What the 1-1 pattern looks like
Picture a wall along the bottom of the board. Two revealed 1s sit side by side along it. In front of them are three unrevealed cells. Two of those cells touch both 1s, and one cell touches only the right-hand 1. Like this (? is unrevealed, # is the wall):
? ? ?
1 1 #
# # #The leftmost cell is safe. The two 1s share a mine — it lives in one of their common neighbours — and that single shared mine is enough to satisfy both numbers. The cell only the right-hand 1 touches is forced safe.
Why the deduction holds
Walk through the logic slowly. The left 1 sees three unrevealed neighbours: the two shared cells and the far-left cell. Exactly one of those three is a mine. The right 1 also sees three unrevealed neighbours: the same two shared cells plus its own extra cell on the right. Exactly one of those three is a mine.
Now here's the trick: the only way to satisfy both 1s with a single mine is to place that mine in one of the two shared cells. If the mine were in the right 1's extra cell, the left 1 would still need a mine somewhere — but its only remaining candidates are shared, and they'd then create two mines total, which contradicts the right 1.
So the mine is shared, both 1s are satisfied by it, and any cell that's only touched by one of them must be safe. Click with confidence.
The four common variations
1. 1-1 on a flat wall (the canonical case)
Two 1s on a horizontal boundary, three unrevealed cells above them. The outermost top cell — the one only touched by the closer 1 — is safe.
2. 1-1 along a vertical edge
Same logic rotated 90°. Two 1s stacked along a side wall, three unrevealed cells to one side. The outer cell that only touches the nearer 1 is safe.
3. 1-1 with one 1 in a corner
When the corner cell is a 1, it has fewer neighbours than the non-corner 1. That asymmetry can shorten the shared region from two cells to one, but the rule still applies: the cell only the non-corner 1 touches is safe.
4. 1-1 with a third constraint nearby
Sometimes a number elsewhere on the board pins down one of the shared cells as a mine in advance. That doesn't break the pattern — it makes it even better. The original "safe" cell remains safe, plus you now know exactly which shared cell holds the mine.
A worked example
Imagine you're mid-board on an intermediate (16×16) game. The lower edge of your revealed region looks like this, with · for safe-revealed empties and ? for unrevealed:
· · · · ·
? ? ? ? ?
1 1 2 1 1 Two pairs of 1-1: one on the left (1 1) and one on the right (1 1). Apply the rule to the left pair: the cell above the leftmost 1 (one above-only neighbour) is safe. Apply it to the right pair: the cell above the rightmost 1 is safe. The 2 in the middle still needs work, but you've just turned two cells from "unknown" into "definitely safe" without any guessing.
How to spot it under time pressure
- Scan along walls first. 1-1 patterns appear most often where the board ends — bottom rows, side columns, corners — because walls reduce neighbour counts and make 1s likely.
- Look for pairs of 1s touching each other. Single 1s mean nothing on their own; it's the adjacency that triggers the deduction.
- Trace each 1's unrevealed neighbours. If they share most of them and one has a single extra outside the shared set, you have the pattern.
- Click the extra cell. No flagging, no thinking. The faster you can run this loop, the more seconds you'll save.
Common misreadings
Skipping the shared-neighbour check
Beginners sometimes apply 1-1 to any two 1s that happen to be adjacent. The pattern only works if the two 1s share every unrevealed neighbour except for one extra. If the two 1s have different neighbour sets — say one of them touches an open corridor — the deduction doesn't hold.
Forgetting the third-constraint check
Sometimes a flagged mine elsewhere has already eaten one of the 1s' mine counts. If the left 1 is already satisfied (a flag sits on one of its neighbours), the pattern doesn't fire — the right 1 might still need its mine in its exclusive cell. Always re-count flags before clicking.
Mistaking 1-1-1 for 1-1
Three 1s in a row is a different shape. The middle 1 is shared with both outer 1s, and the deduction extends: both outer cells are safe and the mine sits squarely in the middle pair. Treat it as two overlapping 1-1s, not one.
When the pattern doesn't apply
The 1-1 needs three ingredients: (1) two adjacent 1s, (2) a clean shared-neighbour set, (3) no flagged mines that already satisfy either 1. Pattern doesn't fire when:
- The 1s aren't actually adjacent — a 2 sits between them, or there's a gap.
- One of the 1s already has a flag on one of its neighbours, so it's already satisfied.
- The unrevealed cells form a more complex region (multiple "extras") — that's a different pattern and needs subset deduction instead.
When in doubt, fall back to the two core checks: "is this number equal to its unrevealed neighbour count?" (all of them are mines) and "is this number equal to its flagged neighbour count?" (all the rest are safe). The 1-1 is a specialised shortcut, not a replacement.
Practice this pattern today
The fastest way to internalise 1-1 is volume. Open a beginner board, deliberately slow your clicks, and look for the pattern after every flood-fill. Once you've spotted ten in a row without missing, move to intermediate. By the time you're running expert boards, the pattern recognition should be instant — no conscious calculation, just immediate clicks on the safe extras.
The daily challenge is a great venue: same seed for everyone, so you can compare your time against the leaderboard afterwards. If you're 2× slower than the top players, it's almost always pattern recognition (not clicking speed) that's the gap.