What these numbers actually mean.
Playing now: the Steam Web API’s GetNumberOfCurrentPlayers response for app 5155360, cached for up to five minutes. The check time belongs to the successful reading, not the moment you opened the page.
24-hour sampled peak: the highest successful reading in our rolling 24-hour window. We aim to collect every 15 minutes; page refreshes can add observations. Until a full day has been covered, or if collection has gaps, the page labels the coverage as partial. A sampled maximum can miss a short spike between observations.
Our recorded peak: the highest sample stored since our own collection began. It is different from the game’s lifetime record. The separately labeled SteamDB number is a dated external reference; it is not fetched as a live official peak.
Concurrency is not sales: this count is not unique daily users, copies sold or the number of players looking for a public match. People in menus and bot matches can be included. We do not infer sales from these numbers.
Outages: if Steam fails, we can show the last successful reading with its original timestamp and a delayed label. Readings older than 24 hours are not displayed as a current count. If historical storage fails, the chart and sampled peaks are withheld rather than presented as complete.
Review totals are fetched separately from the Steam review summary and cached for an hour. They include all languages and purchase types, so they may differ from another site’s filters.
Player-count questions
Is the Hood Warfare 2 player count live?
The current count comes from the official Steam Web API, with a cache of up to five minutes. A visible check time shows when the reading was taken. A delayed label means the reading is older than that target, for example while a refresh is pending or if Steam is unavailable.
Why does a lobby look empty if people are playing?
Steam concurrency does not measure joinable public lobbies. It can include menu activity and bot matches. Region, game version and lobby settings may also affect what you see; we cannot diagnose a specific lobby from concurrency alone.
Why is your peak different from SteamDB?
Our sampled peak uses only our collected observations. Sampling intervals, missing readings and the start of our collection can produce a lower maximum than another tracker. SteamDB’s dated historical record is shown separately.
Does this number prove the game is dead or growing?
One reading cannot establish a trend. Compare the same time of day across several complete days and account for missing samples. We do not label the game dead or growing based on one snapshot.