Thirty-three supplier lists, read 26 August 2026
What a provider count measures once the lists are laid on top of each other
Thirty-three operators publish the studios behind their lobbies. Those lists hold 1,997 entries between them and 454 different companies, so the average studio appears on more than four lists. Two rows compared here, Roobet and BetFury, share 35 suppliers outright.
Catalogue size is the number every casino puts on its front page and the number this index refuses to rank on. The reason becomes obvious as soon as two of the lists are laid on top of each other.
What is on record, and what is empty
Thirty-three of the hundred operators in this library publish the studios supplying their lobbies, and the lists were read on 26 August 2026. Sixty-seven publish none that has been opened.
For Stake.com the cell is empty. The four documents read for it — the terms, clause 4.8, the territorial clause and the register entry for permit OGL/2024/1451/0918 — are contracts and a register entry, and none of them describes a lobby.
Four of the twelve rows compared here carry a figure. BetFury names 73 studios, Roobet 64, TrustDice 43, and Vave was recorded on 25 August 2026 with roughly 2,500 games from 40 studios.
Eight rows carry nothing, which records the state of this reading rather than the state of those lobbies.
Nineteen hundred entries, four hundred companies
Put all 33 lists together and the arithmetic is the whole argument.
The lists hold 1,997 entries. They contain 454 distinct company names. So the average studio appears on more than four operators' lists, and the busiest names appear on most of them.
| Studio | Lists naming it, of 33 |
|---|---|
| Endorphina | 26 |
| Pragmatic Play | 26 |
| Hacksaw Gaming | 25 |
| PG Soft | 25 |
| BGaming | 24 |
| GameArt | 24 |
| AvatarUX | 23 |
| NetEnt | 23 |
| Evolution | 22 |
| No Limit City | 22 |
A property that four fifths of the readable field shares is not a property that separates anybody. Whichever of these operators a reader opens, the games on the first screen were bought from the same short row of companies.
The lists themselves run from 12 studios to 150, with a median of 47.
Two rows from this table, laid on top of each other
The general figure is convincing. The specific one is harder to argue with.
Roobet publishes 64 studios and BetFury publishes 73. Thirty-five names appear on both lists — 55% of Roobet's and 48% of BetFury's — and the two lobbies between them are supplied by 102 distinct companies rather than 137.
TrustDice, at 43, shares 25 studios with BetFury and 20 with Roobet. Its overlap with BetFury is 58% of its own list.
Three operators, three brands, three sets of terms, and a supply base that is largely one supply base.
That is the honest ceiling on what a catalogue comparison can deliver. Two lobbies assembled from the same suppliers will feel different in colour and layout and will hold substantially the same games, and no count of studios detects that. Only the overlap does, and the overlap can only be computed where both lists have been read — which is 33 operators out of a hundred.
Why a title count is worse than a studio count
The number operators advertise most often is not the studio count but the game count, and it is the weaker of the two by some distance.
A studio count has a natural ceiling: there are 454 distinct suppliers across all 33 published lists, so no operator can claim more studios than exist. A title count has no ceiling at all. It counts regional variants of the same game, it counts a jackpot version separately from the base version, and it grows every time a supplier ships an update — none of which changes what a reader can play.
One row here carries a title count: roughly 2,500 games from 40 studios, recorded on 25 August 2026. Divide one by the other and the average supplier contributes about 62 titles, which is a plausible figure and also a demonstration of how quickly the larger number is manufactured out of the smaller one. Sign five more studios on the same average and the headline moves by 300.
That is why the studio figure is the one printed on this index, and the title figure appears only where an operator published it beside the studios rather than instead of them.
What the count cannot be checked against
A permit number can be typed into a register kept by somebody other than the operator. A territorial list can be quoted back from a page the operator publishes at a stated address.
A studio count has no such second source. There is no register of supply contracts, no authority that records which studio serves which brand, and no obligation on anybody to publish either. The list is the operator describing itself, and if it is edited between two visits nothing anywhere records that it changed.
A reader can do one useful thing with it, and it takes a few minutes: open the lobby's provider filter and count the names. That produces a figure from the same source this index used, on the day it is done, which is the only audit available.
Why none of this enters a score
The counts above are printed as descriptions and enter no ranking on this site, for three reasons that stack.
A catalogue is rented rather than built, so it changes with a contract and not with a business. It is unauditable, so a figure cannot be held against anybody. And it is shared, so it fails at the one job a comparison needs it for.
What the index scores instead is the material that takes documents to establish: the list of countries an operator excludes in its own terms, the authority that issued its permit, the company that permit names, and the clause that decides when documents are demanded. Those are slow, awkward and checkable. Which lobby carries which slot is none of the three.
The one part of a lobby that cannot be bought in is the set of games an operator writes itself, and what this reading holds on those is set out on the in-house games. The twelve rows with the counts each was measured on are on the front page.