Take the 2-minute tour ×
Writers Stack Exchange is a question and answer site for authors, editors, reviewers, professional writers, and aspiring writers. It's 100% free, no registration required.

In a murder mystery, most of the story is generally focused on figuring out who the murderer is; in "Fair-Play" mysteries, it's assumed the murderer is a significant character in the book, and the fun is in figuring out which of them it is.

This being the case, readers familiar with the format generally try to second-guess the investigation. That is, not only do they try to solve the case along with the sleuth and the unfolding information within the story (which is fine, that's something I want); their guesses are often based upon their knowledge of the structure and conventions of mystery stories - along the lines of "Mr. X can't be the murderer, because he's the first guy the detective suspects," or "It's got to be Little Miss Innocuous, because she's being portrayed as being innocent and trustworthy - so she's the one the author doesn't want us to suspect!".

Now, plenty of mysteries can withstand this sort of meta-interrogation, and some intentionally subvert these conventions. But I'm interested in avoiding this reaction altogether, or at least minimizing it - I want my readers' wits engaged in a duel with the mystery, not with genre-savvy one-upmanship.

How can this problem be addressed - how can I maintain suspense and suspicion towards a wide range of suspects, and encourage the reader not to second-guess the story?

(Note that I'm not looking for suggestions which will surprise the reader despite his genre-savvy. I'm looking for suggestions which will keep him from relying overwhelmingly on genre-savvy to begin with.)

share|improve this question
1  
Forgive me if this is stupid/cliche, but could you make the detective himself genre-savvy? For me being genre-savvy is just one of the tools I use to solve a mystery with the detective. It's hardly my fault if he doesn't have the same repertoire of imaginary crime knowledge. If the detective did...It could be interesting. Or terrible. Please note that this is a purely whimsical thought. –  kitukwfyer May 6 '11 at 14:11
    
Just don't write 50 books Agatha Christie style, where in every story, it's the least likely person. –  thursdaysgeek May 6 '11 at 22:26
    
@kitukwfyer: It's an interesting idea, but it'd be awfully tricky to pull off. Those tropes don't really make sense if you don't know you're in a detective story - in real life, the most likely suspect is a more likely subject than the least likely subject. :P But that could be interesting - a way to head off the reader right away... –  Standback May 7 '11 at 18:02
    
...As I said, just a passing thought. Hey! I know! Just have the murder occur at a murder mystery role-play-thing (which I admit I know nothing about...)! XP –  kitukwfyer May 7 '11 at 20:54
    
@Standback: However, John Dickson Carr's Gideon Fell was perfectly aware that he was in a detective story, as seen in one chapter of "Three Coffins". –  David Thornley May 10 '11 at 1:38

4 Answers 4

A popular variant of the whodunit structure is the howdunit or the howcatchem, in which the question isn't who committed the crime - it's how he managed to pull it off, and/or how the detective succeeded in conclusively proving the culprit's guilt.

I find that this neatly sidesteps the problem, because the reader is no longer guessing which of a list of people is the murderer. So the howdunit discourages "gaming the system" with guesswork, similar to how an open-form question discourages guesswork far more than a multiple choice question.

A lot of mystery stories, both classic and popular, use the "howdunit" structure - including some Sherlock Holmes stories, and TV shows like Monk. Another TV show, Veronica Mars, does something similar - it frequently shifts focus away from finding the culprit and towards clearing somebody who's been unjustly accused.

On the other hand, I personally find the howdunit variant to be less inherently compelling than the whodunit. Watching the detective fill in the blanks - clever as they may be - when the overall lines are abundantly clear simply isn't as suspenseful as the promise of the killer's identity being dramatically exposed.

share|improve this answer

The murder mystery is generally a mystery for four reasons: who did it, how did they do it, how are they caught, and why did they do it.

If you only focus on "who did it", then I believe you have no choice but to accept that the reader is going to try and figure out who did it the whole way. This is because that is the whole purpose of your story. You've set the question, "Who did it?", and the reader wants it answered, so it's natural for them to try figure out the answer to that question. This happens in all books to a degree: how will the protagonist succeed? Will he overcome this obstacle? Will he get the girl in the end? Even if you focus on the how and why, people are likely to second-guess every step of the way that more information comes to light. ("She was having an affair? Maybe he killed him because he found out!")

There are still various ways you can use to get people focused on the story, however.

The most important is through characterisation because this raises other questions that need to be answered, above and beyond the central one of "who did it".

The old "whodunnit" books hardly ever placed their detectives at the centre of the story. They were merely there to solve the crime. There histories were largely immaterial. However, what turned the genre on its head over the years was this idea of the sleuth's life being integral to the story. They're flawed, have conflicts that affect them, and the people around them, they've got baggage they bring to the case, there's political intrigue within the police department, bad relationships with friends, family, ex-lovers.

