How the Casino.net Algo scores a casino
The Algo is the data input behind every rating we publish.
Twenty-six criteria read more than 100 data points, and none of them asks anyone’s opinion. Pick any casino we have reviewed and watch its score assemble, category by category, from the same records the review uses.
26
criteria
7
categories
236
casinos scored
15
markets
v3.0
algo version
10/10
max score
Every casino review on Casino.net opens with a score out of 10. Part of that number is judgement, and part of it is arithmetic. The Casino.net Algo is the arithmetic: 26 criteria, each with a fixed ceiling, each read from a record in our operator database rather than from an opinion about the brand.
This page sets out the method itself. The criteria below are the ones we run, the weights are the ones we apply, and the demo draws on the records behind every review. Where a criterion does not yet do its job, the page says so.
We only ever compare a casino with the others we rate in the same market. A score is a position among local competitors, not a global verdict. It is also why the same brand can score differently in two countries: different licence, different currency, different games, different peers.
Watch a score being built
Choose a market, then a casino. Seven categories fill in turn, each capped independently, each computed from operator records against that market’s pool of rated casinos. Open a category to see every criterion and what the casino claimed or lost on each one.
Choose a casino above
189 casinos, 11 marketsScored against Algo v3.0
Pick a casino and the Algo builds its score one category at a time.
Choose a casino: the demo uses a stored snapshot and rebuilds its score category by category. Rank and distribution use the published reviews in the selected market. The card links to the full review.
Seven categories, ten points
Each category has its own ceiling. No single strength can carry a casino. Safety and responsible gambling is worth a quarter of the score on its own, and a casino that fails there cannot buy the points back with a bonus.
Read a row across: what the criterion is worth, what a casino typically takes on it, what it leaves behind, and the markets where casinos most and least often take it in full. A row with nothing unclaimed is not separating anyone. That is how we decide what to rework next. A market needs five rated casinos before we name it.
Where scores actually land
If every casino scored nine or more, the score would tell you nothing. We stretch the index deliberately: most casinos land in the middle bands and the top of the range stays hard to reach. The chart follows the casino you picked above, plotted against the rest of its market.
In a market where we rate ten casinos the bands are coarser than in one where we rate a hundred. The pool size is always on the chart.
Those counts cover all 236 casinos we rate and move as reviews publish. The chart above narrows to one market at a time, from the demo’s stored snapshot.
Every criterion and its ceiling
Twenty-six criteria apply to any one casino. The flagship payment row changes with the market. The table lists all of its variants.
| Criterion | Category | Max | How we measure it |
|---|---|---|---|
| Geo licenceBy policy | Safety | 1.5 | A licence that is valid for this market under its own law: locally issued where the country requires one, a recognised international licence where it does not. We do not list a casino without one. |
| Responsible gamblingUnder revision | Safety | 0.5 | Deposit limits, self exclusion and reality checks. Awarded in full wherever the data exists. |
| Safety technologyUnder revision | Safety | 0.5 | Encryption and data protection records. Points follow the fields we hold, not a verified certificate. |
| Number of games | Games | 0.5 | Total game count, ranked against every casino we rate in the same market. |
| Number of slots | Games | 0.5 | Slot count, ranked against the same market pool. Libraries under one hundred slots score the bottom band. |
| Live games | Games | 0.3 | Full points for any live dealer tables on record, none for a casino without them. |
| Software providers | Games | 0.2 | RNG and live studios counted together, then ranked against the market pool. |
| Overall RTP | RTP | 0.8 | The published payout ratio, ranked against the market. Casinos publishing the same figure share a middle rank. |
| Slots RTP | RTP | 0.7 | Average RTP of the slots we hold for that casino, ranked against the market. |
| Welcome bonus value | Bonus | 0.7 | The maximum first deposit bonus, ranked against the same market and read in local currency. |
| Wagering requirement | Bonus | 0.6 | Average wagering across the bonus types on record. A lower requirement scores higher. |
| Bonus minimum deposit | Bonus | 0.2 | The qualifying deposit measured against that market’s own bar, for example ten pounds in the UK or one hundred pesos in Mexico. |
| Payment methods | Payment | 0.2 | The count of distinct deposit and withdrawal methods on record. |
| Fastest withdrawal | Payment | 0.3 | Full points under twenty four hours, partial up to three days, none beyond that. |
| Minimum deposit | Payment | 0.1 | Full points for a minimum deposit at or under that market’s bar. |
| Maximum deposit | Payment | 0.1 | Full points for a deposit ceiling at or above the market median. |
| Cashout limits | Payment | 0.1 | The recorded withdrawal cap against the market median. Unlimited withdrawals score in full, a cap below the median scores nothing. |
| CurrenciesBy policy | Payment | 0.1 | The local currency. Awarded to every casino: we do not rate casinos that will not take it. |
| PayPal | Payment | 0.1 | The market’s dominant wallet, checked in the UK, Ireland, Malta, Spain, Italy and the United States. |
| Interac | Payment | 0.1 | The Canadian flagship method, checked in Canada and Ontario in place of PayPal. |
| OXXO | Payment | 0.1 | The Mexican flagship method, checked in Mexico in place of PayPal. |
| M-Pesa | Payment | 0.1 | The Kenyan flagship method, checked in Kenya in place of PayPal. |
| Crypto | Payment | 0.1 | The flagship method in Malaysia, Nigeria and El Salvador, covering the coins those markets actually use. |
| Support availability | User experience | 0.3 | Full points for round the clock support, half for set daily hours. |
| Support methods | User experience | 0.3 | Full points for chat, phone and email together, partial for some of them. |
| Compatible devicesUnder revision | User experience | 0.2 | Mobile and desktop coverage. Awarded in full while the device records are built out. |
| Languages | User experience | 0.2 | Site languages: three or more scores in full, two scores half. |
| Casino awards | Reputation | 0.4 | Industry awards on record, with two or more scoring in full. |
| Other products | Reputation | 0.4 | Sport, poker, bingo or lotto offered beside the casino. |
| Public company | Reputation | 0.2 | A publicly listed parent, which means published accounts and filings anyone can read. |
What changes from market to market
The criteria are the same everywhere. What varies is the pool we rank a casino against, the currency we read the money criteria in, and which payment method counts as the local flagship.
| Market | What changes | Why |
|---|---|---|
| Ontario | No bonus category at all. The Algo redistributes its 1.5 points across the remaining six categories. | Ontario’s rules prohibit advertising casino bonuses, so scoring them would reward something we cannot show. |
| Canada and Ontario | Interac replaces PayPal as the flagship payment row. | Interac is the method Canadian players default to. PayPal presence would separate operators better, but it would not describe the market. |
| Mexico | OXXO replaces PayPal, and we read the deposit bars in pesos, not converted dollars. | A one hundred peso minimum deposit is a different proposition from a ten dollar one, and cash at OXXO is how a large share of Mexican players pay. |
| Kenya | M-Pesa replaces PayPal. | Mobile money is the default payment rail, so a casino without it is unusable locally. |
| Malaysia and Nigeria | Crypto is the flagship method row. | Card and wallet coverage is thin in both markets, and crypto is what the licensed operators support. |
| El Salvador | Crypto is the flagship method row, and the money criteria run in dollars. | The country banks in US dollars, so the deposit bars need no conversion. Unlike the flagship rows in Canada, Kenya and Mexico, this one still separates operators. |
| United States | We read records state by state. | The Algo scores a casino licensed in New Jersey against New Jersey figures rather than a national average that describes nobody. |
| Every market | Rankings, medians and deposit bars use the pool of casinos we rate there. | In practice a brand operating in two countries is two casinos, judged against two sets of rivals. |
What the Algo is not
The Algo answers one narrow question: what the records say about a casino. It leaves out four things, and the review picks up most of them elsewhere.
Not advice to play
A high score means a casino performs well against measurable criteria, not that gambling there is a good idea.
Not the whole rating
The Algo is one input into the score at the top of a review, alongside our expert assessment, reader votes and third party review platforms.
Not for sale
We record commercial relationships under how we make money, and they never touch a criterion, a weight or a pool.
Not a measure of feel
The expert score covers what a casino is like to use, and we deliberately keep that out of the arithmetic.
Where the numbers come from
Because every criterion reads the same operator database that fills the rest of a review, the score and the tables cannot disagree.
Our own operator records
We enter licences, game counts, payout ratios, bonus terms, limits and support hours against a named field per market, rather than scraping them off the casino’s own page.
Nothing is stored as a score
The Algo recomputes the table each time a review loads, so a corrected record shows up in the score immediately.
An empty field is never a pass
Where we hold no record at all, the criterion scores nothing and the review says so.
The Algo and the score at the top of the review
The number in a review header is not the Algo alone. It combines four inputs, and the Algo carries 30 per cent of it. A casino can be strong on paper and still sit lower once we count experience and reputation.
30%
The Algo
26 criteria read from operator records, reproducible from published data.
30%
Expert assessment
A named reviewer’s judgement of how the casino performs in use.
30%
Reader votes
Ratings left by visitors on the review itself.
10%
Third party platforms
Ratings collected from public review platforms in that market.
Together the three human inputs outweigh the arithmetic by more than two to one.
The other three inputs, and the way the four combine into the published score, are set out in How We Rate.
Until a review has collected reader votes the weighting runs 40 / 40 / 20 across the Algo, the expert assessment and third party platforms.
Where no third party rating exists for a market, we renormalise across the inputs that remain rather than substitute a number, so a missing source cannot drag a score down.
Who maintains it, and how to challenge it
One person answers for the criteria and the weights, and no version reaches a review unsigned.
Katya VasilevaHead of Product
Owns the 26 criteria, the category ceilings and the weighting, and decides when a criterion needs rework.
If a figure about your casino is wrong, send the record and the market. We log any correction that moves a published score.
Frequently asked questions
Can a casino pay for a better score?+
Does publishing the criteria let casinos work the system?+
Why does one brand score differently in two countries?+
Why does nobody score ten?+
How often does a score change?+
Do you rate casinos you have no commercial relationship with?+
Every change, on the record
A change to the criteria bumps the major version, a change to weighting or thresholds bumps the minor one, and we log a change to the data behind a criterion without a bump. Entries that moved published scores say so, including the ones that moved them down.
Version 3.0 has run since 25 August 2026. The rule above is how we version today, and a few entries below carry tags from before we applied it consistently. We have left those as they stand.
Wager free offers stopped scoring at the bottom
The wagering row reads game contribution percentages, so an offer with none at all recorded zero contributions and landed in the worst band. Those offers now take the top band outright.
Cashout limits now measured
The row used to award its point to everyone and state that the casino had no fixed limits, which was untrue outside the UK. It now compares the recorded withdrawal cap with the market median.
Fifteen published scores fell by 0.1
State level records for the United States
Scoring switched to per state records, so a New Jersey casino competes with New Jersey rivals.
Flagship payment method per market, and deposit bars in local currency
We stopped measuring Malaysia, Kenya, Mexico and Nigeria against PayPal and against a dollar bar.
Currencies criterion states what it does
Checking whether a currency symbol existed could never fail. We now award the point openly, and the table lists the row as by policy.
Web ratings fallback and stale score guard
Markets without local review platform data borrow through a documented chain, and a review with no source drops the web component instead of carrying an old number.
A handful of overall scores moved by up to 0.4
Ontario scores without bonuses
We removed the bonus category for Ontario, where the local rules prohibit advertising bonuses, and redistributed its points. Ontario casinos also rank against an Ontario pool.
Local money
Every money criterion evaluates in the market’s own currency.
Fair ranking for tied RTP
Casinos publishing identical payout ratios share a middle rank instead of all falling to the bottom band.
RTP scores rose in every market
Same market comparisons
Five criteria that had been scoring zero for every casino outside the UK started scoring after a defect fix, and every ranking became market scoped.
All published scores were recomputed
Launch with the UK index
Criteria across seven categories, computed from operator records and ranked against the UK market. For the first year the index was UK only, so market scoping had nothing to do yet.