The record the review keeps
The record is the least visible part of a protection duty and the one that decides whether the duty can be shown to have happened at all. It carries the signals that opened the review, the crossing dates, the version of the rule that was applied, whether a policy or a person resolved it, the response, and what came back from the player - and in the sampled desk it exists for all 4,000 reviews, including the 1,600 that ended with nothing done.
- reviews
- 4,000
- records kept
- 4,000
- no-action reviews with a record
- 1,600
- fields per record
- 9
- records available on request
- 4,000
- mean trigger to record
- 2.1 days
Read the signals
Five comparisons the account already supports, run continuously rather than after a complaint. 6,000 flags in the sample, 96.0% of them relative to the account own history.
Choose a response
Nothing, a message, a limit or a break, or a restriction - one of four, graded by how much it takes from the player. 960 of 4,000 reviews in the sample changed access.
Write it down
Signals, crossing dates, rule version, path, response and outcome - one record per review, including the 1,600 that ended with nothing done.
Take the appeal
A route outside the desk that decided. 480 appeals in the sample, 176 of them upheld, which is 36.7% - high enough that the route does real work.
A protection record names the signals that opened the review, their crossing dates, the version of the rule applied, whether a policy or a person decided, the response chosen, its dates, and what came back from the player. The sampled desk kept one for every review, including the 1,600 that ended with no action - and those are the rows that prove the monitoring obligation was performed.
Nine fields, and why each one is load-bearing
| Field | What it settles |
|---|---|
| The signal or signals | that a rule was met rather than a suspicion held |
| The crossing dates | that the two-flag window was live, where that rule was used |
| The rule version | which thresholds applied, since they are revised |
| The threshold and the actual figure | the margin - 4.50x against a 4.00x trigger, for instance |
| The path | whether a policy or a person decided |
| The response | what changed, if anything |
| The dates of the response | the length of a break, and its end |
| The player reply | whether the answer arrived, and on which day |
| The outcome | confirmed, lifted, reduced, or closed with no action |
The fourth row is the one an appeal usually turns on. A ratio that cleared a threshold by a wide margin is a different case from one that cleared it by 0.1x, and only the record shows which happened. Without it, every decision reads as a bare assertion.
The no-action reviews are the important ones
It is tempting to think of the 1,600 no-action reviews in the sample as the waste: the rule fired, somebody looked, and nothing was done. In fact they are the only rows that demonstrate the second obligation. A log containing 2,400 actions shows that something happened; it does not show that the operator considered anything and declined. Only a record of reviews reviewed-and-closed shows that the pipeline has a decision at the end of it rather than a trigger wired directly to a limit.
Worked example / sample J
- reviews in the quarter: 4,000
- records kept: 4,000 of 4,000
- of those, no action taken: 1,600 (40.0%)
- responses that changed access: 960 (24.0%)
- mean time from trigger to record: 2.1 days
- contacts in the same quarter: 2,400
- 2,400 / 4,000 = 60.0% of reviews produced a contact
- 4,000 - 2,400 = 1,600 reviews that produced no contact at all, matching the no-action count exactly
How to ask for it, and how to read it
- Ask in writing for the protection record rather than for an explanation of the decision.
- Ask for the rule version and the threshold, not just the signal, so the margin can be checked.
- Check the dates against your own history; a crossing date that does not match a day you played is a data error, and data errors are 14.0% of the sample false positives.
- Check whether a policy or a person decided, because the two are appealed differently.
- Check the response end date. A break with no stated end is the strongest thing to raise.
- Keep the record. It is the document the appeal and, if needed, the regulator reads.
All of this is a description of a mechanism rather than legal advice, and what a particular operator must provide depends on the licence covering the account. The point of the page is narrower and more durable: a protection decision that cannot be read is a decision that cannot be tested, and the record is what makes it readable.