Usually I shortlist three or four images and then select between them, but this castle was so atmospheric that I couldn’t refuse it. Image by Enrique Meseguer (aka darksouls1) from Pixabay. This small-sized version just scratches the surface, follow the image link to salivate over larger versions.

I love it when I discover something that GMs should be using and aren’t, because it means that I get to analyze the reasons, and devise solutions to whatever is holding people back – and I learn things and develop tools for my own use in the process. By my count that’s win-win-win-win – and in this case, we can tack on an extra ‘-win’ because the players get to experience something that has been rare up until this point.

The latest such discovery is the ‘locked room mystery’ which is WAY underused in general, but especially in Fantasy – D&D and Pathfinder GMs, I’m looking at you here!

Why Are They Underutilized?

I can see five possible reasons. Not all of them will apply in any specific case, Campaign, or GM, but most of them will have some influence over the mindset of the hesitant Gamesmaster.

Those five are (1) Perceived Difficulty, (2) Perception of Antiquity, (3) Fear of Failure, (4) Fear of Boredom, and (5) Fear of Accusation.

If I’m going to offer solutions to each of these, I’m going to have to at least glance at the problems in detail. So, let’s do that.

Problem #1: Perceived Difficulty

Mysteries are often perceived as being harder to create / run than other types of adventure, and some people are of the opinion that the locked room is more difficult than any kind of mystery.

Both are less applicable when it comes to modern day low-tech campaigns, but there aren’t all that many of them rolling around. Instead, one or more of the following apply to most campaigns: (a) High-tech (b) Magic (c) Far Future (d) Recent Past with Weirdness (e) Remote Past, Realistic (f) Remote past, fanciful (g) Remote Past, fanciful low-level (h) Time-travel (i) Interdimensional Travel (j) I-don’t-know-when-its-set-but-it’s-not-now. Any one of these ten can trigger a perceived level of difficulty that may seem insuperable.

The common element is that any of these can introduce capabilities that are beyond the modern-day common standard, and hence either introduce esoteric solutions that have to be meticulously prepped against by the GM, or that make it seem too easy for the GM to ‘cheat’ by employing one of them.

There’s also a perception that the GM has to have analyzed every possible solution to the locked room so that he can rule them out to create the mystery, and that sounds like a lot of work for little benefit.

Problem #2: Perception of Antiquity

I know some people who dismiss the entire concept of a locked room mystery as being ‘old hat’ or ‘Victorian’. There’s a concern that the players won’t become sufficiently invested to make the kind of effort that is needed to solve one, or that there won’t be enough action to keep some of the more unruly player elements satisfied.

In other words, GMs are concerned that the locked-room mystery won’t fit their campaign. And, to some extent, that’s a valid concern – if you were to run a modern-style locked room mystery instead of one modeled to fit the context of the non-modern-day campaign.

I’m not so sure that they are relevant to a correctly-contexted version of the concept – a fantasy-oriented one for a D&D campaign, or a hi-tech one for a sci-fi campaign. But that doesn’t matter; the problem isn’t the reality, it’s the perception of reality by a reluctant GM.

Problem #3: Fear of Failure

If there’s a perception of Locked Rooms being hard, then there is an implied corollary that they it’s easy to make a mistake. Again, there’s a kernel of truth in that assessment – I myself have been caught by having demonstrably ruled out the correct solution to the mystery, leaving no viable solution. At least, none that I could think of.

N heads are better than one

If that specific problem ever happens to you – and it can happen outside of locked-room-mystery context – the best solution is to shut up and listen to the players speculate until they find an improbable but acceptable solution. Steal it, file off a few serial numbers, and present it (or some simple variation) as the correct answer – and compliment them on being so perceptive.

Of course, if this happens regularly, then it seems certain that there will be some pattern to the problem, some element of the GMing style that needs adjustment. The most common form of this problem would be a propensity to run off-the-cuff adventures and want to deny solutions that are ‘too easy’ – so if the PCs guess correctly early on, they have to be wrong, and the solution has to be something else.

As soon as you get clever in rejecting the correct answer in this way, you are headed for trouble; it’s only a question of when, not if. Use this technique as a last-resort escape hatch if you paint yourself into a corner.

