What the remaining mine counter actually tells you
In classic Minesweeper Now, the counter starts at the board’s total mine count and subtracts the number of flags you have placed. It does not inspect those flags for correctness. A displayed zero means you have placed as many flags as there are mines; it does not establish that you have identified every mine.
For logical endgame reasoning, use the total mines minus the number of proven mines. If all your flags were placed from valid deductions, that is the same as the displayed counter. If any flag is speculative, calculate the logical remainder separately or remove the speculative flag before reasoning from the display.
For example, an Intermediate board contains forty mines. If thirty-eight mines are proved, there are exactly two among the unresolved squares. That is a global constraint even when those squares lie in different corners. If you merely placed thirty-eight flags, the same conclusion is not justified until the flags are checked.
Endgames are especially sensitive to one wrong assumption. A mistaken flag can make an otherwise valid mine-count calculation appear to prove several safe squares at once. Before a large clearing move, confirm the chain of deductions behind your flags.
Reserve mines for groups before clearing elsewhere
You do not always need to identify a mine’s exact square to account for it. If a clue proves that A and B contain exactly one mine between them, reserve one mine for that pair. The pair may remain ambiguous, but its total is fixed.
Suppose six covered squares remain. A and B form one group containing exactly one mine. C and D form another, disjoint group containing exactly one mine. E and F have no useful local clue. If two mines remain on the board, both are already reserved for the two groups. E and F are safe.
If three mines remain instead, E and F together must contain one mine. That proves a group count, not which member contains it. If four remain, E and F are both mines. The same local groups lead to different conclusions as the global count changes.
| Mines remaining | Mines reserved in the pairs | Conclusion about E and F |
|---|---|---|
| 2 | 2 | Both safe |
| 3 | 2 | Exactly one mine between them |
| 4 | 2 | Both mines |
The deduction does not depend on where the pairs are drawn. What matters is that their mine counts are proved and their squares do not overlap. If you cannot identify the groups precisely, write down their members before adding the totals.
Check your understanding
Exactly two mines remain. Two disjoint unresolved groups each require one mine. What about covered squares outside both groups?
Choose an answer.
Do not count a shared mine twice
Two clues saying “one mine” do not necessarily identify two different mines. A + B = 1 and B + C = 1 may both be referring to a mine at B. Alternatively, B may be safe and the mines may be at A and C. The equations permit either arrangement.
If you add those clues and reserve two mines automatically, you count the shared square B twice. The correct possibilities have different total mine counts: the B-only arrangement uses one; the A-and-C arrangement uses two.
This distinction is exactly where the global count helps. If one mine remains and no unresolved squares exist outside this group, only B can be a mine. If two remain, only A and C can be mines. Local ambiguity becomes a definite solution when the board-wide count rules out the other branch.
Let the mine count settle an endgame
This is a constraint diagram, not a spatial board. One clue says A + B = 1; another says B + C = 1. The symbols name covered squares. No other clues restrict them.
| Frontier mines | Outside mines needed | Completions | Weight |
|---|---|---|---|
| B | 0 | 1 | 100.0% |
| A + C | -1 | 0 | 0.0% |
Choose A, B, or C to inspect it.
With one mine remaining, only B can be a mine, so A and C are safe. With two, A and C must be mines and B is safe. The local equations have not changed.
The explorer abstracts the position into named squares. It does not claim that every three-square shape has these equations. On a real board, first establish which covered squares each clue touches, then check whether there are other unresolved squares that could absorb the remaining mines.
Include unopened areas outside the frontier
A frontier is the set of covered squares constrained by revealed clues. A completely covered pocket may sit beyond it and have no direct clue at all. Those outside squares are still part of the board and still use the total mine budget.
In the A + B = 1, B + C = 1 example, two remaining mines do not force A and C if there is an additional unconstrained square D. One valid arrangement has mines at A and C; another has mines at B and D. Forgetting D produces a false certainty.
Use capacity checks before eliminating a branch. If a candidate frontier arrangement leaves three mines for an outside area with only two covered squares, that branch is impossible. If it leaves zero mines, the outside area is safe. If it leaves one mine among several outside squares, you know their combined count but generally cannot identify the mine individually.
When several disconnected frontier groups remain, combine their possible counts rather than solving each in isolation. A group may admit one or two mines, while another needs exactly two. If only three remain overall and no outside area exists, the first group must use its one-mine branch.
Audit a suspicious flag before the final chord
A contradiction is a reason to inspect your assumptions. If a clue seems to have more neighboring flags than its number, or no possible mine arrangement fits your groups and total, do not resolve the problem by opening an arbitrary square. Recount neighbors and revisit flags whose proof you cannot reconstruct.
The exercise below contains one speculative flag. A chord can be mechanically available because the number of flags matches the clue, while still being logically unsafe. Use the complete board’s mine count to repair the flag before taking the shortcut.
Repair a flag before chording
This complete two-row board contains exactly one mine. All three bottom clues are 1s. Click a covered square to place or remove a flag, then click the bottom-middle 1 to chord.
One flag was placed without proof. Inspect the clues and the total before using it.
The left 1 sees the two left covered squares. The right 1 sees the two right covered squares. Because the board has only one mine, it must be their shared neighbor: top-middle.
On a large board, a wrong flag may have influenced several later deductions. Simply removing it does not validate all the moves that followed. Trace the dependent conclusions and check whether the remaining clues still prove them.
A reliable endgame checklist
- Count the mines proved so far, rather than the flags you hope are correct.
- List all unresolved covered squares, including pockets away from the active frontier.
- Identify disjoint groups with fixed mine totals.
- For overlapping groups, enumerate compatible arrangements instead of adding clue numbers.
- Subtract each arrangement’s mines from the remaining board total.
- Reject arrangements that leave an impossible number of mines for the outside squares.
- Reveal a square only if every surviving arrangement makes it safe.
Does a zero counter mean I can clear everything?
Only if every flag is a proven mine. The counter itself does not verify your flags. A zero display with an incorrect flag can leave an unflagged mine on the board.
Do I have to resolve every ambiguous group first?
No. A group’s fixed total can be enough to clear other squares. Those new reveals may then supply the information needed to resolve the group.
What if multiple arrangements still survive?
Look for a square that is safe in all of them. Multiple possible full arrangements do not automatically mean every move is a guess. If no guaranteed-safe reveal remains anywhere, assess the risks without relabeling an uncertain choice as a deduction.