Touchscreens vs. traditional controllers in handheld gaming

Abstract

We present a study which compares touchscreen-based controls and physical controls for game input using Ubisoft's Assassin's Creed: Altair's Chronicles. Our study used the Apple iPhone as a representative touchscreen-based controller and the Nintendo DS for its physical control pad. Participants completed a game level four times on each platform. Level completion time and number of player deaths were recorded. Results indicate that physical buttons allowed significantly better performance than virtual buttons. Specifically, the number of character deaths on the iPhone was 150% higher than on the DS, while level completion time on the DS was 42% faster. The learning curve for the touchscreen version of the game was also steeper. Participants strongly preferred the physical buttons of the Nintendo DS. We conclude that either game designers should consider alternative input methods for touchscreen devices, or hardware designers should consider the inclusion of physical controls.

Comparing Order of Control for Tilt and Touch Games

“…Without the ability to inject code to manage an experiment, measurement and data collection are compromised. This occurs since logging user actions, timestamps, etc., is relegated to an external process [6,[12][13][14]22]. In some cases, game scores [12,14] or level completion time [12] can be externally viewed and logged as dependent variables.…”

“…This crude form of data collection is cumbersome, prone to logging errors, and lacks precision. In other cases, experimenters are left manually counting and recording events in real-time using a log sheet [22] or video playback [7,13].…”

“…In such cases, it is impossible to reliably distinguish whether the observed differences were due to the test conditions (e.g., touch vs. tilt) or uncontrolled factors inherent to the games or platforms (e.g., processor or display differences between devices) [7,22]. Such differences across platforms are confounding variables, further compromising internal validity.…”

“…We argue that this allows researchers to "distill" a game to just the elements under study, be they control options, display characteristics, or gameplay elements. Consequently, and unlike other work [6,12,14,22], we developed a game as our experimental framework. Based on the arguments above, this offers a higher degree of experimental control, and may yield more reliable experimental results than previous work using commercial games.…”

“…Physical controls are comparatively rare on mobile devices today, but are ubiquitous on dedicated game devices (e.g., handhelds such as the Nintendo 3DS). Previous work [21,22] has demonstrated that physical controls tend to perform better than touch-based controls for games. However, this work used commercial games, and again suffers from the relative lack of control in the design of the experiments.…”