Payout Timelines Quote the Rake, Not the Withdrawal Clock
Ask a casino support agent about payout speed and you'll get a number: 24 hours, 48 hours, "instant." Ask where that number comes from and the answer is almost always the processor, the bank, the blockchain — anything downstream of the casino's own desk. The clock operators advertise measures the last mile. The rake is collected in the first one, and it's quoted far less often because it's the part that costs you money before your withdrawal request ever exists.
That distinction matters more in the United States than most players realize, because the legal online market here is young, state-by-state, and structurally different from the offshore era. New Jersey launched regulated internet casino and poker play in November 2013. Pennsylvania followed in July 2019, Michigan in January 2021, Connecticut in October 2021, and by the end of 2024 six states ran legal iGaming: New Jersey, Pennsylvania, Michigan, West Virginia, Connecticut, and Delaware. Every one of them taxes gross gaming revenue at a different rate — Pennsylvania's 54% on slots is the highest in the country, New Jersey's 15% is the lowest — and every one of them forces operators to publish return-to-player figures on slot machines in some form. None of them force operators to publish the effective cost of a bonus, a cashout fee schedule, or the house's edge on a "free" bet in a way that's comparable across sites.
That's the gap. Payout timelines are a marketing surface. The rake is a cost structure. Players compare the first and ignore the second, and the operators know it.
What "payout timeline" actually measures
When a site says "withdrawals processed in under 24 hours," it is describing one step in a chain of four or five:
- You request the withdrawal.
- The casino's fraud and compliance team reviews it.
- The casino releases funds to a payment processor.
- The processor moves money to your bank, card, or wallet.
- Your bank posts it to your available balance.
The advertised timeline covers step 2 and sometimes step 3. Steps 4 and 5 are where most of the calendar time lives, and the casino has no control over them. ACH transfers in the US settle in one to three business days under normal conditions; the Federal Reserve's FedNow service, live since July 2023, can move money in seconds, but casino withdrawals almost never route through it because the receiving banks aren't set up for gambling-adjacent merchant codes. Card withdrawals, where permitted, can take five to ten business days to appear as a credit. Wire transfers are fast on the sending end and slow on the receiving end, and many banks flag gambling wires for manual review.
So a "24-hour payout" is really a 24-hour processing commitment plus an unstated three-to-seven-day settlement window. That's not necessarily deceptive — the casino genuinely does release the money in 24 hours — but it's a measurement of the operator's internal queue, not of when you can spend the money.
The rake, by contrast, is deducted continuously and invisibly. On a slot machine, it's baked into the RTP. On a table game, it's the house edge. On poker, it's the per-hand or per-hour fee the room keeps before paying the pot. On sports betting, it's the vig. None of that shows up in a payout timeline because it happens before you have anything to withdraw.
The processing-time numbers that get quoted
The American Gaming Association's 2024 industry report put total US commercial gaming revenue at $71.9 billion for 2023, with iGaming representing roughly $6.2 billion of that. iGaming is a small slice of a large pie, but it's the slice with the most aggressive promotional spend, because customer acquisition in a newly legal state is a land grab. That promotional spend is where the rake gets obscured.
A typical welcome offer in a regulated US state looks like this: 100% match up to $1,000, plus 200 free spins, with a 10x wagering requirement on the bonus and a 1x requirement on the deposit. On its face, 10x is generous — many offshore sites used to demand 30x or 40x. But the requirement applies to the bonus amount only, and slot contributions are usually weighted at 100% while table games contribute 10% or less. If you take the full $1,000 and clear it on slots at an average 96% RTP, you're running $10,000 through the machine. The house keeps 4% of that, or $400. Your expected net after clearing the bonus is $600, before the deposit's own play. That's the rake, and it's not on the payout page.
The rake is quoted in RTP, not in dollars
Slot RTP is the cleanest rake disclosure in the US market because it's mandated. New Jersey requires slot return percentages to be published and reported; Pennsylvania's Gaming Control Board posts monthly slot machine revenue and payout percentages by casino and by machine denomination. In practice, regulated US online slots cluster between 94% and 97%, with Pennsylvania's online slots averaging slightly lower than New Jersey's because of the tax differential.
That 3% to 6% range is the rake. If you play $5,000 through a 96% RTP slot, you expect to lose $200. Play $50,000 over a month and the expected loss is $2,000. The variance around those numbers is enormous in the short run — a 96% RTP machine can pay 300% over 500 spins and 40% over the next 500 — but over 100,000 spins the realized return converges tightly on the published figure. That's the number that should be on the comparison table next to "payout speed," and it almost never is.
Table games are worse for the player in disclosure terms because the house edge isn't published on the felt. Blackjack with perfect basic strategy runs about 0.5% against the player, but most online blackjack is dealt from a continuous shuffler or an RNG with rules that push the edge to 0.6% or higher. American roulette is 5.26%. Caribbean stud is 5.22%. Three Card Poker's ante-play bet is 3.37% with optimal play, and the Pair Plus side bet is 7.28%. None of these numbers appear anywhere on the cashier screen.
