Casino Bonus Grading Methodology: How We Score A-F (2026)

ToolsGambling calculates an approximate expected value (EV) for a bonus using a specified turnover, subtracts two heuristic deductions and assigns A, B, C, D or F through ordered rules. The result is not a forecast of net profit or a probability of completing wagering or withdrawing funds.

The Casino Bonus Analyzer helps compare terms under consistent assumptions and identify restrictions to check before accepting an offer. It does not rank casinos or confirm that a particular bonus is available to you.

This guide explains the implementation inspected on October 9, 2026. The examples use hypothetical terms. The manual form and its results were checked in a local interface—not through a real bonus claim, game session or withdrawal.

What information do you need before calculating?

You need the full offer terms and actual amounts, not just the advertised bonus percentage. Some inputs affect the formula, some affect flags, and others require separate checks.

Select the wagering basis exactly as stated in the terms: bonus only, deposit only, or bonus plus deposit. None is inherently the most demanding in every case; the comparison depends on the amounts.

The manual form has no separate bonus-amount or bonus-cap field. For a capped match, establish the amount actually credited first, then calculate the effective match percentage: credited bonus ÷ actual deposit × 100%. Do not enter the advertised percentage if the cap would make the resulting bonus amount incorrect.

If a term is unclear, ask the operator or skip the offer. A completed field does not turn an assumption into a verified fact. The wagering requirements guide explains how eligible bets count toward wagering.

How are turnover and EV calculated?

Numerical EV is the bonus amount minus expected loss on the specified turnover and two heuristic deductions. The model assumes a fixed volume of bets; it does not check whether the balance will last long enough to place them.

The manual calculation uses these formulas:

• Bonus amount = actual deposit × match percentage ÷ 100.
• Required credited wagering = rollover multiple × wagering basis.
• Turnover = required credited wagering ÷ game contribution fraction.
• Expected loss = turnover × (1 − RTP fraction).
• EV = bonus − expected loss − max-bet deduction − payout-cap deduction.
• EV percentage = 100 × EV ÷ bonus.

A 100% contribution is a fraction of 1; 50% is 0.5. RTP of 96% is a fraction of 0.96. The denominator for EV percentage is the bonus, not the deposit. The percentage is not a probability of success.

For a $100 deposit with a 100% match, the bonus is $100. With 10× wagering on the bonus only and 100% game contribution, turnover is $1000. At the fixed 96% RTP, expected loss is $40. With no additional deductions, EV is +$60.00, or +60.0%.

These figures are the output of the stated formula, not a hidden forecast of your withdrawal. The interface displays money to two decimal places and percentages to one; display rounding does not make the model more precise.

House edge in this formula is 1 − RTP. RTP describes the long-run payout fraction of turnover, not win frequency or a refund of your deposit. See RTP versus volatility for the distinction between payout rate and outcome distribution, and the house edge guide for the expected-loss fraction.

Which deductions are heuristics?

The max-bet and payout-cap deductions are configured approximations, not mathematically calculated costs of those restrictions.

The max-bet threshold of 5 uses raw units of the selected currency, without exchange-rate conversion. A max bet of 5 is still a limit—not unrestricted play or a recommended stake. A lower limit does not itself change house edge or mathematically imply higher volatility.

For example, with a $100 bonus, a $2 max bet produces a $5 deduction. A $100 payout cap produces a $20 deduction: 10% of the difference between $300 and $100. This is not the expected value of capped payouts or a prediction that an uncapped payout would normally equal three times the bonus.

Assessing the actual effects of these restrictions would require the payout distribution, sequence of stakes, cash and bonus wallet rules, cap mechanics and stopping conditions. Those are not modeled here.

How are A–F grades assigned?

The grade comes from the first matching condition in the table below. The function checks for a critical flag first, then considers EV percentage and the number of other flags.

There is no E grade, no set of six weighted factors and no calibrated score from 0 to 100. After the critical-flag override, all other flags count equally, whether labeled major or minor.

The boundaries matter:

Two major flags therefore do not automatically produce F. Exactly one noncritical flag can produce D even at −180% EV. That is a feature and limitation of the algorithm, not an endorsement of the offer.

This is the verified function's sequence. If a tooltip or static grading scale in the interface describes different rules, the actual result follows this sequence.

Which seven restrictions does the analyzer flag?

The detector checks seven configured restrictions. Only a bonus whose principal cannot be withdrawn creates a critical flag. These red flags prompt further checks; they are not proof of fraud or illegality.

At exactly 40×, 7 days, 5 currency units or a maximum payout of exactly 5 × bonus, the corresponding flag does not trigger.

The payout cap has two different thresholds: a flag appears below 5 × bonus, while the numerical deduction applies below 3 × bonus. A cap can therefore create a flag without a monetary deduction.

These are ToolsGambling's local heuristics, not universal industry or legal standards. The detector does not perform twelve automatic checks. It does not use the excluded-games list to create a separate flag; country and payment restrictions, changed terms and missing game-contribution tables still need manual review.

How can you reproduce the manual calculation?

Enter the following hypothetical terms in the manual form to reproduce the verified A result. Use the terms-parser section below the chat, not the separate AI chat.

1. Open the analyzer.
2. Select “Manual Form” in the terms-parser section.
3. Enter an actual deposit of 100, a match percentage of 100% and a rollover multiple of 10.
4. Select USD ($) and the wagering basis “Bonus only.”
5. Enter a max bet of 5, slot contribution of 100% and an expiry period of 14 days.
6. Leave maximum payout and minimum deposit blank.
7. Leave the sticky-bonus and restricted-withdrawal checkboxes off.
8. Click “Analyze Bonus →.”

Manual analyzer form with a 100 USD deposit, 100% match, 10× bonus-only wagering, max bet of 5, 100% slot contribution and 14-day expiry

Local ToolsGambling interface capture. Maximum payout and minimum deposit are blank. The 20 shown in the minimum-deposit field is a placeholder, not an entered value.

There is no separate RTP field: the manual form uses 96%. The calculated bonus is $100, turnover is $1000 and expected loss is $40. There are no deductions or flags, so the result is A, +$60.00 and +60.0%.

Manual calculation result showing grade A, expected value of plus 60 dollars and plus 60.0% under hypothetical terms

Local ToolsGambling result for the stated example. This verifies the manual interface calculation, not a real bonus claim or withdrawal.

What changes the result when everything else stays the same?

Each change below is applied separately to the baseline, not cumulatively. The currency is USD, RTP is fixed at 96%, and all monetary figures are in dollars.

Using bonus plus deposit doubles the amount to which the rollover multiple applies, doubling turnover and expected loss. A 50% contribution also doubles actual turnover, but adds a flag—hence B and C despite the same 20% EV.

A max bet of 2 does not change turnover or house edge; it reduces the result by the heuristic deduction of 5. A maximum payout of 100 creates a deduction of 20 and a flag. These figures agree with independent calculation using the formulas and the tool's verified functions.

Why can F appear alongside positive EV?

A bonus whose principal cannot be withdrawn receives F, but its numerical EV is not recalculated. A positive figure beside F must therefore not be read as an expected amount of withdrawable money.

To check this, restore the baseline inputs, enable “Sticky bonus” and click “Analyze Bonus →” again. The grade changes to F, while +$60.00 and +60.0% remain.

Sticky bonus result showing grade F while expected value remains plus 60 dollars and plus 60.0%

Local ToolsGambling capture: the critical flag changes the grade to F, but does not change the numerical formula.

A sticky bonus may have value under specific rules; a non-withdrawable principal does not prove that it is always worthless. However, this analyzer does not model its actual cash value, so the positive displayed EV is not an expectation of funds available to withdraw.

Where does the model stop being reliable?

The main limitation is that the model calculates a specified turnover without simulating whether the balance runs out first. Positive EV does not establish wagering completion probability, net profit, operator approval, eligibility, legality or withdrawal success.

Consider a hypothetical $200 deposit and $200 bonus, with 35× wagering on their combined amount, 100% contribution, 96% RTP, a $5 max bet and no payout cap. Turnover is $14000, expected loss is $560, numerical EV is −$360 and EV percentage is −180%. There are no flags, and the grade is F.

An initial balance of $400 does not guarantee that $14000 of turnover can be completed. The −$360 figure does not predict losing $360 from a $200 deposit without further deposits. It exposes the limitation of a fixed-volume approximation. This is not a claim about a current 35× offer, and the example is outside the current rule for licensed offers in Great Britain.

The analyzer does not simulate balance survival, repeated random game paths, a paytable or an actual bonus claim. Nor does it accurately reproduce table games, free-spin winnings, cashback, staged bonus release or separate cash and bonus wallets.

If a game's contribution is zero, wagering cannot be completed on that game. Stop the comparison: this formula has no finite turnover in that case, even if the current interface grade looks like D.

How should you treat pasted-term analysis?

Pasted-term analysis is a separate data-extraction route, not confirmation of the terms by the operator. The manual form was exercised for this guide; the pasted-text route was not tested.

When information is absent or interpreted in a particular way, the current parser may use these defaults and approximations:

These are defaults and rough approximations, not verified offer terms. Compare the extracted fields against the full source or use the manual form. Do not present an unknown value as fact. Manual mode avoids an AI text-parsing request.

The offer-URL tab is not functional yet: automatic retrieval of terms from a URL is unavailable. The tool also does not automatically retrieve the active game's RTP or validate every country and payment rule.

What if the actual RTP differs from 96%?

You need a separate sensitivity calculation when RTP differs, because this analyzer does not let you change it. Both input routes use a fixed 96%.

For a separate comparison of turnover and expected loss, the wagering calculator offers editable RTP and game contribution. Keep the other assumptions consistent. Its completion-probability figure is another heuristic, not a validated probability for a specific game; do not use it or its payout-cap calculation as confirmation that wagering will succeed.

What should you save, and when should you stop?

Save the source and full terms before accepting a bonus. If essential information is missing, clarify it with the operator or skip the offer.

For a reproducible comparison, record:

• The exact offer URL, date, language version and applicable country.
• Complete terms, the bonus actually credited and any bonus cap.
• Actual deposit, wagering basis and rollover multiple.
• The contribution table, eligible games and RTP of the active game version.
• Max bet, maximum payout and expiry period.
• Cash and bonus wallet rules, withdrawal conditions and rules for ending wagering.
• Payment and country restrictions, plus KYC requirements.

Compare offers under the same assumptions and save the entered values with the result. A grade does not replace checking for changed terms.

Set money and time limits in advance. Do not raise stakes, play faster or longer, or deposit more to meet a deadline or try to “realize” the calculated EV. Declining the bonus is a valid option.

Which legal rules need separate checks?

The analyzer's thresholds do not establish compliance with law or licence requirements. Participate only where gambling is lawful for adults and permitted by the operator's rules; an English-language page does not determine your jurisdiction.

For UKGC-licensed incentives in Great Britain, the current LCCP 5.1.1(3)(a) limits wagering requirements on bonus funds to a maximum of 10×, effective January 19, 2026. The implementation date appears in the UKGC announcement of March 26, 2025, updated July 24, 2025. This rule does not apply automatically worldwide.

The analyzer's “above 40×” flag does not check compliance with that rule. Separately, UKGC guidance on withdrawals of deposits and winnings from deposit funds, updated May 28, 2024, protects those funds from bonus wagering restrictions, subject to general regulatory obligations. A worked example using a combined basis does not mean a Great Britain operator may withhold your entire cash deposit until that wagering is completed.

When was the method checked, and what is the commercial disclosure?

This methodology is based on an implementation inspection on October 9, 2026, and local observations of the manual form. It explains the current formulas and limitations; it does not validate a particular casino or real offer.

ToolsGambling earns commissions from some casino links. The numerical formula and grading function have no affiliate or commission input. That does not prove that every placement or link order is unbiased, or promise complete independence in all commercial decisions.