How to Solve Suguru: Region Sizes, the Neighbour Rule, and Cascades

Updated August 2026

Suguru — also called Tectonic — splits a grid into outlined regions, and a region of n cells holds the numbers 1 to n, each once. The second rule is the one that makes it interesting: no two identical numbers may touch, not even diagonally. Everything below comes from combining those two rules in the right order.

Start with the smallest regions

A one-cell region is a 1, immediately. A two-cell region holds 1 and 2 in some order, so both of its cells are restricted to two candidates before you do any work at all. Small regions carry the most information per cell, so solve outward from them. The largest regions — the fives — are the loosest and should be left until their neighbours have constrained them.

Once you write a 1 in a one-cell region, apply the neighbour rule at once: no 1 may appear in any of its up-to-eight surrounding cells. That is often enough to force a 1 elsewhere in an adjacent region, which forces another, and so on. The first minute of a Suguru is usually a cascade of 1s.

The neighbour rule works in both directions

Most people use it defensively: after placing a number, cross it off the neighbours. The offensive use is stronger. Look at a region and ask where a particular value can go. If every candidate cell for the 3 in a region touches a 3 that already exists, you have a contradiction; if all but one do, you have placed the 3. Working value by value across a region — "where can the 4 live here?" — finds moves that scanning cell by cell misses.

Pencil marks that stay useful

Write candidates only where they help. A five-cell region with nothing placed has all five candidates everywhere, which is noise. Start pencilling once a region has one or two numbers fixed, or when it is adjacent to a region that does. The cells worth marking first are those touching several already-placed numbers, because their candidate lists are short and shrink fast.

Look at cells between two regions

A cell on the boundary between regions is constrained by its own region's set and by every neighbour in the other region. Those shared borders are where contradictions surface earliest. When you are hunting for the next move, scan the seams rather than the interiors.

High numbers are rarer than they look

Only regions of size five can contain a 5, size four or five can contain a 4, and so on. So a 5 placed anywhere eliminates the possibility from a large neighbourhood — but more importantly, in a board with few large regions, the positions of the high numbers are heavily constrained from the start. Locate every region big enough to hold a 5, and work out where those 5s can coexist without touching. That global view frequently settles two or three cells before any local deduction does.

Cascades, not searches

Suguru is designed so that each placement forces the next. If you find yourself scanning the whole grid repeatedly without progress, you have probably placed something without applying its neighbour eliminations. Go back to your last two or three placements and cross off every touching cell properly — the missed move is usually right there.

A reliable order of attack

1. Fill every one-cell region, then apply the neighbour rule around each.
2. Work two- and three-cell regions, whose candidate sets are tiny.
3. For each region, ask where each specific value can go, not just what each cell can be.
4. Pencil in candidates only on cells touching placed numbers.
5. Scan region seams for cells with a single surviving candidate.
6. Place the high numbers by finding non-touching arrangements across the large regions.

Follow the cascade and a Suguru rarely takes more than a few minutes — and never needs a guess.