What a claim of speed can possibly be a claim about, priced hour by hour

1

The only length in the chain anybody can actually compress

Eighteen hours is the figure set here for the leg where money leaves one institution and reaches another, and it is doing less work than a reader would expect. It is the piece everybody pictures when a payout is described as quick, and it is barely a fifth of the total modelled here. Compressing it to nothing would still leave 74 hours and 51 minutes standing.

Forty-eight hours is the piece that decides the answer, and it is the length nobody prints a figure for. A request sits in front of somebody, or in front of a rule engine, and is released or is not. That interval has no network behind it, no settlement cycle, no clearing house and no timetable. It is an internal queue, and an internal queue is the only thing a desk can genuinely shorten.

So a claim about speed is a claim about the queue, which is the proposition this whole note rests on. Anything else being offered as the reason for a quick payout is either a leg too short to carry the claim or a part of the week that belongs to the banking calendar. Working out which of the three is being described is most of the work of reading the claim.

Nought published figures exist for the hold, and that absence is worth stating before the arithmetic rather than after it. The sheet of regulatory fact this site works to carries no distribution of payout times, bars quoting a payout speed against any named operator, and records as a gap what no primary source publishes. Nothing found elsewhere could be cited in its place, so forty-eight hours is this page’s own figure, chosen as a round central case and defended nowhere.

Two sorts of reader arrive at a page about fast withdrawal casinos and the arithmetic below serves them differently. Somebody comparing one promise against another needs to know which part of the chain the promise touches, and part three answers that. Somebody who simply wants their own money sooner needs part four, where the largest saving on the page turns out to be free and to have nothing to do with the choice of anything.

2

Ninety-two hours fifty-one, assembled from four assumptions

Four lengths added together make every figure here, and this part declares each of them once so that nothing later has to restate it. Forty-eight hours of hold before release. Twenty hours of identity work where that was deferred to the exit. Eighteen hours for the money to travel. Six hours fifty-one minutes of shut calendar, which part five derives rather than assumes.

Ninety-two hours and fifty-one minutes is the sum, which is three days, twenty hours and fifty-one minutes. Against a cashier describing a payout in a day or two, the modelled figure is between 1.9 and 3.9 times as long, and the difference is not deception. It is a different question being answered, because a cashier is usually timing from release while a reader is timing from the request.

Twenty-six hours is the right-hand column of that drawing, being a hold halved, identity settled months earlier, a two-hour movement into a wallet and no banking calendar to wait on. That is 72.0 per cent off the modelled total, and just one of those reductions belongs to the reader. The rest are decisions somebody else makes and does not publish.

Four parties own the four lengths and that is the useful way to hold them apart. The desk owns the hold. The desk also owns when the identity work happens, although the reader can volunteer it early. The payment arrangement owns the movement. The banking week owns the rest and is not negotiating with anyone.

92 h 51 min modelled, of which 18 h is money in transit

Nineteen point four per cent is the movement’s share of the whole, which is the figure to carry into part three. A reader choosing between payment arrangements on the strength of a payout promise is working on under a fifth of their own wait, and the three quarters of it sitting in front of that fifth are decided before any payment arrangement is consulted.

3

How much each piece gives back when it is cut

Twenty-four hours is the largest single saving available anywhere on this page, and it comes from halving the hold. Eighteen hours is what abolishing the movement entirely would give back. The inequality between those two is the whole case, and it survives any reasonable argument about the inputs, because a hold has to fall below 36 hours before halving it stops beating a movement of eighteen.

Twenty-one point five per cent against 19.4 per cent is the comparison in the two middle rows, and it is the sharpest line in the table. Moving the identity work to the day the account opens gives back more than making the payment instantaneous would. One of those two is free and available tonight. The other is physically unavailable to anybody.

Forty-three hours is what the three reader-owned rows come to added together, being twenty for the identity work, sixteen for choosing a two-hour landing place and six fifty-one for timing the request. That is 46.3 per cent of the modelled wait, in the hands of somebody who was told the whole thing was a property of the cashier.

