Lessons from Collaboration 2: Mysteries

In case you missed the last post, this is a sequence of excerpts from conversations I had with a friend who wanted help and inspiration to create their first adventure. I’ve sampled the pieces that are the most useful, and added a few post-thoughts of my own.

Talking about puzzles last time got me talking about mysteries, because the bounds for what a puzzle really is, or really should be get pretty elastic when it comes to tabletop roleplaying games with strategic combat elements. But, I am going to start this lessons from collaboration with a warning not to make the same mistakes that I have. This will probably come up a lot more over the course of these entries.

“I am curious to know if you are planning to make this a mystery. If so, I have some mystery advice if you’re open to it, because I have never thought of myself as much of a mystery writer, but found almost every adventure that I really loved turning into a mystery plot. I sometimes write an adventure, don’t intend it to be a mystery, and it becomes a mystery anyway, where I am not really consciously applying my mystery writing techniques I have learned.”

While I definitely have had people tell me things like “oh, this is such a good mystery,” or “you’re a great storyteller,” in situations like these, I don’t really believe it. This is because I can see the difference between writing a mystery unintentionally and writing one intentionally. As good as some people might say that the “impromptu mystery” was, in all these cases, I have noticed that players end the session asking about missing, key information, or completely off base as to what the mystery was “about.” This isn’t just about the player’s experience, though. Mysteries thrive off of structure, so trying to apply one ad hoc rarely works to the extent one would like it to. These kinds of half-baked mysteries are exceedingly complex, and it’s a level of complexity that doesn’t add any apparent complexity to the actual narrative. In fact, it can even detract from it. By contrast, simplifying the mystery as much as possible, at least, from behind the screen, makes a better, more compelling mystery easier run like a breeze. Before I describe more how to do that, I had some points to make about expectations for a mystery game.

“I’m also curious what the PCs are going to know going in and in creating their characters. For example, in Dimension 20 Mice and Murder (a murder mystery plot), the only PCs that aren’t some variation of bard or rogue is a Paladin, who is the Watson bodyguard to another Rogue Inquisitive player’s Holmes. This makes it less likely you have someone playing a 6 INT Barbarian like I mentioned earlier (unless you’re open to that). But, a sufficiently informed player could also play a Barbarian with the goal of going undercover in a fish out of water experience where they learn to solve problems in their own way.”

In this adventure, the PCs were going to be impersonating a group of missing students to infiltrate an event. While I had brainstormed a lot of ways that this could be implemented, or how the villain would be involved, the PC’s relationship to the law and any official investigations, the meat of this set of ideas was focused on figuring out what information the players ought to know beforehand.

A lot of the time, mystery just looks like a certain structure and set of objectives, because it can be concurrent with other genres. For example, mystery and horror work really well together; even though I’m a horror fan, that’s also how all my stories end up being mysteries. In order for a horror movie to work, the characters’ curiosity has to overpower their sense of risk at going into a haunted house. Mystery is a frequent way to accomplish this, particularly when horror PCs operate with a much stronger sense of self-preservation than slasher movie teens. This can be frustrating because the players behind the PCs, unlike a horror movie character, know they’re in a horror movie. They can hear the audience shout “don’t go in there, you idiot!”

When you’re creating a mystery, it can feel counterintuitive to try to spoil as much as possible about the premise or what’s going to happen to the characters, but it’s this player buy-in, players that come in with the mystery in mind, that makes it easy and fun for them to have their PCs go into scary situations even when their genre sensibilities tell them not to. While player buy-in is a necessary part of any tone and plot concept, it’s particularly important and illustrative in this crossover of horror and mystery. This might seem obvious because it’s just applying logic that goes hand in hand with other genres, but with mystery this took me a long time to figure out, because I wanted to, well, preserve the mystery!

I mention this because it’s an important part of the steps that you take to simplify the mystery experience. If this seems strange, this is because every other piece of media does not need to do this, because the characters and the audience are not the same. In a Holmes story, Watson is just, or more confused than the audience is, up until the point where Holmes makes his reveal, and shows everything the audience and the other characters have missed. By the end of the story, the audience will know every piece of important information, and they will know what they story is “about,” even if they didn’t enter the mystery knowing. Sherlock Holmes doesn’t need to know about the mystery ahead of time to be interested in it, because he’s the greatest detective of all time. It would be unbelievable if he didn’t enjoy it. To put it simply, the PCs are not Sherlock Holmes.

This is where we arrive at the reasoning behind simplifying the mystery, beautifully illustrated by the Alexandrian in this post. They introduce a few different ideas derived from this principle, most notably the Three Clue Rule. I’ll try to summarize them here, but you should obviously read the original post as well.

