Stake Mines, explained by the math, and why a two-cent stake already unlocks a six-figure jackpot

The three decisions that define every round

First, the number of mines. This sets both the shape of the payout curve and the risk of every click. Fewer mines means more safe tiles and a gentler climb; more mines means fewer safe tiles and a steeper climb where each successful click is rarer but pays far more. Second, the stake. Mines allows stakes as low as one cent on Stake, which turns out to matter enormously when you look at the peak multiplier — a 2-cent bet on a full-board clear is a six-figure payout. Third, when to cash out. Unlike Limbo, Dice, or Plinko, Mines asks you to make a live decision several times per round: every safe click is an implicit bet that the next click will also be safe.

Why Mines feels different from other Stake Originals

• Two games in one interface. Low mines and early cash-outs is a wagering engine that feels a lot like low-target Limbo — near-guaranteed small wins at 99% RTP. High mines with a chased clear is a jackpot hunt with slot-tier payouts. No other Stake Original gives you both modes from the same table.
• Real-time decision-making. Limbo and Dice are one-click rounds. Mines is a sequence: stake, click, evaluate, click, evaluate, cash out. The game rewards pattern discipline (sticking to a plan) and punishes tilt (clicking "one more tile" because the previous clears were easy).
• The payout grows non-linearly. On Limbo the multiplier is whatever you set. On Mines the multiplier is computed from combinatorics, which means the jump from "clear 10 with 12 mines" to "clear 11 with 12 mines" is a 4× jump (not a 10% jump). Understanding that shape is the entire strategic component of the game.

The two formulas on one screen

Multiplier after k successful clears on a board with N mines: M(N, k) = 0.99 × C(25, k) / C(25-N, k). Probability of reaching that clear without hitting a mine: P(N, k) = C(25-N, k) / C(25, k). The 0.99 factor is the only edge the house keeps — the theoretical "fair" multiplier would be C(25, k) / C(25-N, k), and Stake scales it down by 1% to produce RTP 99%. No dynamic weighting, no hidden adjustments, no configuration-specific tilts. The formula applies identically from 1 mine to 24 mines, from the first click to the last.

Why every configuration has the same expected value

Multiply the two formulas and watch them cancel: P × M = [C(25-N, k) / C(25, k)] × [0.99 × C(25, k) / C(25-N, k)] = 0.99. Every configuration — 2 mines early cash-out, 12 mines full clear, 24 mines single click — returns exactly 99 cents on the dollar in expectation. A low-mine session delivers that 99% as frequent small wins; a high-mine session delivers it as rare, massive wins. The total cost to the player over enough rounds is identical.

Where the peak multiplier comes from

• C(25, k) is maximized at k = 12 and k = 13. Both equal 5,200,300. This is a pure combinatorial fact about 25-element sets. Every multiplier in Mines is a scaled version of these binomial coefficients, which is why the peak of the curve sits precisely at "clear 12 or 13 safe tiles."
• Peak multiplier = 0.99 × 5,200,300 ≈ 5,148,297. Reached by exactly two configurations: 12 mines with a full 13-tile clear, or 13 mines with a full 12-tile clear. Both produce the same monster number because binomial coefficients are symmetric around the middle.
• Peak probability = 1 / 5,200,300 ≈ 0.0000192%. Or roughly one in 5.2 million attempts. A player running 200 peak attempts per day (aggressive hunt pace) would expect one clear every ~70 years on average — but the distribution is long-tailed, and verified peak hits happen several times per year across the full Stake player base.

Provably fair — you can verify every round yourself

Each Mines round is generated from a server seed (committed in advance), a client seed (you can rotate any time), and a nonce (the round counter). After a session, rotating the server seed reveals the prior seed; feeding it back through the public verification script reproduces every safe/mine outcome exactly. The math that turns a hash into a shuffled board is documented on Stake's fairness page, and the same mechanism powers Dice, Limbo, and Plinko. The 1% house edge is the only built-in tilt and it is exactly 1% — no configuration gets a secret surcharge, no nonce gets a stealth push, and no session gets a "tune" based on balance.

