Play-Persona: Modeling Player Behaviour in Computer Games

This thesis proposes a framework to sharpen the focus on player experience when designing and evaluating levels for single player action games.
This framework is both based on an analysis of what a level is (for the designer and for the player respectively) and a theoretical exploration of modeling of players.
From a developer’s point of view, game levels are the confluence of art, code and gameplay. From a player’s point of view, levels are a succession of spaces that contain events.
Due to its intrinsic multidisciplinary nature, level design has been historically difficult to define both in terms of scope of the discipline (what is the result of the process) and the skill set required from the level designers. In different game studios it is possible to see how the role of level designer can be covered by artists with a good understanding of space, architects unbound from gravity and materials and even the specific subcategory of programmers called scripters1. The core focus of designing game spaces is also often disputed exactly because of the heterogeneous composition of its ranks. Some teams see gameplay as king, and every other decision is secondary; in other contexts it is the story that rules and at times even the code, under the appearance of outstanding technical achievements, can become the compass that orients the design process. Another problem often encountered in level design is the lack of an irrefutable, quantifiable assessment of quality and success ratio. QA departments in leading developer’s studios around the world are starting to devote enormous resources to track players’ behaviour through metric data and provide an unquestionable answer to issues such as “is this level good enough?” In this setting, the main concern seems to deal with which parameters are necessary to monitor in order to insure a proper feedback on questions regarding the success rate of game levels and how to relate the analysis carried out on the raw data.

This project intends to look for an element that can be the pivot, flexible and solid at the same time, around which all the other fields and disciplines will harmonically dance, structuring the game space and providing it with consistency, focus and variety. The result is not a cumbersome normative procedure, nor a merely descriptive framework, but instead it provides mental and practical tools for level designers. That central pivot is the play-persona concept, initially inspired from the field of Human Computer Interaction as a mean to imply players during the design phase of games. Play-personas are ways of modeling player behavior and understanding players’ relationship to the ludic and aesthetic affordances of specific computer games. Play personas are defined as clusters of preferential interaction and navigation; they can be employed as a means of assisting the computer game design process, or as an a posteriori method (lens) for better understanding games. Since there seem to be no single universal experience of play that can fulfil the needs of every player, it is necessary to design for a multitude of approaches that players may bring with them to the game, depending upon their psychological traits. This is the design function that play personas fulfil.
Persona-oriented game design can connect quantitative analyses of player behaviour (game metrics) as well as psychological studies of player types, studies into game design processes, and analyses of level design in computer games.