A Nabner book punishes a bad file in a way no sample puzzle ever shows. Your buyer judges puzzle forty-one, when a line your key swears is sound runs a 3 at one end and a 4 at the other — two cells that never touch, in different boxes, that a classic checker and even an adjacency checker wave through without a murmur, and that your book's one extra law forbids absolutely, because on a nabner line no two digits anywhere on the line may be consecutive, however far apart they stand — or when a line repeats a digit its rules page swore could never repeat, because the generator was built from a summary that quietly allowed repeats and nobody re-read the standard — or when a six-cell gold line sits in a puzzle whose law makes six cells impossible, so the book is false before the buyer places a single digit. Nabner Sudoku, in the form this book sells, is a classic frame with a law of pure spacing: a 9 by 9 grid, digits 1 to 9 once in every row, column and outlined 3 by 3 box, crossed by gold lines, and on every line the digits never repeat and no pair of them, anywhere on the line, differs by exactly one. Underneath it is a contract in five parts, each part exact. First, the row rule. Second, the column rule. Third, the box rule. Fourth, the nabner rule: read the line as a set, not a sequence — no digit appears twice, and no two digits in the set are consecutive numbers, whether they sit side by side or at opposite ends. Fifth, closing the contract, given integrity with uniqueness: the givens are fixed exactly as published, and exactly one completed grid satisfies every part at once, reachable by deduction alone, with no guessing. A puzzle that breaks any part is not slightly harder. It is false as stated. A good Nabner book is a verified stack of sets: every line's digits re-read as a set, on the final file, puzzle by puzzle.
Four warnings belong at the front, because they are the mistakes this genre makes most often, and all four start with puzzles Nabner is not. Nabner is not Renban — it is Renban read backwards, and the name is the warning: a renban line holds a set of consecutive digits with no repeats, every digit shoulder to shoulder with the next, and a nabner line holds a set in which no two digits are consecutive at all; they are exact opposites sharing one clause, the no-repeat clause, and a checker, a generator or a rules page that confuses the two produces a book that is wrong on nearly every line while passing its own audit, because a set that satisfies one law can satisfy the other only when it has a single digit. Nabner is not Non-Consecutive Sudoku: that variant bans consecutive digits only in cells that touch edge to edge, and says nothing about two cells five apart on a line of gold — a non-consecutive grid can run a 3 and a 4 at the two ends of a long diagonal of empty cells quite legally, and a nabner line cannot, so a solver or checker built for the touching rule certifies books the gold law condemns. Nabner is not a German Whisper, though both laws push digits apart: a whisper constrains neighbours, demanding adjacent digits differ by at least 5, and says nothing about cells that do not touch — a whisper line can hold a 1 near one end and a 2 near the other, far apart and perfectly whispered, and it fails nabner on the spot. And nabner, in its standard form, does not allow repeats: some summaries of the gold line say digits may repeat if the other rules allow, and setters will find that sentence if they go looking; the standard rule as the puzzle portals state it — digits on a nabner line cannot repeat and must not include any pair of consecutive digits — forbids repeats outright, and a book generated under the repeat-allowed reading is a different, weaker variant that must not be sold under the standard name. A buyer who has been taught any of these four impostors by a careless rules page will solve your book under the wrong law and write the review your book did not earn.
The book is every pair on every line, not the sample puzzle
A sample puzzle proves almost nothing about a Nabner book, because the law lives in the pairs, and a line carries far more pairs than it carries digits. A line of five cells holds ten pairs; every one of those ten pairs is a separate promise — no equality, no difference of one — and the promises bind cells that never touch, in different rows, columns and boxes, so a single misplaced digit can break pairs against every other digit on its line at once while leaving every house in the grid perfectly legal. That is the shape of the buyer's experience: they do not solve your sample, they solve your fortieth line at midnight, pencil in hand, checking a pair your file may never have checked. When the pair fails, the buyer does not think the puzzle is hard. They think the book is careless, and they are right, because the pair is not an edge case of the product. The pair is the product, repeated hundreds of times per book.
So the unit of quality in this genre is not the puzzle and it is certainly not the listing. It is the line read as a set, counted and verified: every line, every digit, every pair, against the solved key and against the printed page. Everything below — the contract, the audit, the difficulty scale, the print rules — is machinery for keeping that one promise at book scale, because a book is where spacing stops being a clever constraint and starts being a manufacturing problem: hundreds of lines, thousands of pairs, one standard.
The contract, part by part
Part one: the row rule. Every row of the 9 by 9 grid holds the digits 1 to 9 exactly once. No gold line, however elegantly spaced, excuses a doubled digit in a row, and no buyer forgives it: the classic frame is the floor the whole book stands on.
Part two: the column rule. Every column holds the digits 1 to 9 exactly once. A line that runs down a column shares cells with that column's own law, and the two laws agree on repeats inside the column and part company everywhere else — the column cannot see the pair formed by two cells of the line that sit in different columns, and the checker that leans on the column to police the line will miss exactly the pairs the line exists to forbid.
Part three: the box rule. Every outlined 3 by 3 box holds the digits 1 to 9 exactly once. Long lines wander through several boxes, and most of the pairs a nabner line forbids are pairs no box ever contains together; the box rule is necessary, it is nowhere near sufficient, and a book audited only house by house has not begun to audit its lines.
Part four: the nabner rule. Take the digits standing on one gold line, forget their order, and read them as a set. Two demands fall on that set, and both are total. The digits cannot repeat: each digit appears on the line at most once, wherever the classic rules would have permitted a second copy in another house. And the set must include no pair of consecutive digits: for any two digits on the line, the difference is never exactly one — not for neighbours on the line, not for the two end cells, not for any pair in between. Four consequences follow, and a setter who cannot recite them is not ready to generate a book. First, every digit on the line bans three values from every other cell of the line: itself, the digit below it and the digit above it — a placed 5 clears 4, 5 and 6 from the whole line — except at the edges, where 1 bans only 1 and 2, and 9 bans only 8 and 9, which is why edge digits are the line's cheapest tenants. Second, sorted in order, the set climbs in steps of at least two from first digit to last; a sorted gap of one anywhere in the set is a broken line, full stop. Third, length is capped: no line can hold more than five digits, because five is the most numbers from 1 to 9 that can stand pairwise non-consecutive — and a line of exactly five cells has its set forced completely, the odd digits 1, 3, 5, 7 and 9, the one and only five-digit set the law allows. A six-cell gold line in a standard book is not a hard puzzle; it is a printing error with a cover. Fourth, a line of four cells, whose set is not fully forced, is still tightly fenced: its four digits must include one from 1 to 3, one from 3 to 5, one from 5 to 7 and one from 7 to 9, so a four-line whose digits huddle at one end of the number line — 1, 3, 5 and a fourth digit below 7 — is impossible on its face, before any solving begins.
Part five: given integrity and uniqueness. The givens printed in the book are the givens the solver proved unique — no digit added for looks, none removed for elegance after the proof. Exactly one completed grid satisfies all five parts at once, with the nabner law switched on, and that grid is reachable by deduction alone. The control matters as much as the proof: run the same givens with the gold lines ignored and the classic frame will often admit the intended grid among several, which is fine and expected — the lines are load-bearing — but a grid that stays unique only while a checker forgets to forbid repeats, or forgets the pairs that do not touch, is a grid whose proof was run under a weaker law than the rules page promises, and its second solution is waiting for a buyer with a pencil.
Verify like a buyer: the eight-step audit
Run this audit on the final exported file — the file the buyer downloads, not the generator's working grid — and keep the record per puzzle. A buyer checks pairs. So does this audit, in an order that catches the cheap failures before the expensive ones.
Step one: line inventory. List every gold line in the puzzle: how many lines, how many cells each, which cells in reading order. Count the cells twice, once from the puzzle page and once from the key. Any line longer than five cells fails here, instantly and without appeal — the law allows at most five — and any line whose cells on the page differ from the cells in your working file fails with it. Most book-scale errors are born at this step and discovered by buyers because nobody ran it.
Step two: set extraction. For each line, read its digits off the solved key and write them as a set, order discarded. This is the object the law actually governs. An audit that works along the line cell by cell, neighbour by neighbour, is auditing a whisper, not a nabner, and will pass pairs it never looked at.