Wagering-first configurations (2-5 mines, early cash-out)

3 mines with a 3-tile cash-out: ×1.48 at a 66.96% hit rate. Put that in auto-mode and two out of three sessions end in profit; losing sessions bleed a few percent. This is the Mines equivalent of low-target Limbo — a disciplined wager engine that clears bonus rollover and VIP turnover at the 99% RTP floor without forcing your bankroll to ride real variance. 3 mines with a 5-tile cash-out tightens the payoff to ×2 at a 49.56% hit rate, which is almost exactly a coinflip at double — the cleanest-feeling version of the game to most players who want something more engaged than Limbo but not random enough to notice draw-downs.

Mixed-session configurations (5-8 mines, 5-8 clears)

5 mines with an 8-tile clear: ×8.50 at 11.65%. Small hit rate, meaningful payout, moderate losing runs. A 100-round auto session swings through real red and real green but tends to land near break-even with the 1% drag. This is the "I want to actually watch the game" mode — entertaining without destroying your bucket in a single bad run. Pair with a modest stake ($0.10-$0.50 on a $500 session bankroll) and the configuration delivers several hours of engaged play for a small expected fee.

Jackpot-hunt configurations (10+ mines, large clears)

• 10 mines, 5 clears. ×28.20 at 3.51%. A "warm-up" jackpot hunt — occasional clears keep the session alive, losing runs average out to the modest edge, and every so often the payout lands above ×50 when the multiplier ladder cooperates.
• 12 mines, 8 clears. ×831.98 at 0.119%. The transition line from jackpot-adjacent to true jackpot hunting. A clear pays for the next thousand losses — treat every single bet as a ticket rather than a session.
• 15 mines, 8 clears. ×23,794.60 at 0.00416%. Jackpot territory from a 2-cent stake. One clear in roughly 24,000 attempts returns about $476 — more than an entire month's entertainment budget for most disciplined players.
• 12 mines, full 13 clears. ×5,148,297 at 0.0000192%. The peak. A 2-cent stake hitting the peak returns close to one hundred thousand dollars. One in 5.2 million attempts, and real players hit it several times per year across the platform.

Why the cash-out decision matters on high-mine configs

On a 2-mine board the multiplier climbs gently; cashing out at 5 vs 7 vs 10 clears is a lifestyle choice, not a leverage choice. On a 12-mine board the multiplier doubles or triples with every subsequent clear. Cashing out one tile early on a hot streak is a 3-5× payout left on the table; cashing out one tile too late is a total loss. The jackpot-hunt configurations are where discipline — committing to the clear target before the round starts, and actually cashing out when you reach it — earns or burns real money. The math does not care; your bucket does.

The exact peak path

Drop 12 mines on a 5×5 board, leaving 13 safe tiles. Click all 13 safe tiles in sequence without hitting a mine. After each click the multiplier climbs: ×1.06, ×1.21, ×1.40, ×1.65, ×1.97, ×2.41, ×3.01, ×3.85, ×5.06, ×6.83, ×9.48, ×13.66, and finally ×5,148,297 on the thirteenth and last safe click. (The last click detonates the multiplier because at that point only mines remain on the board — a full clear probability of exactly 1 / C(25, 13) × C(13, 13) = 1 / 5,200,300.) Cash out and the payout lands. The formula is public; the Stake UI displays the multiplier-to-be-won at each step; the provably fair engine lets you verify the outcome was unmodified.

What a two-cent stake actually produces

• $0.01 stake × ×5,148,297 = $51,482.97. One cent on the peak is already a mid-five-figure payout.
• $0.02 stake × ×5,148,297 = $102,965.94. Two cents is the first full six-figure threshold — a jackpot number most progressive slots demand dollar-level minimum bets to even enter.
• $0.05 stake × ×5,148,297 = $257,414.85. A nickel on the peak is a quarter-million-dollar payout.
• $0.20 stake × ×5,148,297 ≈ $1,029,659. On Stake's standard max-win caps (typically $1,000,000 per bet across the platform), anything above a $0.20 stake on the peak configuration is capped to the million-dollar ceiling. The configuration pays jackpot-level from pennies and hits the platform cap from a quarter.

How this compares to slot jackpots

Progressive-jackpot slots advertise peak payouts of $1M-$10M but require qualifying bets that are usually one to two orders of magnitude larger than Mines' peak entry. A typical progressive slot demands a $1-$5 minimum bet with a roughly 1-in-20-50 million hit rate. Mines offers a ×5.1M payout from a $0.02 bet at a 1-in-5.2M hit rate — a meaningfully better ticket price at a meaningfully better hit rate. The difference is that progressive slots promote their jackpots aggressively while Mines sits quietly inside an Originals page because the game is nominally a low-edge casual tool. That marketing asymmetry is the only reason Mines is not constantly discussed in the same breath as jackpot slots.

