Limbo Strategy: Targets, Variance, Expected Value
Limbo picks a random multiplier from 1x to infinity. Here is how to choose a target, understand the variance at each level, and not blow your bankroll chasing big numbers.
Limbo sets a random multiplier before each round starts. You pick a target threshold. If the result meets or beats it, you win the multiplier on your stake. The house edge is 1% at quality crypto casinos and it is the same at every target. High targets are not better value, they are just higher variance. The strategy is entirely about variance management: how many losing rounds can your bankroll absorb before the hit you need arrives.
What it is
The screen is clean. A number field where you enter your target multiplier, a stake field, and a button. You set a target of, say, 3x, place your bet, and the round resolves instantly with a revealed multiplier. If the result is 3.00x or above, you win 3x your stake. Below 3x, you lose your stake.
There is no manual cash-out, no building tension. The result is already determined when you hit play. You are betting on whether a random draw from an exponential distribution lands above your line.
Most platforms display the last N results and running statistics. The visual is minimal compared to Crash and Plinko, which makes Limbo the preferred format for players who want to run systematic auto-bet strategies without the distraction of animations. The auto-bet feature is central to how most Limbo volume is played: set a target, set a round count, set a stop-loss, and let it run.
Limbo is one of the nine original casino games tracked on this site and available at every major crypto casino including Stake, Roobet, and Shuffle.
The math
Limbo's underlying distribution is exponential. The probability of the random result exceeding a target multiplier T is:
P(win) = (1 / T) x RTP
At 99% RTP:
| Target | Win probability | Payout | Expected value per unit |
|---|---|---|---|
| 1.5x | 66.0% | 1.5x | 0.990 |
| 2x | 49.5% | 2x | 0.990 |
| 5x | 19.8% | 5x | 0.990 |
| 10x | 9.9% | 10x | 0.990 |
| 50x | 1.98% | 50x | 0.990 |
| 100x | 0.99% | 100x | 0.990 |
| 1000x | 0.099% | 1000x | 0.990 |
The expected value column is identical for every row. The house edge does not change with the target.
What does change is variance, measured as standard deviation per round. At 2x target, standard deviation is approximately 0.995 units per unit staked. At 100x target, standard deviation is approximately 9.95 units. The 100x target is not "better value," it is ten times the volatility for the same expected loss.
The expected number of rounds before a win at target T is approximately T. At 10x, you expect one win every ten rounds. At 100x, one per hundred. At 1000x, one per thousand. These are averages. The actual distribution is geometric, meaning you can easily go 300+ rounds without a 100x hit even though the expected wait is 100 rounds.
Strategy
The core insight: in Limbo, all your decisions are variance decisions. There is no target that has positive expected value. There is no sequence of targets that changes the fundamental math. What you can control is how your bankroll is exposed to the geometric waiting distribution.
Match target to bankroll depth. The geometric distribution of waiting times means high-target strategies require deep reserves. If you are targeting 100x and your bankroll is 50 units, a 100-round drought (which has about 37% probability) will clean you out before the hit arrives. A rough rule: your bankroll should be at least three times the expected waiting period in unit bets for your chosen target.
| Target | Min recommended bankroll (units) | Sessions survived at 1 unit/round |
|---|---|---|
| 2x | 10 | 90%+ of sessions |
| 10x | 50 | ~85% of sessions |
| 100x | 500 | ~80% of sessions |
| 1000x | 5,000 | ~80% of sessions |
These are approximations based on the geometric distribution; actual variance will vary.
Auto-bet with a stop-loss. Set a fixed round count or a loss limit before starting. Limbo rounds resolve instantly and it is easy to click through 500 rounds manually without noticing how much has been spent. Auto-bet with a preset stop-loss is the mechanical floor on session losses.
Low targets for grinding, high targets for spikes. If you want to play for an extended session with predictable drawdown, 1.5x-2x targets give you wins roughly every two rounds and absorb variance smoothly. If you want a session defined by a single big hit, accept that most sessions will end in a net loss and size your stake accordingly.
Bet sizing at high targets. At 1000x target, you are betting that a very rare event occurs. If you bet 1% of your bankroll per round, you have roughly 100 rounds before depletion. The expected wait is 1,000 rounds. The math does not favor chasing 1000x on a thin bankroll. Either size the bet at 0.1% per round or lower the target.
Common mistakes
- Thinking high-target sessions "balance out" after a loss run. They do not, on any observable timeline. The geometric distribution is memoryless.
