Description
Designs in-game events, turns the resulting telemetry into gameplay metrics such as deaths, completion times and paths by level or map location, and analyses them to find where and how player behaviour departs from the design.
Every action in a game runs code that can be logged. Gameplay metrics analysis starts by designing events, such as a player firing a weapon, with related data (weapon, position, match time, outcome) and shared identifiers so that sequences can be reconstructed. Raw telemetry is summarised into gameplay metrics, for example average completion time per level, causes of death per level shown as a progress-frequency chart, or player positions over time plotted on the map. Drachen and Connor distinguish explorative analysis, such as drill-down through finer data and filters until an insight appears, from hypothesis-driven analysis, such as testing whether zombies on Level 10 are over-represented among causes of death; in practice the two alternate. Statistical models and machine learning extend the toolbox, from classification for churn or purchase prediction to clustering of play styles. For user researchers, metrics give precise behavioural records, reach far beyond lab samples, allow live hypothesis testing and connect to business metrics, while lab methods explain why. Charest shows how small studios can start with a few well-documented events around first-time experience, progression, level design and monetisation. The approach suits any product where behaviour unfolds in a space or sequence.
Attributes
- Discipline
- Product Analytics & Strategy
- Loss type
- Siloed
- Stage
- Evaluation, Post-launch
- Data
- Quantitative
- Effort
- Weeks
Steps
- Agree the design and research questions the data should answer.
- Design and document events with context attributes and identifiers that link related events.
- Validate the instrumentation in test builds before release.
- Aggregate telemetry into gameplay metrics by level, map area, encounter or session.
- Explore with drill-downs and visualisations (progress-frequency charts, map overlays) and test specific hypotheses.
- Where needed, apply models such as clustering or churn prediction.
- Triangulate findings with playtests or interviews and feed changes back for re-measurement.
Sources
- Drachen, A., Mirza-Babaei, P. & Nacke, L. (eds), Games User Research (2018)
Also known as
- Game telemetry analysis
- Game analytics for user research
- Gameplay telemetry
- Spatial and progression metrics analysis
Categories
- Analytics & Behaviour
- Product Analytics & Strategy