Limbo Game at Stake: How to Play, 99% RTP, Win Chance Maths
Pick a target multiplier. A number is drawn. If the number is at least your target, you win your stake times the target. That is the entire game, it takes under a second, and the maths is cleaner than anything else in the crash lobby.
Limbo at Stake
The original and still the one with the 99% RTP and the 1,000,000x cap. Crypto only, so check your country page first.
Every crash game is secretly Limbo with a cartoon on top. The crash point is drawn before the round starts, and the only thing the rising curve does is let you react to a number that already exists. Limbo removes the curve. You say the number you want, the game draws its number, and you find out instantly whether yours was lower. Stake built it in 2020 as one of its Originals, and it is the game we recommend to anyone who wants to understand what they are actually buying when they play crash.
What is the Limbo game?
Limbo is an instant game from Stake Originals, the in-house studio behind Stake's Crash, Mines, Plinko and Dice. There is no animation to speak of. The screen shows a target multiplier field, a stake field, a win chance field that updates as you type, and a large number that appears when you press Bet.
You set the target. Say 3.00x. The game draws a result from a distribution that runs from 1.00x up to 1,000,000x. If the result is 3.00x or higher, you are paid 3.00 times your stake. If it is 2.99x or lower, the stake is gone. The draw takes a fraction of a second, and you can bet again immediately. Experienced players use hotkeys and get through several rounds a second.
That speed is the first thing to be careful with. A 99% RTP is excellent, but at ten rounds a minute, the 1% edge works through a bankroll faster than Aviator's 3% does at two rounds a minute.
How to play Limbo on Stake
- Set a stake. Minimum 0.01 in most currencies at Stake, which is a tenth of the usual crash minimum. Maximum depends on the currency and your account limits.
- Enter a target multiplier. Anything from 1.01x to 1,000,000x. The win chance field updates instantly so you can see what you are buying.
- Press Bet. The result multiplier appears. Green means it cleared your target, red means it did not.
- Collect or go again. Winnings are credited immediately. There is no cash-out step because the decision was made before the draw.
Auto mode adds the usual tools: number of bets, stop on profit, stop on loss, and increase-on-win or increase-on-loss percentages. It is a Martingale machine if you want one, and it will show you precisely how Martingale ends.
Target 2.00x, win chance 49.50%, result 3.47x. The 0.50% missing from 50% is the edge.
Limbo win chance and payout table
The maths: win chance equals 99 divided by target.
This is the formula, and it is the whole game.
The probability that the drawn result is at least your target T is 0.99 divided by T. Stake shows it on screen, so there is no mystery. Set a 2x target and the field reads 49.50%. Set 10x and it reads 9.90%. Set 1,000,000x and it reads 0.0001%.
| Target multiplier | Win chance | Break-even chance you would need | Rounds per win, on average |
|---|---|---|---|
| 1.01x | 98.02% | 99.01% | 1.02 |
| 1.50x | 66.00% | 66.67% | 1.5 |
| 2.00x | 49.50% | 50.00% | 2 |
| 3.00x | 33.00% | 33.33% | 3 |
| 5.00x | 19.80% | 20.00% | 5 |
| 10x | 9.90% | 10.00% | 10 |
| 100x | 0.99% | 1.00% | 101 |
| 1,000x | 0.099% | 0.10% | 1,010 |
| 1,000,000x | 0.0001% | 0.0001% | 1,010,101 |
Two things to take from the table. First, the gap between your chance and break-even is 1% of your chance at every target. There is no target that is better value than any other. A 1.01x target loses 1% of turnover slowly; a 1,000x target loses 1% of turnover in long droughts with the occasional enormous win. Choose how you want to lose.
Second, the right-hand column is the one that stops people. A 100x target pays out once every 101 rounds on average. "On average" includes runs of 300 or 400 misses, which are not rare. At 1,000x you are looking at a thousand rounds per hit, and the maths does not care how long you have been waiting. Each draw is independent. The game would be broken otherwise.
The curve is the same shape as every crash game's. Limbo just shows it to you.
Limbo max multiplier: about that 1,000,000x
Stake caps the result at 1,000,000x, which is the highest ceiling of any game in the crash family. The chance of a draw landing there is one in a million, give or take. People do hit it, and the screenshots circulate. What the screenshots do not show is the stake size, which is usually the minimum, or the operator max-win limit that caps the payout. A 0.01 bet at 1,000,000x returns 10,000, which is a lovely story and, over the million rounds it took on average, cost 10,000 in stakes at a 1% edge to produce a net loss of around 100. The lottery has better marketing but the structure is similar.
Limbo provably fair, properly
Limbo's fairness system is the one that other crash games imitate loosely and Stake implements fully. Three inputs make each result: a server seed, which Stake generates and shows you as a hash in advance; a client seed, which you can set to anything you like and change whenever you want; and a nonce, a counter that goes up by one with every bet. The three are combined, hashed, and converted to a number between 1.00x and 1,000,000x using the formula from Stake's own documentation.
When you rotate seeds, the old server seed is revealed and you can recompute every result from that seed pair. Because you control the client seed, Stake could not have chosen a server seed to produce a specific outcome without knowing your input first. It is a better guarantee than the multiplayer crash games give, where the client seeds come from whoever bet first that round. We walk through the general mechanism in how crash games work.
Limbo strategy and auto mode tips
No target beats the 1% edge, so strategy in Limbo is about choosing a shape of losing and sticking to it. Three habits hold up. Fix the target before the session and do not move it after a miss; the player who drops from 5x to 1.5x "to recover" is paying the same edge on more turnover. Stake 1% of the session bankroll or less, because at a 3x target you will see runs of ten misses most hours. And put the stop-loss in auto mode rather than in your head, where it gets renegotiated.
