The players’ trajectories generated by using the tracking software do not allow any statements about the performance or the behavior, respectively. They have to be processed and analyzed with various methods.
For the analysis of the player trajectories an own software is being developed, which provides appropriate functions. The analyses should be automated (as far as possible). Furthermore, the application is used for visualization and comparison of the calculated results.
The functional scope is divided into three functional groups. The first of these groups includes basic analysis that focuses on the “performance” of the players / teams, especially in the movement domain. For instance, these include the movement profiles of the players, the distribution of the movement intensities and the detection of sprints or runs at high speed. It also contains the so-called “heat maps”, which show the stays of the players in the field. Also, team related analyses, such as the “Real Tactical Lineup” (the mean positions of the players during a game) or the determination of the individual parts of the team (e.g. defense, midfield, attack) fall into this group.
A player’s heatmap
Automatically determined team formation parts and their distances
Different distributions of velocities
The second group contains the ball-oriented analyses, which, in contrast to the first group, will only be possible if the position of the ball is determined. In the case of pure GPS tracking, however, this is not the case, since equipping the ball with a GPS receiver is currently not possible. The features that can be calculated using the ball trajectory are the ball possession statistics or the ball contacts of the players. In addition, passes, goal shots, corner balls and goals can be detected.
(Bad) Passes The different pass opportunities of a player The passing history shown as graph
The last group corresponds to the “advanced” analyses. The word “advanced” is intended to express that more complex algorithms are required to calculate the results. Thus, e.g. methods of machine learning or sequence analysis are used to recognize movement patterns in the trajectory data, which cannot be seen at the first glance. These patterns may be typical runs of individual players or typical behaviors of the whole team. In relationship with the ball trajectory, pass sequence patterns or characteristic pass behavior can be determined, too. The calculation of the passing possibilities of a ball owning player is also possible. The patterns obtained in this way contain interesting information about the strategy of a team in different game situations. This kind of information is desired by both, the own coach / team and the counterpart.
Automatically detected repetitive pass sequences Recognized player movement patterns
Relevant news
Demo for the acquisition of trajectories
There is a new demo video. It shows the usage and the different views of the tracking software to track the players briefly. However, this is just a very basic scenario with only one person. A more complex and challenging scenario will ...
Automatic detection of changes in the real tactical formation
The real tactical lineup (RTL) is an analysis option that can be used to illustrate the real tactical action on the field. In contrast to the tactical formation given by the trainer (see Figure 1, left), the RTL (right) reflects the actual ...