LuckyAussieWorld

Independent research. Not a gambling operator. Support: 1800 858 858

LuckyAussieWorld

Australian pokies research desk

The margin we assume, and why the assumption is printed with the figure

Every figure on this site that involves a cost over time rests on an assumed house margin, and that assumption is printed alongside the result rather than buried underneath it. A number produced without a stated margin is a number with no footing, and we would rather show our working than dress a guess as a measurement. This page sets out what the margin is, what it is not, and how we use it.

What a house margin is

The house margin, sometimes called the house edge, is the average amount a game or a product retains from each dollar wagered over a long run of play. It is expressed as a percentage of turnover. If a game retains 4 cents on average from every dollar put through it, the margin is 4 per cent. The margin is a property of the game's rules and paytables, or of the odds offered on a market, and it is fixed by those rules rather than by the fortune of any individual player.

It is worth being precise about what the margin is a percentage of. It is a percentage of turnover, not of a player's starting balance, and not of their winnings. A player who starts with 200 dollars and turns that 200 into 2,000 dollars of cumulative stakes has generated turnover of 2,000, and a 4 per cent margin applies to the 2,000 rather than to the 200. This distinction matters more than any other single point on this page, because turnover is almost always a larger number than the balance a player began with, and the gap between the two is where most cost estimates go wrong.

The margin can be thought of as the mirror image of the player's expected return. It describes the shape of the game over many repetitions. It does not describe what will happen to any one person on any one evening. A machine or a market with a 4 per cent margin will still hand out long winning runs and long losing runs, and those runs are not evidence that the margin has changed. They are the ordinary variance of a process whose average is known and whose individual outcomes are not.

Where margins come from differs by product. In a game of chance the margin is built into the paytable: the prizes are set slightly below the odds that would make the game fair. On a betting market the margin is built into the prices: the implied probabilities of the outcomes, added together, come to slightly more than 100 per cent, and the excess is the book's margin. Both arrive at the same place, which is that the house keeps a small slice of every dollar turned over, on average, across a large number of bets.

Margin and return to player

Return to player, usually written RTP, is the complement of the margin. It is the percentage of turnover that a game returns to players over the long run. If RTP is 96 per cent, the margin is 4 per cent. The two figures are two ways of stating one quantity, and they always sum to 100. A game advertised at 96 per cent RTP and a margin of 4 per cent are the same game described from opposite ends.

This site tends to work in margins rather than in RTP because margin is the figure that plugs directly into a cost calculation. If you want to know what a given amount of turnover will cost in expectation, you multiply the turnover by the margin. You could do the same with RTP by multiplying by the complement, but the arithmetic takes one more step for no gain, and the extra step is an easy place to introduce an error. When a source quotes RTP, we convert it to a margin before using it and we say that we have done so.

It is worth flagging a common confusion. A 96 per cent RTP does not mean that a player who stakes 100 dollars will get 96 dollars back. It means that across a very large amount of turnover the game returns 96 per cent of it. In a short session the return might be 0 per cent, or 150 per cent, or anything in between, because a single large win dominates the arithmetic. The RTP figure describes the machine's design, not the player's experience. The same is true of the margin. Both are averages, and both are averaged over a run of play long enough for the average to assert itself.

Where a product quotes a range of RTP figures, for example across different stakes or different game variants, the margin correspondingly varies. We take the figure that applies to the case being calculated, and if we cannot establish which figure applies, we say so rather than picking the most favourable one. A margin chosen to flatter the result is no better than no margin at all, because it produces a number that looks precise and is not.

A long run average, not a promise for a session

The margin is a long run average, and a session is not a long run. A player who puts 50 dollars through a game with a 4 per cent margin does not expect to lose exactly 2 dollars. They expect to lose 2 dollars on average across many such sessions, and in any one session the outcome will be scattered widely around that average. The scatter comes from the game's variance, which is a separate property from its margin. Two games can share a margin and behave very differently in the short term.

This is the point at which margin calculations are most often misread. A figure that says a player's expected cost over a given turnover is, say, 40 dollars is a statement about the average of many repetitions. It is not a forecast of the session in front of the player, and it does not become wrong if the player wins. A player who wins in a high variance game with a 4 per cent margin has not broken the margin. They have landed on the favourable tail of a distribution whose average is unchanged.

The practical consequence is that margin based figures should be presented as expected costs or average costs, and never as amounts that will certainly be lost. On this site we phrase them that way. Where a calculation produces an expected cost, the text says so. The number is real and it is useful, but it is an average, and the words around it have to carry that.

There is a further subtlety for players who vary their stake. Turnover is the sum of stakes across all bets, so a player who bets small for a long time and large for a short time generates turnover from both. The margin applies to the total, whatever shape the individual bets took, provided the margin is the same across them. If the margin differs by stake size or by game, the total expected cost is the sum of each portion of turnover multiplied by its own margin. Averaging the margins first and then multiplying gives the same answer only if the turnover is split in matching proportions, which is a convenient result and an easy thing to get wrong.

How the margin enters turnover arithmetic

The core calculation on this site is short. A player completing N dollars of turnover, at a margin of m, pays in expectation N multiplied by m. If N is 1,000 dollars and m is 0.04, the expected cost is 40 dollars. Every figure of this kind on the site follows the same pattern, and the margin is always stated as a decimal fraction alongside the result so that the arithmetic can be checked by anyone reading it.

The reason the multiplication works is that turnover and margin are defined so that they can be multiplied. Turnover is measured in dollars. The margin is a fraction of a dollar per dollar turned over, which is dimensionless. Their product is in dollars. This sounds like a triviality, but it is the reason a margin has to be quoted as a plain fraction rather than as an odds price or a house percentage with hidden structure. A figure described as a 4 per cent margin and a figure described as decimal odds of 1.04 are not the same thing, and one cannot be substituted for the other in this multiplication without producing nonsense.

Turnover itself needs care. It is not the same as deposits, and it is not the same as a starting balance. It is the total amount staked, counting each bet separately. A player who begins with 200 dollars and reinvests winnings can generate turnover many times the starting sum. Where a product imposes a turnover requirement, for example a condition attached to a bonus, the requirement is stated in multiples of the bonus or of the deposit, and the expected cost of meeting it is the required turnover multiplied by the margin. That calculation is the one most often done badly, because the multiple applies to the bonus or the deposit and not to the player's total balance, and the resulting turnover figure can be much larger than people expect.

Because the calculation is so simple, the entire reliability of the result rests on the margin. If the margin is right, the product is right, subject to the caveat about averages. If the margin is wrong, the product is wrong in direct proportion. A margin stated as 4 per cent when the true figure is 6 per cent understates the cost by a third. This is why the margin is not a background detail but the load bearing element of every figure, and why we print it rather than assume the reader will know it.

{{qvx-img:qvx-penalty-unit-1}}
The Commonwealth penalty unit, as published: $313 for an offence committed from 1 July 2023, $330 from 7 November 2024, and $364 from 1 July 2026. Claims 20 and 21. The figure is a unit value and not a penalty: no maximum is stated here, because multiplying a unit by a number of units is arithmetic and not a published figure.

Naming the assumption, and the penalty unit trap

Why print the margin at all. A calculation without a stated margin is a number with no footing. It looks like a measurement, and a reader has no way to tell whether it rests on a 2 per cent margin, a 4 per cent margin, or a figure invented to make the arithmetic come out tidily. Printing the margin does two things. It lets the reader check the working, which is the ordinary courtesy of showing a sum. And it lets the reader substitute their own margin if they think ours is wrong. A figure whose assumption is visible can be argued with. A figure whose assumption is hidden can only be believed or disbelieved, and neither is a useful response.

The same discipline applies to every published value the site quotes, and the penalty unit is the clearest example of why. The Commonwealth penalty unit is a value set by government and published as a dollar amount. It was $313 for an offence committed from 1 July 2023, $330 from 7 November 2024, and $364 from 1 July 2026. That is a unit value. It is not a penalty, and no maximum penalty can be read off it alone. To arrive at a maximum, the unit is multiplied by the number of units set for the particular offence in the particular Act. That multiplication is arithmetic performed by whoever is doing the calculating. It is not a published figure, and it should never be presented as though it were one.

The parallel with the margin is exact. A margin of 4 per cent is a published or assumed property of a product, in the same way that the penalty unit is a published value. The product of that margin and a player's turnover is an arithmetic result, in the same way that the product of the penalty unit and the number of units is an arithmetic result. Neither product is itself a published fact. Presenting either as one would be presenting a guess, or at best a calculation, as a measurement. This site names its assumptions rather than doing that: the margin is printed with the figure, the turnover basis is stated, and where a value is derived by multiplication, the text says it has been derived.

That is the whole standard, and it is a modest one. We do not claim to know each reader's margin. We claim only to say which margin we used, to use a defensible figure, and to keep the arithmetic visible so that anyone can redo it with their own inputs. A reader who disagrees with our 4 per cent can multiply their own figure by the same turnover and get their own answer, and both answers are legitimate because both assumptions are in plain view. The alternative, a bare number with the assumption stripped out, gives the reader nothing to work with and invites them to treat an average as a promise and a calculation as a measurement. We would rather be slightly pedantic and correct.

Questions about the margin

What is a house margin in plain terms?
It is the average share of each dollar wagered that the game or market retains over a long run of play, expressed as a percentage of turnover. A 4 per cent margin means 4 cents retained on average per dollar turned over.
How does margin relate to return to player?
They are complements and always sum to 100 per cent. At 96 per cent return to player, the margin is 4 per cent. This site converts any quoted return figure into a margin before using it, and says when it has done so.
If the margin is 4 per cent, will I lose 4 per cent of my stake in a session?
Not necessarily. The margin is a long run average over a large amount of turnover. Any single session will scatter widely around that average because of variance, and a winning session does not contradict the margin.
What is the turnover formula?
Expected cost equals turnover multiplied by the margin, with the margin written as a decimal fraction. For 1,000 dollars of turnover at a 4 per cent margin, the expected cost is 40 dollars. Turnover counts every stake, so it is usually much larger than a starting balance.
Why print the margin instead of just the result?
Because a result without its assumption is a number with no footing. Printing the margin lets the reader check the sum and substitute their own figure. It also keeps a calculation from being mistaken for a measurement, which is the same reason the penalty unit is shown as a unit value and a penalty is never read off it directly.
Scroll to Top