Fair play mystery rules are a promise: every clue the detective uses to name the culprit appears in the text first, in the same order the reader sees it. Break that promise and readers don’t feel surprised, they feel cheated. This guide shows how the rules work in practice, what a fair cast and clue set look like, and how to test your own draft before anyone else does.
Most of the conventions come from two short documents written almost a century ago. Ronald Knox set out ten commandments for detective fiction in the late 1920s, and S. S. Van Dine published twenty rules for writing detective stories in 1928. Both were responses to the same thing: a genre that had drifted into magic and coincidence dressed as genius.
Below is the working version of all of it, in the order I would use it if I were plotting from scratch.
Table of Contents
- What Are Fair Play Mystery Rules?
- The five elements every mystery needs
- How Do Fair Play Mystery Rules Work in Practice?
- The six moves that make a plot fair
- Ronald Knox’s Ten Commandments of Detective Fiction
- What Clues Should a Fair Play Mystery Include?
- How Many Suspects and Red Herrings Are Enough?
- How Do You Reveal the Solution Without Making It Obvious?
- Common Fair Play Mystery Mistakes to Avoid
- Frequently Asked Questions
- Do all fair play mysteries have to be completely predictable?
- What is the difference between a clue and a red herring?
- Should the reader know everything the detective knows?
- How many clues does a fair play mystery need?
- Can a mystery have an unreliable narrator and still be fair play?
- How do I make a fair play twist satisfying?
- Conclusion
What Are Fair Play Mystery Rules?

Fair play is the agreement that the reader gets a fair chance at the solution. Nothing important is held back until the last chapter, no deduction depends on a fact the reader never saw, and no coincidence does the detective’s work. The answer has to be sitting in the book the whole time, waiting for somebody to assemble it.
That is the whole contract, and it is narrower than people assume. Fair does not mean easy. A fair play mystery can be so densely plotted that most readers get the twist wrong on the first pass and have to go back through the chapter looking for the moment that gave it away. Hard is fine. Hidden is not.
The instinct behind the rules goes back a long way. Edgar Allan Poe’s 1841 tale of the murder in the Paris apartment helped invent the closed-circle cast, and G. K. Chesterton pushed hardest for the idea that the detective should win by reasoning rather than by rank or psychic talent. The Detection Club, formed in 1930 by writers including Dorothy L. Sayers and Agatha Christie, took the conversation seriously enough to debate it as a craft problem rather than a matter of taste.
The five elements every mystery needs
- A crime or disappearance. Something concrete has changed and it shouldn’t have.
- A central question. Who did it, and how did they manage it?
- A cast of suspects. People with motives, access and secrets.
- Clues in the open. Declared to the reader, not merely implied.
- A solution that accounts for the evidence. Every loose thread has to close.
Fairness lives mostly in the fourth item, but it fails at the fifth. You can show the reader every clue and still cheat if the answer only works because you invent a fact in the denouement.
How Do Fair Play Mystery Rules Work in Practice?

Fair play is not a mood you sprinkle over a plot at the end. It is a sequence of moves, and doing them in the right order is most of the work.
The six moves that make a plot fair
- Fix the solution first. Know the answer, the method and the motive before you write a single scene. A plot built backwards is far easier to keep honest.
- Name the central question. One question drives the book. Everything else is atmosphere or character work.
- Introduce the cast. Give each suspect a motive, a means and a reason to lie, then put all of them in front of the reader early.
- Plant the clues while planting the story. A clue should arrive inside a scene that has a reason to exist, not as a detail the writer shoved in to justify the ending.
- Build genuine alternatives. At least two other suspects should have enough evidence against them to look guilty for real.
- Resolve by deduction. The detective names the chain of reasoning, ties each link to a specific earlier moment, and shows why every other suspect fails.
Ronald Knox wrote his ten commandments for the Detection Club in 1928, and they were endorsed the following year. They are still the shortest usable summary of what a fair plot has to do.
Ronald Knox’s Ten Commandments of Detective Fiction
- The criminal must be introduced early enough for the reader to understand the crime.
- All clues must be declared to the reader, in a way that keeps them from being obvious.
- No unaccountable intuition: the detective may not solve the case on a hunch he cannot explain.
- No accident or coincidence may help the detective, and no accident may create the situation.
- No secret room or passage may be used, because that hides evidence rather than explaining it.