- Using Martingale on high targets. A 1000x Martingale requires 2^N times your stake on the Nth consecutive loss. After ten losses, that is 1,024x your original bet. Platforms have maximum bet limits. Your bankroll has a floor. The math breaks before the system recovers.
- Playing Limbo without a stop-loss. Fast rounds and instant resolution make it the easiest game to overplay in a session.
- Comparing wins across targets. A 50x hit on a 50x target and a 10x hit on a 10x target have the same expected value. Neither is "luckier" than the other.
- Not checking the RTP before playing on a new platform. Not all Limbo implementations are 99%. Third-party versions can sit at 96-97%.
Where to play it
The 100% RTP Limbo demo on this site runs zero house edge, which means you can test the geometric waiting distribution at any target without losing money to the edge. Set 1000x and run 500 rounds. It is informative.
For real-money play:
- Stake carries Limbo at 99% RTP with full auto-bet controls, session stats, and provably fair verification.
- Roobet includes Limbo with fast round resolution and a clean interface.
- Shuffle offers Limbo in its Originals suite with transparent RTP.
- Rainbet runs Limbo at competitive RTP with a simple round history display.
- Razed is one of the stronger newer rooms for Limbo with rakeback for consistent volume players.
---
FAQ
Q: What is the RTP on Limbo? A: At Stake, Limbo runs at 99% RTP. Most quality crypto casinos match this figure on their Originals. Some third-party Limbo implementations run lower.
Q: How is the Limbo result generated? A: The result is a random number on an exponential distribution from 1.00x upward. It is provably fair, meaning the result is derived from a hashed seed pair you can verify independently after each round.
Q: What is the best target multiplier for Limbo? A: There is no "best" target in a positive-EV sense. All targets carry the same 99% RTP. Lower targets (1.5x-2x) win more often but pay less. Higher targets (100x+) win rarely but pay more per hit. Match the target to your bankroll and variance tolerance.
Q: Is Limbo the same as Crash? A: Structurally similar but different in one key way. In Crash, you hold through a rising multiplier and choose when to cash out. In Limbo, the multiplier is already fixed when the round starts, and you set a threshold before it is revealed. There is no manual cash-out decision.
Q: Can Limbo go to a very high multiplier? A: Yes. The distribution has no theoretical upper bound. Multipliers above 1,000x occur but very rarely. The probability of hitting a 1,000x target is roughly 0.099%.
Q: What does the expected value calculation look like for Limbo? A: EV per round = (probability of winning x payout) minus stake. At 2x target with 49.5% win chance (99% RTP), EV = (0.495 x 2) minus 1 = -0.01 units per unit bet. The 1% house edge is constant regardless of target.
Q: How does Limbo differ from Dice? A: Both use RNG and both show you the exact odds before you commit. The difference is presentation. Dice uses a slider on a 0-100 number line. Limbo uses a multiplier threshold from 1x upward. The math is equivalent.
FAQ
What is the RTP on Limbo?
At Stake, Limbo runs at 99% RTP. Most quality crypto casinos match this figure on their Originals. Some third-party Limbo implementations run lower.
How is the Limbo result generated?
The result is a random number on an exponential distribution from 1.00x upward. It is provably fair, meaning the result is derived from a hashed seed pair you can verify independently after each round.
What is the best target multiplier for Limbo?
There is no "best" target in a positive-EV sense. All targets carry the same 99% RTP. Lower targets (1.5x-2x) win more often but pay less. Higher targets (100x+) win rarely but pay more per hit. Match the target to your bankroll and variance tolerance, not a hot-streak feeling.
Is Limbo the same as Crash?
Structurally similar but different in one key way. In Crash, you hold through a rising multiplier and choose when to cash out. In Limbo, the multiplier is already fixed when the round starts, and you just set a threshold before it is revealed. There is no manual cash-out decision.
Can Limbo go to a very high multiplier?
Yes. The distribution has no theoretical upper bound. Multipliers above 1,000x occur but very rarely. The probability of hitting a 1,000x target is roughly 0.099%.
What does the expected value calculation look like for Limbo?
EV per round = (probability of winning x payout) minus stake. At 2x target with 49.5% win chance (99% RTP), EV = (0.495 x 2) minus 1 = -0.01 units per unit bet. The 1% house edge is constant regardless of target.
How does Limbo differ from Dice?
Both use RNG and both show you the exact odds before you commit. The difference is presentation. Dice uses a slider on a 0-100 number line. Limbo uses a multiplier threshold from 1x upward. The math is equivalent.