If you translated Sherlock Holmes into a DnD PC, he’d have to make a truly incredible number of passed skill checks to get the evidence he uses to make his final set of deductions. This is even assuming that the hypothetical player controlling Holmes knows what things to examine in the room. Returning to structure, the structure of an adventure in which Holmes is a PC is based on him making all those ability checks successfully, and choosing which evidence to seek out or examine based on the information he gets from those checks. Every decision, every die roll, is a point of potential failure. And again, the PCs are not Sherlock Holmes.

In terms of simplifying the mystery structure, this gets us to seven different ideas. First, don’t lay a bread crumb trail of clues. Second, don’t lock any piece of information completely behind die rolls. Third, create multiple fail states rather than one, complete state of failure. Fourth, reward players for the decisions that they do make. Fifth, make mini-mysteries/puzzles. Sixth, keep things connected. Finally, seventh, state your revelations.

That’s a lot! Fortunately, the Three Clue Rule makes all of these easy to do without even trying. The Three Clue Rule is well-named because it stipulates that, for any piece of key information, you create at least 3 different methods of the PCs learning that information. If you want some helpful examples about how you might accomplish this, there are a few in the original post. This rule, when implemented, creates a pool of evidence that guides the PCs towards the information you want them to discover by the end of the story. It effectively eliminates these points of failure.

How does this apply to our ideas around simplifying the mystery experience? First of all, it’s definitely a far cry from the Holmesian bread crumb trail. The Three Clue Rule creates redundancy to prevent each of those bread crumbs from being a point of critical failure, or a sticking point where the PCs will stand around trying to figure out what to do next, or, equally likely, what they are doing there in the first place. If one of them doesn’t land, you can deploy another clue in order to keep the plot moving.

Many people’s solutions in puzzles or mysteries is to say “don’t lock information behind die rolls.” This is true in the sense that it should not be taken absolutely. While die rolls should not be the only way of getting a key piece of information, they can actually be extremely useful for giving you more clues to pad out your Three Clue Rule. This touches on an idea in my previous post where I described the flexible bounds of puzzles. For any problem to be solved, the PCs should have clearly stated, actionable objectives Dice rolls give PCs things to do, and they are an important part of the actual execution angle of problem-solving in a game. The benefit of the Three Clue Rule is that you can use dice rolls when other methods fail, and use other methods when PCs fail dice rolls. Rolling slows down the second to second play but tells the players that their PCs are definitively doing something. Varying gameplay is a key part of pacing, which is the backbone of effective narrative structure in games.

“Multiple Fail States” is another technique that can be used in tandem with the Three Clue Rule to eliminate feelings of failure. This was an idea that I got from Matt Colville that changed the way I play the game. This was my analysis of it in my conversations with my friend I was working on the adventure with:

“The origin of this comes from the idea that it’s more dramatically satisfying if a failed check on getting across a pit doesn’t kill the PC instantly, but they do slip, and oh no, they’re barely hanging on by a loose vine. In mysteries, it’s like this: if the PCs make a check, but they don’t meet the DC, they should still get new information.”

Basically, this sits in line with the principle of the Three Clue Rule, that continued action to attempt to solve the mystery should result in your guidance of the PCs towards the information that you want them to know. Best of all, for the PCs, this looks like complication and complexity, but for you it’s just making it easier to have a smooth and effective mystery experience.

I mentioned in my previous post on this that, much of the time, the best solutions to a puzzle are the ones the players give you. The same works for clues in a mystery. Three clues is only the groundwork from which to draw from, and the actions of PCs or hypotheses of the players can either directly give, or at the very least, help you come up with clues that will become usable in real time. This is rewarding players for decisions, even if it’s not the ones that you planned on them making. Even better than eliminating potential failures through redundancy, sometimes you get success states handed to you that you know the players will pick up as important pieces of evidence, because they’re the ones that guessed they were important in the first place. This also involves the previously stated idea which is, “any time the PCs do something to advance the story, give them new information.”

I’ve been composing the idea that any mystery is composed of a set of key pieces of information. Each of those pieces of key information, what Ben Robbins calls “Revelations,” but each of those revelations is composed of at least 3 clues, potentially ad infinitum. While I wouldn’t call it simple, it’s elegant that the mystery is like a nesting doll of information. One way to keep things even more simple and elegant is to package each clue or each revelation, whatever the case may be, with the same sort of choices that inform normal adventure design. Part of figuring out ways that PCs can get these clues is that, eventually, they will figure out what to do. While “solve the mystery of the _____” is an OK objective, each of the key pieces of information or clues can have much more clearly stated objectives that the PCs choose. “Defeat _____,” “find _____;” these kinds of active, clear objectives help guide the PCs towards solving the larger mystery as a consequence of their continued play. Again, this aids the goal of the Three Clue Rule, which is to eliminate points of total failure and let the PCs move towards success as an inevitable consequence.

