Understand game growth
Compare changes in a game's audience with changes in Roblox's overall player count.
How to use it
- Choose a growth metric and a comparison range.
- Check the direction of change and the metric card's border color.
- Use the Roblox-relative metric to assess the game's share of the audience.
Advice and practical context
Growth relative to platform
Normalization separates a game's audience-share movement from Roblox-wide activity. If game CCU rises 10% while the platform rises more over the same period, the relative position can decline. The reverse can happen during a platform-wide drop.
Matching windows
Use matching timestamps and periods for the game and platform. Different boundaries can manufacture apparent outperformance. Do not call a relative metric retention or attribute the difference to a particular release without additional evidence.
The relative line answers audience-share questions; the absolute line answers operating-scale questions. A game can lose relative share while real CCU rises. Show both when the decision affects staffing, server capacity, or budget.
Technical detailsCalculations, permissions and behavior
Popularity relative to Roblox
Only shared completed hours are compared. Game CCU is the mean of five-minute snapshots with at least 75% coverage; platform CCU is the chosen source's hourly estimate. Daily peaks are not used.
The chart shows percentage change from the first shared point of the selected range, starting at 0%. Share change = 100 × ((game CCU / platform CCU) / initial ratio − 1). Auto selects hour, day, week or month.
Larger intervals keep the last real matched point per bucket, not an invented average index. If the game grows 20% and Roblox grows 10%, relative growth is about 9.09%. This measures share in an external index, not proven advertising or update impact.
The ? icon on each result opens its concise definition, so the interface does not repeat a long explanation under the cards. RobloxCCU / RoTrends and RoWatcher use different methods, not an official census of every player.
- Selected-range calculation
The range ends at the last completed UTC hour. A point is labelled by the hour's end: 1h compares two adjacent completed hourly means, while 30d compares the means at the start and end of a 30-day window. Changing range recalculates the baseline and all three changes.
If shared boundary data is missing, the result uses the available matched observations. The entire requested range stays on the axis; no line is invented beyond observations. All time starts at the first shared point.
- One chart heading
The card keeps one large heading with help, range and interval controls, three result cards and the chart itself. It does not duplicate a source label, partial-range row, latest-interval row or a second internal chart title.
Growth card border colors
In Is the game growing faster than Roblox?, each of the three cards uses its own selected-period change: positive is green, negative is red. Zero change and missing comparable data keep a neutral border. Game share is no longer automatically highlighted in green: a falling share has a red border even when the game's CCU rose.
Useful articles: examples and decisions
The guide covers the steps in the service. These articles explain the metrics, examples, techniques and how to evaluate results.
