This demo pays far more than a real-money version would
145%measured in the lab96–97%published by PG SoftPG Soft · 14 games · 218,502 spins
Every one of the 14 paid back more than it took — and still did at the bottom of its own margin of error.
Read the full explanation ▾Close ▴
We ran 218,502 spins across PG Soft's 14 games in the lab. The weighted average was 145%, against the 96–97% the provider publishes. Put in 100, get back 145 — no real-money slot does that.
So treat everything on this page as a description of how the game behaves, not as what a real-money version would pay. The other figures — how often free spins land, how often a spin pays anything — may have been loosened the same way.
This is PG Soft's own official demo — the launch token our collector uses is the same one PG Soft's own demo page hands out. What we cannot measure is the real-money version, so we are not saying that one is rigged. We are saying the free version people practise on is not the same machine.
Measured return144.2%official 96.8%± 15.6
Hit rate25.1%per 100 spins
"Wins" smaller than the bet26.3%of the spins that pay
Free spins—0 so far — too few to average
Biggest result×612official max 2,500x
Since last free spins21,340spinsLongest drought 21,340 spins
Measured return as spins pile up
The dashed line is the official RTP. The more spins, the closer it should get — and the narrower the shaded range.
Last 30 spins
no winwon less than betwinfree spins
How lopsided the payouts are
11.7×the stake, per spin
Spread of a single spin. The higher it is, the more the return hides in a few rare spins — and the longer the measured RTP takes to settle.
The official numbers — and whether they hold up
Read from the game's own settings
Official figure is outside the range — needs more spins
official 96.81%measured 144.2%
128.6%likely range159.8%
Official RTP is 96.81%. Over 21,340 spins we measured 144.2% — likely range 128.6% to 159.8%. The official figure falls outside that range, which is unusual and worth more spins before we draw any conclusion.
Why the range is so wide: almost all of a slot's payback sits in a handful of rare big wins, so the measured return stays noisy for a very long time. Getting this range down to ±1 point on a high-volatility game takes millions of spins. That's why we lead with hit rate and free-spin frequency instead — those settle within weeks.
Still measuring · a measured return needs 20,000 spins to mean anything
23.2%Hit rate26.5%"Wins" smaller than the bet—Free spins0 so far×200Biggest resultofficial ×1,000
Fortune Dragon19,280 spins ⚠
Still measuring · a measured return needs 20,000 spins to mean anything
20.0%Hit rate30.3%"Wins" smaller than the bet—Free spins215 so far×294Biggest resultofficial ×2,500
Ganesha Gold17,960 spins ⚠
Still measuring · a measured return needs 20,000 spins to mean anything
46.4%Hit rate51.8%"Wins" smaller than the bet—Free spins173 so far×410Biggest resultofficial ×100,000
Mahjong Ways 213,380 spins ⚠
Still measuring · a measured return needs 20,000 spins to mean anything
30.8%Hit rate52.1%"Wins" smaller than the bet—Free spins79 so far×463Biggest resultofficial ×100,000
Looking for the free demo?
Every slot here has an official free demo run by PG Soft itself — that is exactly where our numbers come from. We don't link out to anyone, so here's how to find it yourself: search the game's name together with "PG Soft" and open the result on the provider's own website, where each game has a Demo button. No account, no deposit.
Why bother finding the official one: a casino can license a lower-RTP version of the same game. The demo on the provider's own site is the default configuration — the one our numbers describe.
FAQ
Is this data real?
Yes. Every number on OddsLab comes from spins we actually ran on the provider's own official free demo, recorded by a program that logs every spin and reconciles the balance after each one. If the balance doesn't match the win minus the bet, the line stops and we don't keep the data.
Does a long losing streak mean a win is coming?
No. Each spin is independent. Our hot-slot experiment groups real spins by drought length: the chance of free spins is the same after 10 empty spins as after 500.
Is the demo the same as real-money play?
The game maths in the demo is the provider's default configuration. Casinos can license lower-RTP versions of the same game, so real-money results can be worse than what we measure — never better.
Can I use this to beat the slot?
No. These numbers describe a game's character — how often it pays, how many "wins" are really losses, how rare the big wins are. None of it predicts the next spin, and over time every slot pays back less than you put in.
Why don't you just publish the measured RTP?
Because at our sample sizes it isn't a usable number yet. On a high-volatility slot the measured return can be fifteen points off in either direction after 40,000 spins. We show it with its margin of error and lead with the figures that are already accurate.
Fortune Ox20,940 spins ⚠
Fortune Rabbit20,020 spins ⚠
Fortune Mouse19,740 spins ⚠
Fortune Dragon19,280 spins ⚠
Ganesha Gold17,960 spins ⚠
Mahjong Ways 213,380 spins ⚠