Mines on Stake, explained properly.
Stake's Mines is the reference implementation most other crypto casinos are copying: a 5×5 grid, 1–24 mines, one multiplier step per safe reveal. It's also the one with the most public documentation, so it's the easiest to verify against the Mines simulator — same combinatorics, applied to exactly how Stake sets the board up.
Stake Originals, not a third-party title.
Mines sits under Stake Originals — games Stake built and hosts itself, rather than licensing from a studio. That matters because it means one Mines implementation, one house edge, one fairness system, not a library of differently-tuned versions under the same name.
Choose a bet and a mine count from 1 to 24, then start the round. Stake commits to the mine layout before you click anything — it's generated and hashed up front, so nothing about it can change once the round is live. Each safe tile you reveal moves the multiplier to the next value on the curve; hit a mine and the round ends. Cash out any time after your first safe reveal to bank whatever the multiplier reads at that moment.
You can verify every single round yourself.
Three inputs, one hash
Every round runs through HMAC-SHA256 with three inputs: a server seed Stake generates, a client seed you can set yourself, and a nonce that increments per bet. Same three inputs, same mine layout, every time.
The server seed is committed first
Stake shows you the hashed version of the server seed before you bet — a fingerprint that can only match one exact seed. That fingerprint can't be changed retroactively, so the layout can't be picked to beat you after the fact.
You reveal it to check
In Settings → Fairness you can rotate your seed pair, which reveals the previous (now-unhashed) server seed. Re-run the same HMAC-SHA256 calculation and you get the same mine layout Stake used — match confirmed, or it wasn't fair.
Provably fair proves the round wasn't tampered with — it doesn't remove the house edge. The 1% is baked into every multiplier regardless of which seed pair produced the layout.
The 1% figure above is the widely-cited convention for Stake's Mines and is what every calculation on this site uses. We haven't independently re-derived Stake's exact internal constant round by round — if Stake ever publishes a different number, the underlying formula scales cleanly to match it. There's no site-wide minimum bet — it depends on which currency you're playing with, since Stake supports dozens of crypto and fiat-equivalent balances.
Questions specific to Stake's version.
Does Stake's Mines use the same math as this calculator?
Yes — Stake's Mines is the standard 5×5, 1–24 mine implementation this whole site is built around, at the widely-cited 1% house edge. Every multiplier here is 0.99 × C(25,k) ÷ C(25−m,k), the same formula Stake's published multiplier tables match.
How are Stake's Mines multipliers actually calculated?
Every multiplier comes from one formula: how unlikely your reveal was, repriced to bake in the 1% edge.
The formula
- Win chance for m mines and k safe reveals is C(25−m, k) ÷ C(25, k) — the odds all k picks avoid every mine.
- Fair multiplier is 1 ÷ win chance — what the payout would be with zero house edge.
- Actual multiplier is the fair multiplier × 0.99, which is where the 1% edge gets applied.
Worked example — 3 mines, 3 tiles revealed
- Ways to pick 3 safe of 22: C(22,3) = 1,540
- Ways to pick 3 of all 25: C(25,3) = 2,300
- Win chance: 1,540 ÷ 2,300 = 66.96%
- Fair multiplier: 1 ÷ 0.6696 = 1.49×
- With 1% house edge: 1.49× × 0.99 = 1.48×
Does Stake publish a full Mines payout table?
Not that we've found as a single official document — but it doesn't need to. Since the game is provably fair and the formula above is public, the full table is fully derivable rather than secret — this site computes it directly from that formula for every mine count.
Can I see the mine layout before I start clicking?
No, and that's the point. The layout is generated and its hash committed before your first click — hidden from you but locked in, so Stake can't change it once the round starts and you can't see it in advance either. You verify it was fair afterward, using the seed-checking process described earlier in this guide.
Is Stake's Mines rigged?
The provably fair system exists specifically so you don't have to take that on trust — every round's outcome is reproducible from the server seed, client seed, and nonce using public HMAC-SHA256 math. What it doesn't change is the built-in 1% house edge, which applies whether you win or lose any individual round.
Why does my result feel different from the stated 1% edge?
The 1% is a long-run average across thousands of rounds, not a per-session guarantee. A handful of rounds will always look nothing like the long-run number — that's what "edge" statistically means, not a promise about any one session.
Is there a best mine count to use on Stake?
No — every mine count carries the identical 1% edge, so "best" comes down to what ride you want, not an edge you can find. Low counts mean frequent small wins; high counts mean long droughts and rare, huge multipliers. The full trade-off is broken down on the Mines strategy guide.