Stake Mines strategy: mine counts, odds and cash-out

No strategy changes the maths of Stake Mines. The game is widely reported at 99% RTP, a 1% house edge, whatever number of mines you choose and whatever tiles you click. Every payout is priced from the exact chance of surviving that many picks, then reduced by that 1%. That is why a risky pick that pays 20× is exactly as costly in the long run as a safe one that pays 1.1×.

What you can control is the shape of the risk: how many mines, how many picks before you cash out, and how big the stake is compared with your session budget. This guide shows the actual numbers so you can choose on purpose rather than by feel. If you have not played the game yet, start with the Mines guide.

The board has 25 tiles. With m mines, the chance of picking k safe tiles in a row is the number of ways to choose k safe tiles divided by the number of ways to choose any k tiles. The multiplier at that point is about 0.99 divided by that chance. Real figures:

Table cells show the chance of surviving that many picks, then the approximate multiplier. They are calculated from the combinatorics, and the multipliers in the game may differ slightly in rounding.

Two things stand out. First, one pick with one mine looks almost safe at 96%, but the payout is only 1.03×, which is barely above your stake after the edge. Second, a long climb gets steeply less likely: with three mines, ten picks succeed under 20% of the time.

1 to 2 mines

Very frequent survival, very small multipliers. A round on one mine and three picks has an 88% survival chance for roughly 1.1×. Results move slowly, the 1% edge works steadily, and the long sessions it allows mean it adds up through turnover rather than risk.

3 to 5 mines

The middle ground. Three mines with five picks gives about even odds (49.6%) for around 2×. This is the setting where choosing a target and cashing out feels like a real decision.

8 to 12 mines

Short rounds. Survival odds fall quickly: with ten mines, three picks survive only about one time in five, but pay about 5×. Most rounds end with a mine on the first few clicks.

20 to 24 mines

One or two picks only. At 24 mines the single safe tile is a 1-in-25 chance (4%) for about 24.75×. It is essentially a coin-weighted lottery ticket with a stake the size of one tile.

A higher mine count never raises the return. It only shifts the same 99% into rarer, larger wins.

Because every safe pick is priced at the same 99% return, cashing out early and continuing are equal in expectation. The decision is therefore about variance and discipline.

• Decide the target before the bet. For example, three mines, four picks, cash out. Write it down if needed.
• Do not move the target mid-round. The pull to "one more tile" after a run of safe picks is the mind reacting to a streak, not to any change in odds. Your next tile has the same chance as before, and the risk you carry grows with each one because the stake at risk is the whole accumulated multiplier.
• Keep the target consistent across sessions so you can judge results against the table above. If you target 2× at three mines, expect to succeed about half the time and lose the rest.

The cash-out button works on any safe pick.

Mine positions are produced by the provably fair seeds before you click. Every tile has exactly the same chance of being a mine. That kills several popular tactics:

• Corners are safer. They are not. Corners are chosen as mines at the same rate as the centre.
• Zigzag or spiral patterns. The pattern you click is irrelevant to where the mines are.
• Avoiding tiles where mines appeared last round. Rounds are independent, with a new nonce each time.
• Using auto-pick vs manual. There is no statistical difference.

According to Stake's fairness system, you can verify any round after rotating seeds, which shows the mine layout was fixed before you started. The provably fair guide shows how.

Since the edge cannot be beaten, bankroll rules aim to keep the cost bounded.

1. Set a session budget first and treat it as spent.
2. Stake a small slice of it. At 1% of budget per round you get a long session. At 20% you can lose it in five bad rounds, and with several mines that happens regularly.
3. Choose stop-loss and stop-win levels in advance. Stop when either is reached.
4. Watch turnover. At about 1% edge, every 1,000 wagered costs roughly 10 on average. Quick one-pick rounds pile up turnover faster than you notice.
5. Use account tools. Deposit limits and cool-offs are in settings, and the responsible gambling guide explains them.

The same ideas apply to other Originals. The Plinko strategy guide works through them for a game where the risk is set by the board rather than your clicks.

Martingale

Doubling after each loss assumes you can always cover the next bet. A run of mines is far from rare, especially at higher mine counts, and the stakes escalate quickly. The system turns many small wins into one large loss.

Auto-bet on a fixed number of picks

Auto removes the emotion but not the edge. It helps with discipline only if you set a finite number of bets and a stop-on-loss.

"Safe" one-mine, one-pick grinding

It looks safe because it wins 96% of the time, but a single loss costs as much as roughly 33 wins at 1.03×. The edge is intact.

Chasing a streak

Seeing five safe picks in a row and raising the stake changes nothing about the next tile. It only raises what you can lose.

Copying streamers

What you see is a filtered record of wins. Long losing runs rarely make the highlights. The streamers guide explains how to read that content with care.