Bundesliga unveils ‘Average Positions’ and ‘Expected Goals’ stats

Stats created using machine learning, through a partnership with Amazon Web Services (AWS).

Bundesliga has been working with Amazon Web Services (AWS) to create two new on-screen stats, providing new insight into games.

The new ‘Bundesliga Match Facts’ stats are Average Positions and Expected Goals (xGoals). They are created from real-time analysis of data from the live video feeds, to track players’ locations on the field, and predict the probability of a player scoring a goal.

The video feeds are streamed into AWS’s machine learning and analytics tools to analyse the data and create the stats in real-time. Bundesliga is the first to roll out this type of analysis for football matches.

The stats will be unveiled during highlights of the Borussia Dortmund vs Bayern Muenchen match, shown tonight in 200+ countries. Average Positions will then be available during future broadcasts of Bundesliga matches, while xGoals will be available during highlight matches.

Andreas Heyden, executive vice president of digital innovations for the DFL Group, said: “We at Bundesliga are able to use this advanced technology from AWS, including statistics, analytics and machine learning, to interpret the data and deliver more in-depth insight and better understanding of the split-second decisions made on the pitch.”

Andy Isherwood, vice president and managing director EMEA, AWS, adds: “AWS is helping the Bundesliga enhance the broadcast viewing experience by delivering deeper insights into the game that didn’t previously exist. These two new statistics are just the beginning of what we’ll be able to deliver for football fans as we look forward to unlocking new ways to better educate, engage, and entertain viewers around the world.”

The new stats
Average Positions Fans will now be able to see the positioning of a team’s players on the pitch and gain insight into the team’s intended playing style. Average Positions provides new insights based on analysis performed on data captured from tracking a player’s average location on the field, which is then displayed in real-time. This Bundesliga Match Fact allows viewers to identify the current momentum on the field and understand tactical changes. The new statistic will help pinpoint if a team is setting up in an attacking or defending style, pressing up the middle, or utilising the wings. By displaying the average location and position of each player in real-time, Average Positions helps fans understand tactical changes as the game progresses.

xGoals Using Amazon SageMaker, a fully managed service to build, train and deploy machine learning models, the Bundesliga can now assess the probability of a player scoring a goal when shooting from any position on the field. The goal probability is calculated in real-time for every shot to give viewers insight into the difficulty of a shot and the likelihood of a goal. To calculate the precision of xGoals, machine learning models were trained by analysing 40,000 historical shots on goal in addition to an array of features derived from positional data, including distance to goal, angle to goal, player speed, number of defenders in line of shot, and goalkeeper coverage.