Problem #4: Fear of Boredom

A lot of GMs are also concerned that the game will bog down due to the nature of the mystery, and that players will resort to die rolling when they should be roleplaying and collaborating amongst themselves.

This, of course, is a perennial problem in TTRPGs; why should a locked-room mystery be any exception? But the contention is that such mysteries actively encourage such behavior because the playercharacter; they have different personalities, skills, abilities, drives, and mindsets.

Well, the people who think so are both right and wrong. They are correct insofar as a lack of engagement can be even more lethal than usual in locked-room mysteries (and mysteries in general, for that matter); they are wrong in thinking that either the general form or the locked room are especially prone to this, and/or that the problem won’t be corrected with the same techniques as in any other situation.

Problem #5: Fear of Accusation

Finally, there’s that potential accusation of cheating. Every GM faces this, sooner or later, and it doesn’t have to be in any particular type of adventure, though it can be argued that mysteries in general are more prone to such accusations, and locked-room mysteries more-so than any other. I don’t agree with the premise, but can’t actually refute it.

That makes it the problem with the most potential merit of them all. Fortunately, it’s fairly easily solved.

Solutions

With the five problems explained and their validity either confirmed to at least some extent – even if they are only in the GM’s head, that doesn’t make their presence there any less real – we can turn our attention to solutions. These don’t map one-to-one to the problems – for one thing, there are seven of them!

In some cases, they are only partial solutions; in others, they are applicable to more than one problem.

Approach

The approach you take to designing your locked room mystery is vital. The place to start is always to decide how someone has committed the deed and made it seem impossible for it to be done that way.

That creates the mystery, and it tells you that every other possible approach is not the right answer. All that remains is finding ways for the PCs to rule those wrong answers out of contention, and voila! – the locked room mystery is complete!

Well, aside from descriptions of locations, a sprinkling of distinctive personalities, a red herring or two, and maybe some additional misdirection so that motive or opportunity can’t short-cut the story – the usual prep, in other words.

This not only makes Locked-room mystery plots much easier to create, it makes them much easier to run, too – if you already know how the crime was actually committed and then covered up, any answer the players come up with that you hadn’t thought of is automatically wrong, and you can ad-hoc a reason for that being the case and a plot pathway that leads the investigators to that reason.

Appreciation

That “usual prep” is the key to solving problem #2 – you need to make the trappings wrapped around your mystery interesting enough to sustain player focus while they are engaged in the procedural process of solving the mystery.

A rule of thumb that I like to use is that every suspect the PCs encounter should have their lives changed in some way by either the crime, or the PCs presence, or simply by being there. It doesn’t have to be a big change, and it certainly doesn’t have to be for the better.

What’s more, they often shouldn’t assess those changes completely correctly.

In a traditional locked room mystery,, the suspects (except the villain) are all completely truthful, though any of them may be evasive and attempt to cover their tails with excuses, denials, or simply being mum and playing dumb. They are worried about making themselves look guilty, especially if they had a reasonable motive, and never seem to realize that such behavior actually does make themselves look guilty more than coming clean would have done!

In a more modern variant, this is all taken a step further; most of them will have some reason to lie to the investigators; eliminating them from contention is about finding some reason for the lies and deceptions and lack of cooperation other than being guilty of committing the crime.

Even this can be mixed up by the creative GM, as I have pointed out in the past – the absence of an alibi can be prima facae evidence of innocence (see Taking everyman skills to the next level: The Absence of an Alibi – Dec 2011).

One of the best options, from a GMing / Gameplay perspective, is the suspect who was actually busy committing some other crime somewhere else at the time, especially if their potential penalty is comparable to the penalty for the crime being investigated. This guarantees that they will be uncooperative for “entirely innocent” reasons!

Process

Can you trust the players to come up with their own process for solving the crime? How can you not?

It’s a fact that the most popular television shows of the last 60-70 years have been police procedurals. On top of that, there have been movies, and some players may have read Sherlock Holmes or Agatha Christie. Throw in all the legal-based TV shows like Perry Mason LA Law, The Rockford Files, the NCIS franchises, and so on, and it’s extremely unlikely that any of them will not have the foundations of a process.

