The Y-wing, when it finally clicks

There's a stage in mid-level Sudoku where every new technique feels like one more pattern to memorise. Naked pairs, hidden triples, pointing pairs, X-wings — each one a configuration to recognise on the grid, each one a small unlock when it appears. Then comes the Y-wing, and suddenly the activity changes shape.

The Y-wing isn't a pattern in the same sense as the techniques that come before it. It's a small piece of case analysis — a deduction that sits across three cells and runs an argument about what would have to be true if a particular cell took one value or the other. The shift from spotting-patterns to running-arguments is the first taste of what the upper-tier of Sudoku is, and it's the reason the Y-wing has the reputation for being the technique that finally clicks.

The technique in plain English

A Y-wing involves three cells, each of which has exactly two candidates pencil-marked.

Pick one of them as the pivot — call its candidates X and Y. Find a second cell, call it the first wing, that shares a row, column, or box with the pivot, and whose candidates are X and Z (sharing one digit with the pivot, with one new digit Z). Find a third cell, the second wing, that shares a unit with the pivot but not necessarily with the first wing, and whose candidates are Y and Z (sharing the other digit with the pivot, and the same new digit Z as the first wing).

If you've found this configuration, here's what falls out: regardless of which of the two values the pivot ends up taking, one of the two wings must hold Z. If the pivot is X, the first wing is forced to Z. If the pivot is Y, the second wing is forced to Z. Either way, Z is somewhere in the wing pair.

That conclusion is what powers the elimination. Any cell that sees both wings — that is, any cell that shares a row, column, or box with both wing cells simultaneously — cannot itself contain Z, because the digit is already committed to one of the two wings.

The case analysis move

The earlier techniques are about what's there. You spot a naked pair and the elimination follows from the pair existing. You spot an X-wing and the elimination follows from the rectangle of candidates being where it is. The reasoning is direct: there's the pattern, here's what it forces.

The Y-wing breaks that shape. The reasoning isn't about a pattern that's currently true on the grid — it's about a pattern that would have to resolve in one of two ways, and a consequence that holds in both. You're not reading the grid; you're branching the grid, mentally, into two hypothetical futures and noticing what they have in common.

That mental move — if-X-then-this, if-Y-then-this, therefore-this-either-way — is the foundation of every chain technique above the Y-wing. Skyscrapers, XY-chains, forcing nets, the various coloring methods all use the same fundamental shape, just with longer chains and more cases. The Y-wing is the introductory dose. Get this one comfortable and the harder chain techniques become extensions of a habit you already have, not new techniques to memorise from scratch.

The moment it tends to click

For a lot of solvers, the Y-wing reads as abstract until they've worked it out on a puzzle in front of them. The technique's description sounds like algebra; the technique itself feels different.

The cleanest way to land it is to set up a known Y-wing position deliberately, work through both branches by hand, and notice that the conclusion holds in both. Pick a candidate cell, write if this is X in the margin, and follow the chain forward — the first wing forces to Z. Then write if this is Y and follow that branch — the second wing forces to Z. With both branches mapped on paper, the conclusion stops feeling like a clever algebra trick and starts feeling like an obvious consequence. Of course the eliminations follow; the digit had nowhere else to go in either branch.

After two or three of these worked-by-hand examples, the case analysis collapses into intuition. You stop writing the branches in the margin; you start running them silently. The third or fourth Y-wing on a real puzzle is the one that produces the click.

Where Y-wings appear

In practice, Y-wings show up most often at hard and expert tier, in the second half of a solve when pencil-marks have settled enough that several cells are bivalent. They're rarer than X-wings on raw frequency, but they're easier to spot once you're looking for them, because the building blocks (pairs of bivalent cells) are visually distinctive on a marked grid.

A working scanning habit: once a hard puzzle stalls and the grid is mostly pencil-marked, scan for any two bivalent cells that share a unit and share a candidate. That's a potential pivot-and-wing. If a third bivalent cell exists with the right shared candidates and unit relationships, the Y-wing is in front of you. The scan takes ten or fifteen seconds and surfaces the technique reliably; without the scan, Y-wings tend to hide.

The reason it's the gateway

The Y-wing isn't the most powerful technique in the mid-level toolkit, and it isn't the most common. What it is, is the technique that teaches you the shape of upper-tier Sudoku reasoning. The conclusion it produces is small. The cognitive habit it builds — trace this in two cases, see what holds in both — is what every harder technique uses.

If you've been wondering why master and extreme tier puzzles feel qualitatively different from hard tier, the answer is mostly that they reward the case-analysis habit Y-wings introduce. We've written separately on the difficulty curve and what each tier trains — the through-line that connects mid-level pattern recognition to upper-level chain reasoning runs through this technique. Once the Y-wing clicks, the rest of the ladder is reachable. Before it does, the rest of the ladder feels like a different game.