Methodology

Where the data comes from

Only public, openly available store information. We never log in, never use private APIs and never bypass blocks.

States

• Limited (soft launch): available in at most 30 countries and missing from major markets.
• Expanding: still limited, but added at least 3 countries recently.
• Global: Android — available in 100+ countries; iOS — available in at least 4 of the major markets (US, GB, DE, FR, JP, KR, BR, IN).
• Removed: no longer available in any country we check.

Soft launch types: test market (typical test countries such as CA, AU, NZ, PH, NL, SE, NO, DK…), US first, regional, home market (single country = developer's country).

Not counted as soft launches: long running — a limited release whose store release date is more than 12 months old (a local or regional app that has been out for years, not a test; when the store shows no release date we keep the test-market label); casino — real-money and social casino, betting, poker, bingo and lottery apps, by store genre or by words in the title or developer name. The public top lists also skip one-country releases outside the classic single test markets (CA, AU, NZ, PH, NL, SE, NO, DK, FI, IE, SG) and limited sets wider than 30 countries, show one game per studio and store, and alternate iOS and Android.

Limitations — please read

• iOS has no install numbers. For iOS we use chart ranks and rating counts per country; we do not estimate downloads.
• Android install counts are Google's lifetime totals shown as ranges; our installs/day is a difference of totals, not a measured daily number.
• On iOS we track availability in a fixed set of markets, not every country. A game live only elsewhere can look "removed" until it reaches one of them.
• Store listings in some regions show games that cannot actually be downloaded there; we do not count those as available.
• Studio grouping links store accounts by shared website/email domains and names. It can be wrong; low-confidence groups are marked.
• Genre and mechanic tags may be generated by a language model from the store description.

Traction score

0–100, relative to games of the same age and store. Weights: Android traction 35%, iOS charts 25%, momentum 15%, engagement 10%, expansion 7%, update cadence 5%, on both stores 3%. A game with too little data gets no score instead of a guess.