Auto mode is also where the systems go to die. Set increase-on-loss to 100% at a 2x target and you have Martingale; a run of eleven misses, which happens roughly once every 1,400 rounds, needs 1,024 times the base stake and Stake's max bet will stop you first. Run it in demo for an hour if you want to watch. The stop-on-profit field is the useful one: decide a number, let the game stop you, and leave.
Limbo vs crash: which should you play?
Same distribution, no animation, better RTP. The honest comparison:
| Limbo (Stake) | Crash (Stake Original) | Aviator | |
|---|---|---|---|
| RTP | 99% | 99% | 97% |
| Decision | Before the round | During the round | During the round |
| Max multiplier | 1,000,000x | Uncapped in theory | Operator cap, often 10,000x |
| Round length | Under a second | 5 to 60 seconds | 5 to 60 seconds |
| Other players | No | Yes | Yes |
| Min bet | 0.01 | 0.01 | 0.10 |
If you play Aviator with auto cash-out set and never touch the button, you are already playing Limbo at a worse RTP with a longer wait. The curve, the chat and the live bets are what you give up, and for some people they are the point. For us, Limbo is where we go when we want to test a staking plan quickly and honestly, and crash is where we go to spend an evening. Spaceman and its 50% partial cash-out is the one crash game that offers something Limbo cannot replicate.
Where to play Limbo
Limbo is a Stake Original, so Stake is where the real one lives. Roobet and a handful of crypto casinos run their own clones with the same 99% formula. Both are crypto-first sites with restricted access in many countries, which is why there is no country list on this page: check the country pages for whether a crypto casino is a realistic option where you live, and what the deposit route looks like.
1Stake
The Stake Original: 99% RTP, 0.01 minimum, max multiplier 1,000,000x, provably fair with your own client seed.
2Roobet
Roobet's own Limbo clone runs the same formula at a 99% RTP; worth it if you already have a Roobet balance for Mission Uncrossable.
Limbo bonuses and free play
Stake does not attach a bonus to Limbo specifically. What exists is the site-wide offer set: a welcome or reload bonus with wagering that counts Originals at a reduced weight, the daily races and weekly challenges that pay from a leaderboard, and the rakeback programme on VIP tiers, which on a 1% edge game is a meaningful rebate. At Stake.us the game runs on gold coins for free and on sweeps coins for prizes. For free play anywhere, open Limbo with a zero balance; the formula, the win chance field and the result draw are identical.
Limbo verdict: pros and cons
For it: the 99% RTP, a formula you can read on screen, a provably fair system where you set the client seed, a 0.01 minimum and a result in under a second. Against it: no curve, no decision during the round, no other players, crypto-only access that rules out most countries, and a pace that puts a bankroll through the edge faster than any animated crash game. Limbo is for adults, 18+ or the local age of majority. The pace of the game makes deposit and loss limits more important than usual.
Frequently asked questions
How does Limbo work?
You enter a target multiplier, for example 2.00x, and a stake. The game draws a result multiplier at random. If the result is equal to or higher than your target, you win your stake times the target. If it is lower, you lose the stake. There is no cash-out decision during the round because the round is instant.
What is the RTP of Limbo?
99% at Stake, which is the version everyone copies. House edge is 1%. That beats every classic crash game: Aviator and JetX are 97%, Spaceman is 96.5%. The only crash-family game that does better is Evolution's Cash or Crash at 99.59%, and that one needs a live studio.
What is the win chance for a given target in Limbo?
Win chance equals 99 divided by the target multiplier. A 2x target wins 49.5% of the time, a 10x target 9.9%, a 100x target 0.99% and a 1,000x target 0.099%. The 1% difference between the chance and the break-even figure at each target is the house edge, and it is the same at every target.
What is the maximum multiplier in Limbo?
1,000,000x at Stake. The chance of a draw reaching it is 0.0001%, roughly one round in a million. The game lets you set that as a target, which means a 0.01 stake could in theory return 10,000, and in practice you would be clicking for a very long time. Operator max-win limits apply on top.
Is Limbo better than crash?
Better odds, yes: 99% beats 97%. Less fun, probably: there is no curve, no cash-out decision and no other players. Limbo is what crash becomes when you set auto cash-out and never touch the button, which is the correct way to play crash anyway. If you want the maths without the theatre, Limbo is the honest version.
Can you predict Limbo?
No. Each result is a hash of a server seed, your client seed and a nonce, and the server seed is committed before you bet. Past results carry no information about the next one. Any 'Limbo predictor' is a random number generator, and the only thing you can predict is the long run: 99 back for every 100 staked, across all players.
Can I play Limbo for free?
Yes. Stake lets you open Limbo with a zero balance in demo mode, and the sweepstakes version at Stake.us runs on free gold coins. The demo uses the same 99% formula, so it is an honest way to see how a 10x target behaves over 200 rounds before you risk anything. The one thing it cannot show you is how a losing streak feels with real money.
Is Limbo provably fair?
Yes. Each result comes from a server seed, your client seed and a nonce that increases with every bet. The server seed's hash is shown in advance and revealed when you rotate seeds, so you can recompute any result. You can change your client seed at any time, which is a stronger guarantee than most multiplayer crash games offer.
Daniel Reyes, Editor-in-chief
Daniel has written about online casino games since 2016 and has been playing crash games since the first Bustabit era. He owns the editorial standards on this site and signs off every global page. More from Daniel.