Step three: repeat scan. Check the set for a doubled digit. Any repeat fails the line under the standard law — and if your generator produced the repeat deliberately, under the summary that allows repeats, the failure is not the line's; it is the book's, and the fix is a new generator setting and a regenerated book, not a footnote.
Step four: pair scan. Sort the set and walk the sorted digits once, checking every neighbouring gap in the sorted order: each gap must be at least two. Sorting is the whole trick — consecutive digits in value are neighbours in the sorted set wherever they stood on the line, so one sorted walk inspects every pair the law forbids, including the two end cells a cell-by-cell audit forgets. A gap of exactly one, anywhere, fails the line.
Step five: length and band check. A five-cell line must hold exactly 1, 3, 5, 7 and 9 — compare it digit for digit. A four-cell line must place a digit in each of the bands 1 to 3, 3 to 5, 5 to 7 and 7 to 9. A line that passes the pair scan but fails its band is a line whose set you misread; go back to step two, because one of the two readings is wrong, and the buyer's pencil will find which.
Step six: reverse recount. Close the key, open the printed puzzle page, and recount every house and every line from the page alone, sets re-extracted from scratch. Then compare against step two's sets, digit for digit. The reverse pass exists to catch the export step — the shifted line, the stale key, the puzzle page redrawn from an older grid — and it only works if it is genuinely run from the page, not from memory of the file.
Step seven: uniqueness, law on, plus the impostor controls. Prove the puzzle solves uniquely with the nabner law enforced in full — no repeats, no consecutive pair anywhere. Then run two controls that protect you from your own toolchain: once with repeats allowed, the weaker reading, and once checking only touching pairs, the non-consecutive reading. If the solution changes under either control, your main proof was only as strong as the law your solver actually implemented, and the settings — not the puzzle — are what need fixing before the book ships.
Step eight: the print proof. Print one full page at final size, in greyscale, on the cheapest paper your buyers plausibly own. Gold is a treacherous colour: it prints pale, it fades into cream paper, and a line a buyer cannot see is a law they cannot use. Every line must stay traceable from end to end, distinguishable from the grid, from any renban purple and from each other where lines run close, by weight or pattern and not by colour alone. If the gold vanishes, the book fails step eight no matter how clean its sets are.
Difficulty is spacing work, not given counts
Buyers forgive a hard Nabner book. They do not forgive a dishonest one — a book labelled gentle whose third puzzle walls them for an hour, or a book labelled fierce that folds in ten minutes. In this genre the given count is the weakest lever on the shelf, because the law's power concentrates in the sets, not in the clues. Grade difficulty by the spacing work the solve actually demands, and name the levers in the listing.
The first lever is set narrowing: how far the lines alone carry the solve. A line of five cells announces its entire set before a single digit is placed — five known digits in unknown order — and a book of long lines plays as a game of placing known sets into houses, deep and satisfying but rarely brutal. Short lines say less per line and demand more classic work between them. The second lever is the edge-digit share: givens and early deductions landing on 1s and 9s clear only two values from their lines, while middle digits clear three, so a puzzle whose lines are seeded with middle digits tightens far faster than its given count suggests. The third lever is pair pressure across houses: how often the sorted-gap law, not the classic frame, delivers the decisive elimination — count it in your own test solves, because that count, honestly taken, is the difference between a Nabner book and a classic book wearing gold. The weakest signal, as always, is the raw number of givens, and a listing that grades by it alone is grading by the one number the buyer cannot feel.
So test-solve the whole book, in order, timed, from the printed pages — not from the generator's own estimates, which know the solution too well to judge the search. Grade in bands a buyer can plan an evening around, state the levers you used on the listing page, and let the hardest puzzle in the book set the label: one wall puzzle in a gentle book earns the review, and it will be quoted back at you by name and number.
Print honesty: lines a buyer can actually follow
A Nabner book sells a set law printed as a line, and the buyer's whole experience of the law runs through the drawing of that line. Print is not decoration here; it is the rule made visible, and it has failure modes a screen proof will never show you.
