Mines

How Mines Works

Mines is a five-by-five grid of 25 closed tiles. Before the round starts you set your stake and you choose how many of those tiles are mines — anything from 1 to 24. The rest are safe. You then open tiles one at a time. Every safe tile you turn pushes a running multiplier up, and you can stop and take that multiplier at any point. Open a mine and the round is over and the stake is gone.

That is the whole game. There is no dealer, no card order, no wheel and nothing to read. It arrived from the crypto-casino side of the industry, where it is usually served as a provably fair game: the layout of the board is fixed by a server seed whose hash is published before you touch anything, combined with a client seed you can change, so the board cannot be re-rolled once play has begun.

Two numbers describe any round completely: how many mines are on the board, and how many tiles you open before you stop. Everything below follows from those two, and from nothing else.

Which Tiles You Pick Cannot Matter, and That Can Be Proved

Search for this game and a large part of what comes back is patterns: which squares to open first, which corners are “cold”, apps that claim to read the board. There is nothing there to find, and it is not a matter of opinion.

Put m mines on 25 tiles. Before anything is opened, every tile is equally likely to be a mine, because the layout was fixed before you chose. Open one: it is safe with probability (25 − m) / 25. Whatever tile that was, the board that remains has 24 tiles and still m mines, so the next one is safe with probability (24 − m) / 24, and so on. Multiply those out for k tiles and the answer collapses to a single expression that contains k and m and no reference at all to which tiles were chosen:

chance of opening k tiles safely = C(25 − m, k) / C(25, k)

Corners, edges, diagonals, the tile that “has not come up”: all of them give exactly this number. We checked the identity against the step-by-step product for all 300 playable combinations of mines and tiles, and then again by dealing 400,000 real boards for each of ten settings and opening tiles in a randomly shuffled order rather than left to right. The simulated rates land on the formula every time.

So a pattern generator is selling the one thing this game provably does not contain. The only choice that changes anything is how many tiles you open, and how many mines you asked for.

Choosing 3 Mines and Stopping at 5 Is Literally the Same Bet as 5 Mines Stopping at 3

The expression above has a property that is not obvious from the screen: it is symmetric. Swapping the number of mines with the number of tiles you intend to open leaves it unchanged. C(25 − m, k) / C(25, k) equals C(25 − k, m) / C(25, m) for every pair. We verified all 300 of them.

In plain terms: mines and tiles are interchangeable. A cautious-looking plan and a reckless-looking one can be the identical wager with the identical payout.

The last row is the extreme case and it is the one worth sitting with. Putting a single mine on the board and clearing all 24 remaining tiles — the most patient thing you can do in this game — wins 4.0000% of the time and pays 25× before the house takes its cut. Putting 24 mines on the board and opening one tile — the most reckless thing you can do — wins 4.0000% of the time and pays 25×. Same chance, same payout, same swing. One of them takes 24 clicks and feels like skill.

The Number That Decides What You Pay Is Set Before You Click

Mines is not one game. Several suppliers build it, they all use the same 25-tile grid, and they do not charge the same for it. Two publish the figure on their own sites:

• Spribe — the studio behind Aviator — lists its Mines at a 97% return to player. That is a 3.0% house edge.
• BGaming lists Minesweeper, the same 25-tile game under another name, at 98.4%, released in April 2016. That is a 1.6% house edge.

Nearly twice the cost for the same rules. Nothing you do at the board comes close to that difference, and the choice is made before the first tile, in the lobby, by which version you open. It is worth checking which studio’s Mines a site is running — the supplier’s name is usually on the game tile or in the info panel.

Here is what the two prices look like on the first tile:

The House Edge Does Not Move, Whatever You Do

The multiplier a client shows after k safe tiles is the fair payout reduced by the operator’s margin. Because the fair payout is exactly the reciprocal of the chance of getting there, the two cancel: the expected return of “open k tiles then stop” is the headline RTP, and it is the same number for every k and every mine count. We checked all 300 settings. Every one returns exactly 97.00% at Spribe’s figure.

That is the honest answer to “what is the best Mines strategy”, and it is not the answer the strategy pages give. No cash-out rule is better than any other. Stopping at one tile, stopping at ten, an auto-play ladder, a Martingale on top: all of them cost the same fraction of everything you stake. What changes is the shape of the ride.

The right-hand columns are the same board priced by the two suppliers. Set against a 3-mine board opened five times, the difference is 1.95× against 1.98×.

