Free slots no deposit: one of those two words is doing no work at all

1

Two different free things, and the one that is free in a different sense

The phrase at the top of this card covers two arrangements that have almost nothing in common except the word. The first is a demonstration build: the reel runs, a balance moves up and down, and no money enters or leaves at any point. The second is a credit of real money handed over without a deposit, carrying a condition that has to be worked through before a penny of it can be withdrawn.

The difference between them is not one of degree. On the demonstration nothing exists for a published return to act upon, so the expected close is £0.00 and stays £0.00 however long anybody sits there and whatever multiple a notional condition might ask. On the credit the published return acts on every pound of staking the condition requires, and the result is a figure this card can print: £0.13 out of £10.00 at thirty times, with the published return standing at 96.71.

So the useful finding is a division rather than a number. One of these two things is evidence, and a rather good piece of it, because everything this publisher knows about the machine was read off a demonstration server. The other is an offer, and it is weighed the same way every offer on this domain is weighed. Treating the second as though the first proved something about it is the mistake this card exists to head off.

2

Where every machine figure on this domain came from, and how to go and get it again

This part is method, and it is here because a reader is entitled to check anything on this site rather than take it on trust. Every figure on this domain that cites the game’s own data as its source came out of a single request, and the request can be repeated by anybody.

The maker’s page for a title embeds its free-play demonstration in a frame whose address names the game’s internal symbol, which for the original of this family is a symbol beginning vs10bbbonanza. Opening that address redirects to the game client and issues a session key. Posting an initialisation request with that key and the symbol to the maker’s own game service returns the configuration its server sends the client for that session, and that response is the source behind every machine figure this publisher prints.

What arrives is not a summary. The response carries the return to player, the count of lines, the count of rows, the reel strips themselves, the paytable, the mapping from scatters to free turns, the selectable stake steps and a maximum win for one round. The information screen a player can open inside the game is drawn from the same data. That is why this publisher ranks it above any review: a review reports what somebody understood, and this is the arithmetic the game was handed.

Two honest limits travel with the method and both are recorded. The session used to read the family was opened in pounds, while the original’s own first reading was taken in dollars, and the stake steps came back identical either way. And those steps belong to the demonstration service. They are the game’s own numbers and they are not a statement that any cashier anywhere offers that range, which is a distinction part eight returns to.

3

A build that stakes nothing has an expected close of exactly nothing

The house formula on this site takes a credit, applies the staking a condition demands, and charges the reel’s own share of it: E = B × (1 − W × (1 − RTP) ∕ g). B on a demonstration build is £0.00, because nothing was ever handed over. Multiply a bracket of any size by nought and nought is what comes back, so a demonstration is expected to close at £0.00 at ten times, at thirty times, at six hundred times, and at no times at all.

Hold that for a moment, because it is the one place on this domain where the published return can do nothing whatever. Everywhere else 96.71 per cent is the quantity that decides the outcome. Here it decides nothing, and it decides nothing for the plainest possible reason: a percentage of nothing is nothing, whichever percentage it is.

£0.00 expected from a demonstration build at every multiple there is, because 96.71 per cent of nothing staked is nothing lost and nothing won

One consequence follows, and it is the one most often missed. A demonstration cannot pay, so a large balance on one is a reading and not a return. If the counter climbs to two thousand times what the demonstration called a stake, what has been observed is the paytable behaving as the paytable says it will. Nothing has been won, because nothing was at risk, and the same sequence on a paid build would have cost whatever the stakes behind it cost.

The symmetry holds the other way too, and this is the part that makes a demonstration genuinely useful rather than merely harmless. A run of nothing on a free build has cost nothing as well. A reader curious about the look of the feature round, about how often the money symbols fall, or about what the grid does with four wilds can settle all of it for £0.00, and the only figure that changes if they go on to stake money is the one in the third column of the board above.

4

The other free thing, priced on two titles

Now the credit, which is the half of the phrase that costs something. Take £10.00 with thirty times of staking asked of it. That is £300.00 put across the reel. What the reel takes is a hundred less the figure it publishes, so 3.29 points of £300.00 is £9.87. The credit closes at £0.13.

Thirteen pence is a strange sort of answer and the strangeness is the finding. It is neither a loss nor a gift, but one and three tenths of a per cent of the credit, which is what is left of a ten pound note after a condition has taken the rest of it. B is this card’s £10.00 and W its thirty times, and with g at one, which is the weighting wherever a slot counts in full, the formula lands four tenths of a multiple short of handing back nothing whatever.

