Mines Strategy

Every row is one setup repeated round after round at 97%. Swing here measures how far one round's result typically lands from its average, in stakes, and a losing run is a string of consecutive rounds that ended on a mine. From top to bottom the price hardly moves while the shape of the results changes completely.

A setup near the top of that table wins nearly every time and collects very little for it. At 1 mine, 1 click the board pays 1.01x on 96.00% of rounds, so the charge is collected a sliver at a time and luck never gets enough room to hide it. Over 1,000 rounds from a budget of 100 stakes, the share of players who end up ahead is under 0.1%, and the middle player walks away with 69.6 stakes.

At the far end, 24 mines, 1 click pays 24.25x on just 4.00% of rounds, which leaves a session to a handful of hits. Across the same 1,000 rounds, 57.5% of players lose the whole budget, 32.0% finish ahead and 16.7% at least double it, while the median balance left is 0.8 stakes. Setups in the middle land in between: at 3 mines, 5 clicks, 2.5% go broke and 14.3% come out in front.

The key number: After 1,000 rounds, under 0.1% of players are in profit at 1 mine, 1 click and 32.0% at 24 mines, 1 click, on a game charging both the same return setting.

Neither of the two shapes costs less than the other, so the choice comes down to temperament: a slow, near-certain leak, or long droughts interrupted by the occasional large payout. Whichever suits, setting the budget before the first round, at a sum that could be lost without strain, matters more than the setup.

Turn the mines count and the clicks count around and the chance of a full run does not change. 3 mines, 5 clicks clears the run in 49.57% of rounds, exactly as often as 5 mines, 3 clicks; on a formula-priced board both pay the same multiplier. The reason is simple once spelled out. A round survives only if the squares opened and the squares mined do not overlap, and that question reads identically whichever group is called the mines.

Over every setup a board of 25 squares allows, the 300 pairings of mine count and click count carry just 150 different prices. Most of them come in mirrored pairs, and 6 prices are each shared by four setups, where two pairs happen to land on one chance. Advice that weighs 3 mines against 5 is therefore often comparing a bet with itself, seen from the other side.

Each row is one chance of clearing the whole run, reached by four different setups on 25 squares. The multiplier is the one a formula board set to 97% pays for that chance, so all four setups in a row pay it alike.

A popular habit is to uncover one gem, cash out, and put the whole return on a new board, again and again until a target is met. It feels like one long run cut into steps, yet it is a different and dearer bet. Every new board charges the house share afresh, so 5 one-click rounds at 1 mine, each win carried forward in full, give back 85.87% of the opening stake on average, while one round of 5 clicks on a single board gives back 97.00%.

The difference comes from when the charge is taken: at the start of each round, from the stake, rather than click by click. Why the clicks after the first carry no extra charge is explained on the mines hub; a new board simply starts the meter again. Carried across 10 boards, the return drops to 73.74%. The ridden version even wins a little more often, because each board starts full of gems again, and it pays far less for the privilege.

The rule: One board with several clicks pays the house share once. Several boards with one click each pay it on every board, however the winnings are carried from one to the next.

Return is the average amount handed back per stake on a game set to 97%. In the right-hand column every board carries the same number of mines, gets opened once, and the full balance is staked again after each win.

Stripped of the board, a mines setup is one chance paired with one multiplier, and other quick games sell exactly that pair. 3 mines, 5 clicks wins 49.57% of rounds and pays 1.96x; a limbo bet with a target of 1.96x and a dice bet to roll under 49.57 offer the same chance on a game with the same return. Clicking squares one by one makes the bet neither better nor worse, only slower to settle.

Crash gets to the same place by another road, letting the player pick the exit while the multiplier climbs: a cash-out at 1.95x on a curve set to 97% is reached 49.74% of the time. How those curves are drawn and priced is covered on the crash games pages.

Each row pairs a mines setup with the limbo target and the dice roll that win just as often, all on games returning 97%. The multiplier and the limbo target coincide because each is the game's return figure over one shared chance.

Martingale raises the stake after a loss by just enough that the next win clears the whole series, then drops back to one stake once a win comes in. Played at 3 mines, 5 clicks, starting with 100 stakes, it can afford 6 rounds in a row, the last at 35.76 stakes, and a win at any point recovers the series with 0.96 of a stake on top. Hitting a mine on all 6 costs 69.02 stakes in one blow. That happens in 1.65% of series, roughly one in 61, and the chance of finishing 100 series without it is 19.0%.

A larger budget postpones the blow and makes it heavier. From 1,000 stakes the progression lasts 9 rounds, fails about once in 474 series and takes 597.45 stakes when it does, and only 12.1% of players get through 1,000 series intact. At 24 mines, 1 click the stake barely grows from step to step, since one hit repays a long string of misses, yet the series still runs dry after 39 rounds in 20.35% of attempts.

Under every one of these plans, each stake placed still loses 3.000% on average; only the timing of the losses changes. Why that holds for any staking plan, Martingale's cousins included, is worked through in the guide to betting systems.