EEG-triggered dynamic difficulty adjustment for multiplayer games

doi:10.1016/j.entcom.2017.11.003

Entertainment Computing

Volume 25, March 2018, Pages 14-25

https://doi.org/10.1016/j.entcom.2017.11.003 Get rights and content

Highlights

•
Designing DDA triggering based on EEG readings.
•
Implementing DDA triggering using the Emotiv sensor for a 3rd person shooter.
•
User study I - establishing correlations between EEG readings and game events.
•
User study II - EEG-based triggering improved player experience.

Abstract

In online games, gamers may become frustrated when playing against stronger players or get bored when playing against weaker players, thus losing interest in the game. Dynamic Difficulty Adjustment (DDA) has been suggested as an intelligent handicapping mechanism, by reducing the difficulty for the weaker player, or increasing the difficulty for the stronger player. A key question when using DDA, is when to activate the difficulty adjustment.

In this paper we suggest using the Emotiv EPOC EEG headset to monitor the personal excitement level of a player and use this information to trigger DDA when the player’s excitement decreases in order to ensure that the player is engaged and enjoying the game. We experiment with an open-source third-person shooter game, in a multiplayer adversarial setting. We conduct experiments, showing that the detected excitement patterns correlate to game events. Experiments designed to evaluate the DDA triggering mechanism confirm that DDA triggered based on EEG increases the players excitement and improves the gaming experience compared to the heuristic triggered DDA and the experience of playing a game without DDA.

Introduction

Video games are a popular activity for children and adolescents in the western world today, with almost $97 %$ of the younger US population playing video games for at least one hour per day [20].

While the goal of a gamer in a video game may be to kill enemies, or collect prizes, game creators typically aim to keep players entertained and engaged over a long period of time [42]. It was often observed that playing against the game artificial intelligence (AI) is not as challenging as playing against other players [28], [43]. When playing against other humans, however, it is important to play against players of similar levels of expertise, because when a weaker player plays against a much stronger one, often both players feel dissatisfied — the stronger player is bored, while the weaker player is frustrated. When a gamer wishes to play against one of their friends, matching suitable opponents may become even harder.

Dynamic difficulty adjustment (DDA), in which the players’ abilities to influence the environment are dynamically modified throughout the game, provides a possible solution to this problem [46], [47], [55], [32], [39]. When a weaker player faces a stronger one, the level of difficulty for the weaker player can be reduced, as well as the level of difficulty for the stronger player can be increased. For example, in a first-person shooter game, where the goal of a player is to kill the avatar of his/her opponent, the bullet damage of the weaker player can be increased, and the bullet damage of the stronger player can be decreased. In addition, it may be possible to identify and implement a variety of game-specific adjustments in many games, with the aim of improving the user experience by ensuring that the games difficulty level is optimal for the players. However, one of the most challenging issues associated with DDA is knowing when to trigger such game modes.

Presumably, game modifications should be triggered only when needed, in order to avoid erratic game behavior. In the past, researchers have mainly suggested heuristics based on the game state [3], [27], [53]. For example, in a game that keeps an ongoing score, a decision can be made to apply DDA when the difference in the players’ scores exceeds a certain predefined threshold. This heuristic addresses the situation when one player becomes too strong compared to the other player, based on the assumption that in this situation the stronger player may feel bored, while the weaker player may feel frustrated. The scoring mechanism hence contains clues to the players’ state of mind.

In this paper we suggest a different approach — measuring the players excitement and activating the game modes when the excitement level drops below a predefined threshold. This approach attempts to directly address the core problem of degraded game experience, rather than relying on a scoring mechanism or similar heuristics to determine when the players are no longer excited by the game. We implement a passive feedback/affective state regulation method by using the Emotiv EPOC headset to read electroencephalography (EEG) signals and the Emotiv Affective Suite¹ to translate the signal into an affective state.² Then, based on the affective state, the game mechanism triggers DDA.

We conduct two user studies. The first user study assesses the correlation of the EEG signals to game events, and the results of this study show a good correlation between the Emotiv value for “short term excitement” (STE) and game events. Although, both positive and negative emotions can be utilized to improve gameplay [5], in this paper we attempt to improve the players’ experience by maximizing STE. The second study compares our EEG-triggered DDA to a standard heuristic approach based on elapsed time and game status, as well as to a control game in which DDA is not implemented. The primary hypothesis tested here is that EEG-triggered DDA increases the players’ excitement and improves the gaming experience compared to the heuristic triggered DDA and the experience of playing a game without DDA.