The lower row is the point of the table. The same credit, the same condition, a different title of the same family, and the answer has crossed from thirteen pence to one pound and seventy-six pence owed. Its break-even sits at 25.51 times, so thirty times is already five multiples past the point where the credit stops surviving its own conditions. The title decided £1.89 on a £10.00 credit, which is nineteen per cent of the face value, on 0.63 points of published return.

The lower row is not merely a smaller number either, and the family table says so. That title runs five reels by four rows where the original runs three, it pays on twelve lines where the original pays on ten, its least stake is two pence higher at 0.12, and the biggest round it can pay is 4,000 times the stake against the original’s 2,100. None of that enters the arithmetic above, because a credit priced in money is indifferent to how many rows a grid has. It is here because a reader comparing two rows of one table deserves to know that the two machines differ in four published ways besides the one the table prices.

Both rows are supposition on the offer’s side and published on the machine’s side. The two returns are the maker’s own, taken from the family table behind this site; the credit and the multiple are mine. Nothing here says that any cashier asks thirty times, and a British licence has not permitted that figure since January 2026, when a ceiling of ten times came into force. Thirty is used because it is the shape of thing this domain weighs, and because it lands near the edge of what this reel can carry.

5

The pence that give the published figure straight back to you

There is a small piece of arithmetic in this card that is worth isolating, because it turns the published return into something a reader can hold.

Take the same £10.00 credit and ask ten times of it instead, which is the ceiling a British licence now sets. That is £100.00 of staking and 3.29 points of £100.00 is £3.29, so the credit closes at £6.71. On the other title in the table it closes at £6.08. Set those two beside each title’s published figure, 96.71 per cent and 96.08, and the digits are the same digits.

It is a coincidence of scale rather than a law of anything: a credit of ten pounds at ten times gives a hundred pounds of staking, and a hundred pounds is the base the percentage is quoted on. It is still worth keeping, because it makes the house edge concrete. The 3.29 points nobody can picture turns out to be the £3.29 that leaves a £10.00 credit when a licensed ceiling is applied to it, and the whole of the published return turns out to be the pence that survive.

It also shows how little room there is. Those same 3.29 points, asked three times over at thirty, take £9.87 of the £10.00. A licensed ceiling leaves two thirds of the credit standing. Three times that ceiling leaves one and three tenths of a per cent of it.

6

Whether the demonstration and the paid build are one machine

This is the question the ground for this card puts at its centre, and the honest answer has three parts: what was read, what the reading establishes, and what it does not.

What was read is on the record above. The return to player is not baked into the assets a browser downloads: it arrives as a field in the initialisation response, which the maker’s own server sends as each session begins. For the title weighed here that field came back as 96.71 on every reading, and it came back as 96.71 when sessions were opened against the same server with seven different values of the jurisdiction parameter the launch address accepts. Six of the seven resolved to a United Kingdom identifier inside the game client and the seventh resolved to its own number, and all seven returned the same figure.

What that establishes is narrow and precise. On the maker’s demonstration service, for this title, one figure came back every time it was asked for, including when the question was dressed up as coming from somewhere else. It is also the figure printed on the maker’s public page for the title, so two independent documents agree.

What it does not establish is anything whatever about a build somebody has paid to play. A field that arrives afresh each time sits equally well with a single value that never moves and with one picked for each deployment, and how a figure travels is no evidence about who picked it. The reading reaches that far and this card goes no further. The wider question of what this whole section can and cannot establish about the figures a family of games ships with belongs to the section’s own head page, which is where the argument is made and where the sources for it are weighed.

7

Who would have to tell you, which is where the finding is

If the two builds were ever to differ, the question that matters to a reader is not whether somebody somewhere knows. It is who would be obliged to say. This publisher went looking for that duty in the regulatory sheet behind this network, and what came back is the finding this card was written for.

There is no such duty. The British standard covering a game’s rules and the chance of winning sets out four things, and it joins them with the word or, which means a licensee has discharged it by publishing a house edge, or by giving the chance of a winning event, with no percentage anywhere. For certain other kinds of game the guidance goes further still and accepts an account of the game’s workings and of how a winner is settled. That disjunction is misread across this trade, so it is worth stating flatly: no licensed site is required to publish a return to player figure at all.

If nothing requires the figure, nothing can require a statement about which version of the figure is running. There is no provision anywhere on the sheet behind this network about disclosing a configuration, a version, or a choice among them, and where no source exists this publisher adds none. The absence is the finding, and it is the answer to the question in the heading: nobody would have to tell you.