Nought of the reader’s forty-three hours requires anybody’s permission, and that is the distinction between the two halves of the table. The desk’s rows are larger and they are promises. A promise about a hold cannot be checked against anything, because no figure for a hold is published anywhere a reader could look.

One caution belongs under the table before part four leans on it. The shares are shares of the modelled total, so they move when the inputs move. At a hold of 24 hours rather than 48 the whole wait is 68 h 51 min, and the identity saving rises from 21.5 per cent to 29.0 per cent of it. Cutting one part makes every other part a bigger fraction of what is left, which is why the ranking is given in hours first and in percentages second.

4

The twenty hours available to anybody who asks early

Twenty hours is the figure this page carries for identity work left until the exit, and it splits in two. Some of it is the reader finding a document, photographing it and sending it. The rest is somebody at the other end opening the file and reading it. Neither half is long. The length comes from the two halves happening in sequence, across whatever gap sits between a reader’s evening and a desk’s working pattern.

Nought hours is the same work on a reader who did it when the account opened. The documents were read then, the file is complete, and the request passes through that stage without stopping. Nothing has been skipped, nothing has been hurried and nobody has been asked for a favour. The work simply happened at a moment when nothing was queued behind it.

Two things make this the best value on the page. It is the second largest single saving in the table in part three, and alone among the four largest it can be taken by a reader unilaterally. A desk that halves its hold has done more, but a desk that halves its hold has also made a decision the reader cannot cause, verify or rely on.

20 assumed hours, saved for nothing, by nobody’s permission

Twenty-nine point zero per cent is what the same twenty hours is worth against a 68 h 51 min chain, which is where the model lands once the hold is halved. The saving grows in relative terms exactly as the rest of the chain improves, so a reader dealing with a genuinely quick desk gets more out of this than a reader dealing with a slow one, not less.

One structural point belongs here and it is the reason a British licence is organised the way it is. Identity work placed before play cannot become an obstacle to a payout, because by the time there is a payout to obstruct it is finished. Placed at the exit, the same work is standing between a reader and money they already consider theirs, which is where every argument on this subject starts.

Nothing in this part says a check is avoidable or should be. It says the check is cheap when it is early and expensive when it is late, and the 20 assumed hours are the price difference. A reader who registers, plays, wins and only then opens a camera has bought the expensive version of a thing that was free at the start.

5

Shut hours averaged over a week, and two days to avoid

Forty-eight hours of every 168 are treated by this note as hours in which nothing travels between banks, and that single rule is all the calendar the model needs. It is an assumption of this page. Two sevenths of the week is shut on it, which is 28.6 per cent, and the remaining 120 hours are open.

Six hours and fifty-one minutes is what that rule inserts on average, and the derivation takes two lines. A release landing anywhere in the shut 48 hours waits somewhere between nothing and 48 hours for the week to restart, averaging 24. That happens to 48 releases in every 168, so the average across all of them is 48 multiplied by 24 and divided by 168, which is 6.857 hours.

Nought is the figure for 71.4 per cent of releases, and that is why an average is a poor way to hold this. Most requests meet no shut calendar at all. Just over a quarter of them meet a great deal of it, and the mean sits between two outcomes that almost nobody actually experiences.

Thirty-one hours and twelve minutes is the row worth arguing with, and it is arrived at by asking which delay only one release in ten exceeds. A delay longer than x hours happens to 48 minus x releases out of every 168, so a tenth of releases means 48 minus x equals 16.8, which puts x at 31.2 hours. A reader planning around the average is planning around a figure that a tenth of requests beat by a day.

Sixty-eight hours is the lead this model puts between a request and its release, being the hold and the identity work together, and that lead makes the calendar predictable rather than random. Shifting the request shifts the release by the same amount, so a reader who knows the lead knows which request hours land badly.

Wednesday at four in the morning to Friday at four in the morning is the band to stay out of on these assumptions, because a request made inside it releases into the shut part of the week. A request at four on a Friday morning releases exactly as the week restarts, which on this model is the best-timed request there is. The band is 48 hours wide, which is the same 48 hours seen from the other end of the chain.

