An eighth question, tested and dropped
How BitStarz loyalty levels are worked out, and where that arithmetic is kept
Thirty-seven of the hundred operators advertise a loyalty or VIP programme on their own pages. Zero of the hundred publish the four numbers that would make one checkable, so an eighth box would have come back empty a hundred times out of a hundred.
A loyalty programme is the one part of a casino that is pure arithmetic. Points accrue at a rate, levels cost a number of points, a level returns a percentage. Every term in that sentence is a figure, which ought to make it the easiest thing on the estate to check.
It is the least published.
What the BitStarz documents say about levels
Two documents were opened for this operator and read to the end. The bonus terms carry 15,145 characters and date themselves 25 June 2025. The general terms carry 42,876. Between them they set out a turnover multiple, a bet ceiling, a clock, an abuse rule and a rule about one account per household — all of it numbered, quotable and specific.
Neither document defines a loyalty level.
There is no clause that says what earns a point, none that says how many points a rung costs, none that attaches a rate of return to a rung, and none that says whether points expire. The programme is advertised on the operator's own pages, which is how it was recorded here on 26 August 2026, and the material that decides what it pays was not in the contract.
That is the finding, and it is not a BitStarz peculiarity. It is the shape of the whole pool.
The eighth question that scored zero out of a hundred
The seven questions on this site were chosen because a stranger can settle each one by opening a document and looking. A loyalty programme looked like a candidate for an eighth, so it was tested against the pool before being dropped.
The test asked for four figures:
| what the question asks | operators with it on record |
|---|---|
| the rate at which turnover earns a point | 0 of 100 |
| the points a level costs | 0 of 100 |
| what a level returns, as a rate or an amount | 0 of 100 |
| whether points expire, and when | 0 of 100 |
Four columns, four hundred cells, none of them filled. A question that returns the same answer for every operator measures nothing at all, and a box that nobody can fill is decoration rather than evidence. So the eighth question was dropped, and the reason is printed here rather than left as an omission.
Thirty-seven operators advertise the programme. Not one publishes the sum.
Why the tag was kept even though the box was not
The claim itself is still recorded, because it is a fact about what an operator says. Thirty-seven of the hundred carry a loyalty or VIP line on their own pages, read on 26 August 2026, and BitStarz is one of them. It sits alongside other claims counted the same way: 43 advertise a sportsbook, 47 live dealers, 20 in-house games, 12 instant withdrawals.
None of those counts enters the seven-box tally, and the reason is uniform. A tag records an assertion; a box records a document. Mixing the two would let a marketing line lift a row past an operator that published a clause.
There is a second distortion in the tag figures, and it is large enough that using them without it would be dishonest. A claim can only be recorded where an operator's pages could be opened, and 21 of the hundred would not open at all. So the 37 is a floor and the missing ones are unmeasured.
The same effect shows up in the tally. Operators carrying any tag average 3.94 of the seven boxes; operators carrying none average 1.11. That looks like a finding about marketing and is mostly a finding about which sites answered a request. The narrower version, 4.54 for the 37 with a loyalty claim against 1.87 for the other 63, deserves exactly the same discount.
What a reader loses when the arithmetic is unpublished
A loyalty scheme with unpublished rates is a discount of unknown size. That has three practical consequences, and all three are ordinary rather than sinister.
A rate that is not written down can be changed without anything being breached, because nothing was promised in the first place. A level that took months of turnover to reach can be re-priced upward, and the customer who reached it has no clause to point at. And the value of a rung cannot be compared between two operators at all, which is why no page on this site ranks loyalty programmes and no page ever will on the evidence currently held.
The comparison every reader wants here is not available from documents. It is available from an account, from a year of turnover, and from somebody willing to report the result — which is a different kind of source than this desk uses, gathered by people who did the playing.
Eleven operators of the hundred print the year they were founded. That is the company of facts a loyalty rate keeps: things everybody assumes are published, which turn out to be almost nowhere once somebody goes looking.
What is printed instead
Everything the tally does hold about BitStarz sits in five filled boxes and two blanks: the permit number OGL/2024/165/0185, the holder Gareton B.V., a register entry that resolved Active on 26 August 2026, a schedule of 74 excluded countries and the address it lives at, against no verification figure and no payout ceiling.
A loyalty level would have been an eighth line in that list. It is not there, at BitStarz or anywhere else in the hundred.
Why each of the seven survived the entrance test, and what else was tried and dropped, is set out on the rules of the count. The distribution across the pool, with both totals printed so the arithmetic can be re-run, is on the hundred, counted, and the reading habits behind it are described by the records editor.