The topic for today is mysteries. Not mystery adventures. Mysteries. Yeah, I know that sounds like a bunch of semantical bulls$&%. But it’s not. There’s a difference between designing a dragon — you know, making a stat block — and writing an adventure about slaying a dragon.

You never thought about that, did you?

Thing is few D&D adventures are actually about what they’re about. A dragon-slaying adventure’s only about slaying a dragon for the last half hour or so. All the rest’s about traveling across the wilderness and encountering all sorts of things that aren’t dragons. It’s just surviving until you get to fight the dragon.

But that’s a whole other f$&%ing issue with RPGs like D&D. Hell, with most RPGs. They don’t tell you how to build an adventure that’s entirely about something instead of running a gauntlet and then doing something.

Thing is the mystery in a mystery adventure is like the dragon in a dragon-slaying adventure. Or, put another way, what makes a mystery adventure a mystery adventure is that there’s this puzzle to solve at the end of it. Of course, unlike dragon-slaying adventures, mystery adventures are usually all about solving the mystery.

But, again, I ain’t talking about mystery adventures today. I’m talking about the puzzles at the heart of mystery adventures. The mysteries themselves. Most GMs don’t realize there’s a difference. And since most GMs think they know how to write their own adventures — thanks to the efforts of Messers Dunning and Kreuger — since most GMs think they can write an adventure, they think they can write a mystery.

The real problem’s that most GMs — hell, most people — don’t actually know how mysteries work. And the people who talk about building mystery adventures — and the RPG systems that claim to be about mysteries — don’t bother explaining any of that s$&%. They think building a good mystery is as easy as building a good adventure.

Is this making sense? Probably not. Point is, this article’s about mysteries. About how mystery puzzles work. And how people solve those puzzles. It’s not about building a mystery adventure. But if you don’t get this s$&%, you can’t build a mystery adventure. You can’t run a dragon-slaying adventure if you can’t design — or at least run — a good dragon fight.

Demystifying Mysteries

What is a mystery? Well, there’s two meanings to the word. And I’m only discussing one of them today. First, a mystery is a story or game in which one or more investigators solve a mystery. Second, a mystery is a thing that one or more investigators solve in a mystery story or mystery game. Got it?

When you get down to it, a mystery’s just a puzzle. Which I sort of hinted at in the Long, Rambling Introduction™. It’s a logic puzzle. A specific kind of logic puzzle. And a mystery adventure is about solving just such a puzzle. Today, I’m talking about the puzzles themselves. Not mystery stories or mystery adventures. I’m talking about how the logic puzzle that is a mystery goes together.

More specifically — because people are really f$&%ing careless with their language and use words however the f$&% they want with no regard for clear communication — more specifically, I’m talking about a specific kind of mystery puzzle. The most common kind. I’m talking about Whodunnit mystery puzzles.

A Whodunnit is a mystery puzzle in which something bad’s been done and the investigators — the protagonists — have to figure out who it was that did the bad thing. Ostensibly to serve justice. Or whatever. Doesn’t matter why.

Point is, there’s a crime. There’s a perpetrator whose identity is unknown at the start. And there’s a group of protagonists whose job is to deduce the perpetrator’s identity. That’s the mystery puzzle. Who did the crime? The mystery adventure is all the s$&% the investigators do to solve the mystery. Mostly hunt for information from which to deduce the solution while overcoming obstacles and challenges.

Howdunnit

It’s actually easier to understand Whodunnits if you know a little about how Whodunnit mysteries actually get solved. That is, how does a group of smart and capable investigators — i.e. not your players — actually investigate and solve a Whodunnit.

As I said above, a Whodunnit starts with someone having dun something. Something bad. They stole something or destroyed something or unleashed something or killed someone or defrauded someone or whatever. The thing that got dun is the crime. That’s the word I’m using for it. And the who who dun the crime? They’re the perpetrator.

I’m just laying out some terms here.

So, crime’s been dun by an unknown perpetrator. Then, along come the intrepid incompetent investigators. The heroes whose job it is to figure out who dun it.

