I'm in the process of writing a mystery novel/novella and it conforms to the traditional/classic mystery novel, emulating that of Agatha Christie or A.C. Doyle. I'm having doubts as to the proper way to structure the mystery plot. I know the general principles regarding the incorporation of common plot structures such as Freytag's Pyramid, plot twists, subplots, etc. to mystery novels. But, my question involves the use of scenes, sequels, and MRUs (motivation reaction units).

It is intuitive to use generalized plot elements in a mystery novel such as exposition, rising, middle, climax, and resolution as outlined by Freytag's Pyramid. But, on a microscopic level, it does not make much sense to structure a mystery novel with alternating sequences of scenes and sequels which is composed of MRUs. I bring this up because the sequence of events in a scene is Goal, Conflict, and Disaster. And a sequel consists of Reaction, Dilemma, Decision. This structure is, in my opinion, not compatible with a traditional mystery novel. I don't know if contemporary authors use repeating scenes and sequels in non-traditional mystery novels written today, but in the traditional style mystery novels written today, I don't see this type of plot structure as being conducive to writing an effective mystery or deriving a pleasurable reading experience.

The issue I see is that mysteries and traditional/classic style whodunit mysteries in particular cannot have disasters at every turn. And I don't see how MRUs apply conveniently to mystery stories. How can you have motivation and reaction driving every minute action of the detective? The MRU structure, in my opinion, just doesn't play well with a typical mystery novel, especially traditional style mysteries. I also don't see how sequels fit in nicely with the plot structure of a mystery novel. Moreover, A mystery is like an intellectual game between the reader/detective and the villain/detective. It is race to see who will solve the puzzle first and a game of cat and mouse before the detective brings the antagonist or villain to justice. It is purely an intellectual exercise, and disasters at every turn dilute the main effect of a mystery novel. So are scenes, sequels and MRUs actually incorporated into mystery novels, especially traditional mysteries? If so, how is this done effectively. I haven't been able to find anything online on this topic. Any insight would be appreciated. Thanks.

  • For reference to answers: The structure here referred to (scenes, sequels, and MRUs) is from Dwight V. Swain's Techniques of the Selling Writer. I had not heard of it before this question, but there you go.
    – Weckar E.
    Oct 22, 2019 at 9:27
  • You may also want to have a look at this to determine if that may be a duplicate of your question.
    – Weckar E.
    Oct 22, 2019 at 9:36
  • That question does not discuss whether scenes, sequels and MRUs apply to a mystery novel. But thanks for pointing it out anyways. :) Oct 22, 2019 at 10:08
  • I think that's because most people would be unfamiliar with scenes, sequels and MRUs specifically, or at least by that name :)
    – Weckar E.
    Oct 22, 2019 at 10:22
  • I don't know what this MRU thing is so I don't know how relevant this is, but I will say that there are definitly mysteries with disasters at every turn. Take a look at Christie's And Then There Were None. It's nothing but disaster after disaster as the characters futily try to figure out who on earth the killer is, but fail. There are many times when they think they're getting close to the answer, before yet again disaster strikes and they're back to square one
    – s.anne.w
    Oct 22, 2019 at 23:58

4 Answers 4


... it does not make much sense to structure a mystery novel with alternating sequences of scenes and sequels which is composed of MRUs.

I agree, I dislike this formulaic MRU theory, I don't think it applies, or if it does, it stretches too far the meanings of goal, conflict, disaster, reaction, dilemma, decision.

I do believe there are elements you have to include in a mystery, but you already know that. There has to be tension in figuring it out, there has to be some urgency in getting this done and catching the bad guy, there should be some misdirection or red herrings that seem absolutely plausible but fall apart when pursued.

At the "microscopic" level, your scenes have no particular demands except they matter in some way, even if they are just an example of your theme, or just build a character. Is a sex scene between willing partners supposed to follow an MRU? Is climax a "goal" that causes a "conflict" and then a "disaster"? Does this produce a "reaction", "dilemma" and "decision"? I don't think so, not without stretching definitions until they are unrecognizable, but sex scenes can be absolutely crucial to the plot, they change people. They may relieve sexual tension between two characters, or create an obsessive burden that gets in they way of existing plans, or even be the motivation for abandoning previously made plans. If they consummate love, they may solidify devotion to the point of one character sacrificing their life, something they would not have done without this consummation, or if their partner dumped them for somebody else.