If you haven’t heard anyone mention the quantum ogre problem, the Three Clue Rule provides an elegant and natural solution to it. In the original problem, the PCs come to a fork in the road. Down one path is an ogre, and down the other, no enemies. But, because the GM made the ogre, and an ogre encounter would be more interesting, the GM makes it so that, regardless of which path the PCs pick, that was the path the ogre was on. There’s a lot of issues with this problem in the first place, but this problem becomes much different in a mystery. Assume that the ogre has one of the clues the PCs need to solve the mystery. Now, if the PCs didn’t have any other ways of getting the clue, they need to run into the ogre for the story to work in a much more critical way. Looking at the original post, we find that the Three Clue Rule actually gives players a greater sense of agency within the narrative, and a greater sense of an interconnected world. Your clues are easier to deploy the more interconnected your world is. Imagine that instead of the ogre on the other path, now there’s the ogre’s old campsite, or the ogre’s wife, or someone else also looking for the ogre. Now you don’t need to even see the ogre at all to get the clue; the clue is the piece that’s in a state of superposition. In other words, if it comes to it, you can, as the original post put it, “beat them over the head with it.” But in all the tangle of information, clues, and objectives that are simple behind the screen, there’s no obvious binary that shows you’ve thrown the ogre at them. The PCs still have unlimited agency as far as how to solve the mystery, you’ve just guided them to it so naturally that even players who feel like they’re not “smart enough for that kind of thing” can get to feel like geniuses. Best of all, again, the Three Clue Rule can generate so much inertia that it feels like the mystery writes itself.

Speaking of writing itself, I’ve spent a lot of time parsing my takeaways from the Three Clue Rule, even though, according to me, this rule makes it so all of these ideas can happen almost completely naturally. After all, this is supposed to make things easier, not more complex. The Alexandrian brings up the idea of Revelations as a prep technique. In order to prove that I “literally only needed Revelations and 3 clues to go with each of them,” I used this setup to re-do the mystery for a plotline I was running in one of my ttrpg system tests. All I can say is that the second time around, it was completely different, with an incredible ease of pacing and sense of flow, and a short, simple document to check if I ever got stuck. It was an incredible experience, and went far better than I ever expected. Since I should have gone to sleep several hours ago, and this is way longer than I ever expected, I’ll do a full breakdown of how this worked at a later date. At the very least though, you can see just how simple the idea can get. As you’ll see below, I didn’t even come up with 3 clues for some of the less meaty revelations. Some of them ended up getting way more effective clues through the course of play. This was literally my master writing document for the whole thing:

REVELATIONS

1. The kidnapped people are being brought back housing alien beings.
   1. The victims are not behaving normally. 
   2. The victims have a psychic sense of there being 2 entities. 
2. The alien beings are living inside the kidnapped people because they are susceptible to light.
   1. The victims live with lights off and insist on them being turned off. 
   2. The kidnappings only occur at night. 
3. The kidnappings are occurring as a consequence of an unstable dimensional crossing, where the perpetrators cannot control their crossing and are kidnapping to survive.
   1. One of the victims was a professor, and the victims all live near the lab. 
   2. The victims that have been returned have only been seen or otherwise left their homes in order to visit each other the nights of the disappearances. 
   3. Notes seem to indicate that any existence of an interdimensional portal would be uncontrollable and potentially snatch up any creature
   4. The actual reason for why these disappearances were noticed by the bureau were due to unapproved dimensional research. 
4. The perpetrators are in league with a “talented” terrorist group that is undercover with the police.
   1. Information on the case is being kept by the police, and refuse to investigate genuine ideas such as a serial killer. 
   2. The police seem as though they may have paranormal abilities. 
   3. Members of the police seem to know but do not state that the PCs have paranormal abilities, and are more heavily armed and nervous than one might expect. 
5. The beings are attempting to use the island’s unique geographical properties and the resources of the terrorists in order to return to their own dimension.
   1. There are old houses/ruins on the island that indicate that antiquarians had commented on its geography. 
   2. The Ashford family, for which the island is named, is making deals with the terrorist group and for scientific study on the interdimensional capacity. 
   3. The terrorist group has a house in which they research the beings and their interdimensional moon base. 
6. The terrorists are working with the perpetrators because they have been party to important knowledge and technology from this hyper-advanced interdimensional species. They continue to work together because they each are balancing success and improved capability against each other.
   1. The terrorists have records that indicate that they are sabotaging the beings. 
   2. The terrorists have paranormal weapons research that has rocketed ahead since the disappearances. 
   3. The terrorists plan crossings, trying and failing to predict the ability to cross over. 

That’s (finally) it for this Lessons from Collaboration entry on all things mysteries. I started this series because I thought I had already done a lot of writing for it, but I clearly wanted to re-examine or reframe my previous ideas. Like I said, I’m going to continue this series, but as to what exactly I am going to cover next is a little more dubious. Hopefully I continue to play ttrpg games and get more ideas! As Matt Colville said, “that’s all there is to it! Make art!”