How to write a fair-play mystery
Fair-play mystery promises the reader every clue the detective gets. Knox's ten rules, means, motive and opportunity, and a whodunit that survives a re-read.
Published 26 Sept 2026 · 5 min read
A fair-play mystery is a wager with the reader: every clue the detective sees, the reader sees, and the solution, when it comes, must be derivable from what was on the page. The genre reference calls the whodunit highly codified, and its code is one of the oldest still enforced by readers: Ronald Knox's ten rules for detective fiction, plus the working requirement that every suspect needs credible means, motive and opportunity.
What fair play actually promises
The promise is not that the reader will solve it. Most readers do not, and do not want to; they want to lose fairly. The promise is that they could have won: the solution was available all along, and the detective simply got there first. Every convention of the genre exists to keep that true.
The sin of the genre is the unearned reveal: a culprit who first appears in the final chapter, a clue mentioned for the first time in the denouement, a hunch that happens to be right. Readers forgive a murderer they missed. They do not forgive evidence they were never shown.
The contrast is worth naming because the labels get crossed. Hardboiled and noir traditions trade the puzzle for atmosphere, moral compromise and a corrupt world; their questions are different ones, closer to whether truth matters at all than to who did it. Those are fine traditions. The failure mode is promising a fair-play whodunit and delivering noir with a corpse in it.
Knox's Decalogue
The rules were codified in 1929 by Ronald Knox, and his ten commandments, as Wikipedia's article on the Decalogue presents them, paraphrased faithfully:
- The culprit must appear early in the story, and must never be a character whose thoughts the reader has access to.
- "All supernatural or preternatural agencies are ruled out as a matter of course."
- No more than one secret room or hidden passage.
- No hitherto undiscovered poisons, and no appliance that needs a long scientific explanation at the end.
- "No Chinaman must figure in the story." The article is explicit that this rule targeted the anti-Chinese stereotyping common in 1920s Western fiction: Knox was telling readers to avoid stories that leaned on the stereotyped exotic villain, in language of its era that modern editions do not soften.
- No accident may ever help the detective, and no intuition that turns out to be right without evidence.
- The detective must not himself turn out to be the culprit.
- The detective must declare every clue he comes across.
- The detective's companion (the Watson) must conceal nothing from the reader, and his intellect should be slightly below the average reader's.
- Twins and doubles are forbidden unless the reader has been properly prepared for them.
The same article records two pieces of context worth keeping. The Detection Club, founded in 1930, had its members agree to follow Knox's commandments: this was a professional covenant, not one critic's list. And America produced a parallel code, S. S. Van Dine's "Twenty Rules for Writing Detective Stories" (The American Magazine, September 1928), aimed at the same pulp bad habits from the other shore.
What the rules still teach
Read as history, the Decalogue is a list of the cheating habits of 1920s pulp. Read as craft, it distills to four principles that survive any era:
- Clue transparency. The detective's evidence and the reader's evidence are the same set (rules 1, 8, 9).
- Coincidence discipline. Luck may complicate, never solve (rule 6).
- No cheap disguises. Secret passages, instant twins and unexplained science are withheld information wearing a costume (rules 3, 4, 10).
- A naturalistic world. However baroque the plot, causes stay ordinary ones (rules 2, 5).
Modern practice treats the ten as strong defaults rather than laws, and plenty of excellent mysteries break one deliberately, on the page, where the reader can see the rule bend. What has no modern license is breaking them silently.
Building a puzzle that survives a re-read
- Write the solution first. Then write backwards, deciding for each clue where it hides.
- Give every suspect means, motive and opportunity. A suspect missing one of the three is eliminated by it; the investigation is the process of striking names off, and it only works if each elimination is earned.
- Plant each important clue twice. Once where it reads as noise, once where it nearly tips the reader. Two plantings is the difference between "I never had a chance" and "it was there all along".
- Give every clue a harmless first reading. The reader must be able to misread it honestly; a clue that can only mean one thing is a spoiler.
- Test the reveal against re-reading. The ideal ending sends the reader back to an early chapter where the whole solution was visible. If nothing is visible on the second pass, the puzzle was never fair.
Fair play in serial form
A serialized whodunit has an extra hazard: weeks may pass between a clue's publication and its payoff, and readers will not remember the detail you buried in chapter eleven. The discipline is that cliffhangers may withhold the answer but never the clue: if a chapter ends on a discovery, show what was discovered (see cliffhangers and chapter endings). A steady release schedule helps too, because deduction at a distance needs fresh context.
The clue ledger
Fair play is bookkeeping before it is cleverness. Every clue needs a record: what it is, where it appeared, which characters know it, and when the reader saw it. A detective acting on knowledge she has not earned yet, or a clue the reader saw but the files say nobody found, breaks the contract as surely as an unshown murder weapon. Treat the ledger as part of your story bible, and audit it as you would any continuity error.
How Kanonara approaches this
Facts and who-knew-what state are records with history in Kanonara, and Story Doctor checks new chapters against them, so a detective reasoning from a clue the story says she never received surfaces as an alert with the passages on both sides. See how Kanonara works.
Sources
Craft claims above trace to these sources. Where a point is practitioner consensus rather than settled fact, the text says so.