Skip to content
A clearer perspective on casino play18+ · Information, without the hype
Explore guides
Odds & mathematics GUIDE № 08

House Edge Explained: Expected Loss, Turnover and Real Results

See how house edge is calculated, why turnover matters, and why a small percentage can still mean substantial exposure.

A brass balance scale beside casino chips
Editorial illustration · About our imagery

House edge is the casino's mathematical advantage on a wager, usually expressed as a percentage of the stake. It describes an expected result across repeated play under stated assumptions. It does not mean the casino deducts that percentage visibly from every bet.

A game can produce winners in the short term and still have a house edge. In fact, the possibility of individual wins is fully compatible with the mathematics. To understand the percentage, start with one wager and account for every possible outcome.

Building an expected value

Take a simplified £1 wager that wins £1 profit on 48% of attempts and loses £1 on 52%. The average net result is 0.48 times £1 minus 0.52 times £1, which equals minus four pence.

The house edge is therefore 4% of the £1 stake. The game does not need to take four pence from a winning payout. It gains its advantage through the relationship between the chance of winning and the amount paid.

This calculation is called expected value. “Expected” has a technical meaning here: a probability-weighted average, not the outcome most likely to happen on the next attempt.

On the very next attempt, the result is still either plus one pound or minus one pound. Neither result equals the four-pence average.

A familiar roulette calculation

On a standard single-zero roulette wheel, a £1 straight-up bet covers one of 37 equally likely pockets. A win earns £35 profit, while each of the other thirty-six outcomes loses £1.

The expected net result is (35 − 36) divided by 37, or approximately minus £0.027. That is a house edge of about 2.70%.

If the game instead paid £36 profit for the winning number, this simplified wager would be mathematically fair: one £36 win would balance thirty-six £1 losses across the equally weighted outcomes. The standard payout is lower.

This example also shows why a long payout ratio is not automatically generous. Thirty-five to one sounds large until it is compared with the probability of the winning event.

Turnover determines the scale of exposure

Suppose the relevant edge is 4% and total turnover is £500. The expected loss is £20. That does not say the person deposited £500 or that their actual loss will be exactly £20.

Turnover counts all stakes. A £50 deposit can support more than £50 of turnover if returns are wagered again. Reusing money does not stop it from being exposed to the game's disadvantage on the next wager.

A helpful account review separates deposits, withdrawals, opening and closing balance, total stakes and total returns. Each describes something different. Looking only at deposits can hide how much repeated play took place.

Expected-loss calculations are best treated as a description of exposure, not a spending allowance. Actual losses can be much larger, including the whole available balance.

Time and pace also matter

Imagine a fixed-stake game with a 3% edge. At £1 per resolved wager, one hundred wagers produce £100 of turnover and £3 of expected loss. Five hundred wagers produce £500 of turnover and £15 of expected loss.

Nothing about that comparison recommends either session. It simply shows why a faster interface or longer playing time can change the amount exposed even when the stake button stays unchanged.

Reducing a percentage while dramatically increasing turnover may still increase the expected cash loss. A comparison that discusses only the edge can miss the larger practical change.

The same principle applies to several simultaneous betting positions. Ten one-pound wagers in a round expose ten pounds, not one.

The assumptions must travel with the number

Some games have a largely fixed mathematical edge for a particular bet. Others depend on player decisions or a specific rule set. A blackjack figure calculated for particular rules and accurate decisions cannot be copied onto every blackjack table.

Side bets need separate treatment. A main game with one edge can offer optional wagers with very different probabilities and payouts. Averaging them together requires knowing how much was staked on each.

Even the denominator can matter. Some published comparisons use the initial stake; others discuss total money wagered after optional additional bets. A percentage without its basis is easy to misread.

When a source gives a very precise figure, ask what game, rules, decisions and stake definition produced it.

House edge and RTP are closely related, with limits

For a simple fixed-stake game measured on the same basis, house edge equals one hundred percent minus RTP. A theoretical return of 96% corresponds to a 4% edge.

That relationship becomes less useful when figures refer to different modes, conditional bonuses or different definitions of turnover. Matching the scope matters before subtracting two numbers.

Neither measure describes volatility. A small average disadvantage can coexist with large swings in individual outcomes. The percentage is not a ceiling on what can be lost.

It also does not measure whether an operator is licensed, handles complaints properly or protects account information. Mathematical value and consumer protection are separate questions.

Why systems do not remove the disadvantage

Changing the stake after a loss can change the distribution of results and the speed at which money is exposed. It does not change the underlying winning probability or payout of an independent wager.

A system may produce many small winning sessions and occasional much larger losses. Counting only the sessions that finished ahead hides the size of the losing ones.

House edge is useful because it asks for the complete arithmetic. It prevents an attractive anecdote or a run of wins from standing in for the whole outcome distribution.

Understanding the number is worthwhile. Treating it as a promise that a game will be inexpensive, predictable or profitable is where the interpretation goes wrong.

CHECK THE DETAIL

Sources & scope

Rules and requirements can vary by game, operator and country. Follow the linked source for its current wording and check the terms that apply where you live.

Spotted something that needs correcting?Tell us about it