The line must read as one object. A gold line is a set wearing a path: the buyer must be able to see, at a glance and in bad light, exactly which cells belong to it — where it starts, where it ends, and every cell between — because a cell wrongly included invents pairs the law never made, and a cell wrongly excluded deletes them. Draw lines with a weight and, where your palette allows, a dash or border treatment that survives greyscale printing; gold on white, printed cheaply, is a rumour, not a line. Keep line colour out of the rules text as the only identifier — say gold line and describe the drawing, so a buyer printing in black and white loses nothing but the colour. Where two lines pass near each other, separate them by geometry, not by shades of a colour that prints as one grey. Puzzle page and answer key must draw the identical lines over the identical grids; a key that redraws lines from a second export is a second book, unaudited, hiding at the back of the first. And give every line's cells room: cramped grids shrink digits, shrink lines into the ruling, and turn the pair-checking the genre lives on into eyesight work the buyer never agreed to do.
Licensing, provenance, and whose puzzles these are
A Nabner book is easy to assemble dishonestly and hard to assemble well, and the difference is records. Variant sudoku has a living community of setters whose puzzles circulate in packs, portals and streams, with the gold line's standard wording attached. A puzzle lifted from a portal, re-drawn in your template and sold under your cover is not made yours by the re-drawing, and this community is small enough — and archives its work carefully enough — that provenance is not a courtesy. It is a survival trait.
Keep three records per puzzle and the book can defend itself. The generation record: how this grid and these lines were produced, by what tool, under what constraint settings — including, in writing, which nabner reading the generator enforced, the standard no-repeat law or the weaker summary — because that one setting is where this genre's books quietly diverge. The rights record: written licence or your own authorship, in a form you could show a marketplace without embarrassment. The audit record: the eight steps, dated, against the final file, with the sorted sets kept so a disputed line can be re-walked digit by digit. When a marketplace asks where a book came from — and marketplaces increasingly ask — the seller with records answers in minutes. The seller without them discovers that "my generator made it" is a claim, not a record.
Corrected files, errata, and the delivery rule. Every puzzle book ships with a mistake eventually; the genre only decides whether the mistake becomes a one-star review or a five-star story about a seller who fixed it. The rule is simple and it is not negotiable: when any pair in any puzzle is wrong — a sorted gap of one in a key, a repeated digit on a gold line, a six-cell line that should never have shipped — every buyer gets the corrected file, not just the buyer who complained, and the correction goes out through the same delivery that sold the book, so it lands in the same library, the same downloads page, the same inbox thread the original came from. This is where delivery stops being plumbing and becomes part of the product: a download tool that versions your file and re-delivers the corrected edition to past buyers turns an errata notice from a hope into a fact, and it is why I deliver my own puzzle books through FileFreight — one flat monthly price for download delivery and license keys, the corrected file replaces the old one for every buyer at once, and nobody has to write in and ask. Pair the corrected file with a plain errata note inside it: puzzle number, what was wrong, what changed, in words a buyer can check against their printout. Never silently swap a file a buyer may already have printed and half-solved; a silent fix gaslights the exact buyer who found your error, and they are the buyer your next review is made of. Keep the old version archived, keep the note dated, and let the record show the book getting truer over time.
The launch checklist
Before the listing goes live, walk this list in order and do not skip backward. One: the rules page declares the law — 9 by 9 grid, digits 1 to 9, gold lines whose digits never repeat and include no pair of consecutive digits anywhere on the line — and names the neighbouring variants, renban lines above all, only to exclude them. Two: every puzzle has passed the eight-step audit on the final exported file, with a written per-puzzle record and its sorted sets kept. Three: no line in any key repeats a digit or holds a sorted gap of one, verified set by set, line by line, and no line runs longer than five cells. Four: every five-cell line holds exactly the odd digits, every four-cell line touches all four bands, and forward and reverse recounts agree on every house and every set. Five: difficulty is graded from real solves, and the listing names the levers used. Six: puzzle pages and keys present the identical lines, and every gold line survives printing at final size in greyscale. Seven: all puzzles are your own generation or licensed, records kept, and the generator's nabner setting is written down. A book that can say all seven, truthfully, has earned its listing copy — and the copy should then say the plainest true thing about it: these sets were re-read, one by one, twice.
Top comments (0)