Pointing Pair sudoku technique

Intermediate level Appears in puzzles rated SE 1.7 – 2.3

Look at where a digit can still go inside one 3×3 box. If every remaining spot happens to sit in the same row or the same column, the box has made a promise: the digit will be placed on that line, inside that box. The rest of the line — the six cells running through the neighbouring boxes — can drop the digit entirely.

How to find a pointing pair

  1. Take a box and a digit not yet placed in it.
  2. Mark the cells in the box where the digit could go.
  3. If all of them line up in one row or column, remove the digit from that row or column outside the box.
  4. Repeat per digit; pointing pairs come in clusters and love to chain into singles.
From the table: You do not need full pencil marks for this one. Scanning a single digit across the grid, box by box, finds pointing pairs quickly — which is why this is usually the first technique that feels like "real" solving rather than counting.

The pattern on the grid

555555

Box 1 confines digit 5 to row 2 (blue) — the 5s elsewhere in row 2 (red) are eliminated.

Worked example

In the top-left box, digit 5 is blocked from row 1 by the 5 at r1c7 and from row 3 by the 5 at r3c5. Whatever survives sits in row 2 — say r2c1 and r2c3. The box therefore promises its 5 to row 2, and the 5s pencilled further along that row, at r2c5 or r2c8, are dead.

When it fails

The promise only holds while ALL of the digit's spots in the box share the line. One overlooked candidate on another row of the box and the pointing pair evaporates — recheck the box before erasing outside it.

Related techniques

All solving techniques