A Research Journey Through the World of Computer Games

Send article

Subscribe

Facebook

In his Ph.D. dissertation Play-Persona: Modeling Player Behaviour in Computer Games, Alessandro Canossa offers an analysis of data about computer game players’ actions in the game that establishes a theoretical framework for understanding player behaviour and thus creating better computer games.

By Hans Emborg Bünemann

Computer gamers are a diverse group of people with many different approaches to and motivations for playing, and consequently they display a wide diversity of play behaviour. An increasingly important parameter in the development of games is that gamers should be able to unfold their personal gaming style. With this as his point of departure Alessandro Canossa recently completed a Ph.D. project and defended his dissertation at The Danish Design School. The purpose of the project was to analyse and answer the question of how to develop attractive computer game experiences.

Data Reveal Player Behaviour

A few years ago, some game developing companies began to include a piece of software in the games that lets them collect so-called metric data on player behaviour. Alessandro Canossa analysed and interpreted player behaviour based on data about players’ ways of positioning themselves, using their skills and interacting with other players in the game. For example, this data string

<X=23467 Y=38576 Z=23623 TS=254 MA=3 IW=4 IP= />

means that at a certain time (TimeStamp) the player was at a certain place (X,Y,Z), lurking in the shadows (MovementAttitudes) with his or her map unfolded (InteractionWorld), equipped to strangle someone (InteractionPeople).

Play-Personas – a Tool for Game Designers

The data analysis forms the basis for a categorisation of typical patterns of player behaviour, which in turn was used to develop a set of player archetypes called play-personas. The point of these play-personas is to give game designers a theoretical framework for understanding player behaviour and developing better computer games. In the development phase, this framework should help designers define which play patterns to enable in a given game and design the game with this in mind. Designers can also use play-personas as a design tool for understanding and categorising the accumulated metric data from a large number of players navigating through the game.

Possibility field. Play-personas characterise the game designer’s perception of categories of player behaviour (metaphor) as well as player archetypes that are revealed by the metric data (lens). The field of possibility is the full range of possible actions that the player has to choose from at any given time. Overlaps mean that the designer’s expectations of player behaviour match the players’ actual behaviour. Illustration: Alessandro Canossa

“I constructed the play-persona as a tool for uncovering patterns in player behaviour,” Alessandro Canossa explains. “Play-personas bridge the gap between data analysis and the game designer’s work. They can be used as tools for evaluating games by comparing the designer’s objectives with the player’s. The player has a set of expectations for the game that he embarks on. For example, his objective may be to make it through the game universe to the goal as quickly as possible. The game designer’s objective may have been to facilitate other types of experiences for the player, for example to survive a series of challenging fight situations by drawing wisely on a repertoire of skills.”

Alessandro Canossa suggests that a comparison of the designer’s and the players’ objectives could reveal whether the game design actually supports the personas that the designer had in mind while constructing the game. He believes that a poor match between the designer’s and the players’ objectives are either due to poor communication to the player about the characteristics of the game or a poor game design.

Visualising the Data

In order to make his analysis of the metric data useful to game designers in practice Alessandro Canossa developed a method for visualising the metric data. Visual representations such as the so-called heatmaps (see illustration) make it possible to balance the game’s topology and degree of difficulty and thus eliminate experiences that are frustrating for the player. Such experiences might occur at a particular place in the game where the player dies repeatedly and is unable to move on. The visualisations also offer the game designer valuable information about the type of experience a given game has to offer, for example whether the game is nerve-racking, soothing or aesthetically enriching. Alessandro Canossa says,

Heatmap. Visualisation and analysis can turn metric data into a useful tool for game designers to create better games. For example, a heatmap illustrates places where players have a large (red), a medium (yellow) or a small (blue) tendency to die. Illustration: http://www.bungie.net/Online/HeatMaps.aspx

“My research contribution to the practice of game development is to propose a method for making the metric data meaningful to game designers by visualising and interpreting them and uncovering the patterns in player behaviour.”

Compendium Format Facilitates Journey Experience

Unlike most Ph.D. dissertations, Alessandro Canossa’s dissertation is not a monograph but instead consists of a detailed introduction to the topic and 13 articles, 12 of which underwent peer review as part of the approval process for research conferences. The evaluation committee was impressed that Alessandro Canossa managed to produce and publish so many articles over the three-year duration of the Ph.D. project. The author chose this format for his dissertation in order to offer the reader an insight into the research journey that he undertook.
“My dissertation can be viewed as a form of storytelling. The reader is drawn in almost imperceptibly and makes the journey from my initial idea to the development of the concept of play-persona and the theoretical framework that surrounds it. Through the articles the reader experiences the dynamics that characterises the research process,” Alessandro Canossa explains.

Qualitative Approach in Future Research

The evaluation committee called for a qualitative supplement to the dissertation's quantitative analysis. According to Susana Tosca, an associate professor at the IT University of Copenhagen who was a member of the evaluation committee, the dissertation would have benefited had Alessandro Canossa’s concept of play-personas been related to actual flesh-and-blood computer gamers.
“Ethnographic research has been published on whether the players’ motivation for playing is based on, say, the social element or the competitive element, on gender differences, etc. It would be interesting to combine this knowledge with Alessandro’s abstract play-personas,” says Susana Tosca.
The qualitative aspect is something that Alessandro Canossa intends to look into in the future.
“I would like to study how a qualitative approach could be integrated into my research, for example by using participant observation and interviews with players and game designers. That would make it possible to analyse how designers think when they develop games, and how players perceive their own behaviour in the game,” Alessandro Canossa concludes.