The subject of today’s article emerged during a twitter discussion between John Kahane (@jkahane1), Joe Kushner (@JoeGKushner), and myself (@gamewriterMike) on Twitter last week.

The conversation started when I mentioned to John that I had just started a mystery plotline in my Zenith-3 (superhero) campaign, and that special effort was going to be needed to make the conclusion live up to the buildup, and that one of the ways that I intended to deal with that problem was to make the ultimate killer larger than life. John replied that in his experience, “Murder mysteries are hard to run in RPGs every time, and larger-than-life villains make that tougher.” He then made the point that a general solution needs to work regardless of genre, implying that some genres permit whiz-bangery and over-the-top characters more tolerably than others.

This is an absolutely spot-on point, but one that we didn’t explore at the time, because the conversation moved on with my reply, “[The] Best answer is to make the Mystery the star of the adventure!” and John’s reply “Always is. But never as easy as that! :)”

John continued, “If the mystery isn’t the star of the game, then what was the point of doing a mystery to begin with, right? :)” and Joe pointed out that “In the last few novels I’ve read, the point of the mystery was to immerse the reader into setting characters.”

I responded that this solution doesn’t work very well in an RPG – it can be very dry and non-interactive, just the GM giving page after page of narration to read. A better RPG approach was to “make the plot the key antagonist of the adventure” rather than the villain ultimately responsible in-game.

It was at that point that I suggested that it might be worthwhile writing an article on the subject for Campaign Mastery – and so, here we are.

I should comment further on one of John’s points. He’s quite right in saying that larger-than-life villains can often make mysteries tougher, and that’s because by virtue of being larger than life, they attract attention to themselves – making it much harder to sustain the mystery. It will work in this case because the villain won’t actually appear (if all goes according to plan) until the final confrontation.

The Metagaming Approach to Mystery

Treating the plot as an antagonist is obviously metagaming, and I make no bones about that. Rather than in-game developments driving the plot, the plot is driving in-game developments. But it differs from the most egregious examples of Metagaming in a number of key respects.

1. Plotlines must be dynamic; only the start- and end-points of the plot can be fixed;
2. Characterization cannot be violated;
3. Continuity cannot be violated;
4. Causality cannot be violated;
5. Challenges must continually be presented whose outcome defines the course of the investigation, not whether or not there is ultimately a successful solution/conclusion;
6. The purpose is to provide fun for the participants. The GM’s role is to referee the contest between the metagame “antagonist” and the in-game characters, not to force one particular course of events on the players (which would violate rule #1 above).

Six specific mandates that enable metagaming to work for you, not against you, at least in this particular application.

The problem of Metagaming

Metagaming has a bad reputation as an RPG technique. This largely stems from the use of metagaming to force PCs into following a pre-scripted path between beginning and ending in an adventure, also known as a Plot Train, in which players have no control over the course of events and may as well not be there. It is the dynamic interplay between players and GM that makes an RPG a collaborative effort, with neither in control of the course of events.

Dynamic Plotlines

Mandating that metagaming techniques can only be used to make plotlines more dynamic, changing as needed to make the game more entertaining and more responsive to the activities of the PCs, rather than forcing them to conform to some fictionalized “ideal narrative” such as you would find in a script or a novel, is a key difference between the technique I am suggesting and what most people think of when they speak of metagaming.

The concept is one of making the challenges and opportunities that shape the plotline responsive to the PCs, not the PCs responsive to the shape of a pre-determined plotline. I think you’ll agree that this is quite a significant difference!

Fixed start-point

Almost every plot to which this technique can be applied has a fixed start-point: The PCs are presented with a mystery that they have strong motivation to solve. The essential parameters of the mystery are defined for them – what evidence is at the scene, who the victim is, where the victim was discovered, and so on. In most cases, there are one or more unusual or novel features presented to make the puzzle more interesting. In those cases where the initial mystery appears more prosaic, you can expect to need a plot twist or two to emerge fairly quickly to provide that novel feature – in other words, the initial start-point is deliberately deceptive, and the start-point is not as prosaic as it initially appears to be. So far as this technique is concerned, this initial plot twist is also part of the basic start-point.

Fixed end-point

Most mysteries also have a partially pre-determined end-point, in that the GM should know who has committed the “crime” that poses the mystery, and how. The end-point is the final confrontation between perpetrator and the PCs. Where this confrontation is to occur and how it is to be resolved are usually, but not always, unknown.

Dynamic routing

The path to solving a mystery is never a straight line, because that’s dull. Instead, you need surprises and plot twists and unexpected directions and challenges to be overcome.

Starting from Times Square and heading for Jersey City is fairly straightforward. 7th Avenue to the Holland Tunnel and then the I-78W and you’re right in the heart of your destination. 7.1 miles and 21 minutes estimated travel time.

A more creative and challenging route passes through Albany, up to Quebec City via Sherbrooke, down through Trois-Rovières to Montreal, West through to Ottowa, across the northern side of the Great Lakes to Detroit, then to Springfield, Illinois, to Kansas City, to Salt Lake City, through Reno and Sacramento to San Francisco, south to Los Angeles and San Diego, cross the Mexican border at Mexicali, south down to Guadalajara, North-east to San Antonio and Dallas, East to Jacksonville, south to Orlando, North to Atlanta and Nashville, North-East through Knoxville, Roanoke, Harrisburg and Scranton, East to Middleton,then south to Sparta via Ogdensburg, then East through Clifton on the I-80 E to the I-95, south and then east to approach Jersey City from the West Side. It’s definitely the long way around at 9,916 miles, an estimated 154 hours of driving, but it gets you to exactly the same destination from the same starting point – and will definitely show you things that the direct route won’t.

This is analogous to an RPG in which the first suspect is immediately and obviously guilty and in which the PCs have overwhelming proof of that guilt, an open-and-shut case, as it were, vs. a far more complex path to the final confrontation in which the PCs start looking in the entirely wrong direction, which eventually results in a plot twist that leads them off in other wrong direction, with its own twists and turns, followed by yet another wrong path driven by complications, before slowly heading in the right direction and seeing things fall into place one after another until the final resolution is reached.

Viewing the plot, the mystery, as an antagonist who will respond to PC attempts to move in the right direction with complications and deceptions and red herrings designed to lure them off-course via a more interesting, entertaining route, enables the pathway to success to be as convoluted as is desirable, changing and evolving in a move-countermove pattern in response to the choices, theories, and interpretations of evidence by the players.

The mystery is “willing” to be solved – but it has to be “wooed” first, seduced and cajoled into giving up its secrets. The mystery makes itself as complicated a story as it needs to in order to achieve the maximum entertainment from the journey.

A proportionate response

It might seem that the simplest specific technique for achieving this is a proportionate response – whenever the players propose an action, the GM assesses that action in terms of how directly it can lead to the solution, and inserts difficulties in proportion. A more refined variation might take into account inherent difficulty of the action relative to the capabilities of the PCs.

The problem is that this produces an entirely-too-predictable pattern. “We must be onto something, the GM is trying to make it hard for us”. This is a pattern that interprets the GMs position as hostile, as “the enemy”, rather than the position of neutrality that the GM should seek to adopt.

The GM’s job isn’t to make solving the mystery harder for the PCs; it’s to make solving the mystery more fun for the players.

Success is necessary

The other problem with the “proportionate response” approach is that the PCs need to succeed regularly. If they set out to learn something, they should learn something. If they attempt a chemical analysis or a cross-examination, they should get answers. Things shouldn’t be handed to them on a plate, but they shouldn’t be so difficult that the players feel they are getting nowhere, either. The trick is to get the PCs to ask questions that don’t lead directly to the solution, so that they can be answered without compromising the mystery.

Keep It Entertaining

A better formula is to assess how much fun the GM can provide for the players as a result of the proposed PC actions, and let that determine how easy that action should be as a result. Again, this can become predictable, though – it is to be hoped – with fewer objections on the part of the players!

The best approach is therefore to combine the two approaches. A proportionate response to determine how severe a complication to interject into the process, with the fun factor determining how easily that complication should be overcome.

A useful plotting tool is always to ask yourself how the perpetrator expected to be able to get away with whatever it is that they have done. False alibis, throwing suspicion on someone else, adulterating physical evidence, people who will lie, either for the perpetrator or for their own perceived benefit, and so on – and don’t forget that this is an evolving process; they won’t stop trying to make themselves look innocent.

It’s a maxim of the Columbo television series that the most co-operative individual is guilty until proven innocent. In it’s own way, this is clever, because it means that the perpetrator seems to have a handle on the course of the investigation and can meddle in it as they see fit – though this usually is done so heavy-handedly that it becomes obvious that they are trying to throw the investigation off course. Far better to try and react, and cooperate, to exactly the same extent that you would if you weren’t guilty. Be no more helpful than you would have been, and hinder no more than you would have done, and for the same reasons as you would have done. The criminal who does nothing unusual after the crime is the hardest to detect.

