Players who move between a pub machine in Napier and a pokie on an offshore site usually notice the difference before they can name it. The pub machine feels tighter. The online game advertises a return percentage on its own info screen; the pub machine advertises nothing at all.
Both impressions are basically correct, and the reason is not that one is rigged and the other honest. It is that the two games are built to different economic briefs, imposed by different rules, and the gap between them is the direct arithmetic consequence of those rules.
What a Class 4 machine is required to fund
A pub or club machine in New Zealand is not primarily a game. It is a fundraising instrument that the law permits to look like a game.
The machines are operated by licensed societies, not by the venue, and those societies exist to generate money for authorised community purposes — charity, amateur sport, non-commercial community activity. The regulations set a floor: a minimum proportion of gross machine proceeds, measured GST-exclusive, must go to those purposes. The commonly cited figure is 40 per cent, and it is a minimum rather than a target.
On top of that obligation sit gaming duty, GST, the problem gambling levy, and a regulated payment to the venue for hosting the machines. Every one of those comes out of the same pool the payout percentage is drawn from.
That is the whole explanation for why a pub machine pays what it pays. There is no hidden switch. The money simply has more claims on it before it can be returned to players.
What "40 per cent to the community" does not mean
This figure is misread constantly, usually in the direction of assuming players get the other 60 per cent.
It is not a share of what you put through the machine. Gross machine proceeds are what the machines retain after prizes — turnover minus payouts — not the total amount wagered. Because pokie play recycles, a machine can run a very large amount of money through in an evening while the proceeds figure stays modest. The community share is calculated on the proceeds, and the payout to players happens before the calculation, not after it.
So the 40 per cent is a claim on the house's take, not on your stake. It still matters to players, because a house take that has to cover a mandatory community share, plus duty, plus levy, plus a venue payment is a house take that has to be larger in the first place.
What online pokies publish, and what it means
An online pokie carries a return-to-player figure on its information screen, usually somewhere in the 94–97 per cent band, and it is put there by the game studio rather than the casino.
Two things about it are routinely misunderstood.
It is a long-run average across millions of simulated rounds, not a promise about a session. A game with a 96 per cent RTP does not return 96 cents of every dollar you feed it over an evening; it does so over a volume of play no individual will ever generate. Your actual session outcome is governed by variance, and variance dominates completely at the scale a real person plays at.
And it is calculated on turnover, including money that was won and re-staked. A hundred dollars cycled through a 96 per cent game twenty times loses a great deal more than four dollars, because the 4 per cent margin applies to each cycle rather than to the original stake. This is the single most useful thing to understand about any published RTP figure, online or otherwise.
Why the online figure is visible and the pub figure is not
Online games are supplied by studios that sell the same title into many regulated markets, and publishing RTP is a condition of operating in several of them. Once the number is in the game's help screen for one jurisdiction, it is there for all of them.
Class 4 machines in New Zealand are a domestic product under a domestic regime, and there is no equivalent expectation that the figure be displayed on the cabinet. Minimum return standards exist as a matter of machine approval, but they are not something the player reads off the screen.
The asymmetry is about disclosure convention, not about one sector hiding something the other reveals. It does, however, mean a player can compare online games against each other and cannot really do the same in a pub.
Volatility is the part that actually shapes a session
Two games with an identical RTP can feel entirely different, and the difference is volatility — how the return is distributed.
A low-volatility game returns small amounts often. A high-volatility game returns almost nothing for long stretches and occasionally a great deal. Both can advertise 96 per cent and both can be telling the truth. The high-volatility game will produce more sessions that end at zero and a small number that end very well, and that distribution, not the headline percentage, is what determines whether a given evening feels good.
Pub machines in New Zealand tend toward the lower-volatility end because the stake and prize limits constrain the top of the distribution. That is a regulatory outcome, not a design preference.
What is actually worth checking
For online play, the RTP is on the game's info screen and it is worth looking at — but only alongside the volatility rating, because the two together describe the game and either alone does not. Bear in mind that some titles ship in multiple RTP configurations, and the operator chooses which one to deploy; the number in the help screen is the one in use, which is why it is worth reading at the game rather than from a list somewhere else.
For a pub machine, there is nothing to read, and the honest advice is to treat the comparison as settled: a machine funding a community grant, a duty, a levy and a venue payment out of the same margin will return less than one that funds none of them. That is not a scandal. It is the arrangement, working exactly as designed.
What neither figure will do is tell you how an evening goes. RTP describes the long run, and no one plays the long run.