Building Mysteries

As we mentioned earlier, mysteries make up a much larger portion of adventures than many people realize. Any time players want to find out something they don’t already know, whether that’s solving a crime, uncovering the location of a lost artifact, finding blackmail material on an unhelpful member of the king’s council, or whatever, it’s a mystery.

The Three Clues Method

What most people do realize is that it is very easy to do mysteries very wrong. Just the idea of a mystery adventure conjures up for many an image of the party sitting around for hours trying to figure out what clue they’re missing to solve the puzzle the GM’s put in front of them. A well-designed mystery does not have this problem, and the secret is the rule of three. If you leave three clues in one location, players are probably going to find and correctly interpret at least one of them. If you leave three clues all pointing to the same conclusion, players are probably going to find and correctly interpret at least one of them. This doesn’t mean you should have a set of three clues in one location all pointing to the same, next location, though (unless you want your mystery to be very short, more of an encounter than a complete adventure, in which case it’s fine to have all three clues point to the same conclusion).

Instead, you want to have a starting scene, whether that’s a specific location or a conversation with a specific person or what-have-you. This scene has no clues leading to the ultimate conclusion, but has three clues, each one leading to a different one of three intermediary scenes, which we’ll call scene A, scene B, and scene C. These scenes can be visited in any order and it’s even possible to skip up to two of them, although due to the three clue rule players will typically see all of them (although just because it happens more often than not doesn’t mean that it’s going to happen a lot). The players probably won’t find every one of the three clues in the introductory scene, but they probably will find one of them, and that will give them a lead to follow. If they do happen to find two or all three, so much the better.

Each intermediary scene itself also contains three clues. Two of the clues point to the other intermediary scenes, and the third clue points to the conclusion, the solution to the mystery. So, scene A has clues leading to scene B, scene C, and the conclusion, scene B has clues leading to scene A, scene C, and the conclusion, and scene C has clues leading to scene A, scene B, and the conclusion. Since there’s three clues in each scene, we can be fairly certain that the players will find at least one of them, which will again allow them to continue on the mystery. It’s possible they’ll find the clue leading to the conclusion right away, but it’s more likely they’ll find one of the other two clues pointing to one of the other two intermediary scenes. For example, if they go to scene A, they will probably find one of its three clues. There’s a one in three chance they find the clue leading to the conclusion, but a two in three chance that they find a clue leading to either scene B or scene C.

We’ll assume that the players find the clue for scene B and they go there next, and once again have three clues to find. One of those clues will point back to scene A, which they just came from. So what happens if they find that clue? There’s a one in three chance they’ll find that clue, and if that happens they’ll be stuck, right?

Well, no, because now think of the total number of clues they’ve had for scene C. There was one clue for scene C in the introductory scene, another clue in the scene A, and now they have a third clue for scene C which they found in scene B. Since they’ve had three clues for scene C, odds are they’ve found at least one clue for scene C by now, to they have a lead to follow there. Once they reach scene C, they’ve had three clues for the conclusion, one each at scenes A, B, and C, which means odds are they’ve found and figured out at least one of those clues, which means they can now go to the conclusion of the adventure.

Three Clues in Action

Let’s have an example. Say the players are in an intrigue campaign. Baron von Racist is on the king’s council and is voting down any motion to ally with the goblins. His father was killed by goblins in the war thirty years ago and he won’t let the grudge go even after the players put a new and more friendly king on the throne of the goblin kingdom, so the players have resolved to get Baron von Racist’s cooperation through blackmail. It’s a theme of the campaign so far that basically every noble has some dirty secret that can be used against them, so the first scene, in this case, is just the players sitting down and figuring out how to investigate the baron and find out what his secret is. We anticipate three potential approaches:

First, visit the baron’s castle on some pretense (or just break in, if the party is sneaky) and physically search for evidence of his being up to something he wants kept from the rest of the court.

Second, meet with his serving staff and question them about the baron’s activities, whether using charm or intimidation as leverage.

Third, get a hold of his financial records and see if there’s any unusual purchases. Most stewards keep good records, and if the expenditures don’t add up or some of them have suspicious labels, that might lead the party to any indiscretions the baron happens to have.

