โ—‰ Game Bubbles
โ† Back to the bubbles
METHODOLOGY

How the visualisation is built

Game Bubbles separates discovery, sampling, active tracking and presentation. That keeps the broader catalogue available for discovery while concentrating repeated snapshots on games that are currently useful to track.

1. Catalogue discovery

The discovery process builds a catalogue of Steam applications that can be considered by the site. Discovery is separate from the visible bubble map: appearing in the underlying catalogue does not automatically make an application one of the actively tracked games.

2. Player sampling

Discovered entries are sampled for player activity. This allows the site to distinguish games showing current activity from entries that are not useful for the player-count visualisation at that time.

3. Active tracked catalogue

A smaller active set is promoted from the sampled catalogue. These games receive the repeated snapshots used by the site. The active set can change as the wider catalogue is sampled again, so the visualisation is not intended to be a permanent fixed list of the same titles.

4. Metadata and categories

Metadata is associated with tracked games so they can be searched and grouped. A game can belong to multiple categories. Category counts therefore describe matching catalogue entries rather than mutually exclusive buckets.

5. Ranking and the top 60 bubbles

When no category is selected, the active matching catalogue is ranked by the player snapshot used for the view. When a category is selected, the catalogue is filtered to that category first and the filtered result is then ranked. The bubble map displays up to 60 leading results for readability, while the list underneath can show the complete matching tracked result.

6. Time periods and change

Repeated snapshots provide historical comparison. A current player count describes a moment; a change value adds context by comparing activity across the selected period when sufficient history is available. Longer-period comparisons naturally become more useful as the historical dataset matures.

Important limitations

Game Bubbles is a visual exploration tool, not an audited market dataset. Player counts can change quickly, metadata can change, newly discovered games need time to move through the pipeline, and temporary events can produce unusual spikes. Rankings should therefore be read as snapshots of observed activity rather than statements about quality, sales or long-term success.