EXPLAINER Open Data Published · 5 min read · 1,030 words

Inside Our Low-Volatility Slot Model: 32.6% of Spins Pay and 25% of the Return Sits in the Bonus

Every slot simulation we publish runs on one of three models. This is the low-volatility one, band by band, with what it can and cannot say about a real game.

Inside Our Low-Volatility Slot Model: 32.6% of Spins Pay and 25% of the Return Sits in the Bonus

Only 32.6% of spins produce any payout and the bonus round fires just once every 120 spins on average.

This means a player loses on average 0.035 stakes per spin. In a 1,000-spin session the median result is a 65-unit loss; the bottom 5% lose 276 units or more, while the top 5% can walk away with 303 units or more. Bonus rounds supply 25.3% of the total return, the base game the remaining 74.7%.

WHAT WE FOUND
  • Only 32.6% of spins produce any payout.
  • Bonus rounds trigger on average once every 120 spins.
  • The model’s standard deviation per spin is 5.6 stakes while the expected loss is 0.035 stakes.
  • Bonus rounds contribute 25.3% of total return, base game 74.7%.
  • Wins of 100x or more appear once every 3,010 spins and make up 7.4% of the return.

What the model is made of

32.6% of spins pay something in the low-volatility model calibrated to 96.5% RTP, and the bonus round triggers once in 120 spins on average.

Within the base game, 60% of hits pay less than the stake, which translates to 19.0% of all spins being payouts smaller than the bet that produced them.

The return is split between the two layers of play: bonus rounds supply 25.3% of the model’s return and the base game supplies the remaining 74.7%.

The bulk of base-game hits cluster in the band that accounts for 55.0% of those hits and pays from 0.16x up to 0.75x. The highest-payout band, representing 2.0% of base-game hits, pays from 15.70x up to 47.10x.

Payouts are drawn at random from the published bands; the model contains no reels, paylines, symbols or features and does not replicate the math of any named game.

Where the return sits

Most of the bonuses fall into the first band, which accounts for 70.0% of all bonus rounds and delivers wins ranging from 6.28x to 23.55x the stake.

A smaller slice, 25.0% of bonuses, lands in the middle band and pays between 23.55x and 70.66x. The rarest 5.0% of bonuses occupy the top band, where payouts start at 70.66x and can reach as high as 392.53x.

At the calibrated RTP of 96.5%, the model’s ceiling is a 393x win. Wins of 100x or more appear once in 3,010 spins and they contribute 7.4% of the whole return.

All payouts are selected at random from the published bands; the model contains no reels, paylines or symbols, so the distribution of returns is fully defined by those rows.

Because the bands are the complete definition of the model, any reader can reconstruct the same distribution by using the three rows above and the seed we disclose for each run.

How far a session lands from the average

The standard deviation of a single spin is 5.6 stakes, against an average loss of 0.035 stakes a spin. Those two numbers set the scale for any short run: a single bet can swing wildly while the expected drift is a tiny fraction of a stake.

When we string 1,000 spins together, the distribution stretches. In 5,000 simulated sessions of 1,000 spins at 1 unit the median session finished 65 units down. The unluckiest 5% lost 276 units or more, and the luckiest 5% won 303 units or more. A session that lands near the median is still far from the per-spin average loss, because the variance accumulates.

Over the full run of 5,000,000 spins the model returned 96.56%, against the 96.5% it is calibrated to. The sampling noise at that length is about 0.49 percentage points. That noise figure tells you how much the observed return can drift from the calibrated RTP purely by chance, even with millions of spins.

What the model is not

Lower RTP variants are generated by scaling every payout with a single factor: at 94.0% each payout is 97.4% of its size at 96.5%, and nothing else changes. The model draws each payout at random from published bands; it has no reels, paylines, symbols or features, and it is not the math of any named game. A real slot with the same RTP and a similar volatility label will place its wins differently.

Until 17 November 2026 the calibration of the models left out one factor, and they returned between 0.1 and 0.6 percentage points less than the RTP stated for them.

Repeated by us with that seed, a run gives the same figures. The code that runs the models is not public. The payout bands in the tables are the whole definition of a model, so a reader can rebuild it and will get the same distribution of results, not the same draws.

A base-game hit that pays between 0.16x and 0.75x makes up 55.0% of base-game hits. A bonus round that pays between 6.28x and 23.55x accounts for 70.0% of bonus rounds.

Thus, the model is a stochastic band-draw engine, calibrated to a target RTP, with a single scaling knob for lower RTP, and its full definition is captured in the published payout bands.

THE VERDICT

The model shows that only about a third of spins pay and the average player loses 0.035 stakes per spin. Over 1,000-spin session the median loss is 65 units, with the unlucky 5% sinking 276 units or more while the lucky 5% can win 303 units or more.

The data behind this piece

Every figure in the text comes from the runs below. They are computer simulations and exact calculations, not real-money play, and they can be reproduced: the model parameters are stated and the random seed is fixed.

How the low-volatility model is built

How we ran it: The payout bands, trigger rate and derived shares of the low-volatility model calibrated to 96.5% RTP: any-win frequency 32.6%, bonus round on average once per 120 spins, 25% of the return comes from the bonus round, top payout 393x the stake, computed exactly from the model definition, plus a check of the return over 5,000,000 simulated spins (5,000 sessions of 1,000). Random seed: 39572

Base-game payouts of the low-volatility model (32% of spins without a bonus pay)
Share of base-game hitsPays fromPays up to
55.0%0.16x0.75x
33.0%0.79x3.93x
10.0%3.93x15.70x
2.0%15.70x47.10x
Bonus-round payouts (one bonus per 120 spins)
Share of bonus roundsPays fromPays up to
70.0%6.28x23.55x
25.0%23.55x70.66x
5.0%70.66x392.53x

Questions players ask

How often does the bonus round trigger?

The model triggers a bonus round on average once every 120 spins. Those bonus rounds provide 25.3% of the total return, with most (70.0%) falling in the first payout band (6.28x-23.55x), a smaller slice (25.0%) in the middle band, and the rarest 5.0% in the top band.

What is the expected loss per spin and over a typical session?

Each spin carries an average loss of 0.035 stakes while its outcome varies with a standard deviation of 5.6 stakes. In simulated sessions of 1,000 spins the median result is a 65-unit loss; the worst 5% lose 276 units or more, and the best 5% win 303 units or more.

How likely am I to hit a huge payout like 100x?

Wins of 100x or more appear once every 3,010 spins in the model, and they account for 7.4% of the total return. The largest possible payout is 393x the stake, but such extreme wins are exceedingly rare, making them a tiny slice of overall outcomes.