Nonogram strategy: advanced techniques

Once you know the rules, the real game is technique. These are the deduction patterns that turn slow trial-scanning into clean, confident solving — from overlap math to contradiction testing.

The overlap technique: free cells in every big clue

The most valuable pattern in nonograms, and the one to run first on every fresh line. For a single clue of length k in a line n cells wide, imagine the run pushed as far left as it can go (cells 1 through k) and as far right as it can go (cells n−k+1 through n). Any cell covered by both extremes is covered by every placement in between — so it's guaranteed filled before you know anything else about the line. The overlap exists whenever 2k > n, and it's exactly 2k − n cells in the center.

The multi-clue generalization is sometimes called simple boxes: pack all the clues as far left as they go, then as far right, and fill every cell where the same run covers both packings. A row clued 4 3 in 10 cells packs left as cells 1–4 and 6–8, and right as cells 3–6 and 8–10 — the first run forces cells 3–4 and the second forces cell 8, three free cells from a line you haven't otherwise touched.

Edge and border logic

Borders anchor runs. If cell 1 of a row is filled and the first clue is 4, cells 1–4 are filled and cell 5 gets an X — the run can only extend one way. The same logic applies from the right edge with the last clue, and, crucially, against any X you've placed: an X is a local border, and every technique that works at the grid edge works against it.

The softer version is often called glue: a filled cell near an edge drags its run inward even when it doesn't touch the border. If the first clue is 5 and there's a filled cell in column 3, the run must cover column 3, so it can start no later than column 3 — and every legal placement covers cells 3–5. Fill them.

Gap splitting and segment elimination

Every X you place carves a line into independent segments, and short segments die fast. If the smallest clue remaining for a line is 3, any open segment of length 1 or 2 can be X'd out entirely — nothing fits there. Re-check segment lengths every time a crossing column drops a new X into a row: a segment that used to be viable often just became too small, and eliminating it re-anchors the remaining clues.

X-marking discipline: negatives win puzzles

Beginners chase filled cells; strong solvers chase certainty in both directions. Every X you prove does three jobs at once: it removes possibilities from the crossing line, it creates a border for glue and edge logic, and it shrinks segments toward elimination. Make it a reflex — after every deduction, ask "which cells did this just prove empty?" A filled run that completes its clue, a segment that can't hold anything, a line whose clues are all placed: each one is a batch of X's waiting to be claimed, and each unclaimed X is information you're leaving on the table.

Punctuate completed runs

The moment a run reaches its full clue length, cap both ends with an X — runs can't grow, and those caps are immediate new borders for the crossing columns. When every clue in a line is accounted for, X out everything else in that line in one sweep. Punctuation is the cheapest deduction in the game and the one most often skipped.

Cross-reference rows and columns

Every cell belongs to two lines, so every mark you make is fresh evidence somewhere else. The efficient habit: when a row deduction fills or X's cells, immediately glance down each affected column before moving on. That's usually where the next domino falls — a column that was ambiguous a moment ago now has a glue anchor or a dead segment. Solvers who work line-by-line in isolation do twice the passes of solvers who follow the freshest cells.

The revisit loop: when to re-scan

Lines aren't solved once; they're revisited. A reliable rhythm: sweep all rows, sweep all columns, then return only to the lines that changed since you last looked at them. When a full sweep produces nothing, don't assume you're stuck — recompute your overlaps against the current state of the line. Overlap math tightens as segments shrink: a clue that gave you nothing on the full 15-wide line may force several cells inside the 9-wide segment it's now confined to.

Mistakes to avoid

• Guessing. The cardinal sin. One speculative fill poisons every deduction that crosses it, and you won't find out until the picture stops making sense twenty moves later. If a move isn't forced, don't make it.
• Solving by picture. Filling a cell because the emerging image "looks like it needs an ear there" is guessing in costume. The clues are the only authority.
• Forgetting clue order. Clues describe runs strictly left-to-right and top-to-bottom. A 2 5 line is not a 5 2 line.
• Postponing X's. Every deduction you make and don't record has to be re-derived later — usually at the exact moment you needed it as a border.

Harder techniques for expert grids

Line splitting. When X's have divided a line into segments, work out which clues can belong to which segment. If a clue fits only one segment, treat that segment as its own miniature line and run overlap math inside it. When the segments and clues pair off one-to-one, the whole line decomposes into small easy puzzles.

Contradiction testing. Pick an undecided cell, assume it's filled, and follow the forced consequences for a few steps. If they collide with a clue, the assumption was false — the cell is an X, proven. This is still deduction, not guessing (you never leave a speculative mark on the board), but it's slow, so keep it as a last resort. On PixDojo it's never required: every one of our 700+ hand-crafted puzzles is solvable start to finish with pure line logic — if you're reaching for contradiction testing, there's a simpler deduction you haven't spotted yet.

Where to practice

Technique only sticks with reps. If you're new, start with the how-to-play tutorial, then make the daily challenge your training ground — one fresh puzzle a day, with a streak to keep you honest. Everything plays free in your browser.