Summary
Row/Column Lock is the second technique to reach for when Basic Elimination stalls. The idea is straightforward: if every remaining candidate cell of a color region happens to lie on the same row (or column), then that region's cat is guaranteed to land somewhere on that row (or column). Because Nyangjari allows exactly one cat per row (and one per column), no other region can place a cat there. Every empty cell on that row or column belonging to a different region can therefore be marked X. After the new eliminations, Basic Elimination often triggers fresh cat placements that were previously blocked.
When to use
- Apply Row/Column Lock only after Basic Elimination has run to completion and no further confirmation or elimination is available.
- A region qualifies for a lock when it still has two or more candidate cells but all of them fall on the same row or on the same column.
- After marking new X cells, always re-run Basic Elimination before looking for another lock — a single lock can cascade into several cat placements.
- Because the technique is symmetric, check both axes: a region may not be locked by row but may be locked by column, or the reverse.
Why it works
Nyangjari's rules state that each row contains exactly one cat. Suppose region B still has candidate cells only at positions (row 2, col 2) and (row 2, col 3) — both on row 2. Region B's cat must occupy row 2. Since row 2 can hold only one cat, no other region is allowed to place its cat anywhere on row 2. Any empty cell on row 2 that belongs to a region other than B is therefore impossible and can be marked X.
The critical requirement is that all remaining candidates of the region must lie on the same row or column. If even one candidate lies on a different row, the cat might go there instead, and the lock does not hold. Similarly, the lock applies only to cells in other regions — the locked region's own candidates are untouched.
The same reasoning applies column-wise: if all candidates of a region share a single column, that column is claimed by that region, and every other region's empty cells on that column become X.
Example
▲ Region b's two remaining candidates (highlighted) both sit on row 2. Row/Column Lock fires: the three cells on row 2 belonging to other regions — (2,0) and (2,1) in region a, and (2,4) in region c — are marked ×.
- Starting state: After Basic Elimination, region b's only candidates are (2,2) and (2,3) — both on row 2. No other empty cell remains in region b.
- Lock identified: All candidates of region b are confined to row 2. Region b's cat must go to row 2.
- Row 2 eliminated for other regions: Cells (2,0) and (2,1) in region a and cell (2,4) in region c are marked ×. The highlighted candidates in region b are left unchanged.
- Basic Elimination re-applied: With the new X marks, check whether any region, row, or column has been reduced to a single candidate. If so, place the cat there and propagate eliminations.
- Continue scanning: Look for further row or column locks in the updated board before moving to the next technique.
Watch out for
- All candidates must align — not just some: If a region has three candidates but only two share a row, the lock does not apply to that row. The entire candidate set must lie on one row (or one column) for the lock to be valid.
- Check both axes: A common oversight is checking row locks only. After scanning rows, always scan columns as well. A board that yields no row locks may still have a column lock — or vice versa.
- Lock targets other regions only: Row/Column Lock eliminates cells in other regions. Never mark the locked region's own candidates as X — those are the only remaining positions for that region's cat.
- Re-run Basic Elimination after every lock: A lock adds new X marks, which may create new single-candidate situations. Skipping the Basic Elimination pass after a lock means missing confirmations and stalling unnecessarily.
- Technique order matters: Row/Column Lock is the second technique. Try Basic Elimination first. If Row/Column Lock also stalls, move to Naked Pair or a more advanced approach.