Measuring physical responses during competitive gaming in Fortnite

As part of my master work I wanted to study the physical responses of playing a competitive video game, and especially during encounters with other players in the game. The study was conducted with behavioral measurement tools and demonstrates how these tools can be fruitfully used to observe media during use.
Promotional image from the offical Fortnite website (
Publisert: 3. juni 2019

New equipment for us
In early 2018, we were introduced to the institutes’ new mobile media lab in the course MIX302: Research Design and Field Studies. This is a research kit that includes cameras, a screen capture device, and a computer running Noldus’ software that allows for recording time-stamped video and analysis of the video. The mobile lab also includes the Empatica E4 wristband, that records heart rate, galvanic skin response and movement data, and the Tobii Pro Glasses 2 that records the user’s eye movements as well as video and audio from the user’s perspective.

Gøran A. Slettemark talks about his Fortnite study during a visit at the Noldus IT headquarters the Netherlands in September 2018. Photo: Lars Nyre.

Physical response in competitive games

I wanted to use this equipment to measure physical responses in a competitive video game. After some consideration I decided on a game called Fortnite Battle Royale.

Fortnite had recently become quite popular, and also represented a new genre of competitive games that has not been studied much. This genre is called Battle Royale, after the movies and books that inspired the ruleset. In these games there are usually a large number of players, that compete to be the last player standing. In Fortnite’s Battle Royale mode, up to 100 players jump out of a magical flying school-bus onto an island, where they must find weapons and potions to help them on their path to victory. As time passes, the play area is further restricted by a storm that moves in over the island. Players who are caught in the storm will take damage, providing motivation for moving to the safe zone. A single game takes less than 30 minutes.

Method design for observing Fortnite gaming

I recruited four participants, all who had played the game before. Before the play-session started, I conducted a short interview with each about their experience with Fortnite and other games.

Their level of familiarity ranged from having played some games, to playing more or less every day for the last few months. Those who had played the game for longer reported that the social aspect of the game was an important draw, often playing in groups. When asked about what other games they had experience with, the majority responded with games that could be played socially, either competitively or cooperatively. Some had played competitive shooter games before Fortnite, but not regularly.

While playing for around one hour the participants wore the Empatica E4 wristband and the Tobii Glasses. Because of difficulties with direct screen capture from the Playstation, I decided to use a camera to record the screen. This was required to get a timestamped video that could be used to synchronize the data from the wristband and the glasses.

The recording was done using Noldus Media Recorder and the Tobii recording software. The glasses had to be calibrated, which is done by having the participants look at a calibration card at about arm’s length. The Tobii software made it possible to observe the participants’ gaze as they were playing, which was useful for noticing behaviours. The wristband had to be turned on some time before it would start recording.

Analysis of the observational data

After the observations, the data from the glasses and the wristband had to be transferred to the computer. With the glasses you must go through Tobii’s software. This software allows for analysis on its own, but in this case, it was only used to export the video including the gaze visualization. The data from the wristband had to go through Empatica’s online service, where it could be downloaded as .csv-files.

I used Observer XT to synchronize the data, starting with the time stamped video which was imported from Media Recorder. Then I imported the video from the glasses, which did not come with a timestamp, and had to be manually synchronized. When importing the data from the watch, the timestamp had to be converted using spreadsheet software so that it would line up correctly.

Heart rate within a single game. As we can see, certain aspects of the game can cause a rapid response.
When the participant encountered other players, the response could be plainly seen in the heart rate recording. Galvanic skin response rose steadily during the games, and lowered in-between, which can indicate that making it into the later parts of the game is more stressful (Lazar et al. 2017).
The more experienced players seemed to have less of a response to early game encounters. This could be due to there being less “buy-in” early in the game. As players get closer to a potential win, their competitiveness may increase. The more experienced also players looked at the screen differently, with more quick movements, showing more familiarity with the interface. This is reminiscent of the concepts of flow and immersion, as further described by Nacke and Lindley (2008).

While this study was somewhat simple, it demonstrates how easily tools like eye-tracking can be used to observe media in use in a research environment like Media City Bergen. It could be used more frequently in media development, to compare prototypes against each other or to evaluate existing products. These technologies will become an important tool for designers who wish to understand how their media product actually impacts the user.


Drachen, A., Nacke, L., Yannakakis, G. og Pedersen, A. (2010). Correlation between heart rate, electrodermal activity and player experience in first-person shooter games. Proceedings of the 5th ACM SIGGRAPH Symposium on Video Games – Sandbox ’10.

Nacke, L. Lindley, C. (2008). Flow and immersion in first-person shooters. Proceedings of the 2008 Conference on Future Play Research, Play, Share – Future Play ’08.

Hjortskov, Rissén, Blangsted, et al. (2004). The effect of mental stress on heart rate variability and blood pressure during computer work. European Journal of Applied Physiology, 92(1-2), p. 84-89.

Lazar, J., Feng, J. and Hochheiser, H. (2017). Research methods in human- computer interaction. 2nd ed. p. 369-397.

Publisert: 3. juni 2019
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram