The record
What this index holds and refuses to print
110 separate figures circulate about the thirteen books on this register and are wrong, unsupportable, or true of a different country. They are recorded, they are counted, and not one of them appears on this domain.
This page names no numbers. A warning that reproduced the figures in order to warn about them would be publishing them, and a search engine reading the page has no way to tell the warning from the claim. What is here instead is the anatomy: what kinds of wrong figure this market produces, roughly how many of each, and how they get into circulation.
The reason it is worth a page rather than a footnote is that every one of these numbers looks completely reasonable. A minimum withdrawal is a plausible minimum withdrawal whichever country it was written for. A review score is a plausible score whoever wrote the reviews. None of them can be spotted by reading carefully, only by going and checking, and once one is published, three sites copy it and it becomes consensus.
Thirteen of the fourteen records in the fact base carry at least one refusal, which is a way of saying that every book anybody writes about in this market has something wrong being said about it right now.
The kinds of wrong
- A review score from the wrong country
- The largest single class, and the most mechanical. A brand has one review profile for its global site and none for its Ghanaian one, or one with a handful of entries. The global figure gets quoted as a Ghanaian verdict, and the reviewers behind it are writing about products in entirely different markets. Several of these profiles rest on so few entries that the number carries no information at all, and this index files no review score for any book on the register. 32 refused figures of this kind, across 13 books.
- A cashier figure carried over from another country
- The most dangerous class, because a reader acts on it. Minimum and maximum withdrawals for several of these brands trace back to a help page written for a different market with only the currency swapped, and in one case the operator’s own Ghana path serves another country’s figures. 17 refused figures of this kind, across 11 books.
- Everything else the record refuses
- Claims that do not fall into the classes above: a scraped sentence belonging to a different brand on the same page, furniture read off a block page and mistaken for a product, and figures quoted from a source that turns out to be owned by the company it is describing. 12 refused figures of this kind, across 9 books.
- A margin computed on a flattering sample
- A figure from one top league quoted as a book’s margin. Books price the major European competitions several times sharper than everything else, so a top-flight number makes any book look like the best value in the market, and its all-football figure is usually multiples of it. 10 refused figures of this kind, across 9 books.
- An app figure welded to the wrong platform
- A rating from one operating system beside a download size from the other, or a rating from a neighbouring country presented as local. Store ratings are country-scoped and the same application reads differently a border away, so this is not a rounding error, it is two different quantities under one label. 10 refused figures of this kind, across 8 books.
- A catalogue size nobody counted
- A number of games taken from a platform’s worldwide figure, from a marketing page, or from counting the entries in a machine-readable file that lists the same game once per studio. None of them is a count of what a Ghanaian player can open. 8 refused figures of this kind, across 8 books.
- An allegation about a named person’s money
- Specific accusations that a company took or withheld somebody’s funds, none of them traceable to a regulator action or a named news report, and several of them belonging to a different company with a similar name or to a dispute in another country. Repeating one of these would be defamatory, and the records say so in those words. 6 refused figures of this kind, across 5 books.
- A support channel that reaches another country
- Email addresses and phone numbers presented as Ghanaian support that belong to another market’s desk, to a brand-level mailbox, or to an affiliate programme rather than to players. Somebody following one of these in a bad hour reaches the wrong people. 6 refused figures of this kind, across 4 books.
- A permit number nobody can check
- The Gaming Commission publishes no permit numbers, so every number in circulation is second-hand. Where several versions of a number exist for one brand and nothing corroborates any of them, the record holds all of them back rather than picking the most popular. 5 refused figures of this kind, across 5 books.
- A founding date or a market exit that did not happen
- Brand histories pasted from one company onto another, Ghana entry years that predate the parent company’s incorporation, and at least one claim that a bookmaker has left the market when its site was trading normally on the day the record was compiled. 4 refused figures of this kind, across 4 books.
Where the most is being got wrong
Counted per book, and the count says more about how much attention a brand gets than about how it behaves. The books with the longest refusal lists are the ones enough people write about for the errors to accumulate.
The one refusal that contradicts our own file
One record carries two permit numbers in the field that holds permit numbers and then rules both of them out in the field that lists refusals. Both statements are on the same record and they cannot both be acted on.
This index resolves it towards the refusal, every time. A number held and not printed costs a reader a line of a page; a number printed against a real company on the strength of a source that contradicts itself costs that company something. The register page says which entry it is and shows the shape of the hole rather than filling it.
How to check one of these yourself
Three habits catch most of it. Ask which currency the figure was originally written in, because a limit that reads oddly round in cedis usually started life in another currency. Ask how many independent sources carry it, and then check whether they are independent, because in this market four sites agreeing often means one site quoted four times.
Then ask whether the operator says it. A figure the bookmaker publishes about itself can be held against the bookmaker. A figure only third parties carry is a claim about a company by people with no obligation to it, and this index labels every offer figure with which of the two it is. That distinction is what provenance means here.
Questions this shelf gets asked
- Why not just publish the wrong figures with a correction next to them?
- Because the correction does not travel and the figure does. A number printed on a page gets copied, quoted and indexed regardless of the sentence around it, and a reader arriving through a search box may see only the number. Counting the refusals without reproducing them keeps the warning and drops the payload.
- Does a long refusal list mean a bookmaker is untrustworthy?
- No, and reading it that way would be exactly backwards. The list counts errors other people have made about the company. The brands with the longest lists are the biggest, because they are written about most, and the brand with no refusals at all on its record is the one nobody outside the operator has ever covered.
- Where do most of these wrong numbers come from?
- From other markets, mostly. Several of these bookmakers run in half a dozen African countries from one platform, and a help page or an app listing written for one of them gets read as though it described Ghana. The tell is usually a figure that is suspiciously round once converted, or a support contact with the wrong country’s dialling prefix.