- No unestablished supernatural or preternatural agency may be used to solve the crime.
- That same prohibition applies to a newly discovered poison or scientific method, which was the 1928 version of the same cheat.
- The detective must be given a Watson figure, someone who stands beside him and narrates the case so the reader sees what the reader needs to see.
- No suspect may be allowed to speak or act solely to delay or confuse the reader.
- No twin brothers or doubles may appear, because the identity switch is a surprise with no evidence behind it.
Rules seven through ten read oddly now, and that is useful. Knox was writing into a world where locked rooms, miracle cures and doubles were live publishing gambles. The reasons are gone, but the underlying principle survives: nothing the reader couldn’t have derived from the text gets to do any work at all.
What Clues Should a Fair Play Mystery Include?
Not every interesting detail is a clue, and treating them the same way is how drafts turn muddy. Four categories do different jobs, and you need at least two of the first kind.
Essential clues are the facts the solution depends on. Remove one and the answer stops working. These are non-negotiable and they must appear in the text in a form the reader can actually register.
Corroborating details support an essential clue without carrying it alone. A wet coat, a receipt with the wrong time on it, a window latch that catches. These give the reader the satisfaction of a pattern clicking and they help a clue survive being noticed twice.
Red herrings are honest misdirection. They point at a suspect, a motive or a method that turns out to be wrong, and they are fully present in the text. The difference from a cheat is simple: a red herring is something the reader could have discounted with more thought, while a cheat is something the reader could not possibly have known.
Atmospheric information establishes place, mood and character. It is not evidence and should not be quietly promoted to evidence in the final chapter. If a setting detail becomes load-bearing at the reveal, plant it more deliberately or cut it from the ending.
To make a clue noticeable without making it obvious, give it a reason to be unremarkable. A clue about wet soil matters if a character mentions rain in passing two chapters earlier. A clue about a rope matters if someone uses rope earlier in an unrelated task. Readers store details without flagging them, which is exactly the effect you want.
Then check coverage. Two or three essential clues pointing at the same conclusion is the classic construction, and the strongest version gives the reader one clue that fits the culprit and a second that rules the alternatives out.
How Many Suspects and Red Herrings Are Enough?
Enough to create real alternatives, no more. The test is whether each suspect could be convicted on the evidence alone. If a suspect has no motive, no access and nothing to hide, they are not a red herring, they are furniture, and readers notice the difference.
Four to six suspects works for most closed-circle plots. Below four, the solution is often obvious by process of elimination. Above six, readers start skimming names and your careful clue work stops landing.
Red herrings follow a similar curve. Two or three well-planted ones create a genuinely contested middle. Five or six, and the reader assumes every suspicious detail is a lie, which drains the tension from the real clues.
The hardest judgement call is the Watson figure, the companion who restates and interprets what the detective has found. Some readers find the device tiresome, and the complaint is fair. Its value is that it lets you reveal a clue honestly while explaining its significance later, which is exactly the work fair play needs done. A version of this character is nearly indispensable even when the narration is first person, because something has to carry the reader’s confusion at the moment the clue first appears.
A second debate is whether you can have two culprits. Knox and Van Dine both assumed a single criminal, and their reasoning is sound: a second offender halves the reasoning the reader has to do. Forum writers pushing back on the rule point out that accomplices can work if each one independently fits part of the evidence, and that the person who assembled the plan is still identifiable. That version tends to hold up. Distributed guilt where nobody can be named is where fairness collapses.
How Do You Reveal the Solution Without Making It Obvious?
A fair reveal recaps the chain, not the plot. The difference is easy to feel and hard to describe: the reader should be walked through the reasoning, not through a summary of what happened.
Go back to the planted clues one at a time, in the order you planted them, and name the chapter for each. Then explain what the detective knew and when he knew it. The standard is consistency of character knowledge: if your detective behaves in chapter four as though he saw something he only reads in chapter nineteen, the solution is not fair no matter how elegant it is.
Separate signal from noise out loud. Tell the reader which details were meaningless and why they were not just padding, because that is the moment a red herring confirms it was doing its job. It also earns you the right to have included it.
Finish with verification steps rather than an assertion. Give the reader two or three specific moments they can go back and read, and leave the verdict implied. Readers trust a solution they can check themselves, and they forgive an obvious answer far more readily than they forgive one they cannot verify.
