New online slots: everything a reel publishes on day one, and the one thing it never will

1

Everything a reel publishes about itself on day one

The useful thing about a brand-new machine is how little about it is actually unknown. Open a session on one and the server hands the game a block of figures before a single turn is taken: the return to player, the largest round the machine can pay stated as a multiple of one stake, the number of reels and rows, the line count, the selectable stake steps and the paytable the configuration a game is sent. The page the maker keeps for a title carries the return as well, in a basic game information block that holds that figure and nothing else beside it.

None of those figures improves with age, and none of them is provisional. A return of 96.50 per cent on a page published this month means exactly what the same figure means on a page published in 2020: the share of everything staked that the machine’s own mathematics returns over its whole distribution of outcomes. A reader who waits six months before trusting it is waiting for nothing to happen to it.

So a new reel can be weighed on the day it appears, which is what this card does. A credit of £30.00, a requirement of twenty times a figure of the order such terms carry, a turn priced at the ten pence supposed throughout named again at every sentence that spends it, and a published figure that was inside the machine before anybody had played it.

What genuinely does not exist on day one is the subject of part two, and the precise form of it matters more than the vague one: not a rating, not a reputation and not a consensus, none of which were ever quantities. What does not exist is an observed rate, and the reason it does not is arithmetic rather than impatience.

2

The one quantity that never arrives, however long anybody waits

The regulator’s own wording for what a published return is describes an average theoretical return to player percentage, and the word theoretical is the whole of the matter. The figure is worked out of the machine’s own mathematics across every outcome it can produce, including the outcomes a player will almost certainly never see. An observed rate, by contrast, is whatever happened to a particular run of turns, and it only approaches the published figure once the rare outcomes have had their turn.

Put a number on that, because this family makes it possible. The original publishes one hit rate and only one: its biggest round arrives once in 3,880,481 turns, the one rate this family publishes anywhere the top-round rate the data carries. A licensed build may not run more than 1,440 cycles in an hour, so a single visit from that outcome is 2,694 hours and 47 minutes of unbroken play, which is a little over a hundred and twelve days without sleeping. Any observed rate gathered over anything shorter is missing the largest single component of the figure it is trying to measure.

This is why no page on this site reports what a machine has been doing lately, and why a new reel is not at a disadvantage against an old one. Six years of other people’s play has not produced an observed rate for the original either: nobody publishes one, the maker does not, and the figure everybody quotes for it is the theoretical one that was in the machine on the first day.

The honest summary is short. On day one a reader has the published return, the published ceiling and the published frequency of that ceiling. On day two thousand they have exactly the same three figures. Nothing accumulates, which means nothing is gained by waiting and nothing is risked by not waiting, in the specific sense that the arithmetic a reader can perform is identical either way.

3

The newest page and the first one, priced on one credit

Take the most recently published page of this maker’s thirty-six, and the first of them, and run one offer across both. The newer page carries 96.50 per cent and a ceiling of 20,000 times the stake. The first page carries 96.71 per cent and a ceiling of 2,100 stakes in one round both figures off the family table. Twenty times a £30.00 credit is £600.00 of staking either way, and what each reel takes out of that staking is the complement of its own published figure: £21.00 on the newer one and £19.74 on the older.

The credits therefore close at £9.00 and £10.26, and the house writes the step as E = B × (1 − W × (1 − RTP) ∕ g), where W is the requirement in the terms, B the credit, g the weight a slot counts at and RTP whichever published figure the reader has landed on.

The closing row carries a minus sign and it is not a misprint. On this family, the newest reel is worth £1.26 less of the same credit than the one published six years before it, because its published figure is 0.21 points lower and nothing else in the offer changed. Newness is not an improvement in the quantity that decides the money, and on these two pages it runs the other way.

£600.00 of staking asked of a £30.00 credit, which is 6,000 turns, and £21.00 of expected cost on the newer reel against £19.74 on the older

4

Two dates for one machine, twelve days apart

Here is the oddity this card was built around. For the family’s first title, the maker gives two different dates in two of its own places. Its announcement of the game is dated 14 December 2020. The structured data embedded in its own page for that game gives a publication date of 2 December 2020. Twelve days apart, both the maker’s own, and neither of them labelled a release date by anybody the disagreement as the sheet records it.

The maker’s game page states no release date at all, neither for this title nor for any other of the thirty-six. What exists is the date an announcement was published and the date a web page was published, which are facts about documents rather than about a machine going live anywhere. This publisher therefore prints neither as the release date, and says which of the two it is using whenever it uses one.

