How we read
the evidence.
A small publication should make its sources easier to inspect, not ask for a larger leap of faith.
Start with a bounded claim
We prefer primary sources for claims about a studio’s catalogue, a documented platform capability or an institution’s published research. Each atlas record includes a source URL, a country or cross-market scope, a topic, a review date, an evidence type and a caveat. A source is selected for the claim it can actually support.
First-party evidence is not independent confirmation of every statement a company makes. A studio can establish that it presents a named game; that does not establish the game’s audience size or quality. A platform document can establish an available mechanism; it cannot prove that a particular team uses it successfully.
Keep observation and analysis separate
Our articles combine narrow source-backed observations with original editorial frameworks. Proposed worksheets and invented game scenarios are explicitly labelled as examples. We do not present them as fieldwork, interviews or measured outcomes. No country-wide revenue, player or device-share estimates are inferred from the sources in this edition.
The atlas uses three evidence labels: first-party evidence, documented capability and institutional evidence. These describe source type. They are not probability scores or quality rankings. Cross-market entries refer to general guidance and must not be treated as evidence of local adoption.
Compare the questions, not the headline numbers
The atlas compares the same five criteria for two selected contexts. It shows the complete collection independently of the record filters. An empty criterion means we hold no evidence for that question; it is not a finding that the activity does not exist. Connection counts and household percentages are deliberately not ranked or converted into player totals. The South African household record is explicitly a 2024 historical snapshot.
CSV and JSON downloads contain the displayed public records, source links, dates, units and limitations. They contain no search text or personal information. CSV cells are quoted and leading spreadsheet-formula characters are neutralized. No service receives the export. Read our claim-review worksheet for the difference between capability, implementation and outcome.
A review date is not a publication date
“Evidence reviewed” records when we checked a source. “Source published” records a date the source supplies; where no reliable publication date was established, it remains undated. Publication-date filtering excludes undated entries when a date bound is set. Article dates refer to the preparation of this edition.
Sources can change. Reusing an observation for a business decision requires checking the current source and whether its scope still matches the question. An old observation is not made current by displaying today’s date.
Correct the claim and preserve the reason
A substantive correction should identify the affected statement, the replacement evidence and the date of the change. Minor spelling edits do not need to masquerade as a new investigation. If a source no longer supports a statement, that statement should be removed or clearly qualified rather than silently attached to another citation.
This local edition has no monitored public contact channel. No email address or response promise has been invented. Until an actual editorial contact is established, corrections can be supplied through the person who shared this edition. A public release requires a real, maintained route for reader corrections.
Disclose interests where they matter
No sponsored placement, paid review or affiliate link has been added to this edition. Source links are references, not endorsements. A future material commercial relationship would need a clear disclosure beside the affected content and an editorial review of whether the material still serves the reader.
Arcade Meridian is described as an independent editorial project. We do not invent a legal publisher, named staff, credentials or governance documents. Independence here describes the scope and presentation of this work; it is not a certification of facts about undisclosed ownership.
What the opening edition does not cover
No interviews, representative surveys, legal reviews or device laboratory measurements were conducted. The country dossiers are deliberately limited. Maps are original conceptual artwork, not political boundary datasets or statistical visualizations.
Inspect the source library, compare records in Signals Atlas, or read about the project.