Signality has built an artificial intelligence (AI) platform that gives clubs, leagues and federations real-time data which makes it easier to access statistics from games.
The Swedish-based company has created a system that gives an array of insights to use, free from the hassle of manually setting up cameras or using wearables – saving time and costs on installation that normally come with this type of equipment.
“We have built the world’s first fully automated player and ball tracking platform dedicated for football,” Michael Hoglund said, vice-president of marketing and growth at Signality.
“We don’t require operators manning a system, our AI takes care of capturing and analysing everything from kick-off to half-time, end of the game, and all the 3.5 million data points that occur in between as the game is being played.”
Signality’s data collection takes that responsibility away from coaches and analysts to keep up, capable of recording clearer and more accurate data in the process.
It’s collected in real-time, meaning there’s no need for coding windows, wearables or filming games – it is all done through the power of AI. Signality uses LiveInsight as a way to present all the very best data a coach or analyst could want.
“We are doing individual player tracking in real-time, with a fully automated process,” Hoglund said.
“Our clubs can now access detailed stats and player specific videos within five seconds of an event happening.
“All this is shared as raw data to our customers or through our cloud video and analytics platform, LiveInsight.
“As we collect three and a half million data points per match, LiveInsight is the product we use for clubs to make sense of this data – they determine the information that is displayed there and then the data is collected and displayed automatically – as reports and as playlists with video.
“The massive data-set is made up of the position and identity of all the players (even the referee) 25 times per second.”
Clubs can then have more flexibility about how they use data for all players, as individual performances are analysed.
While it may seem complex to record and track each player’s movement on the pitch, Signality makes this process a whole lot easier.
“We enable the analyst to easily string together complex data-filters that will then automatically populate and generate videos and reports as the game is being played on the field,” Hoglund said.
“For example, an analyst or coach might request for LiveInsight to show video of all the times that their right-back/number two, passed to their striker, number nine, for the last five games.
“LiveInsight will instantly extract those instances with video clips and associated data. We want to enable coaches and analysts to be able to focus more on the insights, rather than spending lots of times doing manual work with tagging, coding, and filming.
“That’s what an AI excels at, automating asks. We want to make analysts and coaches feel comfortable in offloading some of that ‘grunt work’ to us, being that support that frees up time, having their back when it’s crunch time.”
LiveInsight has been built to extract data from automatically recorded video to generate highlights.
It shows exactly how many times a player has touched the ball in different areas of the pitch, giving instant results that analysts and coaches can immediately use.
“For analysts this is a total game changer, saving them hours each week tagging. It also frees them up during games to focus on the game rather than filming it,” Hoglund said.
“Professional football analysts are almost always extremely well educated, and we think our system can make better use of their abundant skill sets.
LiveInsight has proved successful for clubs as they can make full use of their data through automatically tagged video.
“The very nature of ‘machine learning’ means that our product gets better over time,” Hoglund said.
“We’ll be even more accurate than we currently are at measuring player speeds and distances, dribbles, possession per zone, pass success rates etc.
“Human motion analysis directly from the video is another field we see a lot of exciting use cases for it by using joint detection, gait analysis, and player vision field of view.
Many elite clubs in Europe have turned to Signality for data that is less likely to have errors. With automatic insights, it is a more effective alternative than wearables, which can only extract data for a club’s own team, not the opposition or ball.
It leads to much deeper tactical analysis for every second of a match.
“We’re looking to work with innovative club coaches and analysts who want to get better, faster data,” Hoglund said.
“Any club who feel their analysts can make better use of their skills for match analysis and preparation, as well as opposition scouting.”
You can find out more on Signality here.