What is the best target multiplier in Limbo?

In expectation there is none: the chance of clearing target x is 0.99 ÷ x, so chance times payout is 0.99 at every target. A 1.1× target hits 90% of rounds; 1,000× hits 1 in 1,010 rounds and you should expect to wait about 1,010 rounds between hits. The average is identical; only the wait and the size of the payday differ.

Does martingale work on Limbo at 1.1×?

Worse than on a 2× game. Doubling only recovers a loss when the win pays 2×. At 1.1× a win returns 0.1 of the stake in profit, so recovering a lost stake needs ten times that stake next round: the ladder is 100, 1,100, 12,100 after one, two and three losses from a 10 base. A 1,000 bankroll affords 1 recovery, and a run of 2 losses at 1.1× appears in 84% of 200-round sessions.

Is Limbo the same bet as Dice?

Yes, in a different unit. Dice lets you pick a chance and pays 99 ÷ chance; Limbo lets you pick a payout x and wins with chance 99 ÷ x percent. Target 2× on Limbo is a 49.5% chance, exactly the 49.5% Dice roll that pays 2×. Everything in the Dice guide, including the martingale wall, transfers directly.

Limbo guidePublished 2026-09-27

Limbo strategy: the honest maths

Limbo strategy fits on one line: the chance of clearing any target x is 0.99 ÷ x. Multiply that chance by the payout x and you get 0.99, whatever target you type, so no target is better than another in expectation. What the target does decide is how long you wait between wins and how badly a loss-recovery plan scales. This guide computes both for the same formula the free Limbo game runs on, and links every plan to a simulator run you can reproduce.

How Limbo pays

The round draws one HMAC-SHA256 float and computes 1e8 ÷ (float × 1e8 + 1) × 0.99, floored to two decimals and never below 1.00×. If that number is at least your target, you are paid the target. Because the float is uniform, the probability that the result reaches x is 0.99 ÷ x: the 0.99 is the house edge, applied identically at every target.

What the target does

Limbo: chance of clearing each target, the wait, and the return
Target	P(clear)	Clear, one in	P(10 misses in a row)	Expected rounds per hit	Same as Dice at	Return
1.10×	90.0%	1 in 1.1	1.0e-8%	1.1	90%	99%
1.50×	66.0%	1 in 1.5	0.002064%	1.5	66%	99%
2.00×	49.5%	1 in 2.0	0.11%	2	49.5%	99%
3.00×	33.0%	1 in 3.0	1.82%	3	33%	99%
5.00×	19.8%	1 in 5.1	11.01%	5.1	19.8%	99%
10.00×	9.9%	1 in 10	35.26%	10.1	9.9%	99%
50.00×	2.0%	1 in 51	81.87%	50.5	1.98%	99%
100.00×	0.9900%	1 in 101	90.53%	101	0.99%	99%
1,000.00×	0.0990%	1 in 1,010	99.01%	1,010.1	0.1%	99%

Return is P(clear) × target: 99% on every line. The “same as Dice at” column is the win chance that makes Dice pay the same multiplier, which is how the simulator links below are built. At 1,000× the expected wait is about 1,010 rounds and ten misses in a row is close to certain.

What staking plans do at low targets

Martingale is built for 2× payouts, where doubling recovers a loss exactly. At a lower target the recovery stake is larger than a double. To win back L at a payout of x you must stake L ÷ (x − 1), so total exposure grows by a factor x ÷ (x − 1) per loss: 2× at 2×, but 11× at 1.1×.

Recovering losses at a 1.1× target from a 10 base bet
Losses in a row	Lost so far	Stake needed next to get back to even	Fits a 1,000 bankroll?
1	10	100	yes
2	110	1,100	no
3	1,210	12,100	no
4	13,310	133,100	no

Each loss at 1.1× multiplies the exposure by 11. A 1,000 bankroll can recover from 1 consecutive loss; the 2nd is the wall. At a 90% win rate that run occurs in 84% of 200-round sessions.

The simulator doubles after a loss, which at 1.1× does not even recover the previous stake, so its martingale at that target loses slowly on both the wins and the losses. Run martingale at the Dice equivalent of 1.1× and watch the median session drift down; then run it at 2×, where the wall from the Dice guide applies; then flat-bet the nearest Dice equivalent of 1,000× for 1,000 rounds, where most sessions never see a win and a few see a very large one. All three report the realised return on turnover near −1%.

Raising the target after a loss and waiting for a low result before betting high are the other common ideas. Each round is a fresh float with no memory, and every target has the same expectation, so both change the next round's shape and nothing about its worth.

Every Limbo target is the same bet wearing a different number. P(clear x) × x = 0.99, so a round's expected return is 99% of the stake at 1.1× and at 1,000×. A staking plan picks the stake before the round is known, so a session's expected loss is 1% of its turnover regardless of the plan. Low targets make the wait short and the recovery ladder steep; high targets make the wait long and the payday large. Neither moves the −1%.

Understood the maths? Craze runs the real thing.

Same published odds, real stakes. PlinkoLab is sponsored by Craze; whether to play for money is entirely your call.

Play at Craze

Questions

What is the best target multiplier in Limbo?: In expectation there is none: the chance of clearing target x is 0.99 ÷ x, so chance times payout is 0.99 at every target. A 1.1× target hits 90% of rounds; 1,000× hits 1 in 1,010 rounds and you should expect to wait about 1,010 rounds between hits. The average is identical; only the wait and the size of the payday differ.
Does martingale work on Limbo at 1.1×?: Worse than on a 2× game. Doubling only recovers a loss when the win pays 2×. At 1.1× a win returns 0.1 of the stake in profit, so recovering a lost stake needs ten times that stake next round: the ladder is 100, 1,100, 12,100 after one, two and three losses from a 10 base. A 1,000 bankroll affords 1 recovery, and a run of 2 losses at 1.1× appears in 84% of 200-round sessions.
Is Limbo the same bet as Dice?: Yes, in a different unit. Dice lets you pick a chance and pays 99 ÷ chance; Limbo lets you pick a payout x and wins with chance 99 ÷ x percent. Target 2× on Limbo is a 49.5% chance, exactly the 49.5% Dice roll that pays 2×. Everything in the Dice guide, including the martingale wall, transfers directly.

Keep going

Last reviewed 2026-09-27