House Edge Explained: Expected Loss, RTP and Short-Term Results

House edge is a long-run mathematical measure. It describes the expected advantage built into a game, not the amount that a particular person will lose and not a forecast for tonight's session. Keeping those ideas separate helps explain why a game can have a known mathematical advantage while individual results still vary sharply.

What house edge means

The UK Gambling Commission describes the house edge as the casino's average profit from a game. It is usually expressed as a percentage of the amount staked. From the player's perspective, the same percentage represents the expected loss per unit wagered, under the rules and assumptions used for the calculation.

For example, a 4% house edge means an expected loss of £0.04 for every £1 wagered:

• £1 wagered × 4% = £0.04 expected loss
• £100 wagered × 4% = £4 expected loss
• £1,000 wagered × 4% = £40 expected loss

These are mathematical expectations across repeated wagering. They are not promises about a single bet or session. A person who wagers £100 might finish ahead, behind by more than £4, behind by less than £4, or level. The 4% figure describes the average around which results tend to settle only over a very large number of comparable bets.

Expected value is the calculation underneath

Expected value combines each possible net result with its probability. Consider a simplified £1 game with only two outcomes:

• a 48% chance of receiving £2 in total, which is a net win of £1
• a 52% chance of receiving nothing, which is a net loss of £1

The player's expected value is:

(0.48 × £1) + (0.52 × -£1) = -£0.04

The expected loss is therefore £0.04 per £1 wagered, equivalent to a 4% house edge. This toy example is deliberately simple. Real games may contain many outcomes, different prizes, rule choices and conditional probabilities, but the principle is the same.

Expected loss can also be estimated by multiplying total stakes by the house edge. If the same £10 is wagered ten times, total stakes are £100 even though only £10 was initially deposited. At a 4% edge, the expected loss attached to that £100 of repeated wagering is £4. This is why the amount turned over matters more to the calculation than the initial balance alone.

Expected-loss relationship

Total wagering × 4% house edge = expected loss

£1 wagered£0.04 expected loss

£100 wagered£4 expected loss

£1,000 wagered£40 expected loss

House edge and RTP are two views of the same model

For a game where the figures use the same rules and measurement basis, theoretical return to player (RTP) and house edge are complements:

House edge = 100% - theoretical RTP

A theoretical RTP of 96% therefore corresponds to a 4% house edge. The RTP is the share of stakes the mathematical model returns as prizes over the long run; the house edge is the share it does not return.

Our guide to RTP and how it should be interpreted explains the return side in more detail. The Gambling Commission stresses that RTP is an average measured over a significant number of plays and is not the return a person should expect from one gambling session.

The subtraction is useful only when the figures are genuinely comparable. A published RTP may assume a particular rule set or strategy. A contribution rate, side bet or different pay table can involve a different mathematical model. It is safer to compare like with like than to treat every percentage displayed around a game as interchangeable.

Why a session rarely resembles the average

House edge determines the centre of the long-run distribution, while variance describes how widely actual results can spread around that centre. Over a short session, variance can dominate the small expected loss.

Suppose 100 independent £1 bets have a 4% house edge. The expected loss is £4, but this does not mean the final result will be exactly -£4. The available outcomes and their probabilities determine the range. A session containing rare large prizes may move much more sharply than one built from frequent small results, even if both games have the same edge.

Increasing the number of comparable independent bets generally makes the observed average more stable, but it does not provide a deadline by which a game must produce its theoretical result. It also does not create a balancing force within the next few bets. On a random machine, the Gambling Commission says the odds of the next outcome remain constant and are not affected by earlier wins or losses.

See volatility and variance for the distinction between the average result and the dispersion of results.

Rule differences can change the edge

House edge belongs to a defined game and rule set, not merely to a familiar game name. Its calculation can change when the underlying probabilities or payouts change. Relevant differences can include:

• the prize or payout table
• how many winning and losing outcomes exist
• whether optional or side bets use separate maths
• which decisions a player is permitted to make
• whether a published figure assumes a stated strategy

That means a percentage quoted for one version should not automatically be applied to another. Even apparently small differences in payouts or permitted choices can alter expected value. This article does not attach percentages to particular game variants because a reliable comparison needs the exact rules, pay table and calculation method.

Percentage comparisons have limits

A lower house-edge percentage means a lower expected loss for the same amount wagered under comparable conditions. It does not necessarily mean a lower total expected loss in practice.

Consider two hypothetical sessions:

• Game A has a 2% edge and £500 is wagered: expected loss £10.
• Game B has a 4% edge and £100 is wagered: expected loss £4.

Game A has the lower percentage but the larger expected loss because more money is staked. The same effect appears when a fast game allows many more bets per minute, when a person plays for longer, or when the stake per bet increases.

A useful simplified relationship is:

Expected loss = stake per bet × number of bets × house edge

For example, £2 per bet across 200 bets is £400 in total stakes. With a hypothetical 3% edge, the expected loss is £12. Doubling the pace to 400 bets, with everything else unchanged, doubles the total stakes and the expected loss to £24. Pace does not alter the percentage edge; it changes how quickly wagering volume accumulates.

House edge is not the same as observed hold

House edge is a theoretical property calculated from probabilities and payouts. Actual business results are observed over a period and depend on the wagers and prizes that actually occurred.

Terminology around “hold” is not completely uniform, so the denominator should always be checked. The Gambling Commission's formal reporting terms distinguish:

• turnover: the total value of all stakes, including money staked again after being won
• win: prizes returned to players
• gross gambling yield: stakes received minus prizes paid, with specified additions or deductions in the regulatory definition

An observed ratio derived from real turnover and prizes can be above or below the theoretical house edge over a limited period because of variance. It is therefore misleading to use a short-period operational result as though it were the game's fixed mathematical edge.

Common misconceptions

“A 4% edge means I lose 4% every session.” It does not. It is an expected average per unit wagered over repeated play.

“If I am losing more than the edge, a win is due.” Previous independent outcomes do not make a compensating result due. The next outcome keeps the probability defined by the game.

“A higher RTP means I am likely to win tonight.” RTP ranks a long-run mathematical return when figures are comparable. It does not predict an individual session.

“A small edge makes extended play inexpensive.” A small percentage can still produce a substantial expected loss when applied repeatedly to high total stakes.

“Changing bet size removes the edge.” If the game's probabilities and payout proportions remain the same, changing the stake changes the cash amount at risk and expected loss, not the percentage advantage.

The practical reading

House edge is best used as a structural comparison, not a session forecast. It helps show the expected cost of repeated wagering and why bet size, pace and duration matter alongside the headline percentage. It cannot tell an individual when a win will occur, how long a balance will last, or what the next outcome will be.

The long-run figure explains the direction of the mathematics. Variance explains the uneven route, while wagering requirements can increase the amount that must be staked. Together, those concepts provide a more complete view than any single percentage.