Analysing gameplay experiences: Bram Heijligers on the Play-O-Meter Toolkit

Could you briefly explain what the Play-O-Meter Toolkit is and what its purpose is? 

Bram: 'The Play-O-Meter Toolkit is a combination of hardware and software that allows us to comprehensively monitor gamers, what they see, hear and how they process that information neurologically and physiologically, which in turn affects their behaviour. This includes using game controllers, keyboards, or VR systems.’ 

‘We track behavioural responses and how those influence perception, creating a feedback loop. The system measures eye movement, emotional versus rational attention, and physiological indicators like micro-sweat activity, which can predict excitement or stress. We also monitor heart rate and oxygen levels to assess a player's energy and arousal.’ 

‘Facial expressions are analysed to infer emotional states, while audio-visual inputs, what the player sees and hears, are annotated using cognitive, psychological or organisational frameworks. Communication is also measured: tone, volume, and language used, helping us assess the effectiveness of team interactions or pinpoint critical gameplay moments.’ 

‘We are currently using the system with team-based games like Valorant, and plan to expand into individual games like racing simulations. We are even testing applications in the BUas Taste Lab using the same technology.'
 

What prompted you to develop this toolkit? 

Bram: 'My background is in game design, and I always found it difficult to assess whether a game actually worked as intended, whether it taught something, changed behaviour, or simply brought joy. Traditional methods like questionnaires felt too limited.’ 

‘That is what led me to pursue a master’s in Cognitive Science & AI at Tilburg University. In 2021, Frank Peters invited me to teach AI at BUas, and Professor Igor Mayer encouraged me to develop my research into a PhD. I wasn’t ready to give up exploring how to scientifically measure learning and fun in games. So, I started creating a methodology that could work across platforms and genres.’ 

‘My aim is to help improve games, optimise esports performance, and support the industry with better tools. It is also something I genuinely enjoy.' 

How does the toolkit contribute to improving gameplay or user engagement? 

Bram: 'We use it to measure specific aspects like reaction time, for example, how quickly a player visually identifies an enemy or responds with an action like shooting. We can also assess communication effectiveness or detect stress.’ 

‘Players can review critical moments using a dashboard, which helps them and their coaches reflect and improve. Eventually, we hope to build a full AI coach. We are already collecting data to support that.’ 

‘Crucially, we align our biometric and behavioural data with what the player actually saw. Without that visual context, data can be misleading. That is one of the major flaws in existing research methods, too much noise, not enough precision.' 

Is the toolkit aimed more at developers, researchers, or another audience? 

Bram: 'Initially, it was intended for game developers. But obtaining high-quality gameplay data from in-development titles is difficult. In contrast, esports teams train regularly, play consistently, and are goal-driven, making them ideal subjects for structured data collection.’ 

‘So now, we are collaborating with teams like the Breda Guardians who are looking for better coaching tools. We also work with third-year students from the ADSAI programme on both entertainment and serious games, involving them in data engineering or vertical analysis.' 

What challenges have you faced while developing the Play-O-Meter? 

Bram: 'Mostly technical ones. There is no clear manual for building something like this. We handle large volumes of data and sometimes run into computing or network issues.’ 

‘Storing and transferring data securely is another hurdle, especially between our server and the AI training systems, we have had to manually upload data, which is inefficient.’ 

‘There is also the challenge of gaining participants’ trust, some are concerned about being recorded. But overall, most people are enthusiastic and willing to be involved.' 

Have there been any surprising results or insights from your research? 

Bram: 'One of the most surprising findings is how effective even relatively simple models have been. We managed to reach over 90% accuracy in identifying critical moments or communication effectiveness with limited data.’ 

‘With our growing dataset, now over 130 gigabytes across 70 players and 20 matches, we are scaling up to include every major esports event in the Benelux. The more complex the situation becomes, the more potential we see for large language and vision models to generate meaningful insights.’ 

‘For example, games like Valorant, which align more closely with human perception, produce clearer results compared to chaotic games like Marvel Rivals.' 

What impact do you hope your toolkit will have on the games industry or academic research? 

Bram: 'I hope it gives game designers more clarity, helping them define and optimise the intended player experience. Right now, there is often no budget for thorough testing. If we can remove some of that uncertainty, it would be a huge win.' 

‘We are already in conversation with companies like Guerrilla Games and applied game design programmes, but we are not rushing partnerships. The priority is building a stable product before we expand further.' 

What has this project brought you personally as a researcher? 

Bram: 'I really enjoy it. It is a fun and complex puzzle, and I find the gradual unfolding of these challenges deeply engaging. It is the kind of work that keeps me motivated and satisfied.' 

Do you have any advice for students or professionals interested in similar tools or research? 

Bram: 'Follow your passion and intrinsic motivation. It is what sustains you through the complexity and long-term nature of this work. I truly enjoy my job and maintain a good balance, that is essential.’ 

‘If you are interested in game data science, I highly recommend looking into the ADSAI or CMGT programmes at Breda University of Applied Sciences.' 

‘The project also involves close collaboration with several academic institutions: Tilburg University focuses on performance, cognition, and physiology; the University of Twente brings expertise in psychological well-being; and Eindhoven University of Technology advises on human-computer interaction and dashboard usability. At BUas, we lead the applied development, pushing updates to partner institutions and working hands-on with students to explore real-world implementations.’ 
 

Connect with Bram Heijligers on LinkedIn