Flat betting means staking the same amount at the same cash-out target on every round, with no adjustment for wins or losses. It cannot beat the house edge, but it is the lowest-variance way to play a crash game and makes your cost predictable.
If you have read our work on Martingale, you already know that no staking system beats the maths. Doubling after a loss does not change your expected outcome, it just changes how spectacularly you can blow up.
So if every system loses the same amount on average, the sensible question is not which one wins. It is which one loses most gently. That is flat betting, and this guide explains why, with the numbers to back it up.
The 30-second version
Flat betting is the same stake and the same cash-out target every round. It does not beat the house edge, nothing does, but it is the lowest-variance way to play and makes your cost predictable: roughly $30 per 1,000 rounds at $1 a round on a 97% game. The single most important rule is to never raise your stake after a win.
🎮 What flat betting actually is
Flat betting means staking the same amount, at the same cash-out target, on every single round, no matter what just happened. It is the simplest possible staking model, also called level or fixed staking. In a crash game that looks like a $1 stake with a 1.50x auto cash-out, repeated for the whole session.
It is the opposite of a progressive system. Martingale doubles your stake after a loss, Paroli doubles after a win, and Fibonacci and d’Alembert step the stake up and down on a schedule. Flat betting does none of that. The stake never reacts to the last result, because the last result has no bearing on the next one.
🔢 The cost is fixed before you start
Your expected loss is decided by the house edge and how much you wager in total, not by how you stake it. The formula is as simple as it gets.
📖 The formula
Expected loss = stake x house edge x number of rounds, which is the same as house edge x total amount wagered. At $1 a round on a 97% game, 1,000 rounds costs you $1 x 0.03 x 1,000 = $30. On a 99% game it is $10.
The crucial part is that this figure does not change if you switch staking systems. Martingale, Fibonacci, gut feel: all of them carry the identical expected loss, because you cannot combine negative-value bets into a positive-value session. We do not re-prove that here. The full derivation lives in our crash gambling maths guide, and the bigger question of whether any approach can win long-term is settled in can you beat crash games.
“Flat betting is the least losing way to play, not a winning one.”
One thing the formula assumes is a fixed, known RTP. That is the only lever that actually changes your expected loss, so it is worth picking the highest-RTP game you can find. The gap between a 95% and a 97% game is $2 more lost per $100 wagered. Our low house edge crash games guide ranks the best options, and you should always confirm the figure in the in-game info panel, since some titles run operator-configurable RTP.
📊 Why flat betting minimises variance
Among all the ways to stake the same total, flat betting produces the smoothest ride, and there is a clean proof of why. It is the one piece of maths that separates this from generic “flat betting is safer” advice.
Think of a session as the sum of its rounds. If round i stakes an amount s, that round’s variance scales with the square of the stake. Because the rounds are independent, the variance of the whole session is the single-round variance multiplied by the sum of all the squared stakes. The expected value, meanwhile, depends only on the total amount wagered, not on how it is split up.
💡 Key insight
For a fixed total stake spread over a fixed number of rounds, the sum of the squared stakes is at its smallest when every stake is equal. This is the Cauchy-Schwarz inequality, and it is exactly the condition flat betting satisfies. Any progressive system places bigger stakes on some rounds, which raises that sum and therefore the variance, while leaving the expected loss untouched. Same average loss, bumpier ride.
The practical effect is large. A $1 flat bet over 1,000 rounds loses about $30 expected whatever target you pick, but the spread of outcomes around that average depends heavily on the target.
The bottom line: the average loss is identical at both targets, but the 10x target swings nearly three times as wide. Lower targets give you frequent small wins and a tight, predictable session. The minus $30 is a long-run average, not a promise, so any single session can land well above or below it.
📝 For the record: flat betting minimises the variance of your outcome for a given amount wagered. It is not the best route if your goal is to maximise the chance of hitting a specific win target in a losing game, where the maths actually favours betting big and rarely. Those are different objectives. This guide is about the smoothest, most predictable ride, not about chasing a jackpot.
🎯 Flat betting against the other systems
Every staking system carries the identical expected loss. They differ only in how much variance and ruin risk they pile on top, and on that score flat betting is the cleanest of the lot.
The starkest contrast is with the system flat betting is the antidote to. After ten losses in a row, a Martingale player who started at $10 needs $10,240 on the next bet just to claw back to even, and the first long bust cluster ends the session. Flat betting has zero progression risk, which is the whole point of our Martingale guide.
“No staking system changes what you lose on average. It only changes how violently you get there.”
◆
💡 How to flat bet sensibly
Sensible flat betting comes down to one lever that matters, the game’s RTP, and a handful of discipline rules. If you are newer to the format, our beginner strategy guide covers the groundwork this builds on.
Pick the highest-RTP game
RTP is the only thing that changes your expected loss. Choose a 97% or 99% game over a 95% one, and verify the figure in the info panel.
Set a session budget and stake 1 to 2% of it
At 1% sizing you can absorb around 100 consecutive losses, at 2% around 50, at 5% only 20. Crash games throw 7 to 10-round losing streaks regularly, so 1 to 2% comfortably rides normal variance.
Pick one cash-out target and set auto cash-out
A lower target like 1.3x to 2.0x means more frequent wins and lower variance. Auto cash-out removes the timing decision, which matters when rounds last only seconds.
Set a stop and obey it
Stop at your budget or a time limit, whichever comes first, and set a stop-loss at, say, half your session budget. The discipline is the strategy.
⚠️ Important
After a winning round or a hot streak, do not raise your stake. This is the most-violated rule in flat betting and the single most important one. A win does not change the edge on the next round, because each round is independent. Raising stakes on a streak is not confidence, it is tilt with better marketing.
🧠 The real payoff is behavioural
The biggest benefit of flat betting is not mathematical, it is that it removes the decisions that get people into trouble. With nothing to adjust mid-session, there is no decision fatigue and no mechanism to chase a streak.
Because the stake never reacts to “hot” or “cold” runs, flat betting structurally blocks pattern-chasing, which is the trap we cover in our guide to crash game patterns. It also makes the cost transparent. If you lose eight bets in a row, you are down eight units, no more, and you can budget the cost the way you would a cinema ticket.
Chasing losses is one of the clearest signs that gambling is shifting from entertainment to something harder to control. We cover the warning signs, the research and where to get help in a dedicated guide: crash gambling and player harm.