For a television-shaped version, remember the rules are stricter when the audience knows something the detective does not. A show that reveals the culprit to viewers in the first ten minutes is not a fair play mystery in the literary sense, and the audience will say so. Check the Poirot and Knives Out adaptations against a puzzle standard and they hold up. Stack three reversals into a sixty-minute runtime and they do not.
Common Fair Play Mystery Mistakes to Avoid
These are the failures that come up again and again, and nearly all of them are fixable in revision.
Introducing a clue or a motive in the final chapters. This is the single complaint readers raise most often, and it is the one that feels worst. Fix: find the moment the reader needed that information and rewrite that scene earlier.
Detective intuition with no visible basis. The detective announces the answer and the writer assembles the reasoning afterwards. Fix: work out the deduction first, then decide which of your existing clues supports it. If nothing does, you need a new clue, not a longer speech.
Suspects with no meaningful motive. Fix: give every suspect on your grid a want that would survive the crime being exposed, and cut anyone who only exists to be suspected.
Red herrings that touch nothing. If removing a misleading detail changes nothing about the plot, it is decoration, and readers feel the padding. Fix: either connect it to the case so it has to be explained away, or delete it.
An ending that depends on a new fact. A secret twin, a hidden passage, a poison nobody had heard of, a letter that was never mentioned. Fix: the last chapter should contain no information the reader has not already met.
Twist stacking. Three reversals reset comprehension instead of deepening it, and the third one lands on a reader who has stopped trusting the book. Fix: keep one turn, and let the rest of the structure do the work.
One diagnostic exercise worth doing before you show a draft to anyone: read only the last three chapters, then list every piece of information the solution depends on and note the page where each first appears. Anything that first appears in those last three chapters is a cheat, no matter how small.
Frequently Asked Questions
Do all fair play mysteries have to be completely predictable?
No. Fair play requires only that every clue needed for the solution appears in the text before the reveal. It says nothing about how obvious those clues are. Plenty of fair play mysteries defeat most readers on the first attempt and only look obvious afterwards, which is why the genre tolerates complex plotting. Fair means derivable, not transparent.
What is the difference between a clue and a red herring?
A clue is a fact the solution genuinely depends on. A red herring is a fact that points somewhere wrong and is fully present in the text, so a careful reader could have dismissed it. The test is whether more thought would have ruled the red herring out. If the reader had no way to discount it, you have withheld information, which breaks fair play rather than playing with the reader’s expectations.
Should the reader know everything the detective knows?
Almost. Every fact the detective uses at the reveal should have reached the reader too, usually earlier. The usual exception is deliberate narrowing: you can withhold the detective’s interpretation of a detail while still showing the detail, and you can hold back a conclusion that only becomes obvious in hindsight. Withholding the raw fact is what breaks the rules.
How many clues does a fair play mystery need?
Two or three essential clues are enough for most plots, supported by a handful of corroborating details. What matters is coverage rather than quantity: the clues should point at the culprit, eliminate the alternatives, and explain the method. Ten clues that all restate the same fact add length without adding fairness, and a single clue the reader never noticed is worse than no clue at all.
Can a mystery have an unreliable narrator and still be fair play?
Yes, and it can be very strong, provided the unreliability is about interpretation rather than hidden facts. A narrator who misreads a clue, draws the wrong conclusion or describes their own state badly is still telling the reader everything. A narrator who lies outright about a physical fact, with no hint in the text, reintroduces exactly the withholding the rules were written to stop.
How do I make a fair play twist satisfying?
Satisfaction comes from recognition, not shock. The best twists reuse something the reader has already seen, ideally more than once, and reorganise it. Test the turn by asking whether a careful reader could have noticed it in advance. If the answer is no, the twist is a surprise, which is a different effect and usually a weaker one for this genre.
Conclusion
Fair play mystery rules come down to one principle: put every fact the solution needs in front of the reader, then make the deduction hard. Everything else, from Knox’s commandments to Van Dine’s twenty, is a longer or shorter statement of the same promise.
So start at the end. Decide who did it, how, and why, write down every fact that answer depends on, and then work backwards placing each one in a scene that has a reason to exist. Once the last chapter contains no information the reader has not already met, you have a fair play mystery. The rest is tightening, misdirection and deciding which red herrings the cast can plausibly generate themselves.


