For the standard straight 1-2-1 arrangement, the two outer unknown squares are mines and the middle one is safe. That conclusion requires exact neighboring-square relationships. Three matching digits on their own are not enough to justify a click.
Loading the board…
Reveal every safe cell. Each number counts mines in its eight neighboring cells, including diagonals. Classic has a 9 × 9 board and ten mines; Practice has a 6 × 6 board and five mines. Your first cell and its neighbors are safe. On a phone, select Flag before marking a suspected mine; on a computer, right-click. A flag is an assumption until the clues prove it. Some boards can require guessing.
Call three consecutive unopened squares A, B, and C. For this example, the left 1 has A and B as its only unresolved neighbors. The 2 has A, B, and C. The right 1 has B and C. Any other touching squares must already be known safe, or their mine contributions must have been accounted for correctly.
One common layout has the clues in a row immediately below the unknown squares. A cleared area or the board’s edge limits the unknown frontier. Do not assume that limit merely because you recognize the digits. Check the squares diagonally beyond both ends and on the other side of the clues. Those squares can change which arrangements remain possible.
Assume B contains a mine. The left 1 would then require A to be safe. The right 1 would require C to be safe. That leaves only one mine among A, B, and C, but the central 2 requires two. The assumption cannot satisfy all three clues, so B must be safe.
Once B is safe, the left 1 forces A to be a mine, and the right 1 forces C to be a mine. The pair also satisfies the 2. This is the complete reason behind the familiar outer-mine, safe-middle result. You are not relying on the likelihood of a pattern; you have eliminated the alternative by checking every constraint.
Now add an unresolved square D that touches the left 1 but not the central 2. The left clue no longer says there is one mine in A and B; its mine could be in D. The proof you just used no longer applies unchanged. The visible 1-2-1 is still present, but the logical system is different.
You might prove D safe using another clue and recover the simple case. Until then, keep it in the unknown set. This habit matters near corners of opened regions, where a frontier bends and one clue sees squares on two sides. A memorized flat diagram should never replace checking the actual eight neighboring positions.
A displayed clue counts all touching mines, including the ones you have already located. If a displayed 2 has one established mine outside A, B, and C, its remaining requirement in the unknown group is one. Treating it as two would produce a false conclusion. Do the subtraction separately for each clue.
Conversely, larger printed numbers can reduce to remaining requirements of 1, 2, and 1. The same deduction works if the unresolved sets still match A+B, A+B+C, and B+C. A reduction is a calculation, not a new flag placed to make the diagram fit. Keep the evidence for each known mine clear before using it in a second deduction.
Open the safe middle square and use the new clue to continue. You can flag the two outer squares once their status is proved. If later counts disagree, recheck your original neighbor sets and any assumptions used in reductions. Do not keep opening squares just because the first move based on a mistaken pattern happened to survive.
The playable board above offers ordinary Minesweeper practice; it does not guarantee that a particular deal includes this pattern. Start with direct deductions when no suitable 1-2-1 appears. Recognizing the exact case is useful, but inventing one where the geometry differs is worse than leaving an uncertain frontier alone and examining a different part of the board.
In the exact A+B, A+B+C, B+C case, A and C are mines and B is safe.
No. Extra unresolved neighbors or already-counted mines can invalidate the shortcut.
Yes. The neighbor relationships matter, not whether the diagram is horizontal.
Yes, after subtracting only mines whose positions you have actually established.
No. It is a possible logical relationship, not a feature of every layout.