6

A window that shortens one figure and lengthens the other

One mechanism makes a slow payout look quick without touching a single hour of the chain, and it deserves a name. A request is made, and for some period afterwards it can be taken back, with the money returning to the balance it came from. Whatever happens during that period, the chain this page has been modelling has not started.

Twenty-four hours is the window this note assumes for the arithmetic in part seven, and the assumption is the page’s own. It is chosen because a day is the obvious length for such a thing and because it makes the sums legible, not because any desk was observed using it.

One hundred and sixteen hours fifty-one is the real wait with that window in front of the chain, being 92 h 51 min plus 24. The advertised figure does not move, because a desk counting from the moment the request becomes irrevocable is counting from the end of the window rather than from the tap that created it. One number falls and the other rises, and they are measuring different things.

Two readings of the same window are therefore both honest, which is the uncomfortable part of this section. A desk can say with a straight face that payouts are processed within a given time, meaning from the point the request is final. A reader timing from the tap gets a different answer, and neither party is lying about anything.

Nought windows of this kind exist under a British licence, and the rule is unusually flat. A licensed system must not hand the customer an option to call a withdrawal request off, the guidance adds that the sum behind such a request is not to be put back within reach as a deposit, and it goes further still, telling an operator to take friction out of withdrawing rather than build it in. That is a rule about design and not about paperwork.

Three things make the window more than a presentational trick, and part seven prices the third. It delays the chain by its own length. It keeps the balance visible and spendable during the delay. And a balance that is visible and spendable during a delay is a balance that can be staked, which is what the window is worth to the party offering it.

7

What an open window is worth to the party offering it

Thirty pounds an hour is the ceiling on what a returned balance can be staked at, being 300 presses in an hour at a stake of ten pence, on a pace this site assumes and has never measured. One pound twenty an hour is the expectation riding on that, the margin behind it being four pence of every pound at the 96p return held fixed here.

Twenty-eight pounds eighty is therefore the ceiling on what an assumed 24-hour window is worth, being £1.20 multiplied by 24. Seven thousand two hundred presses is the same £720.00 read as a count of presses, and 65.45 goes with a handset at the 110 presses this site assumes a sitting holds, which rounds up to 66.

Forty-four pence is the other end of the honest range and it is a very different number. A reader who takes the money back, plays one sitting of 110 presses and stops has staked £11.00 and expects to lose 44p of it. Nobody presses for a solid day, so the ceiling is a ceiling, and the truthful statement is that the exposure lies between 44 pence and £28.80 depending entirely on what the reader does with an evening.

Two pounds forty an hour is the same line at a twenty pence press, which doubles every figure in the money columns and leaves the hours alone. That is the second line on the plate. Stake size is the one input a reader changes without changing anything else, and it moves the exposure proportionally while moving the wait not at all.

Fifty-seven pounds sixty is what an assumed 48-hour window reaches at ten pence, or what an assumed 24-hour window reaches at twenty pence, and the coincidence is worth naming because it shows what the window is actually made of. Hours and stake multiply. A short window at a large stake and a long window at a small one are the same object.

Nought of this is a charge, a fee or a deduction, and the distinction matters. Nobody takes the £28.80. It is what the arithmetic expects to be transferred if the balance goes back over the reels, which it can only do because the request was left revocable, which is the thing the window exists to permit.

8

Every hour here is modelled, and this is the model

Nothing on this page was timed, observed, obtained from a desk or read off a receipt, and the whole of it is arithmetic over declared inputs. That sentence belongs in the middle of the note rather than in a footnote, because a page about duration reads as though somebody held a stopwatch, and nobody did.

Four inputs and one rule make the model, and they are set out here together so that any figure above can be rebuilt or disputed. The hold is 48 hours. Identity work at the exit is 20 hours. The moving leg is 18 hours to a bank account and 2 to a wallet. The rule is that 48 hours of each 168 carry no movement between banks.

One equation assembles them. The total is the hold plus the identity work plus the moving leg plus whatever the shut rule inserts at the moment of release, and that last term is the only one not fixed in advance, because it depends on when in the week the release falls. Averaged across a uniform week it comes to 6.857 hours, which is where 92.857 comes from.

