Stake Mines: Probability Math, Optimal Cash-Out and Strategy

You choose a bomb count (1–24 bombs out of 25 tiles), then click tiles one at a time. Each safe tile raises your potential payout multiplier; clicking a bomb ends the round and forfeits your bet. You can cash out anytime to lock in the current multiplier × your bet.

The multiplier at each step is calculated to maintain a roughly 99% RTP across the game (1% house edge). With 3 bombs and one tile uncovered, your multiplier is about 1.13x. With 3 bombs and ten tiles uncovered, it's about 7.5x. The closer you push the more you multiply — but the chance of hitting a bomb on the next click rises with each safe pick.

With B bombs and N tiles remaining, the probability that your next click is safe is (N - B) / N. With 3 bombs on a 25-tile grid, the first click has a 22/25 = 88% safety chance. After 5 safe picks (20 tiles remain, 3 still hiding bombs), the next click is 17/20 = 85% safe. After 15 safe picks, it's 7/10 = 70%. Each safe pick reduces your next-click survival rate.

Cumulatively, the probability of getting 10 safe picks in a row with 3 bombs is 22/25 × 21/24 × 20/23 × ... × 13/16 ≈ 7%. That's roughly the multiplier the game pays — about 13x — adjusted for the 1% house edge. The math is self-consistent and provably fair.

In a strict EV sense, every cash-out point has the same expected value (your bet × 0.99). Just like Crash and Plinko, the multipliers are constructed so the house edge is constant regardless of how aggressive you play. There is no mathematically 'better' stopping point.

In a variance sense, cashing out at 2–3 safe picks gives you a high-win-rate, low-payout profile (similar to low-risk Plinko). Pushing to 10+ picks gives you a low-win-rate, high-payout profile. Most disciplined Mines grinders settle around 3–5 picks for the sweet spot between meaningful payouts and survivable variance.

With 1 bomb on a 25-tile grid, almost every pick is safe but multipliers grow slowly — you'd need 20+ safe picks to make meaningful money, and even then you'd cash out at 5–10x at most. With 20 bombs, you have a 5/25 = 20% chance of surviving your first click, but the multipliers explode quickly.

Higher bomb counts produce higher variance and rarer big wins. Lower bomb counts are smoother but cap your upside. The choice is purely a variance preference. The house edge stays constant at roughly 1% regardless of bomb count.

Stake's Mines auto-bet lets you select a fixed bomb pattern (e.g. always click tiles 1, 2, 3) and run hundreds of rounds with stop-loss and stop-profit conditions. This is the only practical way to use Mines for wagering grinding — manual clicking is slow.

The strategic implication: pick a tile pattern with 3–5 picks, run a long auto-bet session at fixed bet size, walk away at predetermined stop conditions. Mines becomes another high-RTP grinding tool with predictable bankroll behaviour.

The bomb placement is randomly determined by the provably fair seed at the start of each round. Bombs are equally likely to be in any tile. Strategies like 'always pick the corners' or 'pick the diagonal' have exactly the same EV as any other pattern — you're just choosing which tiles to reveal.

Confirmation bias makes these patterns feel real because hits feel memorable. Run 1,000 rounds with any pattern and the result converges to the underlying 1% house edge regardless. Pattern superstitions are pure cognitive illusion.

Mines counts as a Stake Original and contributes 100% toward bonus wagering, race entries and raffle tickets. The 99% RTP makes it a competitive choice for grinding alongside Plinko and Dice — better than slots (96–97% RTP) for pure wagering efficiency.

For race targeting, Mines can produce big single-round multipliers when you push to 15+ safe picks on high bomb counts, which can land you on race leaderboards. The probability is brutal but the upside is real. See our originals guide for the cross-game race strategy.