Design and Implementation of a Brain-Controlled Snake videogame based on alpha activity
Exploring modes of interaction in BrainSnake, a cooperative multi-brain BCI game based on alpha activity
Brain-Computer Interfaces is one of the coolest fields of emerging technologies and particularly exciting when it comes of its possible applications for the gaming community. One of the coolest projects I have worked on recently involved the design, implementation and evaluation of a BCI game during my master degree in Human-Computer Interaction & Design at the University of Twente, in the Netherlands. I hereby present you: BrainSnake.
BrainSnake is a BCI version of the classic Nokia game Snake which relies on alpha activity of the player to control its main character (the snake, of course). It was born from the idea of creating the design for a BCI game that could circumvent some common limitations of alpha waves, which allow to track the user’s level of relaxation or focus and is commonly used as a passive BCI input modality (due to the delay it takes to induce and detect alpha activity effectively). We chose to follow typical design patterns of cooperative closely-coupled games, in which the actions of one player significantly affect the experience of the other player, such as complementarity of players actions, and interacting with the same object (in our case, the main snake character). Closely-coupled games were proved to be greatly enjoyable, but also to suffer greatly from the absence of communication between the two players. In our case, we were also interested to see whether players’ communication (or lack thereof) would have an effect on their relaxation levels, and hence performance.
Game design and input modality
We came up with a control mechanism in which alpha waves could be used in a somewhat active way: by having one of the players close their eyes. This is by far the easiest and most time-effective way to detect alpha waves, albeit counter-intuitive for rather obvious reasons (like huh, not being able to see what’s going on in the game). In our experimental design, the snake in BrainSnake constantly moves around the screen until one of the players closes their eyes, gaining control of it, at which point the snake would halt and start rotating up to 90° (one player can rotate it clockwise, the other one anti-clockwise). When the player re-open their eyes, the snake proceeds moving forwad following its new trajectory. Not dissimilarly from regular Snake, the aim of the players is to collect pieces of food and increase their score while avoiding accidentaly biting their own tail or hitting the walls or other obstacles on the screen. We designed these game mechanics expecting players to rely on communication in order to achieve their goal, with one player acting as the blind “controller” and the other one as a “feedback provider”. We also setup an experimental condition in which the two players were not able to communicate in any way, to see how this would affect the gaming experience.
Equipment and implementation
We created BrainSnake sing the Unity game development software and programmed it in C#. Not having any experienced game developer in our team, we started out by following a basic tutorial on YouTube that was meant for a different type of game (Curve Fever / Achtung die Kurve) and adapted it to become our very own version of Snake. Our BCI system comprised a Biosemi Active 2 for acquisition of raw EEG signal data, and the OpenVibe software for data processing and communication with Unity. We filtered the incoming samples of electrical brain signals through OpenVibe to classify the user state as having either high or low alpha activity whenever a certain threshold is reached, which would then translate in a control input for the snake character.
Evaluation of the gaming experience
As I already mentioned, the main focus of our study was to explore different types of interaction and communication between players of BrainSnake. We setup an experimental design in which six pairs of players had to play in two different conditions — one in which they were able to communicate directly face-to-face (“co-present” condition), and one in which they were isolated and prohibited from talking to each other, physically separated wearing earphones (“remote” condition). After each condition, we asked them to fill out a Game Experience Questionnaire (GEQ) which measures seven components: Immersion, Flow, Competence, Positive and Negative Affect, Tension, Challenge. On top of that, we collected post-experiment feedback through individual semi-structured interviews. We then developed a coding schema based on grounded theory to analyse the qualitative data collected through the interviews and generate insights. To do so, we particularly looked into performance metrics for closely-coupled cooperative games, such as instances in which the players would “help each other”, “work out strategies together”, or “get in each other’s way”.
Results from our study drew a rather complex picture of gaming experience. Expectedly, many participants were put off by the counter-intuitive mechanism of having to close their eyes to control the game character (especially in the remote condition), but many also appreciated the added challenge of having to rely on communication with the other player. On the other hand, there have been instances in which players decided not to talk to each other even when they were allowed to, or when communication caused them to get in each other’s way. While this is likely dependant on idiosyncratic qualities of each player and their personal gaming styles and preferences, our data suggests that closely-coupled BCI games may benefit from subtle, non-intrusive ways of communication that have lower chances of disrupting one’s level of focus and relaxation. Interestingly, a few participants explicitely indicated that they would rather use direct communication in a competitive version of BrainSnake, allowing for strategic manipulation of the opponent by using direct communication to disrupt their mental focus. Ultimately, despite our experimental design and unusual game mechanism, BrainSnake was well-received and our participants’ enthusiasm confirms the potential of BCI technologies for the gaming community.
If you wanna know more about our study, our research paper BrainSnake: Exploring Mode of Interaction in a Cooperative Multi-brain BCI Game Based on Alpha Activity has been accepted at ACHI 2018, The Eleventh International Conference on Advances in Computer-Human Interactions that took place last March in Rome, Italy and is available online through the ThinkMind digital library.
M van Almkerk, L Brandl, R Li, P Romeo, M Poel. (2018) BrainSnake: Exploring Mode of Interaction in a Cooperative Multi-brain BCI Game Based on Alpha Activity. The Eleventh International Conference on Advances in Computer-Human Interactions, pp. 39–44
Thank you for reading! If you enjoy this topic, please join in the discussion in the comments below! You may also show appreciation by leaving a few claps tapping the👏 button or by sharing this article with your network.