How We Price Casinos: The Weighted Model Behind Every Score
Catalogue checked:
Every number in this catalogue answers three questions and nothing else: what the price buys, whether the till works, and who is standing behind the counter. Each gets a mark out of 10; the three are weighted 50, 30 and 20; the weighted sum is the score printed on the pricetag. No advertiser input, no popularity allowance, no rounding a favourite upward.
What are the three questions and their weights?
What the price buys carries half the score, because that is what a catalogue is for: the bonus, its wagering conditions, how plainly they are printed, and the range you actually get to play. Whether the till works takes 30% — deposits in, winnings out. Who is behind the counter takes the remaining 20%: the company, its licence entry, its service channels.
| Question | Weight | What sits inside it | Why this much |
|---|---|---|---|
| What does the price buy? | 50% | Bonus size, the wagering multiplier and its clock, how legibly the conditions are printed, and the assortment of games and markets the money reaches | Half the score, because a catalogue answers one question first — what do I get, and what does it cost me. Plain printing is rewarded; conditions that must be assembled from scattered documents are marked down even when the underlying offer is fair |
| Does the till work? | 30% | Deposit rails that function from South Africa, entry thresholds, currency handling, withdrawal behaviour and the public record of contested payouts | An offer is worth exactly what you can carry out of the shop. This line is where official windows, observed behaviour and recurring complaints all land |
| Who is behind the counter? | 20% | The operating company, whether its licence entry can be traced to a register, how long the name has been trading, and the service channels available when something goes wrong | The smallest weight and the one that decides ties. It matters most in the minority of transactions that go wrong — and a label nobody can trace is a finding in its own right |
The arithmetic is deliberately dull. A casino marked 8.4 on price, 9.2 at the till and 8.2 behind the counter scores 0.5 × 8.4 + 0.3 × 9.2 + 0.2 × 8.2 = 8.6, and that is the whole calculation — reproducible by any reader with the three sub-scores printed on the pricetag. All fifteen sub-scores are laid out in the comparison table.
What earns a markdown?
Opacity, above all. A licence number we cannot trace to a public register leaves a blank on the label and costs marks behind the counter. Bonus conditions that resist a single-sitting reading are priced as opacity even when the underlying offer is fair. And a contested exit route — money that reaches the account but struggles to leave it — sets the floor of the shelf.
The complaint record moves scores in one direction only: a clean record earns nothing extra, but a recurring public pattern — payout delays, withdrawals reported as never arriving, accounts closed after wins — is priced into the till line, as Melbet's 3.6 and Betwinner's 4.8 show. Trading history cuts honestly both ways: years on the counter support confidence, which is why a brand that opened in 2024 carries that fact on its label even at the top of the shelf.
What does the desk refuse to do?
Four standing refusals: we do not mark up data that is not visible and checkable; we do not review brands we cannot price by the same three questions; we do not run head-to-head duels, because a catalogue compares the whole shelf in one table or not at all; and we do not adjust a score for commercial reasons.
The catalogue may contain referral links, and a purchase through one costs a reader nothing — but the weights above are fixed before any such consideration and are published precisely so they can be held against us. When the shelf changes, the visible date on each page changes with it. That date is the whole freshness promise: no per-fact hedging, one honest timestamp.
Third-party rating averages are deliberately absent from every pricetag here. They are compiled by other people to other rules, they mix casino and sportsbook complaints without separating them, and a number lifted out of somebody else's model cannot be added to ours. Where a public complaint pattern is real, we describe the pattern and price it; we do not import a score to stand in for the reading.
Prices get re-opened, too. Naledi Khumalo runs the amendment log: an operator that disputes its own entry, or a reader who finds a line that no longer matches the operator's own site, gets the pricetag re-checked and, where the facts have moved, re-marked with a fresh date. How that works is on the about page, and the route in is the correction address.