Applying Aspects of the Expert Performance Approach to Better Understand the Structure of Skill and Mechanisms of Skill Acquisition in Video Games

Abstract

Video games are ideal platforms for the study of skill acquisition for a variety of reasons. However, our understanding of the development of skill and the cognitive representations that support skilled performance can be limited by a focus on game scores. We present an alternative approach to the study of skill acquisition in video games based on the tools of the Expert Performance Approach. Our investigation was motivated by a detailed analysis of the behaviors responsible for the superior performance of one of the highest scoring players of the video game Space Fortress (Towne, Boot, & Ericsson, ). This analysis revealed how certain behaviors contributed to his exceptional performance. In this study, we recruited a participant for a similar training regimen, but we collected concurrent and retrospective verbal protocol data throughout training. Protocol analysis revealed insights into strategies, errors, mental representations, and shifting game priorities. We argue that these insights into the developing representations that guided skilled performance could only easily have been derived from the tools of the Expert Performance Approach. We propose that the described approach could be applied to understand performance and skill acquisition in many different video games (and other short- to medium-term skill acquisition paradigms) and help reveal mechanisms of transfer from gameplay to other measures of laboratory and real-world performance.

Game‐XP: Action Games as Experimental Paradigms for Cognitive Science

“…The authors of this issue of topiCS propose Game‐XP as a promising path, not as a fait accompli . Only time plus a lot of hard work by many inspired researchers will tell. With the exception of the work by Boot, Sumner, Towne, Rodriguez, and Anders Ericsson (), none of the authors have published more than one or two journal or journal‐quality papers on their paradigm. (As a lower limit, let me confess that the paper published by my lab [Sibert et al., ] is our first on Tetris.…”

Allen Newell's Program of Research: The Video‐Game Test

“…Schrodt, Kneissler, Ehrenfeld, and Butz () use artificial‐intelligence methods to build an ambitious cognitive architecture that comes close to an UTC; however, they do not compare the behavior of their program with actual human data. Boot, Sumner, Towne, Rodriguez, and Ericsson () use the time‐honored method of concurrent and retrospective protocols, used for examples by Geyser () and De Groot () to study problem solving and expertise. Interestingly, in sharp contrast with big‐data analysis, Boot et al.…”

“…Interestingly, in sharp contrast with big‐data analysis, Boot et al. () analyzed the behavior of a single participant. Sibert, Gray, and Lindstedt () apply optimization techniques from machine learning to identify the best combination of feature weights for selecting actions in Tetris.…”

“…Finally, Boot et al. () use a softer, but no less effective approach: protocol analysis. In all these papers, the focus on strategies is a goal in itself rather than a means to understand fundamental principles of cognition, as advocated by Newell.…”

Games for Psychological Science

“…Even when behavior is heavily constrained by a well‐defined environment, explicit objective functions, and many hundreds of hours of practice, there is still substantial individual variation in strategy. The work of Boot, Summer, Towne, Rodgriguez, and Anders Ericsson () shows how behavioral variations similar to these might be due to factors such as a weak understanding, on the part of some players, of the rules of the game. In other words, in this case, the development of highly efficient and routinised motor plans is limited by lack of comprehension of the structure of the problem.…”