The investigators usually start by coming up with a shortlist of possible perpetrators. Whos who could have dun it. Those possible perpetrators are called suspects. And the initial list of suspects might have as few as one name on it or as many as lots. Depends on the mystery. Hell, technically, the initial list of suspects could be blank. But I’ll get to that later.

Once they’ve got the initial suspect list, the investigators start digging for information about each suspect. They try to gather enough information to convince themselves — or the court or the king or their employer — that one of the suspect’s probably the perpetrator. At the same time, they also seek information that eliminates any suspects from the list. After all, if there’s only three possible suspects and you can eliminate two of them, you probably have your answer.

This probably isn’t news to you. I’m telling you s$&% you already know. But hold on. Because there’s actually very specific bits of information that almost all Whodunnit mysteries revolve around. And that makes it really easy to design a good mystery puzzle. But that’s coming up in the next section.

Right now, just know that there’s a specific list of three bits of information that establish someone as a suspect. So, after populating an initial suspect list, the investigators focus on qualifying each suspect — by finding those three bits of specific information — or eliminating each suspect — with information that counters one of the three specific bits of information.

Eventually, the investigators end up with a list of probable suspects. Maybe one. Maybe several. And then it’s down to digging up enough information to build a strong enough case against one of the suspects that the investigators feel pretty confident fingering them as the guilty party. And it’s important to note that, rarely do the players actually know for certain that they got the right perpetrator. Oh, sure, sometimes the perpetrator confesses during the final showdown and then gets thrown from a building by accident so the investigators didn’t actually kill him. But often, the investigators are making their best guess based on whatever information they actually found. And if they missed any important information, they might get the wrong person. Hell, they might not feel comfortable making any guess at all.

But then, part of playing a mystery game is the risk of failure. Missing clues or making the wrong deductions and, as a result, throwing the wrong person off the building in the climax. Or not knowing who to throw off the building and never closing the case.

Point is, a Whodunnit’s usually solved in phases. In the first phase, the investigators build an initial suspect list. In the second, they qualify or eliminate the suspects until they’ve got a manageable shortlist of suspects. Prime suspects. In the third, they gather enough information on their prime suspects to pronounce one of them guilty. And along the way, there’s sometimes a troubleshooting phase wherein they realize they don’t have any prime suspects — or any suspects at all — and they’ve got to go back over all the s$&% they’ve already done and figure out what they missed.

Here’s an example.

Someone stole the cookies from the cookie jar in the bakery storeroom. The investigators examine the scene and talk to the bakery owner and the witnesses and come up with a list of four initial suspects: Alice, Bob, Carol, and Dave. They question each of the four suspects and then try to confirm or disprove the information the suspects provide. They investigate each suspect, talking to their friends and families and digging into their personal histories. S$&% like that. As a result, they eliminate Alice and Dave as possible suspects. Left with Bob and Carol as prime suspects, they talk to each of them again and then go back over the crime scene and question a few people again. In doing so, they discover that something Bob said doesn’t match up with the crime scene. They confront Bob with that information, Bob panics, there’s a chase, and then Bob accidentally falls off an overpass directly into the path of a truck delivering baked goods. Ironically. “I guess that’s the way the cookie thief crumbles,” quips one of the investigators while putting on a pair of sunglasses. Then the credits roll.

Formulaic? Yeah. Definitely. But that’s not down to a lack of creativity. Rather, it’s formulaic because that’s the optimal way to solve a Whodunnit Mystery. When you’re doing a cryptogram, you look for the most common letter and assume it’s an e. That’s not formulaic. That’s just how you solve the puzzle.

Now, about those bits of information…

The Three Ms: Motive, Means, And mmm…Opportunity

If you’re a fan of crime dramas and police procedurals, you’ve heard all about motive, means, and opportunity. And there’s a reason. That’s because those are the three necessary pieces of information that show someone could have committed a crime. They don’t prove that someone did commit the crime. Just that they could have. If someone’s got the motive, means, and opportunity, they’re a suspect. And if you can prove someone didn’t have a motive or the means or the opportunity — or any combination thereof — they ain’t a suspect.

