The Comp Rate Is Paid On Your Loss, Not Your Spend

There is a worked example in front of us. Not a casino's. Ours — built from a pocket calculator and the published RTP bands that game studios and certification bodies put on the public record. It is dated to the 2024 certification cycle. What it shows is this: two players can wager wildly different amounts, deposit on different days, sit at different tables, and still be worth almost exactly the same to the operator's loyalty desk. The headline spend number tells you nothing. The number the comp engine actually reads is buried one calculation deeper.

The honest answer to "how does the comp rate calculate by spend bracket" is: it depends, and the thing it depends on is not your spend. So we will not give you a single multiplier. We will walk through three hypothetical players — composite illustrations, no real accounts, no interviews — and run the math on each. By the third one the mechanism is obvious, and you will understand why the bracket you think you are in is rarely the bracket the operator has you in.

Scenario 1: The Weekend Book of Dead Spinner

Let us say you are a recreational slots player. Picture someone who logs in on Friday night, loads an Egyptian-themed title — Book of Dead, Legacy of Dead, one of the Play'n GO Books — and grinds £2 spins until the bottle of wine is gone. Across a month, imagine the total amount cycled through the reels — the handle, not the deposits — comes to £8,000. That feels like a lot. It is not what the operator counts.

Here is the math the loyalty desk runs instead. Play'n GO's slot library, on the studio's published game pages, sits in an RTP band of 94.20% to 96.50%. The house edge is the inverse: between 3.50% and 5.80% depending on which exact title and which licensed configuration the operator deployed. Theoretical loss — the figure the comp rate is actually a percentage *of* — is handle multiplied by house edge.

At the player-friendly end of the band, 96.50% RTP, the math is £8,000 × 3.50% = £280. At the operator-friendly end, 94.20% RTP, it is £8,000 × 5.80% = £464. Same player. Same £8,000 cycled. The operator's expected take swings by £184 purely on which RTP variant the game was licensed at — and the player has no way to see which one they were served.

That variance is the whole game. The same base title can be contracted at different RTP tiers, and the comp engine reads the resulting theoretical loss, not your spend. Whatever percentage the operator chooses to return as tier credits or cashback, it is applied to the £280–£464 number — never to the £8,000 you watched scroll across the screen. Concede the point the marketing makes: yes, high-volume slots play does generate real comp value, because slots carry the fattest house edge on the floor. That concession is true. Everything built on top of it is where the story turns.

Scenario 2: The Live Dealer Blackjack Grinder

Now picture the opposite temperament. Imagine a methodical player who only sits at Evolution live blackjack, plays a near-optimal basic strategy, and treats the session like work. Over the same month, let us say this player pushes £50,000 of handle across the felt. More than six times the slots player's volume. On any spend-bracket logic, this is your whale.

Run it. Evolution's published live-dealer figure for blackjack is an RTP of 99.28%, listed on the provider's own game documentation. House edge: 0.72%. Theoretical loss on £50,000 of handle is £50,000 × 0.72% = £360.

Read that against Scenario 1 again. The blackjack grinder wagered £50,000 and is worth £360 in expected loss. The slots player wagered £8,000 and is worth £280 to £464. The grinder cycled more than six times the money and generated *less* comp-eligible value than the casual spinner. This is the receipt. This is the document we said we had in front of us.

The incentive structure follows directly, and it explains everything operators do without ever announcing it. If comps were paid on spend, the blackjack player would strip the loyalty program bare and the casino would lose money on every reward. So they are not paid on spend. They are paid on theoretical loss, which is why skilled, low-edge games almost universally earn a reduced comp accrual rate — or get excluded from wagering and loyalty counts altogether. The math is not hostile to the grinder out of spite. It is hostile because the grinder's house edge is 0.72% and the slots player's is up to 5.80%, an eight-fold difference in what the operator expects to keep. The desk is not rewarding loyalty. It is reinvesting a slice of a number it already expects to win.

Scenario 3: The European Roulette Tourist

The third profile sits between the other two. Imagine an occasional player who likes the theatre of the wheel — European single-zero roulette, even-money bets, a steady £25 a spin. Not a grinder, not a binge slots session. Picture £15,000 of handle over the month.

European roulette's published RTP is 97.30%, again from Evolution's game data, giving a house edge of 2.70%. Theoretical loss: £15,000 × 2.70% = £405.

So here is the three-way comparison the spend bracket can never show you. The roulette tourist (£15,000 handle) is worth £405. The slots player (£8,000 handle) is worth £280–£464. The blackjack grinder (£50,000 handle) is worth £360. Rank them by spend and the order is blackjack, roulette, slots. Rank them by what the comp engine reads — theoretical loss — and the order collapses into a near tie, with the lowest-spend player frequently on top.

This is not a small distortion. It is the entire reason "spend bracket" is the wrong frame. The brackets that govern your comp rate are theoretical-loss brackets wearing a spend-shaped mask. An operator who runs a £94bn-GGR global market — that 2024 industry figure is on H2 Gambling Capital's published record — has every dollar of that margin modelled by exactly this calculation, scaled across millions of accounts. The loyalty program is not a gift budget. It is a line item that returns a fraction of an already-forecast hold.

