UKGC Uncovers 3,426 Illegal Betting Payments in 157.2 Million Bank Records

Name Matching Failed Before Anything Worked

The panel makes use of YouGov transaction records from 38,000 consenting individuals between July 2018 and July 2026. The UKGC’s researchers used a Yaspa algorithm, set at a threshold of 0.5 probability. Using it, they identified 4.8 million transactions to licensed, non-lottery gambling companies.

Initially, attempts to use fuzzy matching to locate unlicensed operators yielded no results. Researchers had scanned transaction descriptions for known illegal operator names. That suggests such firms disguise their identities from the payment providers they hand data to.

However, two methods of exact-match searching did turn up unlicensed operators.

Searching transaction descriptions against known illegal-operator statements turned up 821 payments. Searching merchant-name fields against suspected operator names flagged 2,628 transactions. With an overlap of 23 between these methods, the UKGC discovered a total of 3,426 unique payments. Those 179 individuals averaged an age of 44, and 115 were male.

The illegal gambling transactions discovered represent just 3,426 out of 157.2 million, equating to 0.0022% of the total. Set against the licensed ones, they equate to just one in every 1,401 gambling-related payments. That is a 15M calculation: 3,426 divided by the 4.8 million licensed gambling transactions.

Among the 146 detected individuals who provided a gambling problem score, nearly half scored 8 or above. That is 45% on the Problem Gambling Severity Index, compared with just 6% of licensed gambling consumers in the same sample. By contrast, 68% of licensed consumers scored zero on the Index, compared with 26% of those detected.

However, 91% of those with repeated detected illegal payments also made transactions with licensed operators. Their behaviour did not significantly change before and after their first illegal payment. The average number of monthly licensed transactions per user decreased slightly from 24.8 to 23.7.

The Commission Has Been Sizing This Market All Year

This study complements other research conducted this year on the unlicensed market, including survey work and web-traffic estimates. It is the Commission’s third approach to analysing this market, following an update on the methodology in April. In comparison, the regulator reported £17.5bn in gross gambling yield for the licensed sector.

The Commission believes this to be a pioneering effort. "It may be the first time such an analysis has been completed by any regulator or research team in the world," wrote Steven Murphy, a Commission data scientist.

The Commission’s limitations: the panel isn’t nationally representative, and undetected transactions are assumed. You also can’t compare the PGSI figures to the Gambling Survey for Great Britain — only to each other.

What This Means

• For operators: the substitution argument is now weaker in evidence. Licensed play does not stop when a customer starts paying an offshore site. Harm cases will not be answered by pointing at the black market;
• For affiliates: anyone sending UK traffic to offshore brands should read the 45% figure as a regulatory trigger, not a market size. The Commission has now shown it can pull named counterparties out of bank data, which makes affiliate referral paths traceable in the same way;
• For providers: payment and merchant-name fields are the weak point the study exploited. Payment firms serving UK consumers should expect requests built on merchant descriptors rather than operator names.

What’s Next

The Commission says it will expand its use of financial transaction data and is inviting work with banks and payment providers. No date is attached.

Source: Gambling Commission, "Illegal gambling – Findings from an open banking data analysis", https://www.gamblingcommission.gov.uk/report/illegal-gambling-findings-from-an-open-banking-data-analysis