Strategy · guides

Reading a fresh board: opening moves

An empty board has no crosses to cash in and no cats casting shadows — so where does the first deduction come from? From the region shapes themselves. Strong solvers spend their first minute reading geometry, not making marks.

Start with the smallest regions

Sort the board by eye: which regions have the fewest squares? A two- or three-square region is the closest thing to a free move the opening offers — it needs only a cross or two before it becomes a forced placement, and its cat is confined to a tiny area from move zero. Everything that tiny area touches is already suspicious: any square adjacent to all squares of a two-square region can be crossed immediately (the color squeeze, available before you've placed anything).

The pink region is the smallest on the board — two squares. Any cat on a highlighted blue square touches both pink cells at once, so all four are dead before you place a single mark.

Find the born-confined regions

Long, skinny regions are the opening's biggest gift. A region that lies entirely inside one row or one column has already made its commitment: that line's cat belongs to it, and every other square in that line is a cross — the confinement deduction, available with zero setup. On many handcrafted boards this is the designed way in: the setter placed a snake-shaped region precisely so the observant get a running start. Scan for it before marking anything.

The purple region lives entirely in row 3, so that row’s cat is purple — and the green and blue squares sharing the row (highlighted) are dead. No crosses needed to know it.

Respect the edges and corners

Corner and edge squares have fewer neighbors, which cuts both ways. A cat in a corner eliminates only three adjacent squares — the cheapest cat on the board in shadow terms — while a center cat wipes out eight. This means central regions interact far more than edge regions: when a center region's candidates are few, its shadow options can strangle multiple neighbors at once, which is where the early squeezes live. Conversely, a region hugging the board's edge tends to resolve late, because so little pressure reaches it.

A corner cat eliminates only three squares; a center cat eliminates eight. That is why central regions interact — and resolve — far earlier than the edges.

Count before you stare

The board's one global fact: N cats, one per row, one per column, one per region. Cheap arithmetic sometimes falls out of it. If three regions between them only occupy two rows, something is wrong — or rather, something is forced: those rows are claimed, and every square of any other color in them dies (the pair/triple version of confinement, and it's spottable on an empty board). Likewise, a row whose squares are all one color settles that region's row instantly, wherever the cat lands in it.

Purple and teal together occupy only the top two rows, so those two cats are purple and teal — and every other-colored square up there (highlighted) is crossed. That falls out of counting alone.

The one-minute opening routine

1. Scan region sizes. Note the two or three smallest — they resolve first and constrain their neighborhoods now.
2. Hunt skinny shapes. Any region confined to a single line → cross out the rest of that line immediately.
3. Check monochrome lines. Any row or column dominated by one color → that region is pinned to it.
4. Ring the tiny regions. Squares adjacent to every cell of a two-square region are dead — mark them.
5. Then start the ladder. With the geometry cashed in, the normal technique ladder has real material to work on.

The routine matters because openings are where guess creep is born: with no constraints on the board, every square feels equally plausible, and impatience fills the vacuum. The geometry pass replaces that vacuum with a handful of certain marks — and certainty compounds.

← Back to the strategy guide