There’s a fourth approach I anticipate from inexperienced investigators, which is to approach the baron directly and hope he somehow lets slip his indiscretion. There’s no cost to roleplaying this scene and in fact it’s a great time to characterize the baron a bit if you haven’t had a scene to establish his character already, but without even knowing which direction to steer the conversation in, players are going to be flying blind. What you can do with this scene instead is to have him drop some clues to other avenues of investigation. Maybe he’ll assert that he doesn’t want anyone snooping around his estate unsupervised (he’ll claim that he’s had a number of attempted thefts recently), that his serving staff keep spreading baseless rumors about everything under the sun, and mention monetary troubles due to…well, due to some expenses. The details are too dreadfully dull to get into, you understand. Speaking with the baron, then, serves as a backup introductory scene if the players can’t come up with more effective approaches to investigate on their own. I expect the approach of snooping around his castle to come naturally to most players, but players frequently defy expectations and not always in a good way.

While snooping around the baron’s estate, the players may overhear serving staff gossiping about the baron’s strange behavior at night, and exchanging theories about what he might be up to. They may also stumble across his steward’s ledgers directly. Finally, they may uncover the hidden passageway that leads to his necromancy laboratory. This is scene A, with its three clues. Odds are, players will find at least one of these.

If the players question the staff, the gossipy maids will pretty freely divulge that he seems to appear and disappear from nowhere, like he can walk through walls sometimes, and is always wroth when a servant catches him skulking in the shadows. He seems to especially skulk in a couple of specific places (where his secret doors are located). They’ll also talk about how they overheard the baron and his steward arguing over the cost of certain “special expenditures.” The baron’s butler is aware of the necromantic laboratory, as it’s left to him to scrub it down, but it takes a particularly high Persuasion or Intimidate check to get him to divulge this secret.

If the players examine the steward’s ledgers, it includes accounts for digging several tunnels and installing several doors a few years ago, even though there’s no other construction listed anywhere in the nearby years and the rooms where the doors were allegedly built don’t actually have doors in them besides their main entrance which, you’d expect, would have been constructed along with the room itself. It also mentions paying a hefty bribe to the butler for “his cooperation in certain matters best left to the imagination.” Finally, it lists the purchase of “discreet materials” from a merchant which, with the right connections (or perhaps if they happen to have stumbled across him before) the characters know to be a black marketeer for necromantic materials.

One way or another, the players learn that the baron is practicing necromancy, which the Church would take issue with. Players might have several options here, blackmailing the baron with this information as originally planned, or offering to bring information on a heretic to the Archbishop – another member of the king’s council – in exchange for a favor, that favor being voting with the players ally with the goblins (bonus points: Baron von Racist is probably getting replaced on the council after the Archbishop sics the inquisition on him, and the new councillor might vote to ally with the goblins or abstain).

So there it is: A complete mystery adventure. What if it doesn’t go according to plan, though? What if players miss all three clues? When this happens, you need to create a new scene with three clues pointing to your scenes A, B, and C for them to discover, preferably with some kind of cost for leaving the mystery unsolved. If it’s a standard murder mystery, what happens is the killer strikes again, leaving a new set of clues, but also having killed someone. Perhaps a contact or ally of the players’ is slain, or if it is specifically their job to catch the killer, the victim may not have been personally important but now their boss is agitated with them for their failure, but there is a new crime scene with new clues to follow.

Playing With the Form

Once you’re comfortable with the three clues method, you can make more complex scene chains. You can add a fourth clue to a scene that leads to a red herring dead end scene, which itself has no clues leading anywhere else. Since this is a fourth clue, its presence won’t disrupt the rest of the scene chain. You can stack multiple layers of intermediary scenes, so for example, scene A1 has clues for scene B1, C1, and A2. Scene A2, then, has clues leading to B2, C2, and the conclusion. Scene B1 has clues for A1, C1, and B2, and scene B2 has clues for A2, C2, and the conclusion, and so on.

Just like dungeon crawls have nested dungeons, you can build nested mysteries. Each five-scene mystery (or eight scene, if you’re going with the double decker plan, with maybe one extra scene added as a red herring, or whatever) can have three extra clues added to it, each one pointing to the introductory scene of one of three completely different mysteries (the conclusion of a mystery can be a good place to put these, and you can put three clues each pointing to a different mystery all in the conclusion scene of your first mystery to keep things self-contained for the first adventure before unfolding).

Much like dungeon crawls, there’s a lot to do with the mystery adventure format once you’re familiar with it.