So Irish Casinos
A LITTLE PROGRESS, EVERY DAY

Your daily adventure

Three small challenges. Complete the set for +100 XP.

Daily missions reset at midnight UTC. Your badges stay with you on this device.
The maths behind the reels · Interactive explainer

Slot RTP explained:
where does the other 4% go?

Follow the money, change the numbers and watch identical RTPs produce very different results.

A slot showing 96% return to player is describing its long-run maths. It is not promising to turn your next €100 deposit into a €96 withdrawal. The missing piece is total money wagered — every stake, including stakes funded by earlier returns.

That distinction changes how you read a slot’s information panel. This guide separates the expected return, the cost of repeated wagering and the variation you can experience along the way.

01 / Understand the percentage

What does 96% RTP actually mean?

Theoretical RTP is an expected payout per unit staked. A 96% game has an expected return of €0.96 per €1 wagered and a 4% house edge. That is a statistical relationship, not a rule that makes every hundred spins settle at exactly €96. The UK Gambling Commission explicitly distinguishes the long-run average from an individual playing session. Source: RTP explained ↗

INTERACTIVE 01

Follow €100 of wagers.

EXPECTED VALUES
100 units wagered1 dot = 1 unit
Expected returnExpected house edge
THEORETICAL RTP96%
Expected return€96
Expected loss€4
A diagram of the expectation across wagers. It does not predict a player’s payout or allocate a fixed quota of winning spins.
Watch the explanation as a short video

Recorded example: the RTP changes from 96% to 92%. Use the interactive controls above to explore other values.

How is RTP calculated?

For a simple fixed-stake game, multiply each possible payout by its probability, add those expected payouts, then divide by the stake. Observed RTP uses a different calculation: total payouts ÷ total stakes × 100. It describes the sample you measured. Theoretical RTP describes the model that generated it. Source: calculating RTP ↗

One deliberately simple model · €1 stake

48% chance of a €2.00 total payout
52% chance of a €0 payout

(0.48 × €2.00 + 0.52 × €0) ÷ €196% RTP

The slider changes the winning payout in this teaching model. Probabilities stay fixed. Real slots can have many payout combinations, bonus states and configurations; this is not the paytable of a commercial game.

02 / Put the difference in euros

Is 96% RTP much better than 92%?

It is four percentage points higher. The corresponding house edge falls from 8% to 4%: half the expected loss for the same amount wagered. That comparison is more useful than a “good” or “bad” badge. The figures below are example settings, not a survey of current slots or a recommended threshold.

INTERACTIVE 02

Four points. Twice the expected loss.

FIXED COMPARISON
Total wagered€500
96% RTP4% house edge
€20
92% RTP8% house edge
€40
EXPECTED LOSS AT 92% VS 96%2×

€20 more expected loss for the same €500 of wagers.

Bars compare expected losses on a shared scale, not predicted session results. Total wagered is stake × rounds; it is not the starting deposit.

Reusing a return creates another wager. For example, a €20 deposit can fund more than €20 in total stakes if earlier rounds return money and that money is staked again. RTP is applied to that accumulated wagering. It cannot tell you how much money will be left when you stop.

You can compare further game rules with our house edge calculator. For a specific slot, start with its own help or information panel and identify the exact game configuration being described.

03 / Separate the average from the experience

Same RTP. Very different journeys.

RTP is one average; it does not describe the whole payout distribution. In the experiment below, one model returns smaller amounts more often. The other returns much larger amounts rarely. Both have exactly the RTP selected in the first visual. Their individual results can still spread widely apart.

INTERACTIVE 03

Six samples. Two payout profiles.

BOTH 96% RTP
Follow a sample
Frequent smaller returns€2.00

48% return a payout · 52% return zero

Observed return (%)06012018024096% target101,000 rounds
Highlighted sample’s observed RTP96.8%
Rare larger returns€24.00

4% return a payout · 96% return zero

Observed return (%)06012018024096% target101,000 rounds
Highlighted sample’s observed RTP79.2%
Solid lines: generated samplesDashed line: theoretical RTPBoth charts use the same axis scale
Sample 1: 96.8% observed return in the frequent-payout model and 79.2% in the rare-payout model. Each has €40 expected loss across €1,000 of wagers, regardless of these individual results.
A fixed sample of independent €1 trials, not a session with a finite starting balance. There is no balance-based stop rule. Results are generated from the stated probabilities, never forced towards RTP.
Show the model and exact results

For each profile, payout = selected RTP as a decimal ÷ probability of a payout. A seeded pseudo-random generator draws independent trials for each sample. Changing the RTP or sample size retains the seed; “New experiment” uses a new seed. The experiment is reproducible, not a simulation of any named slot.

Seed 41. Plots start at round 10 so the first few outcomes do not dominate the scale. Payout labels are rounded; calculations use full precision.

SampleFrequent: paid / stakedRare: paid / staked
1€968 / €1,000 · 96.8%€792 / €1,000 · 79.2%
2€970 / €1,000 · 97.0%€1,128 / €1,000 · 112.8%
3€980 / €1,000 · 98.0%€1,008 / €1,000 · 100.8%
4€982 / €1,000 · 98.2%€1,104 / €1,000 · 110.4%
5€910 / €1,000 · 91.0%€624 / €1,000 · 62.4%
6€954 / €1,000 · 95.4%€864 / €1,000 · 86.4%

Increasing the sample size often makes the measured percentage steadier. It also increases total wagering and the expected loss in euros. It is not a reason to keep playing until your result “catches up”. For independent random rounds under unchanged rules, a losing run does not make the next payout due. Source: random outcomes ↗

These examples differ in their variance as well as their payout frequency. An actual slot’s volatility depends on its full payout distribution; hit frequency alone does not specify it. The regulator’s RTP-monitoring guidance also treats volatility as necessary context when assessing measured return. Source: RTP and volatility terms ↗

Keep the useful part

RTP questions, answered plainly

Does 96% RTP mean a 96% chance of winning?

No. In the two examples above, the chances of receiving a payout are 48% and 4%, yet both start at 96% RTP. Probability and payout size together determine the expected return.

Can I work out a real slot’s RTP from 100 spins?

You can calculate the observed return of those 100 spins. That sample does not establish the game’s theoretical RTP. Large rare payouts make short-run estimates especially uncertain.

Does higher RTP guarantee I lose less?

It reduces expected loss for the same wagering, all else equal. A particular result can be very different. A higher RTP also does not compensate for increasing your total stakes without limit.

Next: read a slot’s information panel or explore roulette at the learning table.

Updated 7 September 2026 · Independent teaching models · Sources linked throughout the article.

18+ · Please gamble responsibly · National Gambling Helpline 1800 936 725