Skill Gaming vs Gambling Banking: When the MCC 7995 Argument Works (2026)

Every week an operator asks us the same thing: can we get banked under a software MCC 7372 instead of gambling MCC 7995 because our product is "skill-based"? Sometimes the answer is yes. Most of the time the argument collapses on contact with a real acquirer's risk team, and the operator ends up with a closed account, a retroactive bill, and in the worst cases a MATCH listing. This guide explains the actual legal tests for skill versus chance in the jurisdictions that matter, when the skill argument genuinely works at the banking layer, and what happens when it doesn't.

Table of Contents

1. Why operators try the skill argument
2. The legal tests by jurisdiction
3. Game-by-game classification matrix
4. The MCC code map
5. Why banks and schemes reject most skill arguments
6. Consequences of MCC misclassification
7. Where the skill argument actually works
8. Borderline cases and how to position them
9. The honesty test
10. Practical recommendations
11. FAQ
12. Related Articles

Why operators try the skill argument

The economics are obvious. A gambling acquirer charges a merchant discount rate of 4.5 to 8 per cent, sometimes higher for offshore-licensed operators. A SaaS acquirer charges 2.5 to 3.5 per cent. Rolling reserves on MCC 7995 typically sit at 5 to 10 per cent of monthly volume for six months; software merchants pay zero or nominal reserves. Add scheme licence fees, integrity-monitoring programmes and the gambling-specific compliance overhead, and the gap between the two classifications easily exceeds 4 per cent of gross volume.

For a borderline product, that gap is the difference between a viable business and a margin-thin one. So operators reach for the cheapest classification their lawyer can defend, present it to the acquirer with a tidy narrative, and hope the underwriter doesn't push back.

The problem is that underwriters now push back routinely. Card schemes have spent a decade tightening MCC 7995 enforcement, and acquirers who let "skill" merchants slip through gambling controls get hit with scheme fines, retroactive interchange and forced book audits. The skill argument has gone from a useful grey-zone tool to a red flag in most onboarding files.

The legal tests by jurisdiction

Banking classification doesn't follow any single legal test, but acquirers lean heavily on the regulatory definition that applies in the merchant's licensing jurisdiction. The four frameworks below cover most cases.

The US "predominance test"

Most US states apply a predominance test: an activity is gambling if the element of chance predominates over the element of skill. A handful of states use the stricter "any chance" test, where the presence of any material chance makes the activity gambling.

Under predominance, chess, billiards, organised sports and most pure puzzle competitions pass as skill. Poker fails in the majority of states despite obvious skill content, because courts find the deal of the cards predominates over the strategic decisions on any single hand. The New York Court of Appeals' 2022 ruling on DFS (daily fantasy sports) confirmed it qualified as a contest of skill under state law; the Massachusetts SJC has gone the other way on related products. There is no federal answer.

Layered on top is the UIGEA (Unlawful Internet Gambling Enforcement Act), which doesn't define gambling but bans the processing of payments connected to bets unlawful under any applicable federal or state law. Acquirers use UIGEA exposure as a hard filter regardless of the underlying skill question.

The UK "skill exception"

The Gambling Act 2005 defines a "game of chance" to expressly include games that involve both chance and skill, which closes most of the doors operators try to open. The Act leaves a narrow path: the Gambling Commission can designate certain skill-heavy games as outside the gambling perimeter, and prize competitions are lawful if they offer a free entry route of equal standing to the paid route (the "WhatsAuction" model).

E-sports illustrate the line cleanly. Betting on an e-sports match is gambling and requires a UKGC licence. Operating the tournament that the bets are placed on is not gambling and needs no gambling licence. Both can sit on the same domain only if the betting product is licensed and operationally separated. The full UKGC stance is documented at the UK Gambling Commission.

EU member-state regimes

There is no harmonised EU position. Member states draw the line individually, and several have created bespoke skill-game regimes:

• Germany: the Federal Court of Justice ruled in 2014 that poker is gambling, and the GlüStV (Glücksspielstaatsvertrag) treaty applies. Skill exemptions are narrow.
• France: similar to Germany on poker; horse racing has a separate regime under PMU; ARJEL/ANJ licences cover poker as a regulated form of gambling rather than skill.
• Italy: operates a dedicated "Giochi di Abilità" (skill games) licensing regime under ADM, distinct from casino and betting licences. Skill games here are formally licensed gambling-adjacent products, not unregulated.
• Spain: the DGOJ (Dirección General de Ordenación del Juego) regulates skill games as a separate vertical with its own licence. Operating without that licence is unlawful regardless of how the operator characterises the product.
• Malta: the MGA offers a controlled skill games licence as well as B2B variants. An MGA skill-games licensee can in some cases secure non-gambling banking, but only because the regulator has formally classified the product.