Twelve days is worth something concrete, which is the only way this card knows how to treat a discrepancy. At the pace ceiling a licensed build carries, 1,440 cycles an hour, twelve days of continuous play is 414,720 turns, and with a turn at the supposed ten pence that is £41,472.00 of staking. The gap between a maker’s two accounts of its own machine is therefore wider, in turns, than most readers will play in a lifetime. The point is not that somebody made a clerical error. It is that the date a machine arrived is not a published quantity, while the figures that decide what it pays are.

That inversion is worth holding on to. The thing a search for a new machine is sorting by, its newness, is the one thing the maker does not state. The things nobody searches by, the return and the ceiling, are stated precisely and from the first day.

5

What a date can be used for and what it cannot

The sheet behind this site carries a page publication date for all thirty-six titles, running from 2 December 2020 to 5 October 2026, and it labels the column for what it is. Sorted by that date the family lines up in a clean order, and the order is good for exactly one thing: telling a reader which pages the maker put up most recently. It is evidence about publishing, not about availability.

Three uses of such a date are legitimate and worth naming. It identifies the most recently published title in a family, which is what a reader hunting a new machine is usually after. It settles which of two titles the maker documented first, which matters when one title’s description borrows wording from another’s. And it dates the reading itself: every figure on this site carries the day it was read, because a published figure can change and a reader is entitled to know how fresh the arithmetic is.

Three uses are not legitimate. It is not a release date, and nothing on this site calls it one. It is not evidence about how long a machine has been available in any particular lobby, which is a fact about that lobby. And it is not a proxy for quality in either direction: the newest page in this family publishes 96.50 per cent and the oldest publishes 96.71, so a reader sorting by date and hoping to sort by return would have sorted it backwards.

The general rule this card follows: use a date to find a machine, never to price one. The pricing is done by the quantities the machine states about itself, and those are the same quantities whatever the date beside them says.

6

What six years actually changed about these reels

Set the family’s first pages against its most recent ones and four things have moved, none of them the thing a reader would hope for.

The ceilings have risen enormously. The first title pays at most 2,100 stakes in one round and three others of the family share that cap, while the most recent pages reach 20,000 and 25,000 times the ceilings in the game’s own data. Where a turn costs ten pence that is a move from £210.00 to £2,500.00 as the largest single round. The published returns have not risen with them: the newest page is 0.21 points below the oldest, and the family’s whole band of figures has not improved in six years of titles.

Purchasable features have appeared. The two oldest titles report a single flat return and no bonus round that can be bought, while thirty-three of the thirty-five slots now carry a separate published return for each bet mode, the bonus-buy options among them. And the shapes have multiplied: five reels by three rows in the early titles, and since then five by four, six reels with as many as seven and nine rows, and one title on a three by three grid.

So newness in this family means a bigger ceiling, more ways to bet and a slightly lower published return. A reader who prefers the chance of something large is being offered exactly that, and paying 0.21 points for it on the two pages this card weighs, which comes to £1.26 on the credit above. That is a trade a reader can make deliberately. What they cannot do is make it accidentally by sorting on a date.

7

What waiting for evidence would cost a reader

Suppose a reader distrusts a figure published by the party that built the machine and wants to see it demonstrated before committing. Price that. To put ten thousand turns through a new reel at the ten pence supposed throughout is £1,000.00 of staking, whose expected cost at 96.50 per cent is £35.00. Ten thousand turns is also six hours and fifty-six minutes at the pace ceiling a licensed build carries, and at the end of it the reader has an observed rate drawn from a sample that cannot contain the machine’s largest outcome, which arrives once in nearly four million turns on the only title that publishes such a rate.

The cost of that exercise is set by the turn price and by nothing else, which is worth seeing in figures. The same ten thousand turns, taken at the £5.00 a licensed cycle will accept, is £50,000.00 of staking, expected to cost £1,750.00 at the newest page’s published figure, and at the £2.00 a player under twenty-five is held to it is £20,000.00 of staking and £700.00 of expected cost. The sample is identical in every case: the same count of turns, the same machine, the same inability to contain an outcome arriving once in nearly four million. Only the bill changes, and it changes fiftyfold across the reel’s own published range of stakes.

Thirty-five pounds and seven hours, then, to learn less than the maker’s own page says for nothing. That is the arithmetic of waiting for proof, and it is why this site spends its effort on published figures rather than on observation: the observation is expensive, slow, and worse than the document it would be checking.