Sizing the jackpot hunt honestly

The honest framing is unchanged from any serious lottery: size the total budget for the hunt, not the per-ticket stake. A player who decides "the Mines jackpot bucket is $50 per month" and runs 2-cent stakes on 12-mine configurations is buying 2,500 attempts per month at 1-in-5.2M odds per attempt. Over a year that is 30,000 attempts, or roughly a 0.6% chance of seeing the peak in that year. Over ten years: about 6%. Over a lifetime of 30-40 years of casual play: 20-25%. The math is honest about what it delivers — a small non-zero chance at a genuinely life-changing payout, paid for by a capped entertainment fee. That is the entire value proposition, and sized correctly it is a reasonable entertainment purchase. Sized incorrectly — paid from grinding capital, stretched with progressive staking, chased with higher mines when the bucket empties — it is the fastest way to torch a Stake account.

Why Mines clears rollover faster and cleaner than slots

• Four-times-lower house edge than most top-RTP slots. Mines at 99% vs typical slots at 94-96%. Same wagering counter, same bonus rollover math, four times less expected cost to complete the wagering requirement.
• No feature-bet or bonus-buy exclusions. Slots sometimes exclude free-spin or bonus-buy wagers from the count. Mines' every cash-out round counts fully.
• Variance control. A slot forces a single fixed-volatility experience per spin. Mines lets you dial the variance by picking the configuration. For bonus rollover, 3 mines with a 3-5 clear target at a fixed stake in auto-mode is a near-flat session line — the rollover clears at roughly 99% return per dollar wagered, and the bucket rarely moves more than a few percent from start to finish.
• Round speed. A slot at 10-20 spins per minute (including animations) lags behind Mines auto-mode at 40-60 rounds per minute on low-mine configs. For time-limited bonuses, that 3-4× speed difference matters.

The practical wagering configuration

Set 3 mines. Set the cash-out target at 5 safe clicks (×2 multiplier, 49.56% hit rate). Use auto-mode with a fixed stake equal to roughly 0.5% of the session bankroll — on a $500 session, that is $2.50 per round. Set auto-mode to run 1000-2000 rounds and enable stop-on-balance at ±20% to protect against the rare long red streak. A single clean session clears $2,500-$5,000 in wagering at a theoretical expected cost of $25-$50 against the house edge — the cheapest clean wagering turnover you can buy on Stake aside from low-target Limbo.

How the wagering bucket feeds the VIP stack

Every dollar wagered in that clean low-mine Mines engine moves the Stake VIP turnover counter forward at face value. Bronze unlocks at $10,000 turnover — about 3-4 auto-mode sessions. Silver at $50,000. Gold at $100,000. Platinum I adds reload at $250,000. The rakeback that starts at Bronze lands in the VIP wallet weekly at a rate set by your level; that rakeback can be withdrawn with no rollover because it has already cleared its own requirement. For the full VIP ladder breakdown, exact turnover thresholds, and how rakeback/reload/weekly boost stack, see the Stake VIP strategy guide — Mines belongs in the same rotation as low-target Limbo as the cleanest low-friction engine to build that turnover.

The VIP rebate stack in one paragraph

Bronze at $10,000 turnover unlocks rakeback and weekly boost — about 1-2 weeks of relaxed low-mine auto sessions. Silver at $50,000. Gold at $100,000. Platinum I at $250,000 adds a daily reload stacked on top of the existing rakeback and weekly boost. Each step up the ladder increases the rebate percentage and the reload amount. At Bronze the total rebate stream returns roughly 0.3-0.5% of all turnover; at Platinum it crosses 1% and in some cases exceeds the game's own house edge — which is the precise moment the Stake VIP machine transitions from "reduces your cost to play" to "actively pays you to keep playing." For the full per-level breakdown and the exact USD thresholds, see the Stake VIP strategy guide.

