House edge measures the casino’s mathematical advantage under a game’s stated rules and play assumptions. It helps explain the average cost of betting, but it does not predict the result of your next round or session.
For New Zealand players, the useful comparison concerns the exact bet, payout and rules. A casino brand does not have one house-edge percentage covering every game, and a familiar game name can conceal several different versions.
This guide explains the calculation with fictional NZD examples, compares specific game rules and shows why turnover, strategy and volatility matter. A lower percentage does not make a gambling budget safe from loss.
What house edge measures
Think of a bet as a set of possible outcomes, each with a probability and a monetary result. The expected result is the probability-weighted average. If that average favours the casino, the player has a negative expected result under those assumptions.
Expected loss is an average, not a bill
For a fictional fixed-stake game with a 2% edge measured on total stakes, NZ$100 of turnover has an expected loss of NZ$2. There is no separate NZ$2 charge taken from every NZ$100 deposit. The percentage describes the game’s mathematical expectation.
Your actual result can be a win, a small loss or the loss of the full amount available to play. Short sessions can differ greatly from the expectation. Winning a session does not disprove the edge, and losing more than the expected amount does not by itself establish a fault.
The betting basis matters
A conventional house-edge figure can be measured against the original bet. Some games allow extra wagers during the round, such as a blackjack double or a later poker-game raise. A percentage based on the opening wager and one based on all money staked use different denominators.
Michael Shackleford’s explanation of house edge distinguishes the initial-bet measure from a measure using total wagers. Check which basis a figure uses before multiplying it by a betting total.
Rules and decisions belong to the percentage
A published edge is meaningful only with its assumptions. These can include the wheel, deck count, paytable, commission and the decisions made by the player. Removing those details turns a useful figure into a vague comparison.
Our casino terms guide explains related language such as stake, paytable, turnover and total return. Keep those definitions separate from the money deposited into an account.
Calculate expected loss with a clear betting total
Where the edge is defined on total stakes, the calculation is:
Expected loss = total stakes × house edge as a decimal.
Convert 2% to 0.02 before multiplying. For fictional turnover of NZ$500 at a 2% edge, the result is NZ$10: NZ$500 × 0.02. This is a model average under the assumptions, not a forecast or a maximum loss.
| Stake and rounds | Total stakes | Assumed edge | Expected loss |
|---|---|---|---|
| NZ$1 × 100 rounds | NZ$100 | 2% | NZ$2 |
| NZ$2 × 100 rounds | NZ$200 | 2% | NZ$4 |
| NZ$2 × 500 rounds | NZ$1,000 | 2% | NZ$20 |
| NZ$5 × 200 rounds | NZ$1,000 | 5% | NZ$50 |
A deposit can produce more than its value in turnover
If you deposit NZ$50, then stake and restake funds, the betting total can exceed NZ$50. For example, 100 rounds at NZ$1 create NZ$100 of turnover if those rounds are played. A starting balance does not guarantee that you will be able to complete them.
Use the actual relevant betting total, rather than assuming that a NZ$50 deposit represents only NZ$50 of exposure. Reinvested winnings are still money being put at risk.
Match the formula to the quoted basis
If an edge is defined per initial bet, add the initial wagers used by that model. Do not automatically multiply that figure by a total including later raises, splits or doubles. For a comparison using all wagers, find an analysis stated on that basis.
The house-edge calculator can help with arithmetic once you have the correct inputs. It cannot identify an unknown game’s rules, select the correct strategy or predict a session from the percentage alone.
Keep currency and payment costs separate
Calculate in a consistent currency. If an account uses AUD while your budget is in NZD, convert the cost using the relevant payment information. A conversion fee or withdrawal charge is an additional account cost, not part of the game’s mathematical edge.
Read the NZD account guide if the balance and payment currencies differ. Do not add a fee percentage to a game edge unless the amounts and betting bases actually match.
Compare specific games and bets
The figures below describe the listed rules, not a casino’s entire lobby. Percentages are rounded. Confirm the exact game’s information before applying a comparison to a table, variant or paytable.
| Game or bet | House edge | Rules and assumptions |
|---|---|---|
| Single-zero roulette | About 2.70% | A standard straight-up bet paying 35 to 1 on a fair 37-pocket wheel. |
| Double-zero roulette | About 5.26% | A standard straight-up bet paying 35 to 1 on a fair 38-pocket wheel. |
| Baccarat: Banker | About 1.06% | Conventional eight-deck rules, 5% commission on Banker wins; ties return the stake. |
| Baccarat: Player | About 1.24% | Conventional eight-deck rules, even-money Player wins; ties return the stake. |
| Baccarat: Tie | About 14.36% | Conventional eight-deck rules with the Tie bet paying 8 to 1. |
| 9/6 Jacks or Better | About 0.46% | The full-pay table, optimal decisions and the qualifying five-coin royal payout of 800 per coin. |
| A fictional 96% RTP pokie | 4% on the matching basis | A hypothetical example, not a named game or a statement about every pokie. |
| Blackjack | Depends on rules and play | Use the exact table rules and the strategy assumptions; there is no universal blackjack percentage. |
Roulette: the zero changes the expectation
A standard single-zero wheel has 37 pockets. On a NZ$1 straight-up bet paying 35 to 1, one pocket produces NZ$35 profit and the other 36 produce a NZ$1 loss. Assuming equal pocket probabilities, the average net result is (35 − 36) ÷ 37 = −NZ$0.027027 per bet.
