Remote Pairs in Sudoku

A chain of identical twins

Sometimes several cells scattered across the grid all carry the exact same two candidates, say a 4 and a 7. Cells with only two candidates are called bivalue cells, and a set of bivalue cells that all share the same pair is the raw material for a Remote Pair. On their own they are just look-alikes. The technique appears when you can link them into a chain, each cell seeing the next through a shared row, column or box, so that a decision in one ripples predictably down the whole line. The links themselves are the ordinary strong and weak connections that power every other chain technique, which is why Remote Pairs make such a natural first chain to learn.

Why the values alternate

The heart of the move is that two cells sharing a unit cannot both hold the same digit. So if the first cell in the chain is a 4, the cell it links to must be the 7, the next one flips back to 4, then 7 again, and so on. The pair alternates like the colours on a chessboard. You may not know whether the chain starts on 4 or on 7, but you know for certain that it strictly alternates from one end to the other.

Opposite ends are guaranteed opposite

Because the values flip at every step, the relationship between two cells depends only on how many steps apart they sit. Cells an even number of steps apart always hold the same value as each other. Cells an odd number of steps apart always hold opposite values. That second case is the useful one. Two cells an odd distance apart are guaranteed to be one 4 and one 7, in some order. Between the two of them, both digits are already spoken for, even though the exact assignment is still unknown.

A worked example

Imagine four cells that all read 4/7: r1c1, r1c8, r9c8 and r9c1, and suppose each links to the next around that loop through a shared row or column. Start a colour chain: call r1c1 colour A. Its partner r1c8 becomes colour B, r9c8 flips back to A, and r9c1 becomes B. Now r1c1 and r9c8 are both A, while r1c8 and r9c1 are both B. The two colours represent 4 and 7 in some order, so one colour is entirely 4 and the other is entirely 7.

The elimination

Now look for any cell that can see two opposite-coloured ends of the chain. Since one of those ends is 4 and the other is 7, any outside cell that sees both an A end and a B end cannot itself be 4 and cannot be 7. You erase both candidates from it. The elimination lands on cells that lie at the crossroads between the two colours, which is exactly where the chain has already committed both values. If nothing on the board sees both colours, the chain still earns its keep, because tracing it sharpens your picture of where each value can live even when it produces no immediate removal.

1. Find several bivalue cells that all share the same candidate pair, such as 4/7.
2. Link them into a chain where each cell sees the next through a common row, column or box.
3. Colour the chain in two alternating colours from one end to the other.
4. Any cell outside the chain that sees both colours cannot hold either candidate, so remove both.

Practising the move in Shapedoku

Remote Pairs are one of the friendliest doorways into chain logic, because the pattern is a single repeated pair rather than a tangle of different digits. Turning on Notes in Shapedoku makes the pairs easy to see, and the smart cleanup keeps your candidate marks honest as you place shapes, so a genuine 4/7 chain stands out. Because each Shapedoku value is a distinct glowing shape rather than a number, colouring a chain in your head can feel more natural; you are following two shapes, not two digits. Build the habit here and the leap to longer chains becomes far less intimidating.

When Remote Pairs fail to appear

Not every board offers a Remote Pair, and that is normal. The technique needs a supply of bivalue cells that happen to share one pair and happen to connect, which is a fairly specific arrangement. If you cannot find one, do not force it; fall back to simpler scanning, hidden singles, or pointing pairs, and come back to chain logic once the grid has thinned out. Techniques are tools, not obligations, and the calm solver reaches for whichever one the current board actually rewards. With practice you will start to notice promising pairs almost passively, the way a chess player notices a loose piece, and reach for the technique only when the grid genuinely offers it.