Why Mines pairs with VIP better than slots do

• Lower fixed house edge. 99% RTP vs 94-96% on typical slots. Same rebate rate, four times less bleed.
• No exclusions. Every Mines round counts fully toward VIP turnover. Some slots partially exclude feature or bonus-buy wagering from progression.
• Adjustable variance. For turnover you want flat variance so your session bankroll survives the stretch needed to accumulate wager volume. Low-mine Mines delivers that; most slots force a much wider variance per spin regardless of what you want.
• Auto-mode that actually produces the volume you target. 40-60 rounds per minute on 3-mine configurations, vs slots at 10-20 spins per minute including animations.

Stacking with the 3× sport multiplier

Stake's VIP program counts sport wagers as 3× toward progression (a structural gift baked into the Stake ecosystem). An optimal weekly rotation for a disciplined grinder is roughly 50% low-mine Mines or low-target Limbo + 25% sport (pre-match favourites, line-shopped) + 15% live dealer + 10% slots for variety. The Mines share keeps the turnover floor stable at 99% RTP; the sport share triples the progression speed per dollar wagered; the rest provides entertainment without meaningfully distorting the math. This is exactly the split recommended in the VIP strategy article — Mines and Limbo are interchangeable as the anchor leg, and many players alternate them session-to-session for variety.

Grind-bucket setup (90% of total Mines bankroll)

• Configuration. 3 mines, 5 safe clicks cash-out. ×2 payout at 49.56% hit rate. Flat-ish session curve.
• Stake size. 0.5% of session bankroll. $500 session → $2.50 per round. Fixed, never progressive.
• Auto-mode. 1000-2000 rounds per session, fixed stake, stop-on-loss at -20% and stop-on-profit at +30%.
• Cadence. 2-4 sessions per week. Every session moves VIP turnover by $2,500-$5,000. In a month the counter climbs by $20,000-$40,000 at an expected cost of $200-$400 against the house edge.
• Feedback loop. Weekly rakeback lands. Monthly weekly-boost stacks. VIP level climbs. The bucket grows slowly, visibly, predictably.

Jackpot-bucket setup (10% of total Mines bankroll, capped)

• Configuration. 12 mines with a full 13-clear target, or 15 mines with an 8-clear target. Both are deep moonshot territory.
• Stake size. Absolute minimum the platform allows. 2 cents per attempt. Never more.
• Session shape. 50-200 attempts per session, manual or auto. Accept that every session almost certainly ends in full loss; the prize is one in 5.2M attempts.
• Monthly cap. Total spend on the jackpot bucket is fixed up front ($20-$100 monthly is typical for disciplined players). When the cap is hit, the hunt stops until next month. Period.
• Expected outcome. Most months: small capped loss equal to the bucket. Some months: a mid-tier clear (×1,000-×100,000) pays a nice surprise. Once in a decade or longer: the peak clear and a six-figure payout.

Rules that keep the two buckets honest

Never fund the jackpot bucket from the grind bucket. Never scale the grind stake after a jackpot miss. Never chase a lost jackpot session with higher-mine attempts. Never merge the two into a single "Mines session" because the configurations are emotionally contagious — a few early jackpot clears will have you climbing the mine count, and a few grind losses will have you chasing recovery on the peak configuration. Both of those moves are terminal. The buckets are separate for exactly this reason.

Settings that should always be on

• Fixed stake size (no progressive). Auto-mode offers martingale-style "increase on loss" and anti-martingale "increase on win" options. Both break the variance math. For grinding, fixed stake always.
• Stop-on-loss at 20% of session bankroll. This is critical. A $500 session with a $100 stop-loss is never going to become a $500 session with a $400 loss. Auto-mode respects the stop exactly.
• Stop-on-profit at 30% of session bankroll. Equally important. An auto-mode run that never pauses eventually gives back its wins to variance. Lock a +30% day and leave.
• Round count cap. Set a maximum round count (1000-2000 is reasonable for 3-mine configs) as a second safety net under stop-loss and stop-profit. If neither triggers, the round count ends the session on its own.

Settings that should always be off

• Martingale (increase on loss). On 3 mines with 5 clears at 49.56% hit rate, a 7-loss streak happens roughly once per 300 rounds and requires 128× the original stake to recover — which almost always exceeds the session bankroll. Do not use.
• Anti-martingale (increase on win). Less catastrophic but still distorts variance. Every win grows into a larger stake which makes the inevitable loss a proportionally larger hit. Fixed stakes beat both progressive formulas in every long-run simulation.
• Unlimited round count with no stops. The fastest way to convert a planned $500 session into a $0 session. VIP rebate does not reward session-ending losses; it rewards consistent turnover with discipline intact.

