Defining Fun
A fascinating project that I have been involved in recently is the concept of defining “fun”. Defining exactly what “fun” is can be especially difficult, as fun is subjective and relative to the individual experiencing it. This shows in the definition of fun: “what provides enjoyment or amusement” (thanks, Websters).
However, understanding, or trying to define exactly what “fun” is might not be a completely intractable problem. Personality theorists struggled with many of the same issues in their attempts to define aspects of personality. Personality tends to also be unique to each individual, and relative to those experiencing it. So perhaps by applying some of the same analytic techniques used to identify the defining elements of personality we can begin to uncover the important aspects of the gaming experience.
The crux of this concept is the lexical hypothesis, or the idea that any characteristic important enough to make a difference in conceptualization is measurable. If there is some phenomenon which is correlated with an experience, then it can be described by some word. By this logic, a corpus of representative words could be used to define virtually all words that relate to the concept of fun. These terms could then be analyzed by various means (for example, clustering, rating, or factored by semantic structure) to determine which terms are most closely related to “fun”.
So how does this tie into games?
The purpose of all games is for the user to have fun. However, the context or goals of a particular game may alter the most highly rated factors of fun (just as different personality types might rate different characteristics of fun as being of higher value). Confounding this is the lexicon unique to gaming, which can limit the meaningful clustering of generalized terms related to fun.
This project, then, attempts to capture a corpus of words that define fun as they relate to games. By sampling all of the words that are used in relation to gaming, we can build a lexicon that can be clustered in relation to the type of game played. By using clustering and scaling techniques as a function of characteristics such as game type (RTS, FPS, etc) or abstract game values (Non-stop Action, Team-Based Strategy, etc.).