Analytics #

Our analytics services capture data about live services and game health. There are two categories of data: metrics and telemetry.

Metrics are aggregated timeseries data representing numeric values. Examples include the number of players currently online (CCU); the number of RPC requests per minute; or the median, 65th, and 99th percentiles of RPC processing duration.

While telemetry applies to many different kinds of data collection, Pragma Engine uses it to describe high-volume, less structured, big data information. Pragma Engine provides a real-time telemetry event collector that terminates data into a data store of choice. Telemetry is useful for collecting high resolution data around specific questions or features, and requires post-processing to produce actionable insights.

Examples:

Category	Metrics	Telemetry
live ops	CPU, memory, requests per second	traces, spans, structured events
game data	CCU, match starts, purchases	enemy encounters, menu iteractions, weapon usage stats

Contents #

Topic	Description
Telemetry	Use telemetry to collect high-resolution data around specific questions or features.
Metrics	Metrics are aggregated timeseries data representing numeric values.