Thirty-seven names, and thirty-seven is the median of this table
Eleven lists, counted and compared
Eleven operators on this site publish a restricted-territory list that could be read. The lengths run from 1 name to 74, and the middle value of the eleven is 37 — which is exactly the number in the clause this site was built around.
A restricted-territory list is the one operator disclosure that every comparison site prints and nobody measures. It appears as a number in a column, longer looks stricter, and the reader is left to guess what stricter means.
Eleven rows on this site publish a list that could be read. Their lengths are 1, 15, 30, 34, 36, 37, 42, 45, 47, 48 and 74.
The middle value of those eleven is 37, which is the length of the clause this site is named after.
What does a list length actually measure?
Drafting, mostly. The two extremes on this table make that plain.
At one end sits a 74-name list. At the other sits a row whose restricted list, as read, names a single country. Between them sit nine operators selling substantially the same product to substantially the same market, and the spread between the shortest and the longest is a factor of seventy-four.
A difference that large across near-identical businesses is not a difference in commercial appetite. It is a difference in how much of that appetite each operator chose to write down.
Which is why the length of a list is a fact about a document rather than a fact about a market.
The subject operator's own clause makes the point from the inside. Its first sentence bars anyone residing in, or temporarily located in, any jurisdiction where online gambling is restricted or prohibited, and names nobody. Thirty-seven names follow that sentence, but the sentence reaches further than they do, and a reader who scans the list for their own country has answered half the clause. The whole thing is printed word for word on the territory clause.
Do these operators agree about anything?
Less than the overlap in their business models would suggest.
Ten of the eleven lists are held here as country codes, which makes them comparable. Across those ten, not one country appears on every list.
North Korea and the United States appear on nine of the ten. Cuba, Curaçao, Iraq and Iran appear on eight. Lithuania and Syria on seven. Then the United Kingdom, Israel, the Netherlands, Sudan, Ukraine, Yemen and Zimbabwe on six, and after that the counts fall away into ones and twos.
Two clusters explain most of what agreement there is. One is the sanctions map, which is why the same handful of names recurs. The other is the map of countries whose own regulators issue permits and police the market, which is why several European states appear repeatedly on lists drafted by operators licensed elsewhere.
Everything outside those two clusters is an individual decision, and individual decisions do not line up.
How much does the subject's list have in common with its neighbours?
Matching the thirty-seven names in clause 3.4 by hand against the coded lists of the other rows gives an overlap that is smaller than the headline lengths imply.
The closest neighbours are Instant Casino, sharing 22 of its 48 names, and Wild.io, sharing 21 of its 45. In the middle sit CryptoLeo at 18 of 74 and Playbet.io at 15 of 30. At the far end, Bets.io publishes a 42-name list that shares only 7 names with the 37 — two long lists, drawn up separately, agreeing on a sixth of their contents.
Two operators can exclude forty countries each and be excluding different countries.
That is the sentence a column of list-lengths hides. Where a specific country matters to a specific reader, no summary count answers the question, and the only reliable procedure is to open the clause and search it for one name. The forty-five-name list belonging to the closest of these neighbours is compared with the thirty-seven clause by clause on the Wild.io comparison.
Which direction do these lists move?
Up, on the only evidence available here.
The equivalent passage in the copy captured on 24 April 2025 named 24 places. The copy captured on 9 August 2026 names 37. Thirteen names joined and zero left across those fifteen months, and eight of the thirteen are European Union member states.
Neither copy carries an undertaking to notify an existing account holder that the list has changed. Both files exist because a public archive took them, not because anybody announced anything, and the same is true of every other list on this table: what is held here is a reading taken on a date, not a subscription to an operator's amendments.
An account opened while a country was absent from a list can continue under a document that names it later. That is the practical exposure, and it has nothing to do with how long the list was on the day of the deposit. The full comparison between the two dated copies is on the two editions.
What the clause settles, and what it does not touch
A restricted-territory clause answers exactly one question: whom the operator says it declines to serve.
It carries no authority over the law where a reader lives, and it makes no statement about it. Those are separate questions with separate sources, and a contract is the wrong document to ask. Where the law of a reader's country stands on an offshore account is a matter for someone working from statute, and this page declines to convert a commercial clause into a legal opinion in either direction.
The consequence written into the subject's clause is worth quoting for what it covers. It states that an account may be restricted where access is attempted from one of the named jurisdictions, and that it is solely the reader's responsibility to ensure their use of the site complies with the laws of their country of residence or location.
Responsibility assigned, in a document the reader had no part in drafting.
How the two captures were obtained, and five places where this reading is weak, are set out on the reading rules. The eleven lists and everything else each operator publishes about leaving sit together on the front page.