The story is not just about the crime any more, but the people involved.

The second trick is to have other events operating in conjunction with the crime that add another dimension to the story.

One of the best examples of this (as well as characterisation) that I've seen in recent years was the excellent Danish crime drama "The Killing", which had a murder entwined with a local mayoral election.

A lot of political drama took place that was influenced by the murder (not least because one of the suspects was a politician).

We also followed the family affected by the murder of their daughter, and how their relationships broke down afterwards. We followed the sleuth, whose own obsession was tearing apart the world around her.

It emphasised the point that murders don't happen in vacuums, and they have far-reaching consequences on those involved. It is the impact of these consequences that help the reader become more interested in the story beyond just "who did it". If the only question you've posited is "who did it", that's the only question a reader will try to answer. If you posit other questions, like, "Will he get elected?" "Will she leave him?" "Will he kill the suspect?" "Will she be fired?" then you've given the reader a lot more to be interested in above and beyond the rather clichéd "whodunnit" formula.

Edit: Okay, so based on your comment below, I think you already answered it in your question: play with the conventions of genre, and the reader's expectations. The best way to do this is to immediately destroy a reader's expectation right upfront, because then they're put off-balance, and will know they cannot rely on the conventions of genre for the rest of the story. Make the criminal an early suspect who is released because the sleuth suspects someone else. Make the cops human, so they make errors of judgement, and incorrect assumptions. Make a character appear innocent, reveal that they've lied, or are hiding something, but then reveal that it's unrelated to the case. (The Killing used this to great effect.) Have other suspects who are guilty of something, but not what we think they're guilty of. Create a situation like "the fog of mystery" for the reader, so they cannot rely on what they know from previous works. Of course, that means you yourself need to be knowledgeable about the genre, and that means a lot of reading!

share|improve this answer
    
I like your new edit; I think that's a good way to go. I'm still not sure where to take that, though - I don't see how I can play with reader expectations, and have them know that I'm doing so, early enough for it to make a difference to the experience of the majority of the story. –  Standback May 6 '11 at 12:32
    
Well, that's the point though: they don't know you're playing with them except when you take the bottom out from under them and they realise they were completely wrong. Do it early enough in the story, and it'll leave them second-guessing everything they thought they knew. –  Craig Sefton May 6 '11 at 13:42

In Role Play we call this "meta-gaming": analysing information as if it is slave to convention as opposed to being part of a fresh and thrilling new narrative.

There are two methods of combating cynical audience members who employ this tactic.

  1. Write the words "All Bets Are Off" on a piece of paper. Affix this somewhere you will see it whenever you pause for thought when writing. Never forget it. There are things that you think cannot happen in your story, people who will not die, people who are above suspicion, events that won't happen. There are a few that are sensible and, for some reason, are the first targets of idiot writers trying to be "edgy" (see "murderer and detective are the same character" for details, not existential, just gratingly moronic). Other than these few fairly obvious aspects, the subverting of which are most often a betrayal of a good reader not a clever twist for a meta-reader, there should be no safe place. As a writer if you don't think: "How the hell am I going to work that?" then how are your audience going to?

  2. Introduce a randomiser. Dice, cards, a tombola, anything. Let chance tell you who a murderer is, what a plot twist will be. Live with the advice of decisions made by blind chance, not too many that your whole story becomes arbitrary, just enough that it is prevented from going stale. If you don't find it difficult to manage to incorporate a stunning twist that Dame Fortune has insisted you incorporate the chances are the stunning twist is predictable. People will not see coming what you didn't see coming, and vice versa.

2 is really a technique for extending 1. So the tldr of the whole thing is:

ALL BETS ARE OFF.

Once the reader knows this they will stop trying to be cute and just read the goddamn story.

EDIT:

Okay, I've been asked for an example of how to do this so that the reader knows that you're not messing around.

The example that springs to mind came in an early episode of one season of 24 (years since I watched the show so can't remember even which season, I think it was 4). There was some extremely high up CIA dude who was making things difficult in CTU and a terrorist demanded that he have a bag put on his head, for him to be brought to a location and for Bauer to shoot the guy in the head. Obviously, I didn't believe that JB would be allowed to shoot some innocent guy in the head at the behest of a terrorist but, no, he did it and the guy resolutely remained dead afterwards.

At that point I no longer trusted my own ability to pick the writing team's choices, so I gave up and just enjoyed the thing.

