Every figure in one place
This page holds the whole model: the ten invented samples the desk is built from, and the arithmetic that turns them into every figure that appears on the site. Nothing here is observed from a live account, and no operator, regulator, charity or person is described - the samples are a worked model of how a protection duty fits together.
- samples
- 10
- accounts
- 40,000
- reviews
- 4,000
- contacts
- 2,400
- breaks
- 480
- records
- 4,000
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.
Every figure on this desk derives from ten invented samples: 40,000 accounts over 90 days, 6,000 signals, 4,000 reviews on three triggers, 2,400 contacts, 480 breaks, 1,600 false positives, 360 missed cases, one worked account and a quarter ledger. Each sample is listed below with the arithmetic that links it to the others, and every figure is illustrative.
The chain, from accounts to outcomes
| Step | Sample | Figure | Derived from |
|---|---|---|---|
| accounts | A | 40,000 | the sampled population, 90 days |
| flagged accounts | A / B | 4,000 | 6,000 flags at 1.5 per flagged account |
| reviews | C | 4,000 | 2,800 + 720 + 480 across three rules |
| no action | D | 1,600 | 40.0% of reviews |
| a message only | D | 1,440 | 36.0% of reviews |
| a limit or a break | D | 640 | 16.0% of reviews |
| restricted or closed | D | 320 | 8.0% of reviews |
| contacts | E | 2,400 | 1,440 + 640 + 320 |
| breaks | F | 480 | 12.0% of reviews |
Read downwards, the table is the whole pipeline; read across, each row is one division. Nothing in it comes from a source outside the samples, and every percentage on the site is one of these divisions.
The divisions, written out
Every percentage on the desk
- flagged accounts: 4,000 / 40,000 = 10.0%
- flags per flagged account: 6,000 / 4,000 = 1.5
- two-flag reviews: 2,800 / 4,000 = 70.0%
- high-severity reviews: 720 / 4,000 = 18.0%
- affordability reviews: 480 / 4,000 = 12.0%
- no action: 1,600 / 4,000 = 40.0%
- accounts changed: 960 / 4,000 = 24.0%
- contacts: 2,400 / 4,000 = 60.0%
- answered within 7 days: 1,680 / 2,400 = 70.0%
- changed on an answer: 480 / 1,680 = 28.6%
- breaks: 480 / 4,000 = 12.0%
- shortest break share: 320 / 480 = 66.7%
- false positives: 1,600 / 4,000 = 40.0%
- a one-off cause: 1,280 / 1,600 = 80.0%
- a data-error cause: 224 / 1,600 = 14.0%
- missed cases: 360 / 36,000 = 1.0%
- missed by a closed window: 288 / 360 = 80.0%
- staked by reviewed accounts: 30.0% from 10.0% of accounts
- still playing at 90 days: 3,424 / 4,000 = 85.6%
- closed their own account: 576 / 4,000 = 14.4%
What the numbers cannot tell you
- They cannot describe your account, your operator or your licence - they are a model, not a measurement.
- They cannot say whether any real rule set uses these thresholds. The multiple, the window and the lengths are invented to make the mechanism legible.
- They cannot say whether a protection contact was justified in any individual case, because a rule reads a comparison and not a person.
- They cannot say whether a duty was performed. Only an operator own record and an inspection can show that.
- They cannot substitute for help. Support routes are on the legal page and are independent of everything on this one.