The GB build: operator-selectable RTP, assessed

A review usually pins a specimen. This one pins a label, because in the GB catalogue the label has become the variable part. A slot with one name and one thumbnail can arrive at two GB-licensed sites as two different machines, and what differs is the return-to-player configuration the operator licensed. Not a secret, not an allegation, not rigging: disclosed practice with a disclosure problem, and the regime is under the glass.

The label, as it stands

Operator-selectable RTP — the specimen

What varies	The RTP configuration; the operator selects which build it licenses
Cause cited by the trade	Remote gaming duty at 40% on or after 1 April 2026, per GOV.UK
Reported shift	Roughly 96% toward 94%, 92% named as the requested maximum — G Games' Helen Walton, quoted by iGaming Business
Where it is disclosed	The paytable or game-information panel where you play
Where it is rarely surfaced	The lobby tile — RTS 3C requires the figure to be easily available before a customer commits to gamble, but does not prescribe the tile
Evidence of divergence	An analysis of 20 titles across UK-facing operators, per IBTimes UK
Assessed here	The disclosure system, not any operator or game

One name, two machines

Studios have shipped multiple RTP configurations of the same title for years. Our Big Bass Bonanza entry carries the standing footnote: the headline figure is the best build, and the version a given site licenses may publish a lower one. Taxonomically this is awkward: a species is defined by its mechanics, but the maths underneath is a variable the taxonomy does not capture. For a decade the footnote was a curiosity. This year it acquired a date.

What April changed

GOV.UK’s betting and gaming duty guidance gives the remote gaming duty rate as 40% on or after 1 April 2026, up from the 21% that had applied since 1 April 2019. Duty lands on the operator, who has a lever the player does not see.

Trade reporting says the lever was pulled. iGaming Business reported in April the assessment of G Games co-founder Helen Walton, quoted saying standard RTPs have moved from around 96% to 94% and that 92% is now the requested maximum. That is one named supplier-side account of prevailing norms, not a measured survey, and it describes requests made of studios rather than any rule. Hedge accordingly.

Separately, IBTimes UK reported an analysis tracking RTP across 20 popular titles at UK-facing operators after the increase: operators had selected differing configurations of the same games, with Hollywoodbets UK reported as holding the highest average. Worth knowing, and worth discounting — the report does not name the firm that conducted it, a real gap for the market’s most-cited comparison. Hollywoodbets’ Don Barker, quoted in the same iGaming Business piece, says the company does not believe that cost should be passed on to players — a stated position, not a recommendation.

What a lower build costs

A published RTP is not an abstraction. Walton, in the same piece, is quoted saying a 4% reduction in RTP materially shortens play time and reduces the likelihood of experiencing bonus features — four points being the 96-to-92 span, not the two-point move to 94%. The number promises returns to nobody, but it does set what an evening at a given site tends to buy. Two builds of one game are the same show at different ticket prices, and one panel tells you which ticket you bought.

Where to look, and what you will not find

The configuration is published. Open the game, find the information panel — the “i” — and the figure for that build sits in it, at that site, on that day. It is the only authoritative copy; a number quoted anywhere else may already have been superseded.

What you will not find is any of it in the lobby. Tile, name and thumbnail are identical across builds. The Gambling Commission’s technical standards do reach this ground: RTS 3C requires the information a customer needs to judge their chances of winning — the return-to-player percentage among the items it lists — to be easily available before they commit to gamble, and the implementation guidance gives game selection pages and menus as one example of where it should sit. What the guidance does not do is prescribe the tile, and the tile is where the choice is made. A disclosure that is complete, accurate, per-game and functionally invisible is a peculiar sort of disclosure.

None of this is a reason to go shopping for a higher build. Fix the evening’s budget and the session length before the first spin, and treat the paytable as a label, not a lever.

Both columns of the ledger

The catalogue entry

Pros

The figure genuinely is published, per game and per site, checkable before a first spin
Multiple configurations are long-standing disclosed practice, not concealment
The cause is dated and documented: GOV.UK gives the 40% duty rate from 1 April 2026
At least one operator has publicly argued against passing the cost to players, so the number can be competed on

Cons

RTS 3C requires the figure to be easily available before a customer commits to gamble, but nothing prescribes the lobby tile — so identical branding fronts different maths
The figure lives behind the game's information panel, not at the point of choice; disclosure that must be hunted for barely functions
Configurations can change without notice: a figure verified in April is not one guaranteed in August
The most-cited comparison of operator builds — the 20-title analysis IBTimes UK reported — names no analysing firm
A lower build costs session length and feature frequency, per the G Games quote

Verdict

A museum that changed a specimen’s label every few months while leaving the old photograph on the door would not be accused of dishonesty. It would be accused of bad curation. Open the “i” panel first, and treat every third-party RTP table as a rumour until the paytable confirms it.