As a more general formulaic rule, you can implement the six elements of the MRU across several scenes. Characters do have Goal, Conflict, and Disaster. Then those are plausibly followed by Reaction, Dilemma, Decision. But those could be six different chapters, or scattered across many chapters.

Trying to drive that kind of overall story structure down into the scene structure is, in my mind, overdoing a good thing. Every scene (or pair of scenes) is not a story in itself, just like every part of a car is not a car in itself.

It is like saying the only way to build a wall is by mortaring together uniform bricks. But you can build a perfectly functional and more interesting wall by mortaring together randomly shaped stones. It might even be stronger than a brick wall.

Different scenes do different work. The rule I would follow is that a scene DOES some kind of work, in character introduction or development, in setting exposition, in plot development, as a turning point (character epiphany or despair or whatever), as proof of something (somebody the reader likes dies, proving this is dangerous territory).

Have a story purpose in mind for your scene, that isn't just "its fun to write".

Do not use MRUs unless you believe a scene-pair written that way makes your story stronger, don't do it because somebody told you it makes your story stronger. When it doesn't make sense, revert to good writing: Sustain reader tension (keep them wanting to see "what happens next" on four levels: What happens by the end of the scene? By the end of the chapter? By the end of the Act (about 25% of the story) and by the end of the Story? Otherwise they get bored and start to skip ahead looking for tension.

Some scenes are naturally "payoff" for this tension, tension relieving scenes. A sex scene is like that, a battle is usually like that. But usually something happens in such scenes, one tension is relieved but the next tension on the same level is introduced: The battle IS what happens next, it has an outcome, and that has repercussions, and those ARE the source of tension moving forward.

The sex scene relieves the tension of sexual suspense (are these two ever going to get naked?!), but creates a relationship that is the new source of tension, moving forward.

The rule to follow, for now and all time, is don't bore the reader, pay attention to sustaining tension on all these four levels. And don't think every scene is formulaic, create scenes that seem like the plausible "next thing" for some characters to do, and accomplish something without boring the reader.

  • Like you, I'm not a huge fan of structuring scenes rigidly to the Goal, Conflict, Disaster sequence. As you said, there must be some variety. For example, you can't have conflict and disaster in the few opening scenes of the novel when exposition is the main focus. It just doesn't make sense. Nor does it make sense to use this structure when the detective is actively analyzing clues etc. The approach I'll adopt is creating "disaster" in a limited number of "scenes", tension in the sense of subtle setbacks, and the "sequels" as the events highlighting the detective's analytical train of thought Oct 22, 2019 at 23:02
  • @RajNarayanan It is also tension to have the detective certain something is a clue, but just can't figure out what the heck it means. e.g. an American detective investigates what looks like an office murder, but nothing was taken, then finds a weird coin in the bottom of the victim's coffee cup, under an inch of coffee. Turns out it is an ancient Roman silver Denarius, worth a few hundred dollars. The detective thinks it means something, he just can't figure it out, nobody that knows the victim has a clue, and it drives him (and the reader) nuts. (and it does mean something in the end).
    – Amadeus
    Oct 23, 2019 at 21:41
  • That's a good example. I also meant to say that the "disasters" could be something as simple as subtle setbacks like the coin example you provided. Disasters don't necessarily have to be another murder, a major road block, etc. Oct 24, 2019 at 1:37
  • @RajNarayanan And that is what I mean. I love language, I wouldn't call finding the coin a "subtle setback", it is a clue, that is a stroke of luck. It is a puzzle and that creates tension, but it isn't a negative incident that forces a change of plans. IMO classifying the finding of a clue under "setback" or "disaster" stretches the meaning of "disaster" beyond the breaking point. And that is what is wrong with the MRU model; it isn't helpful because it isn't prescriptive, it plays so loose with the labels it uses that it doesn't really tell you what to write. (continued)
    – Amadeus
    Oct 24, 2019 at 10:04
  • What is the difference between a goal (finding out what the coin means) and a disaster (the MC doesn't know what the coin means), and a conflict (the coin doesn't fit with the picture presented of the victim)? Where is the conflict in a scene where the MC pokes around the victim's office alone, and discovers a Denarius, apparently hidden in a hurry? There's no confrontation or argument, he gets it in a baggie without smudging any prints and leaves. Stretching MRU to cover this scene means MRU covers anything you write, so it isn't helpful at all in telling you how to write a scene!
    – Amadeus
    Oct 24, 2019 at 10:22

The answer is no. Simply because there are no elements that should or should not be included within a novel, mystery or otherwise. There are no correct formulas. This is evidenced by Dwight Swain's inability to write a successful work of fiction.