You have to, essentially, grab the reader early on with some side example like this (the whole "shoot this dude in the head" bit was a sub-plot anyway) and show them in no uncertain terms that anything could happen. If you make an early point where you bait-and-switch the reader by signposting one thing and delivering another then they will realise pretty quick that any guesses they might make about the conclusion of your opus are just as pointless.

The key is to make readers think, right from the get go: "This dude's just crazy enough to do it, too!"

--END EDIT

FURTHER EDIT

To make this specific to a whodunit.

The standard plot structure of a whodunit is:

a) murder b) investigation c) revelation

One way to circumvent the expectation, particularly in a debut, would be to do something weird like.

a) murder b) quick and obvious investigation c) murder of the murderer d) investigation e) revelation

So the initial murder itself is just an easily solved, by the numbers piece of plodwork. The real investigation is who murdered the murderer? This has the beauty of offering plenty of depth to the scope of the investigation. Is this murder some kind of revenge for the first murder? Or is there another issue at hand?

It doesn't even require you to have an idea about making the first murderer sympathetic in some way. If they are people could be shocked that some "nice guy" committed the dirty deed. If the original murderer's a nasty piece of work, however, then it adds a layer of conflict for the detective who wants to solve the murder but cares little for the victim.

--END EDIT

share|improve this answer
    
This struck a chord with me; that's definitely the type of message I want to convey to the reader. I also remember how powerful "ALL BETS ARE OFF" moments have been in other books I've read. The thing is, you've given suggestions on working by that credo, but how to I convey to the reader that metagaming isn't going to crack this case? What "ALL BETS ARE OFF" moments are possible in the early stages of a whodunit? –  Standback May 6 '11 at 12:35
    
@Standback: See edit. –  One Monkey May 6 '11 at 12:50
    
The most shocking "all bets are off" moment I've encountered is in LOST, when (spoilers?) John Locke knifes Naomi in the back in a season finale. Up to that point the viewer had been taught that Locke was inherently good—like his namesake espoused—and to see otherwise shattered any notions I had of the character and of the directions I thought the show would take. –  kevboh May 6 '11 at 13:19

He who lives by the trope, dies by the trope. If your readers are trying to use metaplot to solve your mystery, they're setting themselves up for a lot of toying with, if you choose to alter the rules they're familiar with.

The first step to playing with tropes and conventions is to become a master of them yourself. Make a list of mystery tropes you are familiar with and have a good look at it, then have a look through the Internet for more (TVTropes.org is a good starting point, but a word of warning: it's incredibly addictive).

Tempting as it may be to break every rule out there, it is simply another path to predictability: the attentive reader will quickly notice that they are reading an "inverse mystery story" and adjust accordingly. Subvert some of the conventions, but play others straight - it will keep your readers on their toes. If you write more stories, change things around so your readers never know what to expect. If after writing several "mixed-up" mysteries you then write one that exploits every trope in the book, you might actually surprise them.

Moreover, it's worth keeping in mind that real-life mysteries aren't anything like literary ones. Terry Pratchett makes this point in one of the City Watch books from the Discworld series (can't recall which one though, anyone remember?) For every single interpretation of a clue, there are countless equally valid interpretations. When you eliminate the impossible, you aren't left with one thing, "no matter how improbable"; you are left with a maze of possibilities of varying probability.

If your story takes place in a world where mystery stories exist, it's likely that your characters will also be genre savvy to an extent. Thus, your detective might point out (in coversation or as part of his musings) that "real world" mysteries (that is, ones he usually deals with) are nothing like the books - but make sure that they really aren't, unless you plan for them to express incredulity at the pat nature of the latest case (hey, stranger things have happened). On the other hand, the wrongdoer might try to play some tropes to their advantage, counting on life resembling literature.

Finally, remember that for a mystery story to be "fair", the only requirement is for the reader to receive information at the same time as the detective (no dei ex machinae) and for the evidence to logically support the verdict.

Addendum edit:

The best way to stop genre-savvy readers from meta-thinking is to send them strong signals that it won't work from the outset. Target a few common tropes and show that they aren't used here. Perhaps it will quickly turn out that Little Miss Innocuous has a watertight alibi. The "apparent-red-herring-but-actually-a-vital-clue" might be shown to be a red herring after all. The evidence against the first person suspected might accumulate, but then some of it might be shown to be false, only for new evidence to appear later.

Once the reader realises that they aren't reading a by-the-numbers mystery story, they'll start paying more attention to what was actually written.

share|improve this answer
    
Subvert the trope of the Super-Observant Detective. For example, the detective thinks that because the suspect has a ring indentation on his finger but no ring, he and his wife are at odds, or the man is cheating. And it turns out that the guy is having his ring cleaned, or resized, or it got lost, or he forgot to put it on that morning. –  Lauren Ipsum May 6 '11 at 23:55

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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