The contributions of the paper are twofold. First, we present a case study for the design of EEG-triggered DDA technique in a modified version of the open-source third-person shooter game, “Boot Camp”.³ Second, we study the user experience, comparing EEG-triggered DDA to standard heuristics triggering, as well as a game without DDA. Our study confirms that gamers enjoyed the EEG-triggered DDA better than the other two options and that this technique significantly increases the player’s level of excitement. Our study further shows that the choice of the triggering strategy is important and significantly impacts the way players experience the game.

The rest of the paper is structured as follows: We begin by providing some background on brain-computer interfaces, EEG, and DDA in Section 2. We review the Boot Camp game that we use in our experiments and discuss additional related work. Then in Section 3 we discuss our EEG-triggered DDA triggering technique, and the modifications that we added to Boot Camp in order to allow DDA. We also explain the heuristic method of triggering DDA which is also evaluated in our experiments. In Section 4 we move to the user study that we conducted, first showing that EEG measurements in this setting correlate well with game events, and then comparing EEG to heuristic triggering of DDA. We discuss the main results of this study in Section 5 and conclude in Section 6 with a summary of this research and an indication of future research directions.

Section snippets

Background

In this section we briefly review the rapidly growing field of brain-computer interfaces and assessment of the affective state. We discuss BCI application to games, dynamic difficulty adjustment (DDA) in games, and the “Boot Camp” third-person shooter game.

Designing DDA triggering strategies

In this paper we design and evaluate EEG-triggered DDA for increasing the players’ excitement. In this section we discuss the DDA features added to Boot Camp and the motivation for each type of game adjustment. We also describe two DDA triggering methods: EEG-triggered and heuristic.

Empirical evaluation

In this section we describe the set of experiments we conducted to compare heuristic-triggered DDA, and EEG-triggered DDA, in third-person shooter games such as Boot Camp. We begin with a preparatory experiment without DDA, designed to establish a correlation between the game events and the affective states of the players. Then we move on to a second experiment, evaluating the applicability and benefits of the EEG-triggered DDA.

Discussion

Our experiments demonstrate that using EEG to decide when to trigger game changes to improve the player’s experience can be useful. Most players reported that they preferred the game where EEG-triggered DDA was used (the E game) to the alternatives.

To support our claim as to the benefit of EEG-triggered DDA in improving players experience, we analyzed a few possible alternative explanations for the preference of the E game over the other two options. First, it might be that the Emotiv EPOC

Conclusion

In this paper, we explored dynamic difficulty adjustments triggered using players’ EEG readings in a multiplayer adversarial environment. We evaluated our suggestions by conducting a user study using the Boot Camp third-person shooter game. We suggested a set of modes appropriate for this game, in various scenarios, based on whether the players are close or far from each other, and on the identity of the better player.

The preparatory experiment showed that the EEG readings from the Emotiv EPOC

References (55)

S.H. Fairclough
Fundamentals of physiological computing
Interact. Comput.
(2009)
A. Nijholt et al.
Turning shortcomings into challenges: brain–computer interfaces for games
Entertain. Comput.
(2009)
D. Afergan et al.
Dynamic difficulty using brain metrics of workload
N.A. Badcock et al.
Validation of the emotiv epoc®eeg gaming system for measuring research quality auditory erps
PeerJ
(2013)
A. Baldwin et al.
A framework of dynamic difficulty adjustment in competitive multiplayer video games
R. Bernays et al.
Lost in the dark: emotion adaption
M.V. Birk et al.
The false dichotomy between positive and negative affect in game play
P.M. Blom et al.
Towards personalised gaming via facial expression recognition
L. Bonnet et al.
Two brains, one game: design and evaluation of a multiuser bci video game based on motor imagery
IEEE Trans. Comput. Intell. AI Games
(2013)
D.P.-O. Bos et al.
Brain-computer interfacing and games

J.T. Cacioppo et al.

Handbook of Psychophysiology

(2007)

M. Csikszentmihalyi

Flow and the Psychology of Discovery and Invention

(1996)

Y.A. De Kort et al.

People, places, and play: player experience in a socio-spatial context

Comput. Entertain.

(2008)

C. Dormann et al.

Understanding game design for affective learning

N. Ducheneaut et al.

Alone together? Exploring the social dynamics of massively multiplayer online games

M. Duvinage et al.

Performance of the emotiv epoc headset for p300-based applications

Biomed. Eng. Online

(2013)

K. Emmerich, M. Masuch, Helping friends or fighting foes: the influence of collaboration and competition on player...

L. Ermi et al.

Fundamental components of the gameplay experience: analysing immersion

Worlds Play: Int. Perspect. Digital Games Res.

(2005)

J. Gonzalez-Sanchez et al.

Abe: An agent-based software architecture for a multimodal emotion recognition framework

J. Gonzalez-Sanchez et al.

How to do multimodal detection of affective states?