The hard part is always keeping those procedures interesting. Momentum, sustained by steady progress, will work for a while. When you detect signs that this is no longer enough, it’s time to confound a basic assumption that most of the players will take for granted without even realizing it – their expectations of a static and stable situation.

I think of these as the “wandering monsters” of the mystery plotline. Someone shows up to complicate the situation, or introduce a new and unrelated problem, or there is an unexpected development in the ‘private life’ of a PC. These always have to be either smaller in scope than the main mystery, or more long-term.

Just as you would do, I prepare a list of these in advance, with at least one ‘targeting’ each PC individually. When the time comes to pull one out of my back pocket, I don’t choose randomly; I pick one that will make life more ”
interesting” for whoever is showing the greatest degree of boredom, even if it’s indirectly (one of my favorite techniques is to inflict a problem in which a different PC is the solution, at least partially.

Yes, this is metagaming, but it’s metagaming of what I consider an acceptable variety, because the purpose is to make the game more entertaining and not to detract from player agency..

Passing A Roadblock

The Phone Rings. (Wait for someone to answer it).

“Hi, this is Monica Desalt down at City Hall, (Wait for reply)

“I’ve been cleaning up in the archives for an inspection next week and came across a box full of records people had discovered over the last 5 or I don’t know maybe 6 years that had been mislabeled or misfiled. Someone decided that putting them wherever they were meant to be was too much trouble. Sorry, I seem to be rambling on a bit. (pause for reply)

“Anyway, when I started doing what whoever it was didn’t do, I happened to notice that you people down at Tanzerbaum had attempted to access one of them, and the system, not finding the right record, had defaulted to an older copy. I’m sure that the work you do, whatever it is, is super important to you, so I thought I’d check with you to see if the updated report might be useful. Do you want me to shoot you a copy, or will you wait 24 hours for it to reappear in the system?” (Pause for reply).

For every important clue or piece of information, I come up with two or three alternate ways to get the information to the PCs, preferably via some form of roleplay, sometimes involving die rolls.

These are “Wandering Monsters” that just happen to be helpful to the party. Some will be reluctant, some will try to avoid giving the information in question and have to be persuaded or forced, others will mention the important bit in a side-comment, one that will have to be correctly interpreted. The example above is a relatively straightforward one, any skill checks required would be fairly easy ones; some of them will be that direct, others will be trickier.

Once, in one of my games, it was a wandering extra-dimensional sphinx, from a place where truth was used to power up transportation jellies. This sphinx, or more properly, this Cryosphinx Astronaut, offered to sell the PCs a straight answer or two for every trio of riddles they successfully solved.

Another time, it was a seemingly normal human, but who happened to be a very baroque dresser. He observed that they were stuck at exactly the point he thought would encounter difficulties, knew the solution to this part of the puzzle, and would trade it if they pulled off a small bank heist for him. Questioning revealed that he had seen the construction of the crime scene in person 200 years earlier because he was a Vampire that happened to live in the area. All he wanted was about 10 pints of Ab negative from the blood bank; he’d get them himself but it would raise questions that he didn’t want the locals asking. Pitchforks and wooden stakes and the like usually followed, a great inconvenience to one of his kind. So they had to obtain the necessary without telling anyone why they needed it… This replaced an intellectual puzzle with one that could be solved using non-intellectual methods.

As you can see, the strange and exotic are fine so long as they match the reality profile of the campaign overall.

“Wandering Monsters” of this kind happen by whenever the PCs start getting so tangled up that they request a skill roll or a save of some sort for their characters to do what the players couldn’t. (When the players already have the information they need, but are failing to add 2 and 2 together to get 4, I’ll sometimes permit one of those, too).

Key points to remember when designing these encounters: How does the WM have the answer? How does the WM know that the PCs need the answer? Is the encounter interesting enough? Does the encounter solve the immediate problem on a silver platter, or does it simply offer a leading hint? Is the encounter too easy or too difficult? Is the quid-pro-quo one that the players should be willing to pay, and if not, how can they solve this new problem? And, above all, does the encounter prevent the direct solution of the problem by roll-playing instead of role-playing?