Auto-pick patterns worth knowing

For grind-bucket auto-mode, the tile pattern is mathematically irrelevant — every safe tile pays the same in a provably fair deal. Pick any pattern (corners, diagonal, a neat row) and commit to it for the session. For jackpot-hunt manual play, a consistent pattern helps with provably-fair verification after the fact and mentally removes the "did I pick the wrong tile" rationalization after a bust. One common pattern on 12-mine hunt: click the four corners, then the four mid-edges, then the four interior cardinal points, then the interior diagonals — a sequence that distributes clicks evenly across the board and makes verification reviews quick.

Mines — the dual-mode engine

Strengths: only Original on the site that pairs low-friction wagering with a six-figure-from-pennies jackpot path. Multiple variance profiles from one interface. 99% RTP across every configuration. Weaknesses: slower round cadence than Limbo (the cash-out decision adds seconds); more clicks per round means more fatigue on long auto sessions. Best used as the anchor of a wager rotation with a side bucket for jackpot hunting.

Limbo — the minimum-attention volume engine

Strengths: the fastest auto-mode on the site (500-1000 rounds/hour at low targets), zero strategy overhead, single-click rounds. Weaknesses: no jackpot path comparable to Mines' peak — Limbo caps at ×1,000,000 but is capped for a full clear at x1,000,000 stake-to-payout which requires a much larger stake minimum to produce jackpot payouts. For the full breakdown see the Stake Limbo math guide.

Dice — nearly identical to Limbo, slight interface variance

Strengths: 99% RTP, published target mechanics, a shade more engaging than Limbo for players who like the over/under UI. Weaknesses: same lack of jackpot upside as Limbo; stays firmly in wagering-engine territory. See the full Dice math guide for the session blueprint.

Plinko — the adjustable-variance engine

Strengths: rows and risk slider let you dial the exact variance you want, from "near Limbo" to "wild as slots." Max multiplier at 16 rows / high risk reaches ×1,000+ but never approaches Mines' peak. Weaknesses: a little slower, and variance is distributed across every ball drop rather than concentrated into a jackpot path — fun to watch but not a jackpot hunt. See the Plinko math guide for row/risk comparison tables.

Mistake 1 — Mid-session mine-count creep

You start on 3 mines for grinding. Fifty rounds in a hot streak pushes the bucket up 8%. You think: "let me lock this in with a few quick ×5s on 8 mines." Now your variance profile is different, your stops were set for low-mine variance, and one bad streak later the session is in the red. The fix: decide the configuration before the session starts and do not change it until the session ends.

Mistake 2 — Using grind capital for jackpot attempts

The single most common terminal mistake. After a few dozen unsuccessful 12-mine attempts the jackpot bucket empties, and the player reaches into the grind bucket "just for another ten attempts." The grind bucket was designed for turnover at 99% RTP with tiny variance; it is not sized for a jackpot hunt's losing runs. One merged session can erase a month of disciplined grinding. The fix: hard wall between buckets. When jackpot money is out, the hunt is over until next month.

Mistake 3 — Chasing cash-out one tile too far

You are three clicks from a full clear on a 12-mine hunt. Every prior click was safe. The multiplier is already at ×1,400, payout already visible. One more click is another 3× on the payout. Every single one of those decisions, on average, gives back the expected house edge — that is how the formula works. The fix: commit to a cash-out tile count before the round starts. When you reach it, cash. If the round continues to clear after, that was the cost of discipline and the price of not going broke by one-more-tile-itis.

Mistake 4 — Treating the peak multiplier as a plan

It is a dream, not a plan. The hit rate is roughly 1 in 5.2M attempts. Sizing it as a lottery ticket works; sizing it as "this is what a path to wealth looks like" turns Mines from a wager engine with moonshot optionality into a disaster engine with wagering overhead. The peak is the bonus at the bottom of the page, not the business plan on top.

Mistake 5 — Running auto-mode without stops

Mentioned above and worth saying twice. Auto-mode with no stop-loss is the fastest way to lose a Mines bankroll. Set the stop at 20% of session bankroll, always. The rebate is earned along the way; the bankroll is what must be protected.