First, there’s motive. Motive refers to a reason to commit a crime. People — especially fictional people — don’t do things without reasons. Humans don’t act at random. Which is why chaotic-neutral is a stupid alignment. Committing a crime is an extreme act. I know most gamers don’t think killing’s a big deal — which scares the s$&% out of me — but it is. So is assault. And theft. Most people don’t commit crimes. When people do commit crimes, there’s a good reason for it. Well, not good. Not morally good. Strong. A strong reason.

Point is, when someone does a crime, there’s usually a strong and personal reason for it. Almost always an emotional reason. And those reasons are actually a very good way to populate an initial suspect list. This is why cops always ask whether a victim had any enemies and they look into the financial status of anyone who had anything to gain from committing the crime.

Next, there’s means. Means refers to the ability to commit a crime. If the crime involved a weapon, the perpetrator must own or have access to the said weapon and must know how to use it. If the crime involved a feat of strength, the perpetrator has to be strong. If the crime involved a lot of planning, the perpetrator has to be smart. If the crime involved esoteric or secret knowledge, the perpetrator has to know it. If the crime was expensive, the perpetrator must have access to money. See how this works? It’s pretty f$&%ing simple.

Finally, there’s opportunity. Opportunity means having the chance to commit a crime. And usually, opportunity’s a matter of someone actually being there when the crime happened or else having access to the scene of the crime. That’s why cops on TV are always trying to “put the suspect at the scene” with physical evidence.

So, motive, means, and opportunity. If somewho had a good reason to dun the crime, if they were able to dun it, and if they were at or near the place where it got dun, they’re a suspect. And somewho that lacks a motive or the means or the opportunity can’t possibly have dun it.

Now, remember that s$&% about how crimes get solved? In the beginning, the investigators have to put together a list of initial suspects. And usually, they’ll stick a person on that list based on just one of the three important bits of information. If someone got dead and you have a reason to want them dead? That’s a motive. You’re on the initial suspect list. If someone got dead at an exclusive party and you were at the party? That’s an opportunity. You’re on the initial suspect list.

Back to the Cookie Crook Caper. The investigators get called by the bakery owner to figure out who stole the cookies from the cookie jar. Only bakery employees have access to the back room. Security camera footage shows Alice was in the back of the bakery alone for most of the night. She had opportunity. Bob is a self-confessed chocoholic who’s been in and out of 12-step programs. He’s got a motive. Carol, meanwhile, is extremely tall. The tallest employee by a good foot. She’s the one who put the cookies on the high shelf to begin with. She certainly had the means to get them down.

Problem with this approach is that you can sometimes end up with a huge-a%& suspect list. Like unmanageably huge. Let’s say the wealthy host of a dinner party got killed. And let’s say that there were a hundred guests and twenty staff all coming and going from various rooms throughout the night. And there’s no security cameras. And the staff all hated their scheming, stingy a$&hole of a boss. And most of the guests hated him too. In that case, the investigators have to cull the suspect list down a bit before they can even start investigating. Like, if the victim was beaten to death with bare fists, maybe they focus on the ex-professional boxer, the tae kwon do expert, and the server with the violent temper and a bunch of prior assault convictions. If none of those suspects pan out, the investigators can widen their let later.

With a list of suspects who each had either the motive or the means or the opportunity to commit the crime, the investigators try to fill in the blanks for each suspect. Or eliminate them. Sure, Alice had the opportunity. But she’s got diabetes and takes excellent care of herself. She’s got no motive to steal cookies. Bob had the motive what with his chasing the chocolate dragon, and it turns out he’s a free-climber in his spare time with great upper body strength. And he has a key to the bakery’s back door. Means and opportunity. While tall Carol wasn’t actually working the night of the theft, a witness saw her leaving the alley behind the bakery that night — establishing her opportunity — and the manager admits she often took home pastries and cakes — establishing a sweet tooth for a motive.