The Asimov Prescription

For a long time, many writers and editors felt that mystery plotlines were impossible in science fiction because it was too easy for the writer to “cheat” the audience by introducing some new technology that could solve the mystery instantly. Asimov thought this was bollocks, and said so, and to prove the point, wrote many mystery novels and short stories set in a science-fiction setting. Others gradually came to the party, such as Niven, but they were basically following the rules that Asimov had laid down.

Mysteries in Fantasy suffer from the same problems, but squared and cubed, because you already have a “technology” (magic) that breaks the ‘normality’ rules and doesn’t even have to make as much sense as a science fiction plot device. But Asimov’s solution to the science fiction problem works equally well for this genre, too, as many authors have shown over the years.

I discussed Asimov’s basic principles in The Butler Did It: Mystery Plotlines in RPGs (July 2012), and his technique in Leaving Things Out: Negative Space in RPGs (Dec 2013), because they work equally well when applied to RPGs as they do in literature.

I specifically looked at the Fantasy RPG Mystery in Boundaries Of The Fantastic (Jan 2021). Readers may also find Ghosts Of Blogs Past: An Air Of Mystery – Using an RPG to write mystery fiction (July 2013) to be worth their time.

Asimov’s prescription was very simple: Don’t cheat the reader. Not once, not ever.

There is a presumption of innocence baked into a reader’s assumptions – that the author is playing fair. That’s why it is so shocking to readers and so heavily condemned when an author violates this principle, even if they’ve never been told about it.

In sci-fi, that means making sure than any new tech is discussed and demonstrated early in the story, and in particular laws and limitations, and then sticking meticulously within those restrictions for the rest of the story.

In fantasy RPGs, it means establishing some reason why the PCs can discount any magical ability that they don’t have and won’t acquire within a level or two, at most. (The best, most common, one is simply that the victim wasn’t important enough to justify using such means; the very existence of some mystery about the crime implies a fear of discovery that only makes sense if the perpetrator is vulnerable to the accusation).

And, if they do have access to something like Teleport or Wish, ensuring that they also have access to some means of detecting the use of such after the fact, and some expert who can provide answers on the limits of those abilities.

This can pay further dividends into the future when the PCs do get their hands on these abilities, by establishing that they have to obey those same restrictions – in effect, introducing house rules to constrain the plot-wrecking easy-answer potential of those powerful magics.

If the GM has established a pattern of never cheating the players, they will be inclined to assume that he’s not doing it in the mystery plotline. It’s that simple, really. Don’t Cheat The Reader (players).

A bible for traditional Locked-Rooms

I usually talk about the inspiration for an article at the start, but in this case, it makes more sense to discuss it here.

There is a book, “Golden Age Locked Room Mysteries” edited by Otto Penzler (Link is to Amazon, I may earn a small commission if you buy a copy).

I obtained my copy through my local library because it contains a Perry Mason short story that I haven’t read, but that’s neither here nor there. The introduction to the collection is the direct source of inspiration for the article. In the course of that introduction, Penzler discusses the 1935 novel The Three Coffins by John Dickson Carr (aka Carter Dickson), published in England as “The Hollow Man” (Amazon link as above, but this new edition won’t be published until Oct 1, 2024. If you want it sooner, try this link instead, which leads to an omnibus volume [in short supply] which includes The Three Coffins). He writes,

“In perhaps the most arrogant display of his command of the locked room mystery, he has his series detective, Dr. Gideon Fell deliver a lecture to a captive audience …

“In this display of erudition, Fell spends fifteen pages enumerating all the ways in which a locked room does not turn out to be impenetrable after all, and in which the impossible is clearly explained. He offers scores of ideas for solutions to the most challenging puzzles in the mystery genre, tossing off in rapid succession a greater cornucopia of invention than most mystery writers will conceive in a lifetime.

“When he has concluded his seemingly comprehensive tutorial, he informs the attendees that none of these explanations are pertinent to the present case and heads off to conclude the investigation.”

My immediate thought was, “what a great reference tool for GMs”. By the time you throw in the other examples offered by the authors and TV shows mentioned earlier, there are plenty of sources of inspiration to draw from.

This wrestles the notion that these plots are hard to write to the ground and pummels it mercilessly into the pavement, directly attacking problem #1 and contributing to all the others as well.

Low-level vs High Level

I already hinted at this in a couple of the preceding sections, but thought I should make a point of highlighting it.

The greater the capabilities that the PCs have access to, the more possible solutions to the Locked Room there are if the GM is playing fair, solutions that will have to be ruled out one by one in the course of the plot. Locked Room mysteries are at their easiest (from the GM’s point of view) with low-level characters – but that only makes the occasional one in a high-level campaign all the more rewarding – when they work.

The number one rule is always to “make the plotline interesting” – no matter what the type of adventure is.

Types of Locked Room Mystery

So let’s look at some ways to do so.

The Traditional (Impossible Murder)

Because locked-room mysteries are rare in RPGs, they are automatically going to be more interesting. Throw in some cleverness (original or borrowed) and you don’t need that much more, really.

But it’s worth noting, as Penzler does in the introduction quoted earlier, that the form<.em> of a locked room can exist without the trappings of the locked room. With sufficient cleverness, you can dispense with the locks. With a little more creativity, you can even dispense with the room.

A snowy field with only one set of tracks, when the victim was known to be alive after the end of the last snowfall can be just as compelling a locked room mystery as any other.

For that reason, he generalizes the sub-genre to “(apparently) impossible crimes”. Thinking about that definition leads you naturally and directly to the design process that I gave earlier.

The more you generalize, the broader the palette of options becomes. The Gods in most fantasy milieu are immortal, only their avatars can be killed (with limited exceptions) – so if one is found, stone cold dead, and not even the Gods can work out how it was done, you have an apparently impossible crime – and that’s a locked room mystery, no matter where it happened! Someone then has the bright idea of calling in some mid-level mortals to solve the crime, because they are more accustomed to doing so, mortals who just happen to be the PCs…

Or maybe it’s one of the Princes Of Hell (same logic, but adds an extra twist)…

The Impossible Larceny

Another variation is to change the nature of the crime. This is basically taking the Heist plot and putting the PCs on the other end of it – having to figure out whodunit.

A Pretty Little Bauble: The Heist in RPGs (Dec 2017) might be useful reference).

The Dungeon Twist

This notion came to mind while reading the Penzler introduction that I quoted from earlier. It basically applies the concept of a layered defense to a generic dungeon.

Let’s start with the concept of the central room of a dungeon being a locked room – one in which something has happened. Each room leading to that central chamber adds a layer of defense to the inner chamber – and (entirely “coincidentally”) rules out one of the possible solutions.

You have a situation in which the entire dungeon creates the “locked room mystery”.

Imagine there’s a famous Wizard who abides in a well-protected tower to prevent his being annoyed by people stopping by to consult him, extort him, whatever. Out of the blue, the PCs receive a letter from said Wizard which starts off, “If you are reading this, then despite my many protections, I have been killed. I charge you with the responsibility of discovering who and how, and bringing them to justice. As reward, should you succeed, you will qualify under the terms of my last will and testament to be my heirs…”

Mission Impossible: The Role Reversal

Or you could turn this on its head, and have an enemy of the PCs protected in every way that he can think of. If the PCs don’t bring an end to him, he will make their lives an unendurable misery, but he’s politically well-connected; any overt means like wide-scale devastation will cause consequences that are undesirable, from the PCs perspective. This creates a ‘locked room’ that the PCs have to find a traceless way into – in effect, charging them with creating the Locked Room mystery.

There are any number of Mission Impossible plots in the TV series that are essentially similar to this premise. Disguises, misdirection, psychological manipulation, and outright deception are the tools – what the PCs do with them is up to the players.

Locked Rooms Should Hold No Fear

I hope that all this has shown that there is more scope and potential to the locked room mystery than a locked room, and that GMs should have no hesitancy in pulling the occasional example out of their back pockets. And if the example plotlines thrown out in the last sections don’t entice you, I don’t know what will.