RTP, variance, and the casino's edge
97% RTP doesn't mean you'll get back 97% over an evening. We explain what the return is, derive the rule 'chance of reaching ×m ≈ 0.97 ÷ m,' show why a cash-out strategy doesn't change the average, and use simulations to demonstrate what variance and the casino's edge are.
'97% return' is the most quoted and most misunderstood figure in crash games. It sounds almost like 'you'll lose only 3%.' In reality it's a statistical property over a huge distance, not a promise for your evening — and it's precisely those remaining 3% that make the casino profitable with a mathematical guarantee.
In this article we cover three things that the whole economy of the game rests on: what RTP and the casino's edge are, why no cash-out moment changes your average, and what variance is — the thing that makes a short distance so deceptive.
What RTP (the return) is
RTP (Return to Player) is the share of all bets placed that the game returns to players over a very large distance. An RTP of about 97% means: for every nominal $100 that passes through the game, on average about $97 comes back — and about $3 stays with the casino. The key words are 'on average' and 'over the distance': in an individual session you can double up or lose everything, and that doesn't contradict 97%.
An important caveat right away, which we'll return to: the 97% is calculated from turnover (the sum of all bets), not from your deposit. If you re-bet your winnings again and again, turnover becomes many times larger than the deposit — and with it grows the amount the casino's edge takes.
The casino's edge
The remaining 100% − 97% = 3% is the casino's edge (house edge). It's built into the very math of the game: the expected result of any bet is negative in advance and equals −3% of its size. On a single bet it's imperceptible (you either get lucky or you don't), but over the distance the law of large numbers does its work — the actual return converges to 97%.
Where this markup in a crash game comes from we saw in the mechanics breakdown: the distribution of crash points is chosen so that a small share of rounds breaks off practically instantly. These instant losses form the edge — the casino doesn't need to 'tweak' the result, the math is already on its side.
Why strategy doesn't change the average
Here's the most important and most counterintuitive conclusion. Take a simple strategy: always cash out at multiplier ×m. You win (receive m times your bet) if the round reaches m, and lose the bet if it doesn't. For the game to have a return of 97%, the probability of reaching m must be such that the average result equals 0.97:
From this immediately follows that very rule from the mechanics breakdown. But something else is more interesting: substitute any target m back in — and the average return always comes out the same, 97%. Whether you aim for ×1.5 or ×100 — the expected return doesn't change. No cash-out moment reduces the casino's edge: you only choose the shape of the distribution of wins, not their average.
| Cash-out target | Chance of reaching | Average return |
|---|---|---|
| ×1.5 | ≈ 65% | 97% |
| ×2 | ≈ 49% | 97% |
| ×5 | ≈ 19% | 97% |
| ×10 | ≈ 9.7% | 97% |
| ×100 | ≈ 1% | 97% |
Variance: one EV, different evenings
If the expectation is the same for all strategies, why are evenings so unlike one another? Because of variance (spread). JetX is a high-volatility game: the result of an individual session can end up far from the average in either direction. It's precisely variance that creates the illusion that 'the strategy works' or that you've 'caught a hot streak' — when it's just noise around a slow negative drift.
To see this in numbers, we ran 200,000 simulations each of an 'evening' — 100 rounds of $100 with a deposit of $10,000. The result is telling: the average remaining balance barely depends on the chosen target and stays around $9,700 (that is, −3% of turnover), while the evening ends in the green only about a third of the time.
| Strategy | In the green after the evening | Average remaining |
|---|---|---|
| Cash out at ×1.5 | ≈ 35% | ≈ $9,703 |
| Cash out at ×2 | ≈ 34% | ≈ $9,702 |
| Cash out at ×10 | ≈ 38% | ≈ $9,695 |
The same average remaining balance with completely different 'feelings' from the game — that's what variance is.
The main mistake about 'only 3%'
Now let's return to the caveat about turnover. The 3% edge is taken from each bet, not from the deposit once. When you re-bet your winnings, your turnover exceeds the deposited amount many times over — and every $100 cycled through the game again loses its 3%.
Simple arithmetic: with bets of $100 you lose on average $3 per round. So a deposit of $10,000 'on average' fully dissolves in about 3,300 rounds of play — with a total turnover of about $330,000. Not because the game is 'unfair,' but because 3% × large turnover = the whole deposit. That's why '97% return' and 'losing everything' get along perfectly.
Let's sum up. A 97% return is an average over the distance, not a forecast for a session. The casino's 3% edge is built into the game and doesn't depend on your cash-out tactics. Variance makes a short distance deceptive and a long one predictably loss-making. All of this is a consequence of the math, not of 'unfairness': how exactly the fairness of each round is verified, we cover in the article on provably fair.
Frequently asked questions
RTP (return) is the share of all bets placed that the game returns to players over a very large distance. An RTP of about 97% means that for every $100 of turnover, about $97 comes back on average, and about $3 stays with the casino. This is a statistical average over millions of rounds, not a promise for your session: over an evening you can both double up and lose everything.
No — this is the most common mistake. The 3% is taken from turnover (the sum of all bets), not from the deposit once. By re-betting your winnings, you cycle amounts many times larger than your stake through the game, and each bet again loses its 3%. With bets of $100 that's about $3 per round, so a $10,000 deposit on average is gone entirely in about 3,300 rounds. '97%' and 'losing everything' are perfectly compatible.
No. Whatever multiplier you choose as a target, the average return is the same — 97%, because the probability of reaching ×m is approximately 0.97 ÷ m. Early cash out gives frequent small wins, a high target gives a rare big one, but the expectation in both cases is negative and the same. The cash-out moment changes variance, not the average.
Variance (volatility) is the spread of results around the average. JetX is a high-volatility game: an individual session can end up far in the green or in the red, even though the expectation is negative. It's precisely because of variance that a short distance is deceptive — it seems like 'the strategy works,' when it's just noise around a slow loss. Over a long distance variance averages out, and the casino's edge remains.
The distribution of crash points is chosen so that a small share of rounds breaks off almost instantly (at ×1.00). These instant losses form the markup: the average result of any bet comes out about 3% less than what was put in. The casino doesn't need to 'tweak' an individual round — it remains random and verifiable through provably fair — the edge is built into the math of the distribution itself.