If you’ve solved a few nonograms, you know the pattern: the first half of the grid falls fast, and then everything grinds to a halt. The overlap trick that got you started stops producing, and you’re left staring at rows that could go three different ways. This article covers the techniques that break those deadlocks — the ones experienced solvers lean on when 15×15 and bigger grids stop being polite. Fair warning: none of these are magic. They’re sharper versions of the same question you’ve been asking all along.

First, a Thirty-Second Recap

This article assumes you know the basics: the clue numbers describe runs of filled squares in that row or column, in order, with at least one blank square between runs, and the overlap technique gets your first squares on the board. If any of that is fuzzy, start with our complete nonogram guide and the step-by-step tutorial — this article picks up where those leave off.

Key point: Every technique below is the same question asked harder: given everything I know so far, where must each run sit — and where can’t it?

Edge Logic: Let the Walls Work for You

The edges of the grid — and every X you place — act like walls that squeeze a run’s options. Say a row’s first clue is a 5, and column work tells you the very first square is filled. That run is now locked to the wall: squares 1 through 5 fill in, and square 6 gets an X, because the run can’t be six long. One filled square just settled six.

It works in reverse too. If the smallest clue in a row is a 3, and there’s a two-square pocket trapped between the wall and an X, nothing fits in it — a 3 doesn’t squeeze into a two-wide gap. X out the whole pocket. In my experience this is the single most common move people miss on hard grids: they hunt for squares to fill and walk right past squares they could eliminate.

Splitting and Joining Runs

Mid-game, filled squares start appearing from crossing columns, and the real question becomes: which clue owns them? If a row’s only clue is a 4 and two filled squares sit with a single empty square between them, they can’t be separate runs — there’s only one run in the row. Join them: the square in the middle fills in. Now flip it. If the row’s clues are 1 1, no two filled squares are ever allowed to touch, so every square directly beside a known fill gets an X. Same evidence, opposite conclusions — the clue decides which way it goes.

Punctuate: Solve With X’s, Not Just Fills

Solvers call it punctuating: the moment a run is complete, cap both ends with X’s before you do anything else. It feels like busywork, because those X’s don’t “solve” anything in the row you’re looking at. But every X you place is a new wall for the column crossing through it, and edge logic feeds on walls. Most stalled grids I look at aren’t missing fills — they’re missing X’s.

The Ping-Pong Habit Every time you settle a square in a row, glance up its column before moving on — and vice versa. Hard grids don’t fall line by line; they fall in chain reactions. One X completes a column, the column fills a square three rows down, and that square unlocks a row you gave up on ten minutes ago.

Contradiction Testing (Careful With This One)

When pure deduction runs dry, pick a square with only two possible readings, assume one, and follow the consequences a few moves. If you hit an impossibility — a run that doesn’t fit, two runs forced to touch — the other reading was right, and that’s a real deduction, not a guess. I’ll be honest, though: on paper this gets messy fast, so keep the test short and pencil it lightly. And if you find yourself needing it on every other row, the puzzle probably isn’t the problem — you’re likely sitting on simpler deductions, usually missing X’s.

Putting It All Together: One Row, Start to Finish

Here’s a ten-square row with the clue 4 3, solved with nothing but the techniques above:

Clue: 4 3   (# = filled, X = blank, . = unknown)

1. Overlap gives three squares:  . . # # . . . # . .

2. Column work rules out squares 1 and 10. With square 1
   gone, the 4 must live inside squares 2–6, which adds
   square 5; the same squeeze on the right adds square 7:
   X . # # # . # # . X

3. Contradiction test: could the 4 sit at 3–6? Then square 6
   is filled, and the 3 would need squares 8–10 — but 10
   is an X. Impossible. So the 4 sits at 2–5.

4. Punctuate square 6, and the 3 has one home left:
   X # # # # X # # # X  — solved.

Notice nothing exotic happened. Overlap, edge logic, one short contradiction test, punctuation — each move fed the next.

The Takeaway

Edge logic, splitting and joining, punctuating, and the row-column ping-pong aren’t really separate skills — they’re one habit: squeeze every line for what it must contain and what it can’t, and write both answers down. Build that habit and grids that used to stall will fall a row at a time. And if you get well and truly stuck anyway, we’ve got an article for that too.