One slot, two returns, and only one of them is yours

Every slot carries a published "return to player" (the share of all stakes it gives back over a very long run), and on a game with a pot bolted on to it that single figure is the sum of two unlike quantities which the supplier adds together without ever saying that it has.

One part is paid out by the reels, round after round, to whoever happens to be sitting in front of them; the other part is set aside for a prize which will be collected once, by one person, somewhere on a "network" that may run across a hundred casinos at a time. Added together they look like one number. Spent, they behave like two.

The formula is short enough to write on a line, and it is the one piece of arithmetic that makes the rest of this subject legible: the "total return" equals the return of the game without the jackpot, which the trade calls the "base return" (the part the reels themselves give back to you), plus the share assigned to the "pot" (the trade calls that second term the "jackpot share"), and nothing else at all enters the sum.

We take that decomposition from a secondary source rather than from a supplier, and we say so plainly, because the companies which publish the total almost never publish the "split", and the few figures that exist for the second term have had to be recovered by subtraction.

What follows from the formula is the most consequential sum on this site, and it is also the sum that no comparison table in the trade performs. A reader who wants to know what a jackpot slot is costing him has to subtract before he compares (total minus base, and nothing more complicated than that), and the numbers he needs are in our table below.

The same network, the same pot, seven points apart

The clearest case in the business sits inside one network, where nine titles share a pool and a contribution and still differ by nearly seven points of total return. We keep that table on the Mega Moolah page, where every row belongs to the network it describes, rather than repeating it here; what matters for this page is the shape it shows.

The base game is where the difference lives, because the "jackpot share" (the slice of every stake that leaves the game and joins the pool) is identical on all nine. A title at 88.12 per cent of base return and one at 87.37 are paying the same 5.30 into the pot, and the player who never wins the pot keeps whatever the base game gives back.

What the player who does not win is really paying

Twenty-nine titles sit on the Mega Moolah network and every one of them feeds the identical pot, so a reader might reasonably suppose that they all cost the same to play; they do not, and the spread between them is wide enough to be the first thing worth checking before you open any of them.

At the kind end of the table the original game and the seven "Aurum Signature" titles return 88.12 per cent from the reels (88.00 on Mega Moolah Goddess, 87.37 on Fortunium Gold Mega Moolah), which with the 5.30 points of "contribution" added back makes a published total of 93.42 (the kindest pair of figures on the network). At the other end Mega Moolah "Megaways" returns 81.63 from the reels for a total of 86.93 (the lowest pair of figures anywhere on the network). Between those two totals there is close to seven points of return, on games which are chasing one and the same prize.

Now take the 81.63 and read it for what it is, because this is where the published figure stops being useful. The 5.30 points which separate it from the total are not lost money: they are real, and they will be paid, in one lump, to somebody who is almost certainly not the reader sitting at the game.

For everybody else they are "return on paper", which is to say a share of the stakes which passes through your account and comes out the other side as somebody else's cheque. That is the trick, and it is not hidden.

Read that way the game has an honest figure, and it is a figure we have not seen printed anywhere: a player who never wins the "Mega", which is practically every player who ever loads the game (the level falls about five and a half times a year across the whole network), is facing a "house edge" of 18.4 per cent instead of the 13.1 the published total implies (the two figures differ by the whole of the contribution). Between those two numbers lies the difference between an expensive slot and a punishing one.

The one decomposition that somebody actually published

Games Global has never published the split of a single title in its catalogue, and neither has Blueprint, nor Relax Gaming, nor Playtech (four suppliers, and not one line of decomposition between them), which means that the whole of this reasoning has to rest on the one game whose breakdown exists in print at all.

NetEnt gives Mega Fortune a total of 96.60 per cent on its own pages (the only figure in this paragraph that comes from a supplier), and the full "breakdown", which reaches us from a secondary source and not from NetEnt, assigns 62.8 points to the base game, 16.8 to the "free spins", 9.3 to the bonus feature and 7.60 to the three "progressives", stripped of its jackpots the game returns 89 per cent.

That last number deserves a moment of attention, because 89 per cent is not a modern figure for a "video slot", while Mega Fortune is a modern game by the standards of this trade (released on 23 January 2009, and still "in distribution" in 2026). Its reputation rests on a pot; its reels have been built to pay for one.

We print the decomposition with its status attached, which is the only responsible way to print it, since the 96.60 total sits on the supplier's own page while the four components beneath it do not. If NetEnt ever publishes the split itself, we will replace the line and say that we have replaced it.

The same game, sold twice, with the price on the label

There is exactly one clean experiment in this subject, and it exists only because Relax Gaming happened to ship the same slot in two configurations (one with five jackpot levels, one with none at all) and to publish a return for each of them.

Strip the jackpot levels out of Money Train Origins and the published return is 96.11 per cent (the plain title is still in the catalogue). Attach them, and the figure on the label of the Dream Drop version reads 94.00. The 2.11 points which sit between the two are the price of the pot, stated by the maker, on one set of reels, with nothing at all deduced by us.

It is the only comparison of its kind that we could find across every supplier in this document, which is itself a finding worth recording, because where a game exists in a single version the reader has neither a control nor a measurement and has to take the total on trust.

The same subtraction, read from the other end, explains the whole of the first generation of these machines. Treasure Nile is published at 92.11 per cent and Major Millions at 89.37 (LotsaLoot, at 89.95, is no better), and because those are totals rather than base figures, the oldest progressives are even dearer than they look.

A pot does not have to mean a poor return

The conclusion a reader might draw at this point is that progressive slots return less than ordinary ones, and it would be the wrong conclusion drawn from the right evidence.

Divine Fortune is published at 96.59 per cent in total (released 24 January 2017) and Mercy of the Gods at 96.64 (17 December 2018), both of them with a progressive level attached; both of them squarely in line with slots that have no pot whatever. Mega Joker goes further still, since it is published at 99 per cent in its "Supermeter" mode (a figure which reaches us from a database and not from NetEnt), with a hit frequency of 17 per cent and a "maximum payout" of 200 times the stake. It is a progressive game with a better return than most "flat" slots on the shelf!

What separates those three titles from Mega Moolah Megaways is not the presence of a jackpot, which they all have. It is how much return the supplier chose to move out of the reels and into the pot, and how improbable the level at the top of the ladder has been made. We know of no table in this trade that shows either!

Put the two readings together and the rule for a reader comes out short enough to carry. The question is never whether a slot has a jackpot; it is how many points of return have left the reels to pay for one, and whether anything the supplier publishes lets you work that out. On five of them (Jackpot King, WowPot!, Dream Drop, Jackpot Play and Jackpot Madness), as what a pot takes from a stake sets out in detail, nothing does.

Three casinos where these titles run

• Jackpot CityAlderney AGCC Category 1, Baytree (Alderney) Ltd · Kahnawake 00892Visit
• The HippodromeGambling Commission account 39372, Betway Limited, white label in the public register · MGA/CRP/130/2006Visit
• Dove SlotsGambling Commission account 39175, Jumpman Gaming Ltd, domain Active in the public registerVisit

18+ · Sponsored link · Terms on the operator's site

Questions