I. Granic et al.

The Benefits of Playing Video Games

(2013)

H. Gürkök, D.P.-O. Bos, M. Obbink, G. Hakvoort, C. Mühl, A. Nijholt, Towards multiplayer bci games, in: BioSPlay:...

H. Gürkök et al.

Brain-computer interface games: towards a framework

A. Harris et al.

Including affect-driven adaptation to the pac-man video game

T. Harrison, The Emotiv mind: investigating the accuracy of the Emotiv EPOC in identifying emotions and its use in an...

R. Herbrich, T. Minka, T. Graepel, Trueskill: a bayesian skill rating system, in: Advances in Neural Information...

C. Hondrou et al.

Affective, natural interaction using EEG: sensors, application and future directions

Cited by (53)

A methodology for using players’ chat content for dynamic difficulty adjustment in metaverse multiplayer games
2024, Applied Soft Computing
Personalization of game difficulty is a critical task in leveraging artificial intelligence (AI) technologies to enhance player engagement in virtual worlds like metaverse. One of the key challenges in this area is developing methods for assessing a player’s perception of game difficulty. This information can be used to dynamically adjust the game difficulty to match the player’s skill level and preferences, which can improve the player’s experience and engagement. The existing approaches have limitations such as relying on costly external devices, requiring time-consuming feedback or questionnaires, and being specific to certain game genres and narratives. In this paper, we propose a new method called ChatDDA for evaluating a player’s perception of game difficulty by analyzing the content of their chat messages. Our method uses a pre-trained language model to extract semantic features from the chat messages, which are then used to train a feed-forward neural network to predict the player’s level of hopefulness or despair about succeeding in the game. Three pre-trained language models—BERT, RoBERTa, and Twitter-roBERTa—are fine-tuned on a purpose-built dataset of player chat messages of the popular multiplayer online game PlayerUnknown’s Battlegrounds (PUBG) tagged as expressing optimism or pessimism regarding game success. The results showed that our method can accurately predict a player’s perception of game difficulty, with an accuracy of 0.953 on the test dataset of player chat messages. This suggests that our method has the potential to enhance player engagement and immersion within the game, ultimately leading to more satisfying and enjoyable metaverse experiences.
Theories, methodologies, and effects of affect-adaptive games: A systematic review
2023, Entertainment Computing
Affect-adaptive games gained in popularity over the last years in human computer interactions studies, promising potential benefits for player experience, performance, and even health. It is however not yet clear how affective games are being evaluated, what the precise effects are, and how they are based on emotion theoretical concepts that are still not universally agreed upon. This systematic review investigated these questions by analysing relevant high-quality evaluation studies of the effect of affect-adaptive video games on various outcomes in regards to their effects, theoretical assumptions, and methodologies. Out of 3,930 papers, 26 studies were included based on preregistered inclusion and exclusion criteria. A high variance regarding theoretical assumptions and methodological approaches was observed, as well as an overall poor methodological rigour, leading to the conclusion that more work is needed in constructing better methodological standards for game evaluation studies and theoretical considerations when developing and testing affect-adaptive video games.
Recognition of EEG based on Improved Black Widow Algorithm optimized SVM
2023, Biomedical Signal Processing and Control
Citation Excerpt :
In the field of gaming, Li et al. [10] developed a P300 game which is highly compatible with the BCI system, making the game fully accessible to disabled person. Stein et al. [11] adjusted the difficulty of the game by analyzing the real-time excitement level of the player through EEG signals, which enhanced the gaming experience for players of different levels. While the enormous potential of EEG is fascinating, the research associated with it is challenging, the inherent characteristics of EEG, including weakness, non-linearity, non-stationarity, and randomness, make it difficult to be processed and analyzed.
As a classifier suitable for nonlinear samples, support vector machines (SVM) is widely used in Electroencephalogram (EEG) signals pattern recognition. The performance of SVM depends mainly on the selection of model parameters. This paper modifies the reproductive formula of the Black Widow Optimization (BWO) algorithm to obtain an improved Black Widow Optimization (IBWO) algorithm, which solves the problem of the restricted search radius of the BWO algorithm. We construct an IBWO-SVM model in which IBWO is used to optimize the penalty parameter $C$ and the kernel function parameter $g$ of the SVM, and the optimized SVM model is applied to EEG signals classification. The effectiveness and superiority of IBWO-SVM is verified by using the BCI competition III dataset IVa EEG public dataset, and compared with the SVM optimized by the original BWO algorithm (BWO-SVM), the artificial fish swarm algorithm (AFSA-SVM) and the particle swarm optimization algorithm (PSO-SVM), The simulation results demonstrate that all indicators of the IBWO-SVM model are better than the three comparison models, in which the average classification accuracy on five subjects reaches 97.29%. In addition, the two-class motor imagery (left arm stretch and right fist clenched) EEG data collected from our experiment is also used to test the performance of the IBWO-SVM model. The average classification accuracy of the final twelve subjects is 5.21%, 4.16% and 1.47% higher than that of BWO-SVM, AFSA-SVM, and PSO-SVM, respectively. Therefore, the IBWO-SVM model can effectively improve the accuracy of EEG signals pattern recognition, which has a very important practical value.
