Crypto Hi-Lo in 2026: Every Card's Odds, Worked Out

Crypto hi-lo is a card game with one question on repeat: will the next card be higher or lower? You start with a card face up, make your call, and if you're right your stake is multiplied. Then you can cash out or keep going, stacking multipliers into a ladder.

It's an old parlour game rebuilt for crypto casinos, with a provably fair deck and a payout formula you can check on every card. Because the odds change with every card shown, it feels like a game of judgement. Mostly, it's a game of arithmetic.

We popped the hood on the card odds, the multiplier formula, the skip button and the ladder, with worked examples you can follow. The free-play demo above uses play money only, so you can test a few ladders before deciding whether hi-lo suits you.

Where to play crypto hi-lo in 2026

Hi-lo is usually an in-house "original" at crypto casinos, listed with dice, limbo, mines and plinko. Some studios also make card-based higher-or-lower games for the slots lobby, and a few live studios run shows built on the same idea.

The list below draws on our full Haxies Index across 17 casinos, with partner casinos featured first and badged. The Index scores the cost of clearing each welcome offer, each operator's stated payout speed and the licensing and ownership details it publishes. It doesn't score hi-lo, and our brand data doesn't confirm which casinos carry it, so check the lobby first.

Two brands in our data market provably fair originals. BC.Game lists 74 of them, including a Crash game at a stated 99% RTP. Betpanda also lists provably fair originals, and the Betpanda review records its published withdrawal caps. Neither entry names a hi-lo game, so treat any you find as something to check.

The casinos with the highest Index scores on our list of the best crypto casinos pair cheap-to-clear offers with quick stated payouts. The hi-lo edge itself comes from the game's multipliers, which you can check card by card.

How does crypto hi-lo work?

In crypto hi-lo, a card is shown and you bet on whether the next card will be higher or lower. Each rank from Ace to King has a fixed chance of being beaten either way. Your multiplier is (1 minus the house edge) divided by the chance of your call. Win and you can cash out or continue; lose and the ladder ends.

That's the snippet. Now the mechanics.

The deck

Most crypto hi-lo games use an endless deck. Each new card is drawn independently from all 52 cards, so the same card can appear twice in a row. That keeps the odds simple: every rank has a 1 in 13 chance on every draw, whatever came before.

A few games use a single finite deck that runs down. Those are rarer online, and the rules panel will say so. Everything on this page assumes the common endless deck.

Ranks, not suits

Ace is usually low and King high, giving 13 ranks. Suits normally don't matter for the higher or lower call. Some games offer side bets on colour or suit, which have their own odds.

Ties

Every game has to decide what happens when the next card is the same rank. The most common design pairs "higher or same" with "lower or same", so a tie wins either call. At the extremes, where one of those would be certain, games switch to "higher" only or "same" only. Read the rules panel once and you'll know which version you're on.

The ladder

After a correct call, your winnings ride on the next card if you continue. Multipliers stack: two wins at 1.5x make 2.25x. At any point between cards, you can cash out the whole running total.

Card odds and multipliers in crypto hi-lo

Here is the full table for the common "higher or same / lower or same" design with an endless deck and a 1% house edge. Ace is 1, King is 13.

On an Ace, "lower or same" can only win on another Ace, so it behaves like a "same" bet at 1 in 13. On a King, the same is true for "higher or same". That's where games swap in a strictly higher or a "same" option.

How the numbers are built

With 13 equally likely ranks, the chance that the next card is the same or higher than a card of rank r is (14 minus r) divided by 13. The chance it's the same or lower is r divided by 13.

Take a 4. Higher or same covers 4 through King, which is 10 ranks. That's 10/13, about 76.92%. A fair game would pay 1/0.7692, which is 1.3x. At a 1% edge it pays 0.99/0.7692, which is 1.287x.

The formula

Multiplier = (1 - house edge) / probability of your call

Turn it round to check any game:

House edge = 1 - (multiplier x probability)

On a 7, the chance of higher or same is 7/13, about 53.85%. If the game pays 1.8386x, the edge is 1 minus (1.8386 x 0.5385), about 1%. If it pays 1.80x, the edge is about 3%. One check on a 7 tells you the price of the whole game.

The ladder: how stacked calls really pay

The ladder is what makes hi-lo feel different from dice. You aren't placing separate bets. You're letting one stake ride through several calls.

Worked example: three calls

You stake $10 and the first card is a 4.

1. You call higher or same. Chance 10/13, multiplier 1.287x. The next card is a 9, so you win. Running total: $12.87.
2. On the 9, you call lower or same. Chance 9/13, multiplier 1.43x. The next card is a 3. Running total: $18.40.
3. On the 3, you call higher or same. Chance 11/13, multiplier 1.17x. The next card is a Queen. Running total: $21.53.

