By Dana Whitlock, Senior Games Analyst
If you take one thing from this article, make it this: return-to-player describes the behaviour of a game across millions of spins made by thousands of people. It says almost nothing about what will happen to you tonight. Both of those statements are true at the same time, and holding them together is the whole skill.
What the number actually measures
Return-to-player is the percentage of all money wagered on a game that the game returns as wins, calculated over the game's complete cycle. A pokie certified at 96.4% is mathematically designed so that, across its full theoretical run, A$96.40 of every A$100 staked comes back out as prizes. The remaining A$3.60 is the house edge — the operator's margin.
That figure isn't marketing. It's derived from the game's paytable and reel weightings, then verified by an independent testing house running the RNG through tens of millions of simulated spins. In Australia the labs you'll see most often are Gaming Laboratories International, iTech Labs and BMM Testlabs. Their certificate is the reason the number can be trusted at all.
The timescale problem
Here's where nearly everyone goes wrong. "Millions of spins" is not a figure of speech. For a typical pokie, the theoretical cycle that produces the certified RTP runs to somewhere between ten million and a hundred million spins.
A busy player might manage 500 spins in an hour. At that rate, personally completing one full cycle of a ten-million-spin game would take about 2,280 years of continuous play. You are never going to experience the certified RTP. You are going to experience a tiny, wildly variable sample of it.
RTP is a statement about the game. Variance is a statement about your session. Confusing the two is why people say a machine "feels cold" — they're describing normal short-run variance in a system that has no memory.
Why volatility matters more than RTP over one night
Two pokies can both be certified at 96.4% and behave nothing alike. The variable that separates them is volatility — sometimes called variance — which describes how the return is distributed across spins.
A low-volatility game pays small amounts frequently. You'll hit something on roughly one spin in three, most of those wins will be under your stake, and your balance will erode gently. A high-volatility game might hit once in six or seven spins and pay almost nothing most of the time, saving its return for rare bonus triggers that pay hundreds or thousands of times the stake.
| Profile | Hit frequency | Typical top win | A$50 at A$0.50/spin |
|---|---|---|---|
| Low volatility | ~1 in 3 | 500× | Often lasts 200+ spins |
| Medium volatility | ~1 in 4 | 2,000× | Typically 120–180 spins |
| High volatility | ~1 in 6 | 10,000× | Often gone inside 80 spins |
| Extreme volatility | ~1 in 9 | 50,000× | Can be gone in 40 spins |
Same RTP, four completely different experiences. This is why "just play the highest RTP game" is incomplete advice. If you have A$50 and you want two hours of entertainment, a 95.8% low-volatility pokie will serve you far better than a 96.6% extreme-volatility one.
A rough rule for bankroll. Budget around 100× your spin size for medium volatility, 250× for high, and 500× for extreme. At A$0.50 a spin that's A$50, A$125 and A$250 respectively. Below those numbers you're not really playing the game — you're buying a lottery ticket on its bonus round.
The thing most casinos won't tell you
Here's the part of the industry that deserves more scrutiny. Studios frequently ship the same game in multiple RTP configurations — commonly 96.5%, 94.2% and 92.0%. The artwork is identical, the paytable looks identical, the name is identical. The operator chooses which band to run, and that choice is usually invisible to the player.
A 4.5-point spread might not sound like much. On A$10,000 of turnover it's the difference between an expected loss of A$350 and an expected loss of A$800. Over a year of regular play it's substantial money.
Most operators run the low band and don't mention it. Some list the theoretical maximum in a help file while running something lower. We take the other approach: MATE29 contracts for the highest available configuration on every title, publishes the live figure on the game tile, and removes anything that drops below 94%. You can see every number on the games page without opening a single help file.
How to actually use RTP
Given all of the above, RTP is still worth paying attention to — you just have to use it for the right job.
- Use it to compare, not to predict. Between two games you'd equally enjoy, take the higher RTP. It's free expected value. Don't use it to forecast tonight's result.
- Pair it with volatility, always. Decide how long you want the session to last first, pick the volatility that fits, then take the best RTP within that band.
- Watch it when clearing a bonus. If you're working through wagering, high RTP plus low volatility is straightforwardly the correct choice — it minimises the chance of busting before you finish.
- Treat missing RTP as a red flag. If a casino won't publish the figure for a game, assume it's running the low band. There is no other reason to hide it.
Four things RTP is definitely not
- Not a schedule. A game that hasn't paid in 200 spins is not "due". Each spin is independent; the RNG has no memory of the last one.
- Not affected by your stake. Betting A$5 instead of A$0.50 doesn't improve your return percentage. It just moves ten times as much money through the same maths.
- Not affected by time of day. There's no quiet-hours setting. Certified RNGs run identically at 3am and 3pm.
- Not a guarantee of anything. It's a long-run expectation with enormous short-run variance around it. Budget for the loss.
The honest summary. Every pokie has a negative expected value for the player. A 96.4% RTP means that over the long run you lose 3.6% of everything you stake — that's the price of the entertainment, and it's not beatable through game selection, timing or staking patterns. Play with money you've already decided you can spend, set a limit before you start, and read our responsible gambling guide if any of this stops feeling like fun.
Where to go next
If you want to see the theory in practice, every title in the MATE29 library lists its live RTP on the tile and runs free in demo mode — spin a low-volatility game and a high-volatility game for ten minutes each and the difference becomes obvious immediately. If bonuses are your next question, our guide to wagering requirements covers how RTP interacts with turnover targets.