Gameplay metrics form the basic pieces of information for working with telemetry data from a game design angle, and it is often in the gameplay metrics that we find the root causes for observed player behaviors. For example, that a specific curve is causing too many crashes in a racing game, that the Nerf gun is overpowered and overused in relation to other weapons, or that players never go to Ye Olde Magic Shoppe, even though an important quest start there.
Game user research experts Alessandro Canossa and Janus Rau Sørensen, and I, wrote a chapter for an up-coming book on game telemetry analysis a while ago, and here we take a stab at formulating some guidelines for the deep basics of working with gameplay metrics, e.g. categorising different approaches. In this two-parter, I will describe the essence of these ideas, which we have also adopted in Game Analytics.
Analysing gameplay metrics can be done in different ways, but fundamentally, is either performed via analysis or synthesis – both classic scientific methods. They are different, but in practice this difference can be subtle and they often go hand in hand:
Analysis is when we break down a complex whole into parts or components. For example, when we break down the action-sequences of the players in a time spent analysis.
Synthesis is is the opposite procedure, i.e. combining separate elements or components in order to form a coherent and complex whole. For example a chart showing the number of daily active users (DAU) is a synthesis of time, number of users, date, etc.
Both analysis and synthesis can be initiated by fairly open-ended (do our players cheat?) or specific questions (does that player cheat by using the inverse-shield duping method?), roughly correlating with the concepts of explorative vs. hypothesis driven research from scientific theory.
What this means is that the analytical methods we use to find the answers to questions are either of a type where we are looking to confirm some idea we have and are looking for confirmation, or have a pretty good idea about the possible answers (hypothesis-driven); or alternatively more open, where we are not sure what the answer to a given question is, or have a hard time predicting the possible answers (explorative analysis).
In the next part we will take this further and discuss explorative vs. hypothesis-driven approaches to game analysis, and why this is important for analysts.