Read the last two columns together. Cashing out on the first tile with three mines wins nearly nine rounds in ten and pays 1.10×; four losing rounds in a row is a 0.0% event. Twenty-four mines and one tile pays 24.25× and loses four in a row 84.9% of the time. The expected cost of both is identical. Only one of them will empty a balance in an evening.

The Rounding Is a Second Charge, and It Is Not in the Headline

Clients show the multiplier to two decimal places, and they round it down rather than to the nearest. That is invisible on a big multiplier and expensive on a small one, because two decimal places are a much coarser grid at 1.25× than at 22.00×.

The worst cases are the short, safe-feeling plans. A 2-mine board opened three times — and, by the symmetry above, a 3-mine board opened twice — should pay 1.2597× once the 3% is taken. It pays 1.25×. The return on that particular bet is 96.25%, not 97%: a quarter of the advertised house edge again, charged silently. 27 of the 300 settings lose at least a further 0.25 percentage points this way, and every one of them is a short plan at a low mine count — which is exactly where a cautious player sits.

Across all 300 settings the rounding costs an average of 0.074% of stake on top of the house edge. It can never work in the player’s favour: rounding down only ever moves money one way.

How Far a Board Actually Gets

Three mines is the setting most clients open on, so it is worth knowing what three mines does if you keep going. Half of all boards are over by the fifth tile.

Clearing a 3-mine board completely — all 22 safe tiles — happens once in 2,300 rounds and pays 2,300× before the margin. At a round every twenty seconds that is about thirteen hours of continuous play for one expected occurrence. Clearing a 1-mine board is a 1-in-25 shot, which is the same 4% that a single pick on a 24-mine board gives you.

What the Edge Costs at a Real Stake

Mines is fast. There is no spin animation to sit through and no dealer, so a round can be over in a few seconds, and the house edge is charged on every one of them. At a ₱100 stake, this is what the two RTPs cost per hour of steady play — not a bad-luck figure, the expected one:

The comparison worth making is against the games we have already costed. Mines at 97% is a 3.0% game and at 98.4% a 1.6% one. Our baccarat page works the banker bet out at 1.06% and the player bet at 1.24%; the perya Color Game costs 7.87%, and Sakla exactly 5%. Mines sits between the table games and the street games, and where in that range it sits is decided by the supplier rather than by the player.

Where the Law Stands

Where you can legally play this for money depends on where you are, and it varies more than most players expect.

In the Philippines, real-money online gaming is licensed and supervised by PAGCOR, and play is restricted to those aged 21 and over. In India, the Promotion and Regulation of Online Gaming Act, 2025 prohibits online money games, and states the prohibition applies whether a game is based on skill, chance or both. In the United Kingdom, online casino games may only be offered by operators holding a Gambling Commission licence, and the minimum age is 18. Rules elsewhere differ again, and they change.

Whatever the position where you are, the arithmetic above does not change with the jurisdiction. Nothing on this page is legal advice; check the current rules that apply to you.

Operators

Five of the operators we review are listed below, in the order the site uses for readers across South-East Asia, with the welcome offer each currently advertises.

The offers are the sites’ general welcome bonuses rather than anything specific to this game, and bonus funds usually carry wagering requirements that count instant-win titles at a reduced rate or not at all. The terms on the operator’s own page are the ones that apply.

Other Games We Have Covered

Aviator is the other Spribe title built on the same provably-fair mechanism, and the same argument about predictors applies to it. Keno and the lottery pages cover the other draw-style games where the whole result is settled by a single random draw. For the games where a decision genuinely changes the return, see blackjack and baccarat. The casino games hub lists the rest, and if you play through a phone wallet our GCash guide covers how that rail works.

Before You Play

There is no system here and nothing on the board to find. The layout is fixed before the first tile, every tile is the same tile, and the only two decisions available — the mine count and the stopping point — are interchangeable and cost exactly the same. The one choice that changes what you pay is which supplier’s version you open, and that is made in the lobby. In Philippine-facing lobbies that supplier is usually JILI, whose Mines is one of 35 arcade-type titles on PAGCOR’s approved list.

Decide before you start what the session is worth to you, treat that as the price of the entertainment, and stop when it is spent. Gambling is not a way to make money and never a way to recover a loss. If it is causing harm to you or to someone close to you, GambleAware offers free and confidential support. 18+, or 21+ depending on where you are.

For a live game with a presenter rather than a solo round, Crazy Time runs on the same principle that each spin is independent of the last — which is why its results history cannot forecast anything either.

FAQ