The dangers of the big picture

I’m a big advocate of always keeping an eye on the bigger picture. This is, however, one occasion in which this practice can lead you astray. Don’t let your long-term plans get in the way of an entertaining game for the players in this adventure. If the course of the mystery brings the PCs into contact with someone who is not who they seem, it might be tempting not to risk exposing them as someone of interest, preserving their perceived innocence until that part of the big picture comes to fruition.

Don’t Do It. That eventual plot twist will have even more power if the PCs suspect the double-agent and then clear him or her as a result of their investigations! Don’t be afraid to complicate your future if the result is a better adventure right now!

Characterization cannot be violated

Of course, if you had a completely wide-open field of possibilities, it would be easy. Every time the PCs think they have their man, throw in a complication that shows they could not have done it. Keep going until the clock indicates that it’s time for the adventure to end, then make the next suspect the real perpetrator.

This throws internal logic to the winds. Characters change as necessary to make the plot more complicated. Nothing is predictable any more, and this quickly becomes boring and unrealistic to the players.

There can be rare occasions to take this approach, but as a general rule, it is not acceptable. (I have in mind a specific adventure I ran in which a suspected double-agent was in fact a mercenary, selling his talents to whoever would pay his price. So he started as a Bulgarian who was secretly a Russian Spy, then was revealed as a triple-agent working for the Americans, then as a mole within the American organization for an international organized crime cartel, and then as having infiltrated that crime ring on behalf of the Russians. No matter who came out on top, he was able to claim loyalty to the cause. On top of all that, he was being framed for the crime which brought him to the attention of the PCs in the first place by someone who had no idea of any of this!)

As a general rule of thumb, personalities, once allocated to an NPC, should not change. If you need them to behave in a way contrary to the way this personality suggests they would, you need to arrange circumstances so that the NPC is put into a position where they think this is their only alternative, within the bounds of their personality.

A great example is of someone who thinks a son is guilty of the crime (possibly falsely) and who therefore does everything they can to misdirect the investigation even though they are clearly a good person who would normally be cooperative, even to the point of trying to make themselves look potentially guilty through suspicious behavior. Or of someone who is guilty of an unrelated crime who destroys or degrades evidence of the main crime to protect their real secret.

Continuity cannot be violated

No retro-active rewrites of events are permitted (except in a time-travel plotline of course – which is a whole separate headache). Once a relevant piece of information is in the hands of the PCs, you can’t change the events that led to that piece of information being relevant. If an individual becomes obviously guilty in the eyes of the players because he committed the crime as the GM originally envisaged it, you can’t retrospectively rewrite the crime scene to have someone else commit the crime. Instead, you need to shed doubt on one or two of the key pieces of evidence pointing to the guilt, sending the PCs off on a wild goose chase – but that doubt needs to result from actions that can be performed now, not that are being inserted retroactively as having happened then.

This comes back to the impartiality of the GM and avoiding a “him vs them” situation at the table. You have to be able to lay every event at the feet of an NPC and have it be reasonable that the NPC has caused that event. Violating continuity essentially means deciding retroactively that you were lying to the players – and that’s not on. The NPCs are your agents, and you can introduce more if you need to – they can and should lie when its appropriate.

The times when you are most tempted to rewrite the past happen when there’s a plot hole in your very clever plan, and the PCs a climbing down it. The correct way of handling this problem is to ask yourself (a) which NPC could plug that plot hole now (not back then) and (b) why they would do so, within the boundaries of any established personalities and capabilities.

Causality cannot be violated

If the PCs do something which should reasonably have a particular outcome, it should still have that outcome. Players need to be able to predict the outcome of their actions. Unlikely as it might seem, I have seen GMs try to violate this action because the normal outcome would be detrimental to their plot. This is “bad metagaming”.

The urge to do so has two origins – the first is failing to anticipate what the PCs might do or attempt to do in a particular situation, and the second is by presenting subjects of possible activity in a provocative manner.

Overlooking a probable action

• “I mindscan for an aura of guilt”.
• “I cast a truth spell on them all and ask each if they did it.”
• “I listen for changes in their heart rate to find out which one’s lying.”