Most mysteries / thrillers rely on skills such as misdirection, selective POV and psychological manipulation of the reader.

There would be no mystery about 6th sense if it was revealed the protagonist was dead in first scene.

I strongly suggest you read a few Agatha Christie novels and analyse them, if that is the writer you wish to emulate.

  • Your'e right. Mystery novels shouldn't have to rigidly conform to common elements of fiction writing, which is why I'm going to vary my approach that's appropriate to the writing task at hand. I'm also in the process of studying Christie's and Doyle's writing styles. Thanks. Oct 22, 2019 at 23:34

It should be pointed out that most of those Mystery Series work because the mystery was acceptably unusual and above the norm for the setting. Agatha Christie's most famous work relied on the killer being unable to leave the social gathering and thus, still among the core group. Sherlock Holmes was a consulting expert that Scotland Yard came to when they were clueless (at the time written, the were among the best at what they did, so if they didn't know, it was a stumper). Similarly, Encyclopedia Brown was a fountain of useless knowledge and most of the cases sent his way already didn't pass the smell test with the client, but the client didn't know what was wrong. Brown's own father, the chief of police, would always be the first client of any book, and it was the benefit of his son's attention to detail and being a fresh pair of eyes, coupled with being a wellspring of factoids that made him the case solver... similarly many neighborhood kids figured something was wrong because of the reputation of the character saying it (Bugs Meanie was doubted because he was the neighborhood bully and Wilford Wiggin's reputation as a Snake Oil Salesman was so well known that Brown not being allowed to watch the pitch would have been a bigger clue that the new scheme was a fraud than anything in the newest pitch).

Modern series tend to look to real life stories for their inspiration. Law and Order typically will source it's plot from a recent headline at time of production, while CSI had career investigators on the writing staff, who were often writing bizzare stories from their own cases OR would find a subculture and plot out a murder based on the subculture's inner drama. They were also fond of the victim's odd death initially looking like it was done by the guy who found the body... only to reveal something else was in play long before they actually died... one memorable episode turned out that the "crime" wasn't a murder, which was staged to gain access for a vault theft, which also wasn't a crime since that was staged to hide an insurance fraud scam (Grissom even pointed out he can't prove the scam to a criminal court beyond reasonable doubt... but the insurance company was well within their right to not pay out on the claim based on the evidence he submitted to them. A Crime mystery with no proven crime.).

Another great way to do Agathe Christie/Holmes is to find a mystery dinner group. Mystery dinners are small parties where the party goers act out a role of a suspect in a game. The game sets out that the players are all suspects in a murder in a closed scene and are give a character sheet with their own character and his/her quirks and witness evidence... like seeing another player-character with the victim before the party, but with another player-character at the time the body was discovered, thus providing an alibi to counter yet another player-character, who notes she didn't see him at the time the body was discovered (two separate players will confirm that that the third was accounted for against the fourth). Depending on how the game was written, the murder will be one of the players, who may or may not be provided the knowledge of his characters role (I've played one where my character was the murderer, but I had no knowledge of it). The non-murders will win if they can sucessfully accuse the player who is the murder, while the murder can win by avoiding any accusations.

The game "Werewolf" uses a similar mystery set up, but here, there are typically three murders (the werewolves) and the remaining players are townspeople. In this set up, each player is secretly assigned a character role that acts and reacts in certain ways and no player is privy to any other players role. The game takes place in rounds, broken into night and day. During the night, the werewolves will silently decide to kill one of the townspeople and the players who are not killed may perform secret actions, and during the day, all players will vote on the person who they believe is a werewolf and if there is a significant majority, they will hang the accused.