There is one thing the demonstration does about all that, and it is the best argument for reading a free build rather than a review. Because the response can be read, the maker’s published page can be checked against the maker’s own software, title by title. Across the thirty-five slots read, the two agreed exactly on thirty-three. On two of them they did not: one Megaways title is published at 96.70 and reports 96.72, and one Halloween title is published at 96.04 and reports 96.06. On the £300.00 of staking this card supposes, two hundredths of a point is £0.06, which is what a conflict between a maker’s own two documents is worth in money, and nobody would know of either conflict without a free build to read.

8

Four things a demonstration cannot settle, listed so nobody leans on it for them

A free build is the best source this publisher has about the machine. It is a poor source about almost everything else, and the four limits below are worth having explicitly, because each one is a place where a reader could reasonably assume it had told them something it has not.

The first is the stake. The range this publisher read runs from 0.10 to 250.00 on the original of the family and from 0.12 to 240.00 on the other title in the table above, whose twelve lines put its least stake two pence higher. Those are the demonstration service’s steps. What a reader can actually stake in a particular lobby is a fact about that lobby, and a licensed British one could not offer the top of either range at all, because a cycle there is capped at £5.00.

The second is the condition. A demonstration has no terms, so it says nothing about what multiple a cashier asks, what a game is weighted at while the multiple is being worked off, what the maximum withdrawal is, or how long any of it is open. Those belong to whatever document the offer is published in, and this domain’s whole method is to read them from there and weigh them against the machine.

The third is presence. A title running in a demonstration frame on the maker’s page is no evidence at all that any particular lobby carries it, and the maker’s own licence on the public register, which authorises it to supply software, settles nothing about who has licensed that software or on what terms.

The fourth is the build. This is part six’s answer, restated as a limit rather than as a question: the reading was taken on the demonstration service, so it is a fact about the demonstration service. Where a paid build would be read, nobody publishes a way of reading it, and nobody is obliged to describe it.

9

What the free build is worth, measured in the one way that counts

So what is the demonstration worth? Its expected close is £0.00 by construction, which makes the usual measure useless. The measure that works is what it tells a reader that nothing else would, and on that measure it is worth a great deal.

It is the only public route to the inside of the machine. The grid of five reels by three rows, the ten paylines, the whole paytable including its top four paying rows, the strips the symbols sit on, the mapping that turns three, four or five scatters into ten, fifteen or twenty turns, and the biggest round the game will pay at 2,100 times one stake are all in that one response. Not one of them is on a review, and every one of them is checkable by anybody who repeats the request in part two.

Set against that, consider what the paid side of the same machine publishes about itself: a percentage on the maker’s page, which the demonstration confirms, and nothing else that this publisher could find. So the free build is not a lesser version of the paid one as a source. On everything except the question of which build is running, it is the better source by a wide margin, and it is free in the sense that matters to a reader who wants to check something rather than win something.

That is the sense in which the phrase at the top of this card is worth searching for. A free slot with no deposit is a way of establishing what a machine does, at a price of nothing and with no condition tied to it. A credit with no deposit is an offer, it has a condition attached, and the condition is where all the arithmetic lives.

10

Both free things, totalled

The demonstration build: nothing staked, nothing paid, expected close £0.00 at every multiple that could be asked of it, and the source of the published grid, paylines, paytable, scatter mapping, stake steps and 2,100 times ceiling that every page on this domain spends. Its limits are the stake range belonging to the service rather than to any cashier, the absence of any terms to read, the absence of any evidence about which lobbies carry a title, and the absence of any way to read a paid build.

The credit: £10.00 at thirty times is £300.00 of staking, 3.29 points of which is £9.87, leaving £0.13 expected where the published return is 96.71 and £1.76 owed where it is 96.08. Break-even stands at 30.40 times on the one and 25.51 on the other, so one title is just inside the edge at thirty and the other is five multiples beyond it. At the ten times a British licence allows, the same credit leaves £6.71 and £6.08. The title decided £1.89 of a £10.00 credit, and the condition decided £6.58.

The finding, for anybody who came here for one sentence: the free build is the better source, the credit is the half of the phrase with a price on it, and a reader who treats the first as a reason to accept the second has let a reading do the work of an offer document. Everything the demonstration proves is about the machine. Nothing it proves is about the condition, and the condition is where the money goes.