And that’s all down to actual investigative legwork. It’s down to gathering information.

Leads and Clues and Other Essential Mystery Crap

Establishing someone as a suspect, building a case against them, or even eliminating them from consideration is all down to finding information. See, there’s a logic puzzle at the heart of every mystery. But all the information you need to solve the logic puzzle? That s$&%’s scattered around the world. Thus, a mystery is one part logic puzzle to one part scavenger hunt. And the object of the hunt? Well, those can be clues, evidence, leads, red herrings, false leads, and dead ends. And probably a bunch of other s$&% too. But that’s a good enough list to start with.

Clues and Evidence

Any piece of information that suggests or disproves that a suspect had the motive, means, or opportunity to commit a crime is a clue. It’s as simple as that. Clues can be big or small. Big, smoking gun type clues establish motive, means, and opportunity by themselves. Sometimes two or three of them at once. Small clues merely hint at motive, means, or opportunity. Sometimes only in combination with other clues.

The witness that said he saw Carol leaving the bakery in the middle of the night? That’s a big clue. By itself, it establishes Carol’s opportunity to commit the crime. She was where the crime happened around when it happened. But if all the witness saw was a really tall person or a woman or someone with auburn hair, that’s not enough by itself to point a finger specifically at Carol.

Likewise, if the ill-tempered server was overheard at the party b$&%ing about his boss and saying things like, “I ain’t taking this crap anymore” and “someone needs to take that guy down” or “the next time he talks to me like that, I’m going to punch him,” that’s a big clue. But the ex-boxer’s money troubles — his empty bank account and his pending eviction — don’t establish a motive. Not by themselves. But if the party’s host had once rigged a championship bout — the bout that ended the ex-boxer’s career — now there is a motive. Revenge.

Clues are just bits of information. They come in lots of forms. Including deductions and leaps of logic. Evidence is the term for specific things that demonstrate the information in question. The information that an auburn-haired woman was seen leaving the bakery at 11:03 PM? That’s a clue. The statement from the witness saying so? That’s the evidence. Depending on the setting and the goal of the actual mystery adventure, figuring out the answer might not be enough. The investigators might have to provide enough actual evidence to prove the answer in a court of law.

Leads

A lead is a piece of information that doesn’t really help solve the mystery. Instead, it provides a direction. Something or someone the investigators can check out to turn up a clue or find some evidence. Or even to find another suspect.

At the scene of the Cookie Crook Caper, the investigators find a Chocoholics Anonymous keychain lying forgotten on the ground. That doesn’t give them a clue, but it does send them to the local CA chapter to see if any of the members have a connection to the bakery. Sure enough, it turns out one of the bakery’s employees, Bob, was a member. Likewise, if the party discovers the victim secretly had a kid out of wedlock, they’re going to want to question him. Again, that’s the whole “did the victim have any enemies” question. Or the “who stood to benefit from this crime?”

If you think of an investigation like a dungeon crawl, the clues and the evidence are the magical items that let the heroes beat the final boss. The treasure. The leads are the hallways between the rooms. The visible exits. And the secret doors too. But that’s more about building a mystery adventure, not about the mystery itself.

Red Herrings, False Leads, and Dead Ends

Sometimes clues aren’t clues at all. I mean, just because someone had a motive, that doesn’t mean they’re actually guilty. It might seem important that the tae kwon do expert is secretly the victim’s kid by a secret mistress, but if it’s the violent server that actually did the victim in, that clue’s just a distraction. Sometimes, clues are lies. The witness that saw Carol leaving? Turns out, that was Bob’s friend from Chocoholics Anonymous. He and Bob are both struggling to stay on the program and they both needed a chocolate fix. Bob stole the cookies and split them with Dave and Dave, meanwhile, described Carol leaving the scene.

These Red Herrings, False Leads, Phony Clues, and Dead Ends all make mysteries harder to solve. They give the investigators the wrong answers. Or they distract them from the right answers. And they waste their resources. In the dungeon analogy, they’re the traps and cursed items and natural hazards. The more of them scattered around, the harder the adventure.

