Abstract.
The visual analysis of soccer match data can help soccer data analysts, clearly and intuitively. Discover hidden patterns in the match, and has attracted the attention of many researchers.
Soccer is one of the most popular sports in the world and attracts many participants from all over the world. Thousands of professional soccer players participate in professional soccer matches. And many major soccer events (e.g., FIFA World Cup, etc.) are watched by millions of fans.
Visual analytics can provide a clear and intuitive presentation of the analysis process and results. Support users to explore the data interactively, and are widely used in different data analysis fields.
In the field of visual analytics of sports data, visual analytics of soccer match data has attracted. The attention of many researchers. For example, soccer match data sites such as ESPN, and WhoScored. And football extensively uses visual elements such as icons and timelines to display. The course of the match, and use statistical charts such as bar charts and radar charts to compare and analyze player statistics.
Customized visual analysis systems are widely used in professional soccer match data analysis. Which helps data analysts visually analyze and explore match situations and tactics adopted during the match.
Soccer match data mainly includes statistical data, event data, track data, etc.
For different types of match data, the related visual analytics work can be divided into different categories based. On the analysis tasks. For example, for statistical data, it can be divided. Into visual analysis of match ranking and visual analysis of statistical indicators, etc. For event data, it can be divided into visual analysis of key events and visual analysis of delivery events, etc. For trajectory data, it can be divided into visual analysis of match video. Visual analysis of match time and space trajectory, visual analysis of match team shape, etc.
Soccer Match Data
This elaborates on the existing soccer match data types and summarizes the common soccer data collection methods. Commonly used data types for soccer matches include statistical data, event data, track data, etc. Early soccer match data analysis work mainly focused on the analysis of statistical data. But due to the development of soccer fine-grained data collection technology in recent years, recent soccer. Match data analysis work gradually focused on the analysis of event data and track data. The above data types will be introduced in detail later.
Statistical data
Statistical data in sports matches mainly refers to the basic statistics in a sport. In the field of soccer, statistics can be divided into team statistics and player performance statistics. Among them, team statistics mainly include the number of goals scored, goals conceded, shots on goal, fouls, red and yellow cards. Offsides, corner kicks, saves, etc., as well as the team goal difference, team points, team ranking, etc. after each match. The statistics for players mainly include the number of appearances, goals. Assists, red and yellow cards, running distance, sprint distance, etc.
There are many ways to obtain statistics in soccer matches. Many popular soccer data websites (ESPN, WhoScored, football etc.) will publish the status of points ranking of each league and detailed statistics of popular teams and popular players, and data analysts can get the data directly from the websites. Meanwhile, the live video of most matches will show the statistics of two teams in a certain match, and data analysts. Can record the corresponding statistical values directly in the live video of the match.
FootballAnt Soccer Data
Here for those who just learn soccer data analysis, we recommend using football. The data visualization is more intuitive and the amount of data is also higher, the only defect is that the website is updated very fast. Each update will be lagged, we recommend using APP, here is the download address (App: https://dlinks.24h-sports.com/) download)
Event data
Event data in sports mainly refers to specific events that occurred at a particular moment in time. It is more granular than statistical data and contains temporal and spatial information during the match. In soccer specifically, event data is used to describe passes, carries, fouls and other actions between players. Analysts can analyze event data to understand the tactics and other information adopted by teams to draw valuable conclusions.
For a given soccer match, common event data are mainly records of events that occurred during the match. Each event record includes information such as the timestamp of the event, the type of event, the player corresponding. To the event, and the spatial location where the event occurred. Common basic event types include passes, balls, goals, goals, fouls, etc.
For ongoing matches, open the FootballAnt interface to visualize goals, corner kicks, red and yellow cards, etc.
More detailed classification of basic event types is also possible for analysis purposes. For example, the most common passing event records typically include the type of passing event, when the pass occurred, the passing player, the receiving player, the spatial location of the passing player, the spatial location of the receiving player, other details, etc.
The event data used for soccer match data analysis work is primarily collected from commercial soccer data companies (Wyscout, Opta, etc.) or manually using custom data collection systems. Commercial soccer data companies are able to provide a large amount of detailed and comprehensive soccer match event data, but their drawback is the high cost of data purchase. Researchers also typically develop soccer match data collection systems that help collectors manually tag events in live match videos by providing an interactive interface, and data collectors can also use the interactive system to record events in matches, tagging information such as where the event occurred and the type of event.
Data analysts can choose different data collection methods according to different analysis needs. The disadvantage of this approach is that it requires multiple collectors to label the data, and manual labeling is less efficient and time-consuming. Techniques such as computer vision can also be used to improve the acquisition process and reduce the workload of data collectors.
Trajectory Data
Trajectory data in sports mainly refers to the running trajectory of players and the trajectory of the ball during the match, which contains the most detailed information in the match. In soccer, the trajectory data includes the coordinates of each player’s position on the field and the ball’s position at different moments. Analysts can analyze the trajectory data from multiple perspectives, and can also use techniques such as data mining to discover the patterns of players’ running in the match and get insights related to the match tactics.
The trajectory data in a given soccer match mainly includes the two-dimensional coordinates of the positions of the players of both teams at each time stamp in the match and the two-dimensional coordinates of the positions of the ball. The trajectory data used in soccer match data analysis are mainly from commercial data companies (Prozone, etc.), GPS acquisition, video annotation system acquisition, etc. As with event data, the trajectory data from commercial data companies are of higher quality but are also more expensive to acquire. The methods of systematic acquisition mainly include manual acquisition and semi-automatic acquisition, etc. Among them, the manual acquisition has a high acquisition effort and a long acquisition time.
Therefore, the researchers introduced computer vision technology into the soccer track data acquisition process, using computer vision technology to identify the position of each player on the field, while supporting data acquisition personnel to make manual adjustments and corrections to improve data acquisition efficiency.
A visual analysis system, ForVizor, for visual analysis of soccer formation changes. Details a method for semi-automatic soccer match trajectory data acquisition. The method uses particle filtering technology for player position tracking and formation detection. The system uses an interactive player-tracking method based on a color histogram. And particle filtering, which is a common tracking method in computer vision. First, the collector selects a specific player in the first frame of the soccer match video as the tracking target. The position of the tracked player is displayed. In the field, and the collector can confirm it during the acquisition process. Tracking will stop in 3 situations.
The first one is the switching of the tracking target. When the collector finds that the tracking box is shifted from the target player to another player, he/she can stop the tracking. Click the Modify button and click the target player again for correction.
The second type is tracking target occlusion. The target player may be completely obscured during the match when the collector can set the player position manually.
The last one is the low confidence in the tracking results. For each tracking frame, the particle filtering algorithm calculates the confidence level of the tracking result. If the confidence level is below a pre-set threshold, the system stops tracking and requests a manual correction. The system further maps the player tracking. Results to a 2D plane to obtain the 2D coordinates of the player on the court.
In recent years, with the development of visual analytics, and visual analysis of soccer matches. Data has attracted the attention of more and more researchers.
I still prefer to use FootballAnt because it is easier to get started. It does deep mining of known data and uses artificial intelligence operations to make predictions.
It also has an interesting feature of predicting shutouts. This allows me to not only see the live scores of soccer matches anytime and anywhere but also not to miss a single match!
FootballAnt Soccer Shutouts
Whisper it to you: using football live score, you can see every potential outcome of the 2022 World Cup in Qatar.
You can also download the football app from this
