Player Profile Models: A Research in the Phenomena Player Types
Author
Rueløkke, Jonatan
Term
4. term
Education
Publication year
2014
Submitted on
2014-05-27
Pages
75
Abstract
Denne afhandling undersøger 'spillertyper'—en udbredt måde, som forskere og spildesignere bruger til at beskrive, hvordan mennesker foretrækker at spille—og spørger, om sådanne modeller kan bruges til mere end blot tjeklister. De fleste eksisterende modeller kan ikke måles, og de få, der kan, dækker kun en begrænset del af spilleradfærd. På baggrund af en omfattende litteraturgennemgang udvikler vi The Player Profile Model, der skal fungere både som tjekliste og som værktøj til at matche spillere med spil. For at afprøve modellen opstillede vi et testsetup med tre spil og et spørgeskema, som blev sendt til en bred gruppe deltagere. Spørgeskemaet gav hver deltager en individuel spillerprofil, som vi sammenlignede med deltagernes vurderinger af flere AAA-titler (store, højbudgetspil) og de tre udviklede testspil. De statistiske test kunne ikke afvise nulhypotesen om modellens forudsigelsesevne, hvilket betyder, at undersøgelsen ikke fandt evidens for, at modellen pålideligt kan forudsige, hvilke spil folk foretrækker.
This thesis examines 'player types'—a common way game researchers and designers describe how people like to play—and asks whether such models can do more than serve as checklists. Most existing models cannot be measured, and the few that can cover only a narrow slice of player behavior. Based on an extensive literature review, we developed the Player Profile Model to function both as a checklist and as a tool to match players with games. To test the model, we created a study setup with three games and a survey, which were distributed to a broad group of participants. The survey produced an individual player profile for each person, and we compared these profiles with participants’ ratings of several AAA titles (big-budget mainstream games) and the three study games. Statistical testing did not allow us to reject the null hypothesis about predictive ability, meaning the study did not find evidence that the model can reliably predict which games people prefer.
[This abstract was generated with the help of AI]
Keywords
Documents