That gives an edge of about 2.70%. On a 38-pocket double-zero wheel with the same payout, the calculation becomes (35 − 37) ÷ 38, giving about 5.26%. Extra zeros matter because the number of losing outcomes increases without a corresponding increase in this payout.
Read the roulette guide for bet types. Special payout or zero rules need their own calculation; do not transfer the standard figure to every feature or side bet.
Baccarat: compare the individual bets
The Banker, Player and Tie bets have different probabilities and payouts. The table uses the conventional eight-deck analysis, including 5% commission on winning Banker bets and an 8-to-1 Tie payout. Ties return the stake on the Banker and Player bets in that model.
The eight-deck baccarat return tables show the calculation. Commission-free variants can change particular winning payouts, so removing a named commission does not by itself establish a better return. Our baccarat guide explains the game format.
Video poker and pokies require the exact paytable
The quoted Jacks or Better figure assumes the full-pay table and optimal decisions, including the qualifying royal-flush payout. The full-pay strategy analysis gives a return of about 99.54%. A different paytable or different decisions changes that result.
For video poker and pokies, inspect the game you actually open. A percentage attached to another title, configuration or stake condition is not enough.
How rules and strategy affect the edge
Blackjack needs more than a game name
Record the natural-blackjack payout, dealer soft-17 rule, deck count, doubling and splitting options, and surrender rules. The strategy assumptions need to fit those conditions. A general “blackjack house edge” cannot describe every combination.
For a simple payout illustration, a NZ$10 natural paying 3 to 2 earns NZ$15 profit, while one paying 6 to 5 earns NZ$12 profit. The NZ$3 difference on that event is not a three-percentage-point change to the game’s total edge; the whole calculation includes all possible hands and decisions.
Our basic strategy guide states the rule assumptions for its charts. Using a chart for a different table can change the result. Strategy can affect expectation without guaranteeing a profit.
Side bets have their own maths
An optional side bet is a separate wager with its own outcomes and payout schedule. Its return need not resemble the main game’s return. A large top prize alone tells you little without the probability of receiving it.
Include every wager when you assess how much money is at risk. Increasing the number of optional bets can increase the total committed per round even when the headline table minimum stays the same.
Live and virtual describe the format
A live-dealer table and a virtual table can use similar payouts or different rules. The presence of a dealer on video does not by itself remove the mathematical advantage in the payout structure.
Read the exact rules and bets in either format. The live-dealer blackjack guide covers the practical format questions, including table limits and betting windows, separately from the house-edge calculation.
A game’s theoretical return is not your demonstrated skill
In games with decisions, an optimal-strategy return is a benchmark under the model. It should not be treated as a personal return simply because you recognise the game. Mistakes, unfamiliar rules and different paytables can change the expectation.
House edge, RTP, volatility and win frequency
These measures answer different questions. Read them together rather than using one percentage to stand in for the whole experience.
RTP and the matching betting basis
When theoretical RTP and edge use the same rules, strategy and wagering basis, they complement each other. A fictional 96% RTP game then has a 4% edge. A change from 96% to 97% RTP reduces that matching edge from 4% to 3%.
Check the basis before applying the subtraction to a game with additional wagers. Our RTP guide explains theoretical versus observed returns. A short record of your own wins and losses does not establish the underlying percentage.
Volatility concerns the spread of results
Two games with the same expected return can distribute prizes differently. One may have many smaller payouts; another may concentrate more of its return in rarer large prizes. Short-session swings can therefore differ even when the theoretical average matches.
The Gambling Commission’s explanation of RTP and volatility distinguishes the designed return from measured performance. See our volatility guide for the player-facing concepts.
Win frequency is not expected profit
A paying result can return less than the total stake. For a fictional NZ$1 round returning NZ$0.40, the net result is a NZ$0.60 loss even if the game presents a payout animation. Count the full wager and amount returned.
Likewise, a bet that wins infrequently can have a different expectation depending on the payout. Probability and prize size both belong in the calculation. A high frequency of small payouts does not establish a low house edge.
Several bets on the same round
Covering more outcomes can change the pattern of results without removing the edge. Consider a standard single-zero roulette wheel with no special zero rule, and fictional NZ$1 bets on both red and black. The total stake is NZ$2.
On a red or black result, the winning even-money bet returns NZ$2 in total while the other NZ$1 bet loses. Across the two bets, you break even on that round. On zero, both bets lose and the net loss is NZ$2.
With equal pocket probabilities, the expected loss is NZ$2 divided by 37, or about NZ$0.05405 per round. Relative to the NZ$2 staked, that is still about 2.70%. The arrangement produces many break-even rounds and an occasional loss; it has not removed the mathematical cost.
The two bets also depend on the same spin. Do not treat them as two independent results when thinking about volatility. You can add their expected monetary results, but understanding the range of possible outcomes requires considering how the bets interact.
Long-run averages do not create a recovery deadline
There is no number of rounds after which a particular player must recover earlier losses. The model average does not create a debt that the game owes you. Playing longer increases the amount you put through the game rather than making a required win arrive.
Bonuses, betting systems and repeated play
A bonus adds conditions around the game
A promotion does not automatically change a game’s outcome probabilities. It adds conditions about eligibility, balances, qualifying betting, expiry and withdrawals. Those conditions need a separate assessment from the underlying edge.
In a fictional NZ$20 deposit plus NZ$20 bonus with wagering of 30 times the combined amount, the qualifying requirement is NZ$1,200. Game contribution determines how much actual betting is needed to reach that figure.
Low contribution can require much more betting
If a permitted game contributes 10%, each NZ$10 stake adds NZ$1 of qualifying turnover. A fictional NZ$1,000 requirement would then need NZ$10,000 in actual stakes if the same contribution applied throughout.
At a hypothetical 1% edge defined on those total stakes, the theoretical game cost would be NZ$100. This calculation does not establish the chance of finishing the requirement, the final balance or the amount eligible to withdraw. Read the bonus terms guide and bonus wagering guide before using the figures.
Changing stake patterns does not change an independent bet
Doubling after losses, raising after wins or switching bet positions changes how much you commit. It does not change the probability and payout of an independent roulette spin. Progressions can produce many small results and occasional much larger losses without removing the underlying negative expectation.
Bankroll and table limits also constrain a progression. Do not treat a theoretical sequence of unlimited future bets as something a real account can complete. Changing the pattern is different from changing the game’s rules or probabilities.
More turnover can outweigh a lower percentage
For fictional fixed-stake comparisons, NZ$100 of turnover at 4% has an expected cost of NZ$4. NZ$1,000 at 1% has an expected cost of NZ$10. A lower edge can still accompany a larger expected loss when substantially more money is wagered.
Compare the relevant total, not just the percentage. Do not increase play to make use of a supposedly favourable figure or to pursue an advertised reward.
Use the figures without treating them as a prediction
Start with the game name, version, rules and particular bet. Find the return information or an analysis that matches them. Note whether the figure assumes a strategy and whether it is based on initial wagers or all stakes.
Check the inputs to a comparison
Record the stake, the number of rounds used in a calculation, any additional wagers and the account currency. If you use an estimated round count, remember that a balance may run out earlier and game speed may vary. An assumed total is not a promised session length.
Keep mathematical game cost separate from payment fees and promotional conditions. The how-to-play guide and casino glossary can help you locate the relevant rule details.
Keep money and time boundaries separate from expectation
An expected loss of NZ$5 does not make NZ$5 the most you could lose. Set a maximum based on your circumstances, keep essential spending outside gambling and stop at the boundary you chose. The gambling-limits guide explains different account controls.
If you find yourself continuing because a game seems due to return money, step away from the calculation and the game. Our responsible gambling information provides support routes. The maths can explain an average; it cannot make chasing a previous loss reliable.
House-edge FAQs
Frequently asked questions
What does a 2% house edge mean?
It means a theoretical average loss of 2% of the relevant betting amount under the stated rules and assumptions. For a fictional fixed-stake game measured on total stakes, NZ$500 of turnover at 2% has an expected loss of NZ$10. It does not mean you will lose exactly NZ$10 or that your loss is capped at that amount.
Is house edge always 100% minus RTP?
That relationship works when both percentages describe the same game, strategy and betting basis. A 96% theoretical return then corresponds to a 4% edge. Be careful with games that add wagers during a round: an edge measured against the initial bet cannot automatically be compared with a return measured against all money staked.
Does a lower house edge mean I win more often?
Not necessarily. House edge measures average monetary expectation, while win frequency concerns how often an outcome pays. Prize size, pushes and volatility also matter. A game can pay frequently yet have an unfavourable average return. Check the payout rules as well as the percentage.
Which casino game has the lowest house edge?
There is no
