How often do you see it, a player holding his hamstring, signalling that he just can’t go on? Depending on how severe the injury is, players will miss weeks or maybe even months of action. Ousmane Dembélé couldn’t help his team from Barcelona in the CL semi-final return leg as they were incredibly beaten 4:0 by Liverpool. Or take Newcastle United, who had just made Miguel Almirón their record signing in January, only for him to sustain a hamstring injury in April, which ruled him out for the rest of the season. The fate is shared by Arsenal’s Aaron Ramsey, who will therefore not only miss a nice send-off before his free transfer to Juve, but also be unable to play in next week’s Europa League final – which might also have an impact on their biggest game of the season. What if you could not only make decisions for injured players during the game, but also prevent them from being sidelined at all, based on tons of data and AI algorithms?
Injury prevention on another level
It may cost clubs a pretty penny, if their star players – or any player – get injured. It may be during the season’s run-in, ahead of the World Cup or in the run-up to pre-marketed friendlies, for which tickets need to be sold. While injuries are bad to clubs and for managers, football teams do have the players to compensate; just not always on the same level. For individuals, though, injuries can be much more of a personal catastrophe. They could miss out on the World Cup or the Champions League final. Furthermore, they could be denied bonuses for goals scored or games started and such existentially important matters like contract extensions or transfers to another club might be jeopardised by injuries. Just think of players, who are known to be “injury-prone“. Even if they perform well, suitors could be looking for alternatives.
There are hundreds of reasons why injury prevention is so important. And its importance grows with the number of games in the modern game and the successively increasing intensity. That’s why Zone7 does immerse into injury prevention. The company takes into account 5 million hours of human performances in sports and operates with AI patterns in order to identify break points or potential problems for the players and athletes. Their evidentially accurate system reduces injury days out by an astonishing 75 per cent, while injury rates themselves drop by 70 per cent.
The advantages of the Zone7 system
Zone7 supports clients, who play in the MLS, La Liga or the Champions League and also keep MLB players from being out of action. As of TechCrunch, the company this year raised 2,5 million US dollars in a seed round.
Their system is based on data and AI-powered recognition, so therefore up to date by default. It’s easy to use, since data can be generated via wearables or video technology. And the analysis of millions of data points enables medicals or even managers to make decisions for their training schedules or on the pitch, too. Moreover, these decision can be tailored for every athlete and thereby offer the chance to prevent injuries and optimise individual training practice.
Injury detection and prevention will be high on every club’s medical agenda and is just as important for any and all individual athletes. So, using Zone7 should be considered in order to minimise risks and maximise performance levels – and eventually even the profit. To show you the benefits of the systems in different kinds of situations, we spoke to CEO and Co-founder Tal Brown – founder of Salesforce’s first AI team by the way –, who is just the person to explain and highlight these.
The interview
Spielmacher: Injuries do cost clubs and players playing time and real money, too. Can a more holistic prevention system be a boost for financial resources as well?
Tal Brown: Yes absolutely. The financial benefits of reducing injuries are first and foremost direct in that fewer injuries immediately translate to reduced medical costs. However, many studies show that reduced injuries are strongly correlated with winning and that translates to additional revenue from tickets, endorsements and in some cases – revenues from broadcasting and the league prize money. A good example is Getafe, who are having an extremely successful season (#5 in La Liga, lowest budget, dramatically over-performing vs their budget). One of the financial benefits of such a season is participation in either the UEFA Europa League or the Champions League, estimated to generate a dramatic increase in revenue.
Spielmacher: Fitness coaches, the medical team etc. are there to minimise the injury risks for the club’s players. What’s the unique additional value that Zone7 can offer them in order to do so?
Tal Brown: Zone7 offers two key things: first – we help the medical staff dramatically reduce the amount of time spent on reviewing data. The process of looking at performance, medical and health data every day for 25 players in a squad, and using this data to identify risk and optimal workload is lengthy, so we help complete this much quicker. Also, Zone7 offers accuracy. The mathematical/Artificial Intelligence process we use to estimate risk and identify optimal workload for each player is based on millions of football games and training sessions. This process is statistically quantified to determine accuracy and we ensure this accuracy is continuously improving as we learn from more teams and real-life data.
Spