What the function is for
An integrity function inside an operator has a narrow remit and it is easy to lose sight of it.
It exists to detect suspicious betting on the operator's own book, to decide what to do with the markets and accounts involved, to report to the bodies entitled to know, to cooperate with what follows, and to inform the operator's market design. It does not exist to investigate sport, to reach conclusions about participants, or to be a public voice on corruption.
Functions that stay inside that remit are trusted by the sports and regulators they deal with and are effective. Functions that drift outside it produce accusations they cannot support, and lose the trust that makes their reports useful.
Where it sits
Three placements are common and each has a characteristic weakness.
Inside trading. The traders see the markets first and have the deepest feel for what is abnormal. The weakness is that trading's objective is the book's margin, and an alert that suggests a price was wrong sits awkwardly with the people who set it. Integrity conclusions from within trading are also easily read as commercial decisions dressed up.
Inside compliance or risk. Independent, with the reporting culture and the regulator relationship. The weakness is distance from the markets: compliance sees the alert after trading has already reacted, and lacks the market feel to triage well.
Standalone. Rare below a certain size. Independent and expert, and at risk of isolation from both trading and compliance.
The arrangement that works is usually a small dedicated function, reporting outside trading, with a formal daily relationship to the trading floor and a defined route to compliance for reporting. Trading raises, integrity assesses, compliance reports, and the record is shared.
The structural minimum, whatever the placement: the decision to report a pattern must not require approval from anyone whose commercial position the report affects.
What the function needs
People who understand betting. The core skill is reading a market. Analysts drawn from trading, or trained by it, triage well. Analysts drawn from compliance alone, without market literacy, over-report.
Baselines. Built and maintained, per competition and market type, as the previous lessons set out. Without them the function has opinions rather than measurements.
Account-level tooling. The ability to move from a market anomaly to the accounts behind it, with linkage, quickly. This is the operator's unique contribution and it is where tooling investment pays.
Cross-operator visibility. Membership of an integrity association or a monitoring service, actively used.
A reporting route with prepared templates. To the regulator, to the association, to sports, each with a form of words that describes betting without asserting conclusions.
A lawful basis for data sharing, documented in advance. Under UK data protection law an organisation must determine its lawful basis before it starts using personal information, and must document it; integrity reports often involve suspected offences, which also need a documented condition for processing criminal offence data. The Council of Europe's Macolin Convention likewise requires that exchanges of information about manipulation comply with personal data protection laws and go no further than the necessary minimum.
Retention beyond the ordinary period for anything reported, because hearings are slow.
A voice in market design. The function that sees the alerts should be in the room when competitions and market types are being added.
Working with sports and integrity bodies
The relationship is where an integrity function is valued or ignored, and the currency is the quality of what is sent.
Be specific. Fixture, market, time, volume, price movement, account characteristics within the rules. Not "unusual activity was observed".
Be timely. Before the event where possible, during it where necessary, immediately after otherwise.
Distinguish observation from inference. What the betting showed, separately from what it might mean.
Report the explained ones too, briefly, with the explanation. It calibrates the recipient's trust and builds the shared baseline.
Follow up. Send outcome information after the event. Ask what happened to the report. Learn from the answer. The Macolin Convention expects this to work both ways: on request, the recipient of information should say what follow-up it gave.
Contribute beyond alerts. Tell the sport which of its competitions and market types are attracting disproportionate interest. That is intelligence the sport cannot generate and it directs education and monitoring where they are needed.
Respect the sport's role. It holds the disciplinary jurisdiction and knows the participants. The operator's report is an input to its process, not a substitute for it. In Britain, reporting to the sport is not only good practice: operators taking bets on events governed by the sport governing bodies listed in Part 3 of Schedule 6 to the Gambling Act 2005 must provide that body with relevant and necessary information they know or suspect may relate to a breach of its betting rules (licence condition 15.1.2).
An operator that does these things becomes the one the sport calls when it has a concern of its own, which is where the genuinely useful cases start.
Working with the regulator
Licensed betting operators generally have a reporting obligation to their regulator, and the relationship has a particular shape. In Britain, licence condition 15.1.2 requires betting licensees to give the Gambling Commission, as soon as reasonably practicable, any information they know or suspect relates to an offence under the Gambling Act 2005, with sports betting integrity information going to its Sports Betting Intelligence Unit. The Macolin Convention requires each party to oblige sports betting operators to report irregular or suspicious betting without delay to the betting regulator, another responsible authority or the national platform.
The regulator wants to see that the obligation is understood, that reports are made when the threshold is met, that they are made in the required form and time, and that the operator's own market decisions are consistent with what it reports. An operator reporting a fixture as suspicious while continuing to take unlimited bets on it will be asked why.
The regulator also assesses whether the operator's market design is responsible. Pricing competitions that fail any reasonable governance test, or offering the full proposition set on fixtures nobody supervises, is increasingly a licensing question rather than a commercial one. The Macolin Convention envisages that betting authorities may limit the supply of sports betting, particularly excluding competitions designed for under-18s or where the organisational conditions or sporting stakes are inadequate.
Record every report, its timing, its content and any acknowledgement. Late or missing reports are a finding regardless of what the pattern turned out to be.
Measuring the function
Activity measures describe effort: alerts generated, reports sent, markets suspended. Effect measures are harder and more useful.
Alert precision. What proportion of alerts survive triage, and what proportion of those survive analysis. A function whose alerts all close at triage has thresholds set wrong.
Report outcomes. What happened to what was reported: acknowledged, investigated, charged, sanctioned, explained, nothing. Tracked over time, this is the only feedback that calibrates the function.
Market design decisions taken. Competitions and market types added or removed on integrity grounds, with the reasoning. A function that has never influenced a market decision is not in the room.
Timeliness. Time from pattern to report, and whether the report preceded the event.
Cross-operator confirmation rate. How often a pattern the operator reported was also seen elsewhere, which validates both the detection and the association membership.
Voiding and settlement-hold decisions, with reversal rate and dispute outcomes. A high rate of holds released without action, or of voids overturned in dispute resolution, indicates decisions being made on the wrong evidence.
The failure modes
Each of these is observable from outside the operator.
Over-reporting. Every busy fixture reported, with no triage. The recipients stop reading.
Under-reporting. Patterns explained away by trading because the book was on the right side of the money. The regulator notices eventually.
Conclusion inflation. Reports asserting that a match was fixed. Undermines the case and exposes the operator.
No follow-up. Reports sent and forgotten. The function never learns what it got right.
Market design absent. Everything priced, everything monitored after the fact, the highest-risk product offered on the least supervised competitions.
Voiding as confiscation. Winning bets voided on suspicion, losing bets settled. In Britain, terms must not be unfair within the meaning of the Consumer Rights Act 2015, and customers can take unresolved disputes to an independent ADR entity; one-sided voiding is hard to defend there, and the pattern becomes a reputational problem.
Tipping off. An account restricted with a message that discloses the report. A prejudiced investigation and, in some frameworks, an offence: in the UK, for example, a person who knows or suspects that a money laundering investigation is under way commits an offence by making a disclosure likely to prejudice it.
No lawful basis. Data shared without one, or refused without reason. Either an unlawful disclosure or a useless one.
The standard to hold to
Sports integrity is one of the few areas where the gambling industry is unambiguously part of the solution. Operators see what nobody else can, and the detection capability that exists across the sector has surfaced cases that would otherwise never have come to light.
That position depends on a discipline that is easy to state and takes real effort to keep: describe what you saw, decide what you can defend, report what you should, and do not claim what you do not know. An operator that holds to it is trusted by the sports it prices and the regulators it answers to, and its reports move cases. One that does not is producing noise, or worse, producing accusations, and both are quickly recognised for what they are.
The unusual betting pattern is a hypothesis about a market. The people behind it are real, and the sport that decides what happened to them needs the operator's evidence to be exactly as strong as it is, and no stronger.