Here the game works by ensuring at least one elimination per round (the werewolves always kill, while the townspeople do not need to kill if there is no confidence among them), the werewolves may vote with the townspeople (thus may throw suspicion off of their pact) and each player has a secret role that is typically to the towns benefit, but may cast suspicion on them or ire by the werewolves. For example, one role that is frequently given out is "the child" who is permitted to secretly witness the werewolves but may not publicly accuse anyone during the day out of fear that the werewolves will kill her the next night, but she may vote with the townspeople. If the child can clue in another character about their role, it can cement votes as the child will always vote for a real werewolf and never for an innocent. However the threat is that if a werewolf finds out, they will likely kill her and by not voting if between two innocents, the child might be confused for a werewolf as well. No player is allowed to publically reveal their role, but some players may have an ability to secretly discover another role or do so by looking for patterns (For example, a seer may be able to see one character's role every night and needs to quietly influence to vote to protect innocent characters who are accused, but not alert the wolves, who will target the seer if they figure it out.).

The game is won by the towns people if at any time the entire wolf pack is eliminated and by the wolves if any one wolf surives the final vote (If there are two villagers and a wolf in the final round and the villager is eliminated during the day voting, then the wolfs win (as the remaining villager will die that night). This means that the wolves also have to be wise about their next kill... if they target someone who was never accused of being a wolf, then it narrows the pool of suspicious villagers to be accused. If the wolves leave them, they can later cast doubt on them the next day. There also need not be as many players as available village roles, so each game will be different as not every character is in play (and one died in opening round, though a ref might be sacraficial lamb so the innocent villagers are all given one round to use their skills, and the ref can act as a debate moderator).

This is a great game to play for a mystery writer as it requires players to be stratigic both in what they reveal and what they don't reveal to the crowd writ large as at least one werewolf is present and may target them to eliminate their befits to the villagers.

If you don't have a wide pool of friends, fear not, as these games can be made for single players. While the mystery need not be nefarious (a popular scenario is figuring out Secret Santas) it involves three lists of uniform length and a common relationship between exactly three items on the list. Typically in secret Santa format, it will be a person, a gift that person wants, and a list of Secret Santas (typically a second copy of the first list) and your objective is to find out who got what gift, and who gave it to them. Typically you are given five hints that will eliminate certain combinations. These hints will make a comment most conditions but not all and contain logical operations (Bob did not get the train or the bear). Typically only one complete elimination is made in the hints (Sarah wanted a Book) which can eliminate it for four, and a general rule (Typically, there is only one loop, so Bob can not get Sarah a book if Sarah gave Bob a gift). In secret santa, the player can elimate Bob as his own secret santa.

Another format is Gift Wrap, where five people (Alice, Bob, Charlie, David, Emily) buy five gifts (A book, a train, a bear, a flower, and figurine) and wrap them in five differently patterned wrapping paper (circles, squares, triangles, solid color, stripes) in which case the self referencing becomes harder as all lists are exclusive, and you must figure out which person gifted which toy and what wrapping paper did they use to wrap it.

In either format, the player is still playing detective and wins by making eliminations by the five clues given and the inferences between specific and general clues. The strategy is to take each clue and combine them to make more eliminations (knowing that Alice wrapped her gift in solid color paper and Bob did not give the train of the bear as a gift) will mean we can Bob (and anyone not Alice) from any associations with gifts with solid wrapping. Additionally, we can't yet say Alice's gift is the train or the bear, but if either is her gift, we can elimate the other from associations with Alice (she can't give two gifts), Bob (who didn't use either gift) and solid wrapping paper form an association with Bob (since Alice used it) and from the gifts not used by Alice (since we have her gift).

  • My goal is to write cozy mysteries and I want to incorporate the story structure of the classic cozies (e.g. Christie), but at the same time incorporate modern elements of cozy mysteries, e.g. show don't tell (something Christie and Doyle constantly violated), modern dialogue, modern sentence structures for pacing etc., modern techniques of characterization and crime investigation, and so forth. Any advice on this approach would be appreciated. Oct 23, 2019 at 21:40

Try not to understand "disaster" in the real sense of objective disaster in the real world. The term "setback" might be more helpful in this case. Every scene ends with a setback for your character(s). This setback relates to the scene goal. Put differently: You write "try-fail cycles". Character tries to achieve something, fails, and tries it a different way. Them not achieving the goal is the "disaster". So it's not a story event. It's not plot. It's character point of view. In summary, every outcome of every scene is the answer of the question "Will the protagonist succeed?" The answer will always be one of three things: a) yes, but ... b) no, and furthermore ... c) no ... This is true regardless of genre. It's the stuff dramatic fiction is made of.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.