I’ve seen mysteries fall apart through all of these actions because the GM didn’t remember that he had already let the PC do this in an earlier adventure, and plan accordingly. And I’ve seen the GMs in question announce that it doesn’t work, despite having already established that the PC could do so. In one case, the GM attempted to cover his tracks by having the NPC have taken precautions against the ability in question even though there was no reasonable way for him to know that he needed to do so. In the others, they simply rewrote what the power could and couldn’t do.

Neither response is acceptable. If your mystery is about to be undone, you need to come up with a reason for it not to work that doesn’t violate causality or continuity.

• “They all feel guilty about something, not necessarily the same thing.”
• “They all say they didn’t do it while under the spell’s influence. The perpetrator probably doesn’t think he did anything, he blames the victim for causing it to happen.”
• “You hear no changes in heart-rate that indicate a lie” (because the person lying has a pacemaker that keeps his heart-rate even).

Okay, the last one’s a bit of a fudge, but it gets the job done. It’s not something that the criminal did to defend against the unusual investigative technique, he didn’t know that he would need to protect himself against – it simply changes the circumstances so that the results don’t give the game away, introducing a complication. As soon as the PCs find out about the heart surgery, they will discover how they were misled by this test, but in the meantime, they have prima-facae evidence of innocence, and can be led off down the garden path for a while.

Listing elements in the wrong order

If you want to draw attention to something, mention it last, or first, or immediately after something that’s especially attention-getting, like first mention of a dead body. If you don’t want something to get a lot of attention, bury it, and put something more interesting to one side of it. If the item is sufficiently part of the expected landscape, you can even get away without mentioning it specifically, so long as you emphasize that everything you would expect to find is present.

For example, if the criminal gained access to the scene by means of an air vent, or simply left by means of such a vent, don’t call attention to the vent – but don’t try to be too obvious about hiding it, either. Mention it too prominently, and someone will look into it; when that happens, there is an instinctive response to try to change what the PC would reasonably expect to find if the air vents had been used for congress to or from the scene in order to preserve the mystery. This is entirely the wrong way to fix the problem! Far better to come up with a reason for whoever discovered the body to have concealed the evidence of the vent having been used, thereby giving a good reason for the PC to find nothing.

In fact, because the GM could never be sure that the PCs would not pick up on mention of the air vents, he should have had something of the sort planned anyway!

Challenging the PCs

The PCs should encounter various challenges that have to be overcome in order to solve the mystery, but the eventual solution to a mystery should never reduce to the success or failure of such a challenge. Success or failure changes the course of the story, but not whether or not it reaches an end!

It follows that every failure must result in the information that a success would have delivered being provided to the PCs eventually, anyway. It might be that this happens too late for the PCs to do anything about it – the criminal might have gotten away in the meantime – but the loose ends should be wrapped up.

There are several ways to challenge the PCs “safely”. In general, these are all about changing a “What” or “Who” question into a “How and when” question.

How information is revealed

For example, the challenge does not determine “What information is revealed?” but instead determines “How and when will information be revealed”. There must always be another path to the answers that can and will eventually be followed. Success can indicate immediately, or at the conclusion of a defined process; failure might mean waiting until someone else looks at the question. It might or might not even be obvious that success was not achieved. “You find nothing, but are not completely certain that there was nothing to find.” “A power surge wiped the hard disk”.

How information is hidden

The converse is to change “What information is hidden?” into “How and when will any hidden information be revealed?” This is more a matter of the questions the GM should ask himself, than one that is likely to be asked by a PC, but you never know…

Coincidence & Human Nature

Coincidences do happen. Human Nature will always be a factor. Take advantage of both of thees facts to create more challenging circumstances as necessary.

If you have, for example, an uncooperative witness, overcoming the challenge should never be just a matter of succeeding in a Skill Check. At most, a skill check should reveal that the witness knows something that might be helpful but has reason not to volunteer it. The PCs need to identify that reason and find a way to overcome it, ‘unlocking’ the witnesses knowledge. Once you have that key, that’s when a Skill Check can be used to actually obtain the information, for whatever it’s worth.

The Hurdles to be overcome

It’s a fact that the GM needs to be mindful of: Mystery plotlines will suit some players and some PCs better than others. Things can get awkward when the situation suits one and not the other – a player who’s good with mysteries but whose character is not, or vice-versa. Careful attention needs to be paid to ensuring that everyone gets their share of screen time, both as players and PCs. The challenges that are to be overcome can be a great way to move the spotlight onto someone else. The combat monster gets to conduct a raid in order to secure evidence that the detective needs, for example.

