Independent guide · Casino

RTP vs House Edge: Math, Examples & Risk

Clarify the inverse link between RTP and house edge. See how small percentage shifts alter expected losses on fixed wager totals.

House Edge vs RTP: Mathematical Link and Examples

A 96% RTP does not mean a €100 deposit will return €96. It implies an expected €4 loss per €100 wagered under the stated game assumptions. Turnover, payout distribution and the exact game version determine what that figure can tell you.

How are RTP and house edge related?

House edge and Return to Player (RTP) describe the same expected wager result from opposite sides. On the same wager basis, house edge = 100% - RTP. The edge is expected player loss per unit wagered, before separate costs; it is not the operator’s net business profit.

The relationship is strictly linear and inverse. If the RTP is 96%, the house edge is 4%. If the RTP is 97.3%, the house edge is 2.7%. This calculation assumes a consistent game version and payout structure. Changing the basis, such as comparing different bet types or rule variants, invalidates the direct subtraction method.

    What does 96% RTP imply for €750 wagered?

    Consider a hypothetical slot with a 96% RTP, implying a 4% house edge. Assume a player wagers a total of EUR 750 over a session. The expected payout is calculated as 96% of EUR 750, which equals EUR 720. The expected loss is the difference between the total wagered and the expected payout, which is EUR 30.

    EUR 30 is a mathematical expectation for the stated wagers, not a guaranteed session loss. The game’s bonus features may already be included in its quoted RTP. What the percentage does not specify is the distribution of outcomes: how often payouts occur, how large they are, or how far a session may depart from the average.

      What does 97.3% RTP imply for €2,000 wagered?

      Now consider a slot with a 97.3% RTP. The house edge is 2.7%. Assume a player wagers a total of EUR 2000. The expected payout is 97.3% of EUR 2000, which equals EUR 1946. The expected loss is EUR 54.

      Comparing this to a 96% RTP game with the same EUR 2000 wager, the expected loss would be EUR 80. The difference is EUR 26. This illustrates how small percentage changes in RTP translate to concrete monetary differences in expected loss over larger turnover amounts.

        How do percentages differ from percentage points?

        A 1.3 percentage-point difference in RTP is not a 1.3% relative improvement in winning chances. It is a fixed reduction in the house edge. Players often confuse relative and absolute percentage changes. The impact is linear with respect to total wagers, not exponential or relative to the initial stake.

        For a fixed round count and fixed rules, raising the stake raises monetary turnover and the monetary expected loss proportionately. It does not improve the accuracy of the quoted probability or force actual outcomes closer to expectation. The per-unit edge remains the same under these assumptions.

          What does the simple calculation leave out?

          RTP and house edge do not provide hit frequency or variance. A high RTP game can have high volatility, leading to long dry spells. A low RTP game can have low volatility, with frequent small wins. These metrics alone do not predict the probability of a profitable session or the required bankroll.

          Strategy-dependent returns require specific assumptions. For games like blackjack, the quoted RTP depends on the player's strategy. Comparing games with different strategy requirements or payout conventions is misleading. Always verify the specific rules and version of the game before applying these calculations.

            How do I check the game version and the calculation?

            The RTP Loss Calculator takes stake per round, rounds and RTP. To reproduce the EUR 2000 example, enter EUR 2, 1000 rounds and 97.3%. Total wagers are 2 * 1000 = 2000; expected gross payouts are 2000 * 0.973 = 1946; expected loss is 2000 - 1946 = 54. These inputs describe an illustration, not a suggested session.

            Always check the game's specific RTP version. Different versions of the same slot can have different RTPs. Use the RTP Slot Version Verification Guide to understand how to identify the correct version. Ensure the RTP figure matches the specific game variant you are playing. This prevents applying the wrong mathematical basis to your session.

              Which total belongs in the RTP calculation?

              A €100 deposit is not automatically €100 of stakes. Suppose a ledger records €200 of settled stakes and €170 of credited prizes, including returned stakes where applicable. Observed RTP is 170 / 200 × 100 = 85%; the net result on those rounds is €170 − €200 = −€30. Re-staked winnings count again in turnover. Deposits, withdrawals and an unsettled round do not belong in the same numerator and denominator.

