Characterizing player responses to surprising events in 2D platform games

Abstract

This paper presents an analysis of how players perceive and react to surprising events in a 2D platform game, Spelunky. We present the results of a 34-participant study that focuses on investigating players’ emotional responses to generated events in the game that are designed to evoke a specific emotional response (surprise) from players. Facial recordings, player responses, electrodermal activity (EDA), and interbeat interval (IBI) recordings are correlated to explain the divergence in the perception of surprise across different players. We found that many of the self-reported surprising events did not elicit any physical response from players, and EDA and IBI recordings collected using Empatica E4 could detect only 56% of reactions. However, data aggregated from chosen sources was able to capture 82% of all self-reported surprising events advocating the case of an automatic surprising detection framework for video games. We propose a framework for categorizing surprising events into three distinct categories and show that surprise can have varied responses from a player based on prior gaming experience, which had a negative correlation with the number of perceived surprises. Additionally, we provide suggestions on how to extend the existing theory of surprise in games toward automated game design.