Sound of Football Fans

In every football stadium in the world, a very unique soundscape is created by the fans to motivate their team. Only the loud expression of the fans‘ feelings in the arenas makes the team sport a social experience for both, the fans and the players. The fan chantings generate a special room atmosphere in the stadium building. We make the fan emotions at FC Bayern Munich’s stadium visible! In this season, our acoustic engineers will be attending five home games at the Allianz Arena as part of Siemens' and The Economists‘ REIMAGINE THE GAME project. Using the Acoustic Camera, we measure the sound of the chanting fans and map the sound data to a 3D model of the stadium. The combination of an acoustic and a statistical analysis allows the investigation of the relationship between the sound of fans and the action on the field. You can watch a video about the project and the making of here: VIDEO ABOUT THE PROJECT

After calculating the correlation function, the similarities (correlating components) between the reference channel and microphone signals can either be emphasized or deleted. Fig.2 shows the acoustic photo after deleting the correlating parts between the microphone array signals and the virtual reference channel. As expected, there is a high percentage of correlation between the dominant source and the virtual reference at the exact position of this source. So, with deleting these components from the microphone array signals the source from the first loudspeaker should not be visible anymore. The result shows three other sources: the source from the subwoofer, the reflection from the subwoofer on the table and the 23 dB quitter source from the second loudspeaker. As you can see the emission of the subwoofer is not correlated to the main tweeter source as they differ in the frequency band.

Thus, the noise of the fans seems to play an important role in the positive outcome of the game result.

This is underlined by the gesture of big football clubs such as FC Bayern Munich. Because they honor their fans by officially reserving the back number 12 for them. The questionable influence of the so-called „12th Man“ on the football team was first investigated in 1900 at the University of Minnesota. Nowadays, in 11-man team sports, the term is commonly used for the fans of a team. In other words, the concept of the "12th Man" deals with the influence of the fan noise on the stadium atmosphere and thus on the score.

The connection between sports and the digital world with IoT (Internet of Things) as well as artificial intelligence for game analysis is not a new topic. However, the data analysis and visualization in sports are mostly limited to the field, especially the players. Player performance and game data are already being analyzed in most football stadiums. For this purpose, football players wear tracking devices during the game, which provide data of their movements. In addition, modern football stadiums are equipped with technologies that monitor building parameters such as light, wind, and humidity. The records are analyzed on an ongoing basis so that the calculation of recommendations for care and maintenance is possible.

In this project, however, the focus is on the fans: For the first time ever, an attempt is made to quantify the influence of the chanting fans. The project "REIMAGINE THE GAME" was initiated by Siemens and The Economist. Signal Noise, an agency for data visualization, uses the gfai tech Acoustic Camera as an innovative measuring instrument to measure the stadium atmosphere in Munich. Beamforming is used to record and visualize the fans' reactions to rallies, refereeing decisions, goals and missed opportunities. The result is a visual representation of the stadium atmosphere created by the noise of the fan community. Due to the combination of the results of the Acoustic Camera with a statistical analysis, the relationship between the "12th Man" and the actions on the field are examined. Not only the sound of the fans, but also their influence on the players and the most-acclaimed players are analyzed. The measurement results allow statistical "storytelling" because multiple games are measured and patterns can be identified. Correlation analysis enables the identification of fan-related relationships to influence factors both inside and outside the stadium.

On the Internet, numerous rankings of the loudest fans and loudest stages can be found. While the volume in the stadium can be measured comparatively easily with a smartphone, measurement and analysis using beamforming are very complex. This is due to the fact that the Acoustic Camera captures every football fan as a single data point in a three-dimensional measurement! Our spherical microphone arrays have been developed especially for three-dimensional interior measurements. First, a 3D model is created as the basis for all measurements. This is followed by temporal, spatial and frequency-selective analyzes in conjunction with the acoustic data.

As an interesting fact, we conducted an acoustic measurement inside a stadium already in 2010. At the Michigan Stadium (American Football), we recorded the highest number of noise sources ever measured by beamforming (109,901 viewers = 109,901 individual sound sources). On our YouTube channel you can find a video showing this measurement. It needs to be mentioned that we only used the Acoustic Camera for a 4-minute half-time measurement. On the contrary, we have now recorded the fans in the Allianz Arena throughout the whole game. In the first step, our acoustics engineers used a laser scanner to capture the Allianz Arena in 3D. Subsequently, they created a 3D point cloud of the stadium based on the scan data. After the acoustic measurement, the acoustic hotspots in the stadium were mapped on the 3D model.

During the game against the TSG 1899 Hoffenheim, the acoustic data of 75000 fans was recorded. For this purpose, two Sphere Arrays served for the data acquisition of the crowd noise during the entire game. Secondly, the acoustic data was used to create an acoustic video by mapping the sound sources to the 3D model of the stadium.

As a result, the evaluation results of the Acoustic Camera serve further data visualization and statistical analyzes. While the sound source mapping in beamforming is classically represented as color gradients, for so-called "data scientists" every fan sound is a single data point. These data points are the basis for pattern recognition and correlation analysis, which are performed throughout all measurements in the season. From "Where in the stadium are the fans the loudest" to "In what weather are the fans quieter?" To "Which teams have louder fans than others?" Many questions are possible. After that, predictions about the behavior of the fans and the influence of the chanting on the game score are the next step.

After all, the visualizations of the results including sound recordings of the game FC Bayern Munich vs. Hoffenheim and vs. Borussia Mönchengladbach are available on the project website. A "dashboard" illustrates the results of the data analysis in four categories.

1. Key moments such as goals or fouls can be selected on the timeline of the game: Catching the sound at the exact time. In addition to the acoustic experience through the speakers, a video shows the change of the volume of the fans in different parts of the stadium.
2. The so-called "fan energy score" represents a comparison value for the volume of the fans for individual players, as well as at the time of specific attack and defense actions, or between different football games. The decibel range of the fans in the stadium is measured and combined with the sound pressure level data to calculate an average score of fan volume between 1-100.
3. The analysis of key moments in the game allows a direct comparison of the fan sounds for different highlights like goals, red cards or fouls.
4. Additionally, there is a player analysis: Which players have caused the greatest enthusiasm? The visualization of the fan sounds shows a ranking of the „fan energy scores“ for Bayern's top players in terms of volume reached when they were on the ball.

For the first time in the history of football, a game can be seen in retrospect through the voices of the fans. We make the soundscape of football fans visible! As can be seen, the use of the Acoustic Camera allows the visualization and analysis of the stadium atmosphere, which is generated by chanting fans. In addition, the relationship between the fan story and the score can be analyzed by data scientists. In addition, the statistical game analysis opens up new strategic possibilities for the strategic planning of the team as well as for the optimization of the stadium building acoustics. Clever acoustics planning based on data analysis can further enhance the community experience in the stadium. "For the first time, we can quantify the effect of the fake sounds by looking at the action on the field and the sounds of the 75,000 fans in the Allianz Arena and analyzing emerging patterns," says Signal Noise Chairman Matthew Falla.