Building a Case

I made a crack above about the Cookie Crook Caper ending with Bob breaking under interrogation, running, and then getting himself killed. Remember that? That’s a great way for a movie or TV show to end. Or an action-adventure game. It leaves no doubt about who the guilty party is. Hell, even if the investigators weren’t sure when they started leaning on Bob that he did it, they’re sure now.

But real-life mysteries? And those in more grounded investigation-based games? That’s not usually how it happens. And frankly, that’s a pretty crappy ending. A good, real mystery — a challenging mystery adventure — doesn’t end with a eureka moment and a smoking gun and a confession and a shootout. The investigators might never know if they actually got the right person. How?

Well, the investigators ran around gathering whatever clues they could find. And making whatever deductions they could. The more clues they found, the better the deductions, right? But what if they missed some important clues? Or what if they made the wrong guess? Or what if the mystery has three good suspects and none of them can be eliminated and they all protest their innocence? Then it’s down to the investigators looking at the clues and evidence they do have and deciding which suspect they’ve got the strongest case against. And deciding whether that case is strong enough to actually lock the suspect up. Or execute him. Or whatever. Or else, the investigators have to have a strong enough case to prove to a disinterested party like a magistrate or a jury that the suspect probably did it.

This s$&%’s all down to how the adventure’s structured. There’s genre and setting considerations and theme and tone and all that crap. So it’s kind of beyond the scope of this article. But I’ve got to mention it because sometimes solving an investigation is just about piling up as many clues as possible and making an educated guess.

I’ve also got to mention it because there’s lots of GMing advice — and even entire f$&%ing game systems — that suggest the GM’s job is to make a mystery the players can’t fail to solve. One with a definitive answer that reveals itself if the PCs just stick it out long enough. Or one in which the players can’t possibly miss any of the important clues. Which basically means the GM’s just doling out a logic puzzle in pieces. Which is just another kind of railroad.

In short:

Why, yes, commenter, I have heard of the Gumshoe system. And no, I do not recommend it as a resource for GMs who actually want to learn how to structure a good, challenging mystery adventure. But thanks.

Putting The Pieces Together (and Tweaking the Difficulty Knob)

Now I’ve given you a nice survey of all the different bits and pieces of a mystery. But how do you assemble those pieces into a mystery puzzle? And how do you tweak the difficulty of the mystery puzzle to suit your players and your game? Well, let me end this article with a few examples of different kinds of mystery puzzles that show you how to cram this crap together.

The Simple Whodunnit

The simplest mystery structure there is. There’s a small list of suspects, but only one of them had the motive, means, and opportunity to commit the crime. The investigators come up with an initial suspect list and then hunt down the information needed to establish the perpetrator’s motive, means, and opportunity and to eliminate the other suspects.

Someone stole the cookies! Alice, Bob, and Carol are bakery employees. Alice is visible on the security camera footage prowling around the back room at the time of the theft. Bob’s a recovering chocoholic. Carol is tall enough to reach the cookies. Three suspects established, the investigation starts. Alice can’t eat chocolate due to an allergy. No motive there. Carol was out of town the night of the crime with friends who vouch for her. Alibi means no opportunity. But Bob’s a free-climbing enthusiast — he’s got the means to get the cookies from the high shelf — and the log shows he stopped in the night of the crime to pick up his paycheck — so he had the opportunity. Bob dunnit.

The Advanced Whodunnit

A more complex mystery involves multiple suspects, not all of whom can be eliminated. So, it falls to the investigators to decide who’s the most likely perpetrator by building a case against each. And if they can’t build a strong enough case to indict one of the suspects, they lose.

Someone stole the cookies. Alive, Bob, and Carol are suspects. But Alice has allergies. No motive. Bob and Carol, it turns out, both stopped by to pick up their paychecks the night of the crime. Bob’s a free climber and Carol is tall. Bob’s a struggling member of the local Chocoholic’s Anonymous chapter with numerous lapses. But his sponsor says Bob’s really trying to overcome his demons and he’s getting better every day. Still, he’s got a motive. Carol has frequent bouts of hypoglycemia. According to her doctor, when her blood sugar level is low, she can become irritable and impulsive. She’s got a motive too. And the party can’t break either Bob or Carol under interrogation.