Diversifying dynamic difficulty adjustment agent by integrating player state models into Monte-Carlo tree search
2022, Expert Systems with Applications
Citation Excerpt :
Simple heuristic measurements are often too limited to satisfy the wide breadth of player expectations. Some researchers have developed DDA models to estimate players’ affective states using sensors (Afergan et al., 2014; Chanel et al., 2011; Fernandez B. et al., 2017; Kivikangas et al., 2011; Kneller et al., 2012; Parnandi & Gutierrez-Osuna, 2015; Rani et al., 2005; Stein et al., 2018; Tognetti et al., 2010). Instead of measuring player proficiency, they focus on how their DDA model improves the actual PX.
Game developers have employed dynamic difficulty adjustment (DDA) in designing game artificial intelligence (AI) to improve players’ game experience by adjusting the skill of game agents. Traditional DDA agents depend on player proficiency only to balance game difficulty, and this does not always lead to improved enjoyment for the players. To improve game experience, there is a need to design game AIs that consider players’ affective states. Herein, we propose AI opponents that decide their next actions according to a player’s affective states, in which the Monte-Carlo tree search (MCTS) algorithm exploits the states estimated by machine learning models referencing in-game features. We targeted four affective states to build the model: challenge, competence, valence, and flow. The results of our user study demonstrate that the proposed approach enables the AI opponents to play automatically and adaptively with respect to the players’ states, resulting in an enhanced game experience.
Improving Learners’ Assessment and Evaluation in Crisis Management Serious Games: An Emotion-based Educational Data Mining Approach
2021, Entertainment Computing
Citation Excerpt :
Furthermore, we plan to exploit our evaluation results to adapt various aspects of the game to the detected learners' affective profiles to regulate their emotions when they are feeling some negative emotions. In fact, recognizing the players' affective states allows the researcher/instructor to adjust the challenge level of the game according to their competencies level basing on their current emotional states (engaged, confused, bored or anxious) [37]. For example when players feel anxious, a decrease of the challenge level seems to be effective in helping them to transit into the Flow state again.
For several years, there has been growing interest in the development and use of serious games to improve individuals' quality of life and behavior. In particular, Crisis Management Serious Games (CMSG) have shown their potential for teaching people both technical and soft skills related to managing crises in a safe environment while reducing training costs. To improve their effectiveness, several evaluation approaches of CMSGs have been proposed. However, despite its interest, the learner's emotional state is often neglected. As a result, learners may end up with a deep frustration or boredom. Therefore, we propose an Educational Data Mining (EDM) approach to evaluate learners' affective states in collaborative CMSGs. This approach is applied to assess learners' engagement during a game-based evacuation scenario aiming to train and to raise awareness among students of a university on evacuating the present persons during a fire emergency situation. The experimental results confirm the major predictions of learners' affective dynamics and uncover novel findings namely the presence of transitions from boredom to frustration and from frustration to confusion. Our study shows that combining gaming and emotion aspects under an EDM approach to evaluate CMSGs is interesting since it gives reliable results in a less invasive way.
Assessing the effectiveness of virtual reality serious games in post-stroke rehabilitation: a novel evaluation method
2024, Multimedia Tools and Applications

View all citing articles on Scopus

View full text

EEG-triggered dynamic difficulty adjustment for multiplayer games

Highlights

Abstract

Introduction

Section snippets

Background

Designing DDA triggering strategies

Empirical evaluation

Discussion

Conclusion

Interact. Comput.

Entertain. Comput.

Dynamic difficulty using brain metrics of workload

Validation of the emotiv epoc®eeg gaming system for measuring research quality auditory erps

PeerJ

A framework of dynamic difficulty adjustment in competitive multiplayer video games

Lost in the dark: emotion adaption

The false dichotomy between positive and negative affect in game play

Towards personalised gaming via facial expression recognition

Two brains, one game: design and evaluation of a multiuser bci video game based on motor imagery

IEEE Trans. Comput. Intell. AI Games

Brain-computer interfacing and games