Two-String Kite
A row link and a column link joined in one box remove a digit.
A Two-String Kite is a single-digit pattern that links two "strings" for the same candidate: one strong link running along a row and one running down a column. The Two-String Kite sudoku technique fires when the inner ends of those two strings land in the same box, because that shared box forces one of the two outer ends to hold the digit — letting you erase that digit from any cell that sees both outer ends.
Worked example
On this board digit 1 has only two homes in row 1 — R1C1 and R1C5 — and only two in column 4 — R3C4 and R7C4 — giving one strong link along the row and one down the column. The inner ends R1C5 and R3C4 both sit in box 2, so they cannot both be 1; that welds the two strong links together and forces the 1 onto one of the free ends, either R1C1 or R7C4. R7C1 sees R1C1 down column 1 and sees R7C4 along row 7, so whichever free end turns out to be the 1, R7C1 cannot be — the 1 is eliminated from R7C1.
When to use it
Reach for a Two-String Kite once singles, locked candidates and pairs have run dry and you have started scanning one digit at a time for chains. Pick a candidate that appears exactly twice in some row and exactly twice in some column, then check whether one end of the row pair shares a box with one end of the column pair — that shared box is the hinge that makes it a kite rather than two unrelated pairs. When it does, the two remaining loose ends are your free ends, and any cell that sees both of them is a safe elimination.
How to use it
- Find two strings. Find a row and a column where the digit appears in exactly two cells each.
- Join them in a box. One end of the row and one end of the column must share a 3×3 box (the kite’s knot).
- Eliminate at the far ends. One far end must hold the digit, so remove it from any cell that sees both far ends.
Frequently asked questions
What is a Two-String Kite in Sudoku?
A Two-String Kite is a single-digit elimination pattern built from two strong links on the same candidate: two cells in a row where that digit can only go in those two spots, plus two cells in a column with the same property. When one end of the row link and one end of the column link fall inside the same box, one of the other two ends is forced to carry the digit, so any cell that sees both of those free ends can have the digit removed. On the example, that forces 1 into R1C1 or R7C4 and clears it from R7C1.
How do I spot a Two-String Kite?
Work one candidate at a time. Find a row where the digit has exactly two possible cells and a column where it also has exactly two, then look for an inner pair — one cell from each string — sharing a box. If that shared-box hinge exists, mark the two outer cells as free ends and scan for a cell that sees both of them; on the example board R1C5 and R3C4 hinge in box 2, the free ends are R1C1 and R7C4, and R7C1 sees both and loses its 1.
How is a Two-String Kite different from a Skyscraper?
Both use two strong links on a single digit and both eliminate from a cell that sees both free ends, but the shape of the join differs. A Skyscraper connects two parallel links — two rows, or two columns — whose near ends share a line; a Two-String Kite connects one row link and one column link whose inner ends share a box. In the example the row link R1C1–R1C5 and the column link R3C4–R7C4 meet through box 2, which is the kite's defining hinge.