The investigators go back over the crime scene. Because of where the cookies were located and the layout of the room, they conclude Bob would have had some serious climbing to do. He’d have had to move some stuff and then move it all back into place. Lots of work. And he’d have time, while doing that work, to catch himself and talk himself out of doing what he was doing. CA mantras and s$&%. Carol, meanwhile, could have just grabbed the cookies and gone. Since Bob and Carol are both struggling with impulses and compulsions, it was probably a quick and dirty crime of opportunity, not a premeditated thing. So they call for Carol’s arrest. She’s dragged away by the police, protesting her innocence all the way.

The Impossible Crime

Impossible crimes are crimes without suspects. Something about the crime eliminates all the possible suspects. Take the classic locked room crime. Something happens in a room that’s locked from the inside. No one has an opportunity to commit the crime because no one has access to the crime scene. Or there’s the beloved murder victim who had no enemies and whose death benefits no one. A crime without a motive. Or there was break-in in which nothing was stolen or damaged and no one was hurt.

Obviously, something’s not as it seems in an impossible crime scenario. The investigators have to figure out what it was. The rare antique was swapped for a fake, which is why the robbers didn’t appear to take anything. There’s a secret door that provides access to the locked room. Or the victim was killed with a slow-acting poison that took effect after they locked themselves in for the night.

Impossible crimes usually have just one good suspect. The twist — the thing that ain’t what it seems — either points to the suspect or else the suspect points to the twist. The one provides the lead for the other. When the party discovers the secret door, they investigate the contractor who did some repair work in the adjoining room or the architect who built the house. People who’d know about or stumble on the secret door. If the prime suspect works at a medical lab that suggests maybe a poison was planted ahead of time.

The Frame Job

The frame job’s a mystery with a prime suspect that had a strong motive, the means, and the opportunity to commit the crime. There’s just one twist. The prime suspect’s innocent. It might be just by luck that the prime suspect looks guilty. They don’t have an alibi, they have a motive, they were just in the wrong place at the wrong time. Meanwhile, the guilty party covered his tracks just enough to make him a less likely suspect than the prime suspect. Sometimes, the frame job’s intentional. The perpetrator not only concealed their own guilt but also fabricated phony clues to make someone else look guilty. Sometimes, the frame job is the real crime. That is, the perpetrator had no motive to commit the initial crime. Their intention was to, say, ruin the prime suspect’s life or get revenge by making them appear guilty of a crime.

Serial Crimes

Finally, serial crimes are a special kind of impossible crime. Someone’s committed multiple crimes. But there’s no obvious connection between the crimes. No apparent motive. And that’s because the motive is usually psychological. The perpetrator suffers from a mental illness that manifests as criminal behavior. They might be a compulsive thief driven to commit elaborate crimes to demonstrate their superior intellect. Or they might be a killer with a psychological need to avenge themselves on people of a certain description or social class or position or something.

Serial crimes are, for obvious reasons, extremely difficult to solve. Especially without modern psychological profiling tools to help narrow down the investigation. Otherwise, the trick is to find the pattern in all the crimes. All the victims had red hair, for example. Or all the robberies involved artifacts insured by the same underwriter. Often, in the case of serial murders, the spree is touched off by a single traumatic event. That is, the first in the series of serial murders involves a victim the perpetrator had a connection to. The subsequent crimes are against victims that have something in common with the first. Hence, investigators often focus on the crime that came first, chronologically. Of course, if that crime hasn’t yet been discovered, the investigators might focus on the wrong first crime.

Serial crimes are tremendously complex. Unless you’ve got a really skilled group of players that are really into crime drama, they’re not great fodder for RPG adventures.

But, of course, this isn’t about RPG adventures. This is just about mysteries. And I’m sure that’s good enough, right? You don’t want to hear any more about this crap, do you?