It follows that whenever you contemplate inserting a complication because things are happening a little too easily for the PCs, you should evaluate the opportunities for such spotlight-sharing, choosing the nature of the complication and its eventual resolution accordingly.

Never a dead-end, only a roadblock

The final point that I want to make under the heading of the Challenges to be overcome by the PCs is that they should never result in a dead-end where all avenues of investigation are blocked. An individual course of inquiry might reach a dead end, but there should always be a “next step” for the PCs to make, and the players should always know about it before they encounter the roadblock. How easy that “next step” is expected to be is an entirely different saga!

Metagaming to maintain challenge

The main directive to follow under this approach is to invent complications as necessary to increase the fun – and the difficulty – of solving the mystery, but to do so within the scope of specific parameters. You must never violate the “truth” of the solution (obfuscating it is entirely acceptable, however), must never violate the personalities of the NPCs involved once they have been established (but you can manipulate responses by altering the circumstances in which they find themselves), must never retcon your way out of a problem (but can introduce new factors that have the same effect), and must never violate the players’ trust that you will adjudicate their efforts fairly. So long as you do not violate these restrictions, anything’s fair game.

Roleplaying the “Antagonist”

With so many options open to the GM, how do you make this sort of thing manageable? Well, you can just go with whatever comes to mind first, but that’s not all that reliable as a means of making good choices, just of making easy ones. The best technique I have come up with is to treat the “Antagonist” as a Mastermind, plotting against the PCs for its own motives, and with a distinct personality of its own complete with things that it will and will not do.

This focuses and characterizes the options that are open to the GM into a relatively confined range, making selection of obstacles and twists easier, but it has a side-benefit of considerable worth: using a distinct and different personality each time gives each adventure a quite different flavor.

This Antagonist is not malicious, or vindictive; it toys with the PCs for its own reasons, helping here and hindering there. It is, however, normally omniscient and omnipresent, though you can further refine the personality by limiting those aspects of its capabilities.

Don’t make it too easy

As a general principle, don’t make it too easy for the PCs; there’s nothing worse than a color-by-numbers mystery, where all the players have to do is keep themselves awake and wait. Nor should it be enough for the players to “phone in” their participation; activity needs to be purposeful and not be happening for its own sake.

Don’t make it too hard

There comes a point at which the GM has to accept that the PCs are going to start closing in on the solution. You should never obstruct them until there is only enough game time left to achieve the solution and final confrontation; it’s far better to allow a little time for some minor reverses in the course of that “closing in”. I never introduce a complication without some idea of how long it will take to resolve, so that I leave enough room for the end of the adventure.

You may be tempted to make the mystery the focus of the day’s play to such an extent that the actual confrontation is delayed until the next game session. I’ve tried this a few times, and while it’s not the worst approach in the world, I have found that it makes the final confrontation far more anticlimactic than it would otherwise have been. In general, unless that confrontation is itself intended to be a springboard into some bigger plotline, I will strive HARD to avoid this, and even then, it’s better to end play with the confrontation and discovery of the bigger plotline as a cliffhanger than the other way around, if time can be manipulated to permit it.

Progress must be continuous

I hinted at this when stating that “Success is necessary,” above. At any arbitrary point, the PCs should be able to say that they are closer to a result than they were, that they have eliminated incorrect solutions and, in general, “made progress”. While there are times when you can and will handicap and counter the players with glee abandon, there will also be times when you may need them to have a stroke of good fortune, or to dangle a carrot in the form of at least part of a solution in their direction. You need to monitor the mood at the table closely, and at the first hints of frustration, act immediately to provide demonstrable progress.

Maintaining neutrality

Above all, the GM must maintain neutrality, favoring neither the “Antagonist” nor the PCs. This can be very difficult to do when engaging in a battle of wits – and that’s what this metagaming technique ultimately boils down to – and so special attention needs to be paid by the GM to his neutrality. If the players do something clever, intuit something, deduce something from information available to their characters, they need some progress as a result. Adhere to the principle of generating the maximum level of fun from the adventure, rather than trying to force the adventure down some pathway to a solution; let the path between puzzle and solution grow organically from the soup of ingredients you’ve put in place, and this technique will reward you.

I should conclude by pointing readers to the other articles on this subject: The Butler Did It: Mystery Plotlines in RPGs and its sequel article, The Jar Of Jam and The Wounded Monarch: Two Mystery Examples. Readers might also find Ask The GMs: Penetrating the veil of mystery to be useful, looking at why Mystery adventures can be so hard to create, and how to remove some of that difficulty.