, sometimes just GIVE the players clues. Sometimes don't require a skill check. Sometimes just let the clue be sitting right out in the open, so all the players have to do is say they want to look at it. Sometimes let the witness be perfectly willing to talk, so all the players have to do is say they want to talk to them. Sometimes, the barrier of the players having to notice that they want to look at something or talk to someone is going to be enough without getting the dice involved at all.

Alternatively, if you're going to require a skill check to find the clue, then consider just TELLING the players what it means. You want to be a little careful with this, because you don't want to rob your players of the chance to exercise their player skill at solving mysteries, BUT if you're going to require a skill check to FIND the clue in the first place, then maybe don't require a second check to discover the meaning of the clue.

Always be careful not to set your skill check DCs too high, and be DOUBLY careful not to make the checks too difficult by requiring multiple rolls to succeed. What sounds like "this is an appropriate test of skill" to a person just READING the adventure will often turn out to be too difficult to people actually playing through it. What sounds like "this is way too easy" to someone who's just reading will often turn out to be appropriately difficult for actual players. Set your DCs for players, not for readers. And wherever you set your DCs, make the reward proportionate to the difficulty. If you need one check to find the clue and another to research it, then the reward for those paired successes had better be a REALLY GOOD CLUE so that the players' efforts are worthwhile.

5) Fifth, give the players multiple opportunities to find and interpret each clue. If they fail once, give them a second try. If one approach comes up short, let them attempt another.

Use these multiple attempts to create the narrative of the adventure. Maybe the first time the PCs search a room, they try just looking around very carefully during a house party. If that fails, they can try searching a second time, but they have to try a different approach. Perhaps they try breaking in and tearing the room apart looking for secrets. Perhaps they hire a professional burglar to search the room for them. Make sure there are narrative consequences for whatever approach they choose. The first attempt requires getting invited to the house party and roleplaying interactions with the other guests. The second attempt is sure to tip off the house owner that somebody's on to them. Hiring a burglar is going to require using criminal etiquette to make contact with the local underworld.

If a character can't interpret the meaning of a clue, let them try again if they can get access to a library or a lab. Or let them find an NPC who can interpret it for them. NPCs don't need to make skill checks. Picking the right NPC to ask, and using your etiquette skill to ask them, is difficult enough. There's no reason to add another chance of failure by making the NPC roll the dice as well.

The point is, failing once shouldn't mean failing forever. Players should have multiple clues they could find, multiple ways to get information out of each clue, and multiple ways to "get help" if they find they can't do it alone.

6) And finally, what happens if they can't find any of the clues? Let them fail. Give the whole mystery some kind of time limit. Create consequences for failing to solve it. And if the players fail, let them fail. The killer keeps on killing. The burglar pulls of their heist. The sorcerer summons the monster. The monster destroys the city and slinks back into the ocean.

If you've provided lots of clues, made them easy to find and easy to interpret, allowed second chances for anything the players want to try again at? Then let them fail. Just make sure their failure is legible. At least let them understand the solution to the mystery when they see what happens as a result of them not stopping it. Nothing's going to be less satisfying than having the mystery end and STILL not understanding what happened. If the players don't stop the villain, then at least let them watch the villain take their mask off, or gloat in triumph, or commit one final crime right before their eyes.

At that point, you've created a recurring villain, and a chance for your players to shout "I'll get you next time, my pretty!" When the villain DOES recur, the players have a much better shot at stopping them the second time around. Or, if they're very proactive, they can start planning to bring the pain directly to the villain's doorstep. Either way, failure in one case can be made to simply raise the stakes and make another case more interesting.