A Comparison of Text-Annotated vs. Purely Visio-Spatial Instructions for Video Game Input Sequences

Abstract

Animated tutorials for controller input sequences in video games will be easier for developers to localize, and therefore more universally accessible across multiple nationalities, if they are designed to rely on purely visio-spatial, text-free communication. However, it has hitherto been unclear whether a lack of text in such tutorials may be associated with lower effectiveness in overall teaching outcomes, given that text may augment the available information. With this in mind, we conducted a between-subjects study in which 42 subjects each played one of two versions of a custom, stand-alone game control tutorial designed to teach character moves for an action/fighting game; both versions contained animated visuals, but one also included onscreen text instructions. We recorded players’ performances in each version by measuring elapsed times, failure counts, demonstration replay counts, and skip counts using software-based logging and combined each player’s data in these categories to create an overall effectiveness index to measure and compare teaching efficacy associated with each version. We compared the means of our data between the version groups using t -tests with results suggesting that people playing the non-text-annotated version performed at least as well, all around, as those playing the text-annotated version and even better in the areas of elapsed time, failure count, and overall score. Though a significant co-factor—player skill/experience—is likely an influence that should be further delineated in future studies, our findings clearly demonstrated that text-free game control tutorials are as good as, if not better than, text-annotated ones.