Poker is the most transparent of the verticals, because the rake is charged per hand and players can calculate it. A typical US-regulated online poker room takes 5% of the pot up to $3, or a time charge of $0.50 to $1 per player per hand at higher stakes. If you play 300 hands of $1/$2 no-limit in a session and the average pot is $18, the room collects roughly $0.90 per hand, or $270 across the table — about $30 from you if you're in a third of the pots. That's the session's rake. It's invisible in the payout queue and decisive in your results.
Where the bonus math goes wrong
The rake and the bonus interact in a way that most players don't model. A bonus with a 10x requirement and a 100% slot weighting on a 96% RTP game has an expected cost of 40% of the bonus amount, as calculated above. A bonus with a 30x requirement on the same game has an expected cost of 120% of the bonus — more than the bonus is worth. Operators know this. That's why the headline number is always the match percentage and never the expected value.
There's a second layer. Many US-regulated bonuses carry a maximum cashout cap — often 10x the bonus, sometimes 5x — that converts the bonus from a cash-equivalent into a lottery ticket. A $1,000 bonus with a 10x cap pays at most $10,000 regardless of how much you win. If you hit a $50,000 jackpot while clearing, the excess evaporates. That cap is disclosed in the terms, usually in paragraph 14 of a document nobody reads, and it's the single most consequential number in the entire offer.
The withdrawal friction that isn't disclosed
Once the rake has been collected and you've cleared whatever requirement applies, the withdrawal process introduces its own costs. These are smaller in dollar terms than the rake but more visible, which is why they generate complaints.
Most US-regulated operators offer one free withdrawal per week or per month and charge $10 to $25 for each additional one. Some charge a flat fee on every withdrawal below a minimum threshold, typically $100 or $250. A few pass through the processor's fee directly. These charges are disclosed in the banking terms, but they're not on the payout timeline page, and they're not included in the "24-hour" claim.
The more significant friction is the reversal window. Many operators give you 24 to 72 hours after requesting a withdrawal to cancel it and return the funds to your balance. During that window the money is in limbo — not in your account, not available to spend. The reversal window exists because a meaningful fraction of players reverse their own withdrawals, and the operator keeps the money in play. It's a rake-extraction mechanism dressed as a convenience feature. New Jersey's Division of Gaming Enforcement has taken a harder line on this than most regulators, requiring clear disclosure of reversal terms, but the practice remains widespread across all six iGaming states.
Crypto withdrawals, where available, sidestep the banking system entirely and settle in minutes to an hour, but they introduce a different cost: the spread between the casino's quoted exchange rate and the market rate, plus the network fee, plus the tax treatment of any gain. For US players, crypto gambling is a tax-reporting minefield, and the IRS treats gambling winnings as ordinary income with no offset for losses beyond the amount won. That's a rake of a different kind, collected by the federal government rather than the operator.
What the regulators do and don't require
State gaming regulators require operators to disclose the rules of each game, the RTP of each slot, the terms of each bonus, and the fee schedule for deposits and withdrawals. They do not require operators to publish a single, comparable "cost per $100 wagered" figure that would let a player rank sites by total expected cost. That figure would be trivially easy to compute — it's just the house edge times the amount wagered, plus fees, minus bonus expected value — and its absence is a policy choice, not a technical limitation.
Pennsylvania's Gaming Control Board publishes the most granular slot data in the country, including monthly payout percentages by machine type. New Jersey's DGE publishes annual gross revenue and tax collections by operator. Michigan's Gaming Control Board publishes similar figures. None of them publish bonus-adjusted effective cost, because that would require a standardized methodology that the industry has resisted.
The result is that the most useful number for a player — the expected cost of playing at Site A versus Site B — is nowhere in the public record, while the least useful number — the advertised processing time — is on every homepage.
What a player can actually do with this
The practical response is to build your own comparison. Three numbers do most of the work.
First, the game's RTP or house edge. For slots, it's published. For table games, it's widely documented and stable. For poker, it's the posted rake structure. Multiply by your expected handle to get the session's expected cost.
Second, the bonus's expected value. Take the bonus amount, multiply by the house edge, multiply by the wagering requirement, and subtract from the bonus. If the result is negative, the bonus costs more than it's worth. If it's positive, check the cashout cap and the game-weighting table before you accept it.
Third, the withdrawal's total cost. Add the fee, the reversal window's opportunity cost (if you'd otherwise have the money working), and the settlement delay's effect on your cash flow. For most players this is a rounding error compared to the rake, but it's the number that gets quoted, so it's worth knowing what it actually represents.
The operators have no incentive to make this easy, and the regulators have no mandate to force it. The number that would change player behavior most — effective cost per dollar wagered, bonus-adjusted — is the one nobody publishes. Until that changes, the payout timeline will keep doing the work of a cost disclosure it was never designed to be, and players will keep optimizing for the wrong variable.
The open question is whether any state will require it. Michigan's 2021 law gave its gaming board broad rulemaking authority over "responsible gaming" disclosures, and Pennsylvania's board has shown appetite for granular data publication. A single state mandating a standardized effective-cost figure would force every multi-state operator to compute it, and once computed, it would leak. The first operator to publish it voluntarily would have a genuine differentiator — assuming the number is good. The ones with the worst numbers have the most to lose from a level comparison, and they're the ones currently spending the most on "fast payout" advertising.
That's not a coincidence. It's the rake, working as designed.