There is a second kind of waiting, which is waiting for other people’s reports, and it costs nothing but is worth about as much. Part eight is about what those reports can honestly contain. The short version is that the figures in them are the same published figures, and the judgements in them are not quantities at all.

None of this is an argument for playing a new machine. The expectation on the credit above is positive only because the requirement is modest: at forty times rather than twenty, the same £30.00 closes below nothing on both reels in the table, and no amount of newness or age moves that. It is an argument about where a reader’s attention is worth spending, which is on the published figure and the requirement rather than on the age of either.

8

What a review can add on day one, and what it invents

A description written on the day a machine appears can usefully carry three things: the published figures, correctly attributed; the mechanics, which the maker describes in prose on its own page; and the arithmetic somebody has performed on those, which is what this card is. All three are available immediately and none of them improves with time.

What such a description cannot carry is a measured dispersion, and this is where most of them go wrong. No volatility is published anywhere in this family: the maker’s pages never use the word, nothing in the data a game receives makes room for one, and the maker’s announcements state none. The three reviewers who do state one for the original give three different answers, and one of the three contradicts itself inside a single page, calling the machine high in a table and something milder a few lines later in its own prose.

A figure invented on day one does not become sound by being repeated for six years, which is the quiet risk of a new machine: the first description of it sets the vocabulary everybody else borrows. The sheet behind this site records where sources disagree precisely so that a disagreement survives repetition, and the three-way split above is the clearest example it holds.

So a reader meeting a new reel should treat a description as a route to the published figures and not as a substitute for them. The useful question to ask of any account of a new machine is which of its numbers came from the maker and which came from the writer. On this card the answer is printed throughout: the returns, the ceilings, the grids, the rates and the trigger come off the sheet, and the credit, the requirement and the ten pence a turn are this card’s own suppositions.

9

The newest reel in this family, read off the sheet

For completeness, here is everything this publisher holds about the most recently published page of the thirty-six, which was fourteen days old when the sheet was built. Its published return is 96.50 per cent, a figure eleven titles of the family share. Its ceiling in a single round is 20,000 stakes, which at ten pence is £2,000.00 and at the £5.00 a licensed cycle permits would be £100,000. It runs five reels by three rows across ten paylines and sells a turn from 0.10 to 250.00, every figure of it taken off the family table the sheet’s own column for each.

That is the whole of it, and it is enough to price any offer attached to it. What this publisher does not hold: a release date, because none is published; a volatility, for the reason in part eight; a base-game hit rate, which no page in the family publishes; and any statement about which lobbies carry it, because the sheet names no operator and this card will not either.

Notice what is missing from that list of absences: nothing a reader needs in order to do the arithmetic. The return and the stake range between them fix what a credit becomes, and the ceiling fixes what one round could return. A machine fourteen days old is as weighable as one six years old, which is the finding this card exists to establish.

One caution on the count. The thirty-six are the titles with an English game page of the maker’s own, read on one day. The maker’s announcements name further Bass products without such pages, including one that is not a slot machine at all, so this card says the maker’s site lists thirty-six Bass pages rather than that the series has thirty-six titles.

10

What being new is worth, totalled on one credit

Totalled on the credit this card has carried throughout, being new is worth minus £1.26. The newest page in the family publishes 0.21 points less than the first one, the requirement is identical, the staking is the same £600.00 and the turn count the same 6,000, so the gap between them is entirely the figure the maker printed before either machine was ever played.

Being new buys something else, and the card has priced that too: a ceiling of 20,000 times the stake rather than 2,100, which at ten pence is £2,000.00 against £210.00 in a single round. That is a real difference in what the machine can do and it costs 0.21 points of return to have. Whether the trade is worth making is the reader’s business, and it is a decision they can make on the first day a machine exists, because both figures are published then.

What being new does not buy is any additional certainty later on. The published return is the same document on day one and in year six. The observed rate that would check it needs a hundred and twelve days of unbroken play to see the machine’s largest outcome once. And the day the machine arrived is not published at all: for the family’s first title the maker’s own two sources stand twelve days apart, which is 414,720 turns at the pace ceiling a licensed build carries.

The practical instruction, then, is the same for a reel published this month as for one published six years ago. Open the page the maker keeps for that exact title, read the return out of its basic game information block, find the ceiling and the stake range in the game’s own data, and weigh the offer against those. Nothing in that sequence requires the machine to have a history, which is the one genuinely good piece of news on this card.