Strip the three scenarios down and the same skeleton is underneath each. Comp rate = (operator-chosen reinvestment percentage) × (handle × house edge). The reader controls handle. The game controls house edge through its certified RTP. The operator controls the reinvestment percentage and — quietly — which RTP variant of a game it licensed in the first place.

That last lever is the one almost nobody watches. The certification bodies test that a game's RNG and RTP match the configured paytable — the Gaming Laboratories International certificate scope covers exactly that statistical validation. What the certificate does *not* do is force every operator to deploy the same variant. The same Egyptian-themed title can run at 96.50% at one site and 94.20% at another, both fully certified, both honest, and your theoretical loss — therefore your comp accrual base — differs by over 60% between them on identical play.

All three players also share a blind spot: they watch the wrong number. Deposits, withdrawals, total spend, session length — none of these is what the engine reads. It reads expected loss. Once you internalise that, the program stops looking like a reward for being a good customer and starts looking like what it is: a partial rebate on the margin the operator has already priced into the game you chose.

Which Scenario Is You

Find your real bracket in four steps, not one. First, estimate your monthly handle — total amount wagered, not deposited. Second, identify the certified RTP of the games you actually play; the studio pages and the GLI certificate scope are public. Third, multiply handle by (100% − RTP) to get your theoretical loss. That is the number the comp rate is a percentage of. Fourth — and this is the only step the operator hides — ask what reinvestment percentage they apply to it, because that figure is rarely published.

If you are the slots player, you are worth more per pound wagered than you assume, and the program likes you. If you are the blackjack grinder, you are worth less than your volume suggests, and reduced accrual on low-edge games is not a bug aimed at you. If you are the roulette tourist, you are closer to the slots player than to the grinder. Spend bracket is a story the dashboard tells. Theoretical loss is the story the ledger tells.

FAQ

Is the comp rate calculated on how much I deposit?

No. Deposits are the most visible number and the least relevant to comp accrual. The loyalty engine reads theoretical loss, which is your total handle (amount wagered, including re-wagered winnings) multiplied by the game's house edge. You can deposit £500, cycle it through £8,000 of slot spins, and generate theoretical loss based on the £8,000 handle — not the £500 deposit. Operators reinvest a fraction of that expected-loss figure, never a fraction of your deposit.

Why do slots earn comps faster than blackjack?

Because slots carry a far higher house edge. Published RTP bands from studios like NetEnt (94.00%–96.70%) and Play'n GO (94.20%–96.50%) imply edges of roughly 3.3% to 6.0%, while Evolution's live blackjack runs 99.28% RTP — a 0.72% edge. On identical handle, the slots player generates up to eight times the theoretical loss. Since comps are a percentage of theoretical loss, the high-edge game accrues far faster. It is arithmetic, not favouritism.

Can two casinos give different comp value for the exact same game?

Yes, and this is the least-watched lever. The same base title can be licensed at different certified RTP variants — for example a Book-series slot at 96.50% on one operator and 94.20% on another. Both are validly certified by bodies like GLI, whose scope covers RNG and RTP validation against the configured paytable. But your theoretical loss, and therefore your comp accrual base, differs by over 60% on identical play depending on which variant you were served.

How do I calculate my own theoretical loss?

Three numbers. Take your monthly handle (total wagered, not deposited), find the certified RTP of the games you play from the studio's published pages, and compute handle × (100% − RTP). A player cycling £15,000 through European roulette at 97.30% RTP has a theoretical loss of £15,000 × 2.70% = £405. That £405 — not the £15,000 — is the figure any comp rate is applied to. The reinvestment percentage itself is the only variable the operator typically does not publish.

Does a higher spend bracket always mean a better comp rate?

Not reliably. Bracket thresholds are usually framed in spend or points, but the underlying accrual is driven by theoretical loss. A disciplined low-edge player can wager six figures and sit in a lower effective-value tier than a casual slots player wagering a fraction of that, because the slots player's expected loss is higher. The bracket label tracks spend; the money the operator actually books tracks loss. Those two diverge sharply for skilled or low-edge play.

Are these three player scenarios based on real accounts?

No. All three — the weekend slots spinner, the blackjack grinder, the roulette tourist — are hypothetical composite illustrations built to demonstrate the math. We did not interview players or observe accounts. The only real figures used are the publicly certified RTP bands from the game studios and the published industry GGR data. The personas are scaffolding for the calculation; the calculation is the grounded part.

Where can I verify the RTP figures used here?

The RTP percentages come from the game studios' own published documentation — Evolution's live-dealer game pages for blackjack (99.28%) and European roulette (97.30%), and the slot RTP bands published by NetEnt and Play'n GO. The certification scope that validates those figures is described on the GLI certificate resource page. What no public source discloses is the operator-side reinvestment percentage applied to your theoretical loss — that figure stays internal.

---

This piece does not cover the published tier-credit conversion formula of any single branded loyalty program — MGM Rewards, M life, or any specific operator's named scheme — because that proprietary conversion rate is not in our dataset, and we do not invent rates we cannot ground. It does not address land-based slot floor mechanics, where coin-in tracking differs from online handle. And it does not cover how bonus-funded wagering is weighted differently from cash wagering in comp accrual, which varies by operator and deserves its own teardown. Each of those is a separate calculation, and we will not fake the numbers to close the gap.