Uniformity across the week is the model’s weakest assumption and it should be named as such. Requests are probably not spread evenly: a Saturday evening after a win is a more likely moment than a Tuesday at dawn. Weighting requests towards weekends would raise the 6 h 51 min, so the figure printed here is, if anything, the generous one.

Nought allowance is made for anything the model does not contain, and the omissions are worth listing honestly. A request that fails a check and has to be repeated is not here. A desk that answers in minutes on a Monday and in days in the week before Christmas is not here. A limit that splits one payout into several is not here, and it would multiply the hold by the number of pieces.

Two uses survive all of that, which is why the model is printed rather than abandoned. It ranks the parts against each other, and the ranking is robust to large changes in the inputs. And it converts a vague adjective into a quantity a reader can check against their own experience, which is the only way anybody will ever find out whether a claim of speed was true.

9

A chain with nobody standing at the end of it

A speed claim at this end of the market issues from a desk outside the British licence, and what bites on a page about waiting is not a rule about the wait. It is the absence of anybody to take the argument to when the modelled hours pass and nothing has arrived.

One sentence from the British regulator settles what it will not do: it is not an ombudsman. It publishes that settling a gambling complaint and recovering anybody’s money are outside its powers, that an argument over a payout, or over a game behaving oddly, falls outside what it resolves, and that it cannot see inside a customer account to find out what occurred. Its instruction to a player is to start with the business itself.

Nought comfort is in that for a reader inside the register, and less for a reader outside it, which is why this page sets it out rather than implying something warmer. Even where a licence exists, the body issuing it is not the body that gets the money back. What a licence supplies is a complaints procedure with a shape and an approved route beyond it, and what it does not supply is a person who will intervene in one payout.

Ninety-two hours and fifty-one minutes is the modelled wait, and the question that matters is what the ninety-third hour is worth. Inside the register it is worth the start of a timetable. Outside it the ninety-third hour is worth exactly what the hundredth is worth, because nothing attaches to either, and a reader with no leverage is a reader whose only remedy is to have chosen differently earlier.

Two duties, then, and they work in opposite directions on the same chain. One shortens the chain by forbidding a delay dressed as a convenience. The other does nothing to the chain at all and instead supplies a timetable for arguing about it. Neither is a promise of speed, and a reader looking for the licence to make the payout quick has misread what a licence is for.

Nothing in this part should be read as saying the offshore side never pays. It says the hours are unenforceable and the figures are unpublished, which is a different and smaller claim, and the right response to it is the one in part ten: ask the questions that have checkable answers and ignore the adjective.

10

Five questions worth more than the word itself

First, from what moment is the quoted figure counted. A payout described in hours from release and a payout described in hours from the tap are different measurements, and on this model the gap between them is 68 hours before the movement even begins.

Second, can a request be taken back after it is made. An assumed 24-hour window adds 24 hours to the real wait, subtracts nothing from the advertised one, and puts up to £720.00 of staking within reach at ten pence a press. A system built to a British licence may not offer the option at all.

Third, when does the identity work happen. Twenty assumed hours sit on that answer, which is more than making the payment instantaneous would give back, and it is the one saving a reader takes without asking anybody.

Fourth, what day of the week is it. Six hours fifty-one minutes is the average cost of the shut calendar, nothing at all for 71.4 per cent of releases, and 31 h 12 min or worse for one in ten. On a 68-hour lead the band to avoid runs from Wednesday at four in the morning to Friday at four.

Fifth, and last, is any figure for the hold published anywhere. The answer is no, on this site’s own sheet of regulatory fact and on everything that could be cited beside it, and that single absence is why a claim of speed cannot be checked before the event. It can only be checked afterwards, once, by the reader it was made to.

Ninety-two hours fifty-one against twenty-six hours is the span this note has modelled between the ordinary case and every saving taken at once, and the honest summary is that a reader owns 43 of the hours between them. The word on the banner owns none of them. A page that told a reader to look for the word, rather than to settle identity early, choose where the money lands and pick the hour of the request, would have cost them the only part of this they control.