Stake Limbo: how it works, odds and targets

Stake Limbo is an instant multiplier game. You choose a target multiplier — say 2× — and place your bet. The game produces a random result multiplier. If the result is equal to or higher than your target, you win your bet times the target. If it is lower, you lose the bet. The whole round takes a fraction of a second.

Limbo is often described as single-player Crash with the cash-out fixed in advance. There is no graph to watch and no button to press mid-round: the only decision is the target. Mathematically it is also a close cousin of Dice, just expressed in multipliers instead of a slider.

If your usual Stake address is blocked or slow on your network, the live mirror list shows working addresses with response times measured in your browser.

1. Enter your bet amount.
2. Set the Target Multiplier. Type any value from 1.01× upwards. The Win Chance field next to it updates instantly; you can also type a win chance and let the target update instead.
3. Click Bet. A result multiplier rolls up on screen. Green means it reached your target and you are paid at your target; red means it fell short.

Note that you are paid at your target, not at the result. If you target 2× and the result is 57.3×, you win 2× your bet. Higher results above your target do not pay extra — that is the defining trade-off of Limbo.

The relationship between target and win chance in Stake Limbo is:

Win chance (%) = 99 ÷ target multiplier

In a game with no house edge, the numerator would be 100. The 1% gap is the edge, and it is identical at every target.

Whatever target you choose, the average return per bet is 99% of the stake. A 2× target and a 1,000× target cost exactly the same amount per $100 wagered over the long run — around $1. The difference is entirely in how the results arrive.

The waiting-time rule for high targets

The "average bets per win" column hides how uneven waiting times are. For any target, the chance of going without a hit for:

• as many bets as the average is about 37%,
• twice the average is about 14%,
• three times the average is about 5%.

So at 1,000×, there is roughly a one-in-seven chance of 2,000 bets passing without a single win, and a one-in-twenty chance of 3,000. That is normal behaviour, not a cold machine. If you aim high, your budget has to cover those droughts — or you have to accept that it may not.

Low targets are not "safe"

A 1.01× target wins 98% of the time, which looks safe. But each win pays 1% of the stake and each loss costs 100%. One loss erases about a hundred wins. Over time the 1% edge applies just as it does at 1,000×.

Because each Limbo round is instant, auto-bet can place a very large number of bets in a short time. The Auto tab offers:

• Number of Bets, to cap the run.
• On Win and On Loss, each set to reset to the base bet or increase it by a percentage.
• Stop on Profit and Stop on Loss, to end the run at a cumulative figure.

Two practical points. First, turnover rises quickly at instant speeds, and expected loss is about 1% of turnover. Ten thousand $0.10 bets is $1,000 wagered and roughly $10 expected loss, whatever the target. Second, high-target auto-runs combined with an On Loss increase escalate fast: a 1% increase per loss at a 100× target roughly doubles the bet after about 70 straight losses, which is a common outcome. Set Stop on Loss before you press start.

Target ladders

Starting at a low target and stepping up after each win, or down after each loss. This rearranges bet sizes and targets but each individual bet still returns 99% on average, so the whole sequence does too.

Progression on high targets

At 10× or 100×, some players raise the stake by a small percentage after each loss so that a hit "covers" the drought. The maths works only until a drought outlasts the plan — and by the waiting-time rule above, long droughts are not unusual.

Martingale at 2×

Doubling after every loss at a 49.5% win chance. The same objections as in Dice: rare-but-certain streaks eventually exceed your balance or the bet limit.

Reading the history

Recent results do not affect the next one. Each result is calculated from your seeds and a nonce, independently of what came before.

Hunting the maximum

The 1,000,000× target has a win chance of roughly one in a million. Treat it as a lottery ticket, and be aware that Stake's maximum profit per bet may cap the payout before you reach the full multiplier at higher stakes.

No strategy changes the 99% return. A method can make results steadier or more extreme, nothing more.

Limbo uses the same provably fair system as Stake's other single-player Originals:

1. Server seed. Stake shows you a hash of its server seed before you bet, committing to it without revealing it.
2. Client seed. You can set your own client seed at any time in the Fairness settings.
3. Nonce. A counter that increases by one with every bet.
4. Result. These values feed a published HMAC-SHA256 calculation that produces a number, which is converted into the result multiplier with the house edge applied.
5. Verify. Rotate your seed pair to reveal the old server seed. Check that its hash matches what you were shown, then recompute any past result with Stake's documented method or an independent verifier.

The glossary has short definitions of each term if you are new to provably fair games.

• Match the target to the budget. As a rough guide, keep enough budget for at least three times the average bets per win at your chosen target and stake.
• Use the win chance field. It turns any target into a plain probability — read it before you bet.
• Limit auto-bet speed by limiting runs. Use Number of Bets and Stop on Loss together.
• Do not chase with higher targets. Moving from 2× to 20× after losses increases variance without improving the return.
• Keep winnings separate. Moving profits into Stake's Vault makes it easier to stop.

Stake's deposit limits, break and self-exclusion tools are explained on our responsible gambling page. 18+.