Other jurisdictions worth knowing

Canadian provinces broadly follow predominance. Australia's Interactive Gambling Act treats most online wagering as gambling regardless of skill content. India is fractured by state, with Tamil Nadu and Andhra Pradesh banning DFS as games of chance and Karnataka and Kerala oscillating with judicial reviews.

Game-by-game classification matrix

The table below shows how the major products are classified for banking purposes across the jurisdictions operators most often pick. "Skill" means standard SaaS / recreation banking is realistically available; "Gambling" means MCC 7995 is mandatory; "Mixed" means jurisdiction-specific and likely to require a bespoke licence.

The pattern is consistent: if there is a house bank, randomisation infrastructure or a rake on player-versus-player stakes, the product is gambling. If revenue comes from entry fees or subscriptions on competitions where outcomes are dominated by player skill and the operator takes no position on outcomes, the product is skill.

The MCC code map

MCC (Merchant Category Code) is set by the acquirer at boarding, validated by the card schemes, and effectively immutable for the merchant. Three codes dominate the skill-versus-gambling question.

Operators occasionally try MCC 5816 (digital goods) or MCC 7993 (amusement, recreation services). Schemes treat both as red flags when used for money-prize products and reroute them to MCC 7995 during scheme audits.

The merchant doesn't pick the code. The acquirer's risk team picks it from the description of business, the website, the terms and conditions and the chargeback history. Once it's set, the schemes can override it.

Why banks and schemes reject most skill arguments

The skill-versus-chance debate looks like a legal question. From an acquirer's desk it's a unit-economics question, and the legal label rarely changes the answer.

The "sweat the players" test. If your business model only works because most players lose money, you are running a gambling business regardless of the skill content of the underlying game. Acquirers ask for the P&L. If the gross margin line is dominated by the negative expected value paid by losing players rather than by entry fees, subscriptions or tournament rake on skill-validated competitions, the file is rerouted to gambling underwriting.

Scheme-level classification overrides the merchant. Visa and Mastercard publish MCC definition lists and they audit issuer and acquirer activity against those lists. The Visa Gambling Merchant Programme and Mastercard's equivalent integrity monitoring catch high-volume merchants on the wrong code. When they do, the scheme can force reclassification and bill the acquirer for the interchange difference. The acquirer passes that bill straight to the merchant, with a closure notice attached.

Reverse engineering from the product mechanics. Underwriters look at the rake or commission line. If the platform keeps a percentage of each pot or stake, the product walks like gambling regardless of marketing copy. Geo-blocking patterns are another tell: if you block the US under UIGEA but tell the acquirer your product isn't gambling, your own compliance behaviour contradicts your boarding narrative.

Marketing copy is evidence. "Win big", "jackpot", "double your money", "real cash prizes" and bonus-style promotional mechanics all read as gambling language to a card-scheme audit. The acquirer's compliance team screenshots your site before boarding and again at every review.

If you are running a high-risk product and trying to navigate this honestly, our high-risk merchant account guide and iGaming acquirer guide walk through the underwriting questions in more detail.

Consequences of MCC misclassification

When an acquirer or scheme rules that you've been boarded on the wrong MCC, the consequences stack quickly.

Forced reclassification and retroactive interchange. The scheme calculates the difference between gambling and non-gambling interchange for every transaction in the look-back window (commonly six to twelve months) and bills the acquirer. The acquirer passes it through to the merchant or deducts it from settlements and reserves. Six-figure clawbacks on mid-volume merchants are routine.

Acquirer fines. Direct fines in the 5,000 to 50,000 USD range are typical for first-time misclassification, layered on top of the interchange differential. Repeat or wilful misclassification escalates fast.

Account closure with reserve hold. The acquirer closes the merchant account immediately. The rolling reserve is held for the standard 180-day chargeback window, often extended to 270 or 365 days if the file is being investigated for misrepresentation. The iGaming rolling reserve guide covers the mechanics.

MATCH listing. This is the most damaging long-term outcome. MATCH (Member Alert to Control High-Risk Merchants) is Mastercard's blacklist of merchants closed for cause. Reason code 04 (Excessive Chargeback) and reason code 12 (Merchant Misrepresentation) are the ones that bite here. A MATCH listing under reason code 12 essentially blocks the named principals from boarding any Mastercard acquirer for five years. We cover the survival path in the MATCH list iGaming guide.

Chargeback exposure spikes. Once you're closed and reclassified, the chargeback ratio you were managing under SaaS thresholds (around 1 per cent) is judged against gambling thresholds and historical performance — but the punitive scheme programmes still apply on the way out. Operators routinely discover they were running 2 to 3 per cent dispute ratios that the SaaS acquirer never flagged because it wasn't watching. See the iGaming chargeback management guide for the operational fix.

Licensing knock-on. Gambling regulators (the UKGC, MGA, ADM, DGOJ) require licensees to maintain compliant payment processing. Misclassification is reportable, and unlicensed gambling operation in a regulated market is criminal exposure on top of the banking damage. The FATF updated its 2025 guidance to call out skill-game misclassification as a money-laundering typology, and the FATF recommendations now appear in most acquirer's onboarding training.

Where the skill argument actually works

It does work. Here are the cases where standard SaaS or recreation banking is genuinely available.

Chess platforms run on MCC 7372 without controversy. Chess.com and Lichess both bank under standard software classifications. Paid tournaments with cash prizes don't shift this as long as the operator doesn't run player-versus-player wagering and earns revenue from subscriptions and entry fees rather than rake on player stakes.

Esports tournament organisers (FaceIt, ESEA, BLAST) bank as event organisers or SaaS. Prize pools paid out from sponsorship and entry fees are not gambling. If the same operator builds a betting product alongside the tournament, the betting product needs its own gambling-classified processing stack, hard-separated from the tournament wallet.

US Daily Fantasy Sports has a bespoke regulatory regime that varies state by state but is broadly outside MCC 7995 in the states where it is licensed as skill. Acquirers willing to board DFS operate purpose-built programmes with elevated compliance but lower MDRs than full gambling. Outside the US, DFS is gambling almost everywhere — operators sometimes route it through MCC 7995 voluntarily for clarity.

UK skill prize competitions with a genuine free entry route bank as e-commerce or competition organisers. The free route must be operationally real (postal entry of equal standing, no skewed odds) for the classification to hold under UKGC review.

Pure subscription play-money platforms. Poker apps that don't offer real-money cashout, with revenue from subscriptions, cosmetics or advertising, bank as SaaS without difficulty.

For comparison with regulated-gambling banking on the other side of the line, see the iGaming banking requirements guide and iGaming business bank account guide.

Borderline cases and how to position them

A few products live on the line.

DFS outside the US. The cleanest position is to seek either a gambling licence in the operating market (Italy, Spain) or to voluntarily classify under MCC 7995 with an acquirer experienced in DFS. Trying to defend SaaS classification for cash-prize DFS in EU markets without a licence creates regulatory exposure on top of banking exposure.

Esports with skill-pick prediction layers. If users stake money to predict outcomes, that's betting regardless of skill content in the prediction. Classify the prediction layer as gambling, keep tournament operations on the SaaS rail with operational separation.

Mahjong and backgammon cash platforms. These are gambling under predominance in most major markets despite skill content. Position them as gambling and licence accordingly.

Trading-style skill apps. Skill-based binary-options apps and "trading tournaments" tend to fail two tests at once: they look like gambling to gambling regulators and like unauthorised investment activity to financial regulators. The forex broker bank account guide covers the investment-regulation side.

The honesty test

Strip away the legal vocabulary and ask four questions about your own product:

1. Do you make money when players lose? If yes, this is gambling regardless of skill content.
2. Do you make money on subscriptions or fees that don't depend on player outcomes? If yes, this is genuinely skill.
3. Do you pay out more to skilled players than unskilled players in a way the operator profits from? If yes, gambling-adjacent — treat the file accordingly.
4. Does the platform take rake or a commission on player-versus-player money? If yes, this is almost always gambling in the eyes of acquirers and schemes.

These four questions match how an acquirer's risk officer reads the file. If your answers don't line up with the classification you're applying for, the boarding will fail or — worse — succeed and unwind later.

Practical recommendations

Stop fighting the MCC. The classification follows the product, not the marketing. If your product is gambling, get a gambling licence and board on MCC 7995. The route is harder up front and cheaper over five years than the misclassification path.

Document the skill test if you genuinely qualify. Variance analysis showing top-decile player consistency over thousands of hands, simulations that strip out chance and demonstrate stable rankings, and academic-style win-rate distributions all help if an acquirer or regulator queries the classification. The work needs to exist before boarding, not after a reclassification notice arrives.

Separate skill and gambling products operationally if you run both. Different entities, different domains, different acquirers, different wallets. Mixed flows are the fastest way to drag the clean side of the business into a gambling reclassification.

Use the right licensing infrastructure. If you're skill in a market that licenses skill (Italy ADM, Spain DGOJ, Malta MGA skill licence), apply for the licence. The licensed-skill route gives you both a defensible legal position and an acquirer programme designed for it.

Don't rely on offshore licensing to defend a skill claim in onshore markets. A Curaçao or Anjouan licence doesn't make a product skill — it makes it offshore gambling. The classification follows the product, not the licence. The Curaçao gaming licence guide and Anjouan licence guide cover what offshore licences actually do for banking.

Build the file before you apply. Underwriters reward operators who turn up with a settled position. The US Department of Justice publishes guidance on the federal layer at the US Department of Justice, and UKGC, MGA and DGOJ publish their own classification guidance — cite the applicable one in your boarding pack.

FAQ

Can I call my poker app a "skill game" to get a software MCC?

No. Online poker is classified as gambling in every major banking jurisdiction including the US (under most state law), UK, Germany, France, Italy, Spain and Malta. Acquirers and card schemes treat poker as MCC 7995 regardless of the skill content of the game. Boarding under MCC 7372 with a real-money poker product will be reversed at the first scheme audit, with retroactive interchange and potential MATCH listing under misrepresentation.

Is daily fantasy sports gambling?

In the United States, DFS is classified as skill under federal UIGEA carve-outs and in the majority of states where it operates. New York and most other states' courts have confirmed this. Outside the US — UK, Germany, France, Italy, Spain — DFS is generally classified as gambling and requires the appropriate gambling licence. Banking follows the licensing classification in the operating jurisdiction.

What MCC code does a chess platform use?

A chess platform — including paid tournaments with cash prizes — typically banks under MCC 7372 (software) or MCC 7997 (recreation clubs) where membership-funded. This works because the operator earns from subscriptions, entry fees and advertising rather than from a rake on player-versus-player wagering. If a chess platform introduces head-to-head money matches with operator rake, the classification shifts toward MCC 7995.

Why does my acquirer reclassify me as gambling after months of trading?

Reclassification usually follows a scheme audit, a spike in chargeback ratios, a compliance review triggered by marketing copy, or an MCC monitoring report from Visa or Mastercard. The acquirer's discretion is wide, and contractually the merchant agreed at boarding to abide by scheme rules including MCC assignment. Once reclassified, retroactive interchange, fines and reserve adjustments apply for the look-back window.

Can I avoid MCC 7995 by getting a skill games licence?

Sometimes, in specific jurisdictions. An Italian ADM skill licence, a Spanish DGOJ skill licence or a Malta MGA skill games licence creates a regulated skill category that acquirers experienced in that market can board under non-gambling MCCs. In jurisdictions without a formal skill regime — UK, most of the US, Germany — there is no licensing route that delivers a non-gambling MCC for a product the regulator considers gambling.

Is rake gambling?

Operationally, yes — taking a percentage of player-versus-player stakes is the defining commercial feature of gambling. Acquirers and schemes use the presence of rake as a primary signal during boarding and audit. A platform that takes rake will be classified MCC 7995 regardless of how the underlying game is described.

What's the difference between MCC 7995 and high-risk SaaS?

MCC 7995 is the gambling-specific code with mandatory rolling reserves, gambling-licence requirement, scheme-level integrity monitoring and MDR in the 4.5 to 8 per cent range. High-risk SaaS sits under MCC 7372 or sector-specific codes, has elevated underwriting but lower MDR (typically 2.5 to 4 per cent), and doesn't trigger gambling-specific scheme programmes. The two are not interchangeable and the merchant doesn't choose.

Will I be MATCH-listed if my account closes for MCC misrepresentation?

There is a meaningful risk of MATCH listing under reason code 12 (Merchant Misrepresentation) if the closure investigation concludes the merchant knowingly applied for the wrong code. Reason code 12 is one of the harder listings to challenge and effectively blocks Mastercard acquirer boarding for five years. The full mechanics are in our MATCH list iGaming guide.

Related Articles

• iGaming Banking Requirements: The Complete Data Reference
• Sports Betting Business Bank Account
• The MATCH List and iGaming: How to Stay Off It
• How to Choose a Payment Acquirer for iGaming
• High-Risk Merchant Account Guide
• iGaming Chargeback Management: The Operator's Complete Guide