Mines Strategy (2026): Every Depth Returns the Same 99%

The table that ends the argument

On a three-mine board, for each depth: the chance of getting there, what the game pays, and the two multiplied together.

Cash out after	Rounds that reach it	Payout	Expected value
1 tile	88.0000%	1.1250x	0.990000
3 tiles	66.9565%	1.4786x	0.990000
5 tiles	49.5652%	1.9974x	0.990000
10 tiles	19.7826%	5.0044x	0.990000
15 tiles	5.2174%	18.9750x	0.990000
20 tiles	0.4348%	227.7000x	0.990000

DERIVED from payout(m, k) = 0.99 / [ C(25 – m, k) / C(25, k) ]. The last column is the point: it does not move.

STAKEThe build behind these numbers

Code PLAYSTORIES: rakeback on every wager, win or lose

See provably fair →

Disclosure: this is an affiliate link, so we may earn a commission if you sign up. It never changes a number on this page. 18+. Play responsibly.

Any plan built on cashing out at a particular tile is a plan about variance. That is a legitimate thing to choose, and it is worth choosing deliberately. It is just not a way to make the game pay more.

What a deep target actually costs you

Reading the second column as a bust rate makes the trade concrete. A ten-tile target on three mines busts in roughly four rounds out of five. A twenty-tile target busts in more than 995 rounds out of 1,000. Neither is a mistake and neither is clever; they are two different distributions of the same expected loss, and the deep one needs a bankroll that can survive long dry runs without changing your stake.

Mine count is the same trade in a different dress

Raising the mine count raises the payout at every depth and lowers the chance of getting there by exactly the same factor. Five mines and three revealed pays 1.9974x at 49.57%. Three mines and five revealed pays the identical 1.9974x at the identical 49.57%. The board is a dial for variance with no value setting on it.

The two mistakes that genuinely cost money

The first is raising your stake after a bust to recover it. The edge is charged per round on whatever you stake, so a larger stake buys a larger expected loss and nothing else; the board has no memory of the round that just ended. The second is playing a client whose payouts do not match the formula. That one is checkable in seconds and it is the only check on this page worth running before you play: multiply the offered multiplier by the chance of reaching that depth and confirm you get 0.99.

What a good session plan looks like

Pick the depth for the distribution you want, not for the return, because the return is identical.
Fix the stake before you start and leave it fixed, since no depth or streak changes what the next round is worth.
Check one payout against the formula on the client you are using, before the first real round.
Count the cost per hour, not the edge: a 1% edge on a fast game can outspend a 4% edge on a slow one.

Quick answers

What is the best Mines strategy? There is no best depth for return. Choose the depth whose win frequency you actually want, and keep the stake fixed.

Is it better to cash out early or go deep? Identical in expected value at 0.990000. Early wins often and small, deep wins rarely and large.

Does raising the mine count improve my chances? No. Payout and probability move by the same factor, so the expected value stays at 0.99.

Can a losing streak be recovered by increasing the stake? No. Each round is independent and the edge applies to whatever you stake, so a larger stake only raises the expected loss.

References

Derived from the rules of a 25-tile board; formula and tables published above, re-derived independently in exact fractions on 10 September 2026.
Our verification rules and open data: playstories.co/methodology.

Reviewed by the PlayStories editorial team. Last updated September 10, 2026. Numbers verified per our methodology. Informational only, not betting advice. 18+. Play responsibly.

How the numbers on this page were produced

payout = (100 - house edge) / win chance

Dice and Limbo payouts follow the published identity: multiplier equals 99 divided by the win chance in percent for a 99% RTP build. Expected value is therefore the same at every win chance, which is what fixes the house edge at 1%.

Full verification method · Editorial standards · Corrections are logged on the method page.

Primary sources checked for this page

Stake publishes its house edge and provably-fair method on its own site, which blocks automated retrieval, so we do not link it as a source. Every figure here is instead re-derivable from the published payout identity, and any finished round can be verified by the player from the seed pair.