Your combined multiplier is 1.287 x 1.43 x 1.17, about 2.153x. The chance of getting through all three calls was (10/13) x (9/13) x (11/13), about 45.06%.

The cost of each step

Multiply those two together: 2.153 x 0.4506 is about 0.970. Your expected return on the original $10 is $9.70, an expected cost of about 3%.

That's 0.99 x 0.99 x 0.99. Each step charges 1% of whatever is riding at that moment. Three steps cost about 3% of the starting stake. Ten steps cost about 9.6%, because 0.99 to the power of 10 is about 0.904. Twenty steps cost about 18%.

Why that matters

A long ladder is not a cheaper way to reach a big multiplier. Reaching about 2x through one call on an 8 (2.145x) and reaching it through five calls at 1.17x each (about 2.19x) cost different amounts. The single call charges 1% once. The five-step route charges about 1% five times on a growing balance, roughly 4.9% in total.

The lesson is simple. Each extra card is another bet of the whole running total. If you want a big multiplier, fewer, riskier calls are cheaper than many safe ones.

Safe steps compound slowly

Calling higher or same on a 2 wins 92.31% of the time and pays 1.0725x. Five such calls in a row reach about 1.42x, with a 67% chance of getting there. The expected return is still 0.99 to the power of 5, about 0.951. Safe steps feel free. They're the most expensive way to grow a balance per unit of multiplier.

Skipping cards

Most crypto hi-lo games have a skip button. It discards the current card and draws a new one without a bet. Some games allow unlimited skips, some limit them.

What skipping does

Skipping lets you choose which card you bet on. A 7 is the least pleasant card in the deck: about 53.85% either way. Skip it, and the next card might be a 2 or a Queen, where one direction is very likely.

What skipping doesn't do

Skipping doesn't change the edge. Every call, on every card, carries the same 1% because each multiplier is built from the true probability. A 92% call on a Queen pays 1.0725x for exactly that reason.

Think of skips as a menu, not a discount. You can choose between a likely small win and an unlikely big one on any card. You can't choose a cheaper bet.

When skipping is useful

Skipping helps you keep your ladder to a shape you planned. If you want each step to be about 75% or better, skip any card from 5 to 9 and bet only the safer direction on the rest. If you want fast, big steps, skip extreme cards and bet the long side. Neither plan changes the price per step.

Cash-out discipline

The cash-out button is the most important control in the game. Each card you leave your balance on is another full bet at the house edge.

Pick your exit before the first card

Choose a target before you start: a number of steps, or a multiplier. "Cash out at 2x or after three wins, whichever comes first" is a good plan. Write it down if you need to. In the moment, a 2.15x balance on a 6 always looks like it wants one more call.

Use autobet ladders

Some games let you set an automatic pattern: a call rule for each card and a cash-out target. That removes the in-round temptation, which is where most hi-lo losses really come from.

Expected cost per session

Your total expected cost is roughly the sum of every amount you had riding times the edge. Short ladders with frequent cash-outs keep that number close to stake times rounds times 1% per step. Long ladders make it grow faster than it looks.

Crypto hi-lo vs other instant games

Hi-lo has relatives across the instant lobby. The main difference is that you make a decision between every step.

Mines is hi-lo's closest cousin. Both ladders charge the edge each step and tempt you to take one more. Limbo and dice charge it once per bet, which is why reaching a given multiplier in one limbo bet is cheaper than climbing to it in hi-lo. The spread of edges across crypto casino games is wide, and instant originals at around 1% are among the cheapest bets on offer.

Provably fair hi-lo: checking the deck

Many crypto hi-lo games are provably fair, so you can prove each card wasn't changed after you made your call.

• Server seed: secret, from the casino. You see only its hash before play.
• Client seed: yours to set.
• Nonce: a counter that rises with each bet on that seed pair.

The game hashes these together and turns the output into a sequence of cards. Within a round, a cursor usually moves along the output so each new card comes from a fresh slice of the hash.

How to verify

1. Copy the hashed server seed before playing.
2. Set your own client seed.
3. Play your rounds.
4. Rotate the seed pair to reveal the old server seed.
5. Hash the revealed seed with SHA-256 in an independent tool. It must match.
6. Run the seeds and nonce through a verifier and compare each card with your history.

What the proof covers

It shows the cards were fixed before your calls and not altered. It doesn't tell you the multipliers are fair. The check on a 7 earlier on this page does that. Solvency and payout speed are separate questions again, which is what our Index measures.

Strategy myths

Hi-lo's in-round decisions make it feel like a skill game. A few beliefs follow from that.

"Card counting works"

On an endless deck, every draw is independent, so there's nothing to count. On a finite single deck there would be something, but those are rare online and the multipliers would then update to reflect the remaining cards.

"Always pick the likely side"

Picking the likely side wins more calls, but each win pays less. The expected return per call is 99% either way. Calling the likely side is a variance choice.

"I'm on a streak"

Five correct calls don't change the odds on the sixth. The next card is drawn exactly as the first one was.

"Martingale on the first call"

Doubling your stake after each lost ladder has the same flaw as everywhere else: the edge is untouched, and a run of losses gets expensive very fast. At a 50% call, ten straight losses cost 1,023 units on a $1 start.

Bonus wagering and hi-lo

Hi-lo is a poor way to clear a welcome bonus. Most casinos class it as "other games", and contribution is cut hard. In our brand data, Bets.io counts slots at 100% and other games at 5%, and Betpanda counts other games at 0%.

Worked example

A $100 bonus at 40x on the bonus means $4,000 of turnover, about $80 of expected cost on a 2% slot. At 5% contribution you'd need $80,000 of hi-lo turnover, about $800 at a 1% edge. At 0% it never clears. The wagering calculator runs the numbers for any offer.

Terms also tend to flag low-risk play as abuse, and repeatedly calling 92% sides is a good example. Clear bonuses on slots and play hi-lo with cash. Wager-free rebates are a better fit: Cocobet states 50% cashback on first-week net losses paid wager-free, and Winna offers instant rakeback. The full list of crypto casino cashback offers compares them.

Coins and cash-outs

Hi-lo stakes can be tiny, so pick a low-fee coin before you deposit. TRC-20 USDT and Litecoin are common low-fee choices, and Lightning cuts Bitcoin fees at the two brands in our data that confirm it, BC.Game and Betpanda. Our list of USDT casinos shows who confirms Tether.

After a long ladder pays out, withdrawal speed matters. Cybet, Cocobet, Winna, TrustDice, CasinoCrypto.io and Crypto-Games.io state instant payouts, and Jack states under 5 minutes. The ranking of fast payout crypto casinos compares every brand on stated speed and flags published caps. A first withdrawal may trigger verification even where signup didn't, so plan for it.

Mistakes to avoid

• Skipping the edge check. Look at the multiplier on a 7. It should be about 1.84x at a 1% edge.
• Climbing with safe calls. Many small steps compound the edge on a growing balance.
• No exit plan. Decide your cash-out point before the first card.
• Thinking skips are free value. They change which card you bet on, not the price.
• Playing on bonus money. Contribution is low and low-risk patterns can void winnings.
• Wrong network deposits. Check chain and address every time.

Is crypto hi-lo legal for you?

It depends on where you live, and sometimes on your region within a country. Each casino also publishes its own restricted-country list. Check both before signing up.

Licences differ across our 17 brands. Several publish Anjouan or Curaçao licence numbers, Winna lists the Tobique Gaming Commission, and Betpanda publishes no gambling licence, only a Costa Rica company registration. If verification matters to you, our list of no KYC crypto casinos records what each operator says about checks at signup and at withdrawal.

Keeping hi-lo fun

Hi-lo's rhythm, call after call with a growing balance, is designed to keep you in the chair. Crypto deposits remove the pause a bank transfer would give you.

Set a session budget and an exit rule before you start, and use deposit limits and cooling-off tools. Good habits for responsible crypto gambling start before the first deposit, and Gambling Therapy offers free, confidential support in many languages.

Crypto hi-lo questions

What is crypto hi-lo?

It's a card game where you guess whether the next card will be higher or lower than the current one, played with cryptocurrency. Correct calls multiply your stake, and you can cash out between cards or keep climbing.

How are hi-lo multipliers calculated?

Multiplier equals (1 minus the house edge) divided by the probability of your call. On a 4, higher or same has a 10/13 chance, so at a 1% edge it pays 0.99 divided by 0.7692, about 1.287x.

What is the house edge on crypto hi-lo?

It's set by the operator, and many originals run around 1%. Check it on a 7: the chance either way is about 53.85%, so a multiplier near 1.84x means 1%, and 1.80x means about 3%.

Does skipping cards improve my odds?

No. Skipping changes which card you bet on, but every call carries the same edge because each multiplier is built from that card's true probability. Skips are useful for keeping a ladder to the risk level you planned.

When should I cash out in hi-lo?

Decide before you start. Each card you continue on is another full bet at the house edge, so long ladders cost more. A fixed target, like a number of wins or a multiplier, is easier to stick to than a feeling.

Can you count cards in crypto hi-lo?

Not on the usual endless deck, where each card is drawn independently from all 52. Finite-deck versions exist but are rare, and their multipliers update to reflect the remaining cards.

Is crypto hi-lo provably fair?

Many in-house versions are. After rotating your seed pair, check the revealed server seed against the hash shown before play, then recompute each card from the seeds and nonce. That proves cards weren't changed, not that the multipliers are generous.

Does the demo on this page use real money?

No. The free-play demo above uses play money only. It's a safe way to see how ladders build and how often a long run ends on one unlucky card.