How it works
Hidden pairs are the inverse of naked pairs. Instead of looking at cells with two candidates, you look at digits that can only go in two cells.
The reasoning is symmetrical. If two digits can only go in two specific cells of a unit, those two cells must hold those two digits between them. Anything else those cells contain as candidates can be removed.
Hidden pairs are harder to spot because the cells themselves might list many candidates. The technique only becomes visible when you scan digit by digit and track positions, not cell contents.
When to look for it
When naked pair searches come up empty. Flip the search: for each pair of digits, are both confined to the same two cells in a unit?
Tips for spotting the pattern
- Pick two digits that already appear in many places. Their remaining positions are tightly constrained.
- Hidden pairs in boxes are the most common; check those first.
- After finding a hidden pair, the cells often collapse to a naked pair, which then triggers further eliminations.
Common mistakes
- Looking at cell contents instead of digit positions. The whole point is to scan differently.
- Stopping after spotting the pair. Strip the other candidates from those cells to actually use it.
- Forgetting that hidden pairs require the digits to appear in only those two cells and nowhere else in the unit.