              At a theoretical RTP of 96%, fixed turnover of €200 has an expected loss of €8; at 98% it is €4. Those are expectations, not a promised balance or a forecast for that ledger. A €100 starting balance also does not guarantee that you can complete €200 of planned stakes. Stopping when funds run out changes the realised exposure. The UK Gambling Commission’s live-monitoring guidance distinguishes theoretical RTP from observed prizes divided by turnover.

                Can two 96% RTP games have different short-run risk?

                Consider two invented games, each costing one unit per round. Game A pays 1.20 units gross with probability 80% and zero otherwise: 0.80 × 1.20 = 0.96. Game B pays 12 units gross with probability 8% and zero otherwise: 0.08 × 12 = 0.96. Both have a 96% theoretical RTP and a 4% house edge. These are mathematical examples, not descriptions of actual slots.

                Assume independent rounds, an unchanged payout distribution and enough money to fund exactly ten rounds. The chance of receiving no payout at all is 0.20^10 = 0.00001024% for A, but 0.92^10 = 43.4388% for B. The expected net loss over ten one-unit rounds is still 0.40 units in both. Zero-payout probability is not the same as the probability of ending down: A can return some money and still produce a net loss.

                This comparison is why RTP alone cannot rank short-session safety. Independence is an assumption of these toy games; do not transfer it automatically to compensated machines, stateful features or every real product. A high RTP cannot tell you hit frequency, payout size, the risk of exhausting a balance or whether a session will finish ahead. Our expected-loss calculator calculates an average cost, not that distribution.

                  What evidence should you keep before questioning an RTP?

                  Save the game name, provider, version or rules screen, displayed RTP, currency, stake level and timestamp. Keep the settled round IDs and distinguish cash play, bonuses, jackpots and cancelled rounds. An advertised percentage for another configuration is not a reliable benchmark. The RTP version checklist helps identify this mismatch before comparing figures.

                  An observed 85% in a small ledger does not by itself prove a fault against a 96% theoretical target. Nor does it prove that a recovery is due. Monitoring tolerance depends on the game’s payout distribution and sample size; no universal number of spins establishes fairness. UKGC monitoring obligations concern relevant British licensees, not every site worldwide. If records do not reconcile, use the operator’s formal complaint route and retain the evidence rather than staking more to test convergence.

                    Does 96% RTP mean I get €96 back from €100?

                    No. It describes the theoretical average return relative to stakes under the game’s stated configuration and assumptions.

                    Your personal result can differ sharply. Repeatedly staking returned money increases turnover, so a €100 deposit is not the denominator for every later round.

                      Is house edge always 100% minus RTP?

                      Yes when both describe the same wager, payout basis, strategy assumptions and game configuration. On that basis 96% RTP corresponds to a 4% house edge.

                      Do not subtract figures from different versions or mix a conditional bonus return with the base game. The RTP and expected-return guide explains the distinction between an average and a personal outcome.

                        Will playing longer bring my losses back to the RTP?

                        There is no repayment mechanism in a theoretical average. The UKGC’s player guide explains that previous wins or losses do not alter the next win probability on random machines.

                        That statement should not be generalised to every compensated or stateful machine. More turnover can increase expected loss; it does not create an obligation to return your previous losses. Stop at a pre-set affordable money and time limit rather than chasing a percentage.

                          What does 2% house edge mean?

                          On the same wagering basis, a 2% house edge corresponds to 98% RTP and an expected loss of €2 per €100 wagered, before separate fees.

                          It is not a €2 charge on a deposit or a limit on what one session can lose.

                            Is higher or lower RTP better?

                            Under comparable rules, a higher RTP means a lower expected loss per unit wagered.

                            It does not guarantee a winning session or tell you the payout distribution. Check the exact game configuration rather than a headline percentage for an entire casino.

                              Practical checklist

                              • Confirm the RTP percentage and house edge are calculated on the same wager basis.
                              • Verify the game version and payout rules match the quoted RTP.
                              • Calculate expected loss as total wagers multiplied by the house edge percentage.
                              • Remember that expected loss is a statistical average, not a guaranteed outcome.

                              Sources and further reading

                              Fact-checked: . External sources open in a new tab.

                              1. Return to player: gaming machines (updated 2021-06-16; checked 2026-10-29)UK Gambling Commission
                              2. Live RTP performance monitoring (checked 2026-10-29)UK Gambling Commission
                              3. Testing strategy: live RTP monitoring (updated 2023-12-07; checked 2026-10-29)UK Gambling Commission
                              ← More guides