Atlas data

How the Bookmaker Atlas is built

Every fact on every leaf traces back to a public source you can verify yourself. Trust scores, licence badges, and country access are computed from that data, not invented to sell signups.

1,806 bookmakers tracked50+ public data sources6 score components2026-06-10 last verified

From public source to score

  1. A public source publishes

    A regulator register entry, corporate filing, enforcement bulletin, or security scan goes live.

  2. Atlas ingests and normalises

    The datapoint is stored with its URL, timestamp, and publisher. Nothing operator-supplied.

  3. Six components score independently

    Licensing, corporate, reputation, operational, security, and threat intelligence each score 0 to 100.

  4. One weighted number publishes

    Components combine by fixed weights; red flags cap severe cases. Recomputed monthly.

The full weighting and cap rules live on the methodology page.

Where the data comes from

Every datapoint that drives a trust score is pulled from a public source you can check yourself. No private databases, no operator-supplied data, no language-model-written prose.

  • Regulator registers

    Every market we track has its own official licensee register. Licence status, scope, suspensions, and revocations come straight from the regulator's public-facing register.

  • Corporate ownership

    Public corporate-registration databases for the licensee company, the parent group, and the ultimate-owner chain. Sanctions exposure pulled from the same public databases regulators use.

  • Enforcement bulletins

    Public fine notices and licence-action publications from each regulator. Linked to the original press release, not a third-party summary.

  • Security posture

    Independent web-security scans, transport-layer posture, certificate-transparency activity, hosting-tier signals, and infrastructure-exposure checks.

  • Public reputation

    Aggregate public-reputation and complaint-outcome signals plus app-store ratings. Weighted by sample size so a 50-review brand does not outweigh a 5,000-review brand.

  • Threat intelligence

    Independent phishing, malware, sanctions, anti-spam-zone, and vulnerability feeds. Anything that flags the brand or its infrastructure as compromised feeds the score.

Each leaf links the specific source for every claim it makes. Sparse data is shown sparse: when a brand has no public reviews, the page says so. Nothing is invented to fill space.

Trust-score weights

Every bookmaker scores 0 to 100 across six components. The component weights are fixed by the methodology, not editorial preference.

  • Licensing

    30%

    How strong the strongest active licence is, plus breadth across regulators. Top-tier markets like the UK and Malta score near maximum; light-touch offshore licences score low; no live licence scores zero and caps the overall score.

  • Corporate transparency

    20%

    Whether the legal entity behind the brand is identifiable: disclosed parent, verifiable registration, traceable ownership chain. Hidden or dead-end ownership lowers it.

  • Reputation

    15%

    Aggregate public reputation, weighted by sample size so a 50-review brand does not outweigh a 5,000-review brand, plus app-store ratings on verified listings.

  • Operational

    15%

    Brand longevity and reach: domain age, real-world popularity rank, hosting tier, accepted-country count, and archived activity over time.

  • Security

    10%

    Independent technical scans: web-security grades, transport-layer protection, DNS hygiene, and infrastructure resilience.

  • Threat intelligence

    10%

    Independent phishing, malware, sanctions, anti-spam-zone, and vulnerability feeds, plus parsed regulator enforcement actions.

Computed from facts, not invented to sell signups

Most “best bookmaker” sites are paid placements dressed up as editorial. This is the opposite: every score is the deterministic output of public inputs you can check yourself.

What a score actually means

A score of 85 from this atlas does not mean a human spent two hours auditing the bookmaker. It means the inputs (licence strength, corporate transparency, public reputation, operational track record, security posture, threat intelligence) yielded 85. The same code runs on every brand.

When you should look deeper

For high-stakes decisions (deposits above $1,000, registering as a matched bettor at a new book, moving funds across jurisdictions), treat the atlas as a starting point. Read the source citations linked on each leaf page.

Worked examples

Three brands at different points on the score range. The inputs that drove each result.

Refresh cadence

  • Monthly: full trust-score recompute across every brand.
  • Weekly: regulator-register refresh, enforcement actions, public reputation trends.
  • Daily: security, DNS hygiene, threat-intelligence checks.
  • On change: corporate ownership updates when public corporate-registration databases publish a change.

Last full recompute: 2026-06-10.

Author

The atlas is authored and maintained by Erik Andersson, Content & Marketing Specialist at Sharkbetting. Erik also writes the Sharkbetting blog.

Every component of the score is documented in full on the methodology page. Corrections and source disputes go through our contact page.

Frequently asked questions

What public sources does the Atlas pull from?

Local regulator registers (the official licence database for each market), corporate-ownership and sanctions databases, enforcement bulletins from each regulator, independent web-security scanners, app-store ratings, public reputation aggregators, and threat-intelligence feeds. Every claim on a leaf links to the specific source that backs it.

Is this a manual review or an automated score?

Automated, computed from public sources. A score of 85 does not mean a human spent two hours auditing the bookmaker. It means the inputs (licence strength, corporate transparency, public reputation, operational track record, security posture, threat intelligence) yielded 85 under the same code that runs on every brand. For high-stakes decisions, treat the score as a starting point and read the source citations on the leaf page.

How often is the data refreshed?

Trust scores are recomputed monthly across every brand. The underlying feeds refresh on their own cadence: threat-intelligence signals weekly, licensing and enforcement data weekly to quarterly, public-reputation signals monthly, security scans monthly to quarterly, and corporate-ownership records quarterly.

What if a bookmaker has sparse data?

Sparse data is shown sparse. When a brand has no public reviews, no security audit, or no regulator registration, the page says so explicitly with the audit trail. Nothing is invented or interpolated to fill gaps.

See also: the editorial policy on independence and corrections, and the full Trust Score methodology.

Go deeper or start browsing

The full methodology walks through the scoring math. The alternatives hub groups every brand by parent operator.

Full methodologyBrowse by parent group