Does the paytable match the game?
Symbol values are compared with the in-game screen at two different stakes, so that a value quoted as a multiple of the bet really scales the way the page says it does.
Tobias Wraxall spent eleven years in independent game-testing laboratories, comparing submitted maths documents with the paytable files inside the games themselves. He now applies the same suspicion to every figure published on this site, from RTP profiles to the price of a bonus buy.
Six questions are put to every page, and a page is not signed off until each one has a written answer.
Symbol values are compared with the in-game screen at two different stakes, so that a value quoted as a multiple of the bet really scales the way the page says it does.
A page must say which RTP it quotes and mention that lower profiles may be in use, because the difference between profiles is a real cost for the reader.
Spins per balance, buy costs and percentage gaps are rebuilt from scratch in a fresh sheet, not copied from the draft's own arithmetic.
Stake ranges, caps and buy prices are checked against the demo at the stake shown and against the rules that apply in the market the page addresses.
Hit frequency, volatility and top-prize odds are three separate ideas. Any sentence that lets one stand in for another is rewritten.
Warnings sit next to the claim they qualify, at the same volume, and not in a footer where nobody reads them.
In a testing lab a number is never simply a number. It belongs to a build, a jurisdiction, a bet level and a document version, and an auditor's first question is which of those it comes from. Tobias spent his early career re-running simulations, matching paytable configurations to the maths sheets a studio had submitted, and writing up every place where the two did not quite agree.
That training shaped how he reads gambling content. RTP is a claim about a specific profile, and slots often ship in several. Hit frequency describes how often something pays, not how much, and it says nothing about the odds of the top prize. A buy feature price is a multiple of a stake that changes with the stake. Each of these differences is easy to blur in a sentence, and blurred sentences are where readers get misled.
His role on this site is narrow on purpose. He does not judge whether a game is fun. He decides whether every figure on the page is attached to the right version, whether every derived number survives a fresh calculation, and whether the wording around a risk is as strong as the wording around a reward. Pages are published only after that sign-off.
A number is only as good as the version it belongs to, so every figure on a page has to say which one.
each figure is filed as provider page, in-game info screen, game data or second-hand, and second-hand figures need a stronger reason to stay
every derived number is rebuilt in a new sheet by someone who has not seen the draft's arithmetic
where sources differ, the difference is written down and one figure is chosen with a stated reason
the page goes live together with a short record of what was checked and when
Tobias reads what Imogen Pallister writes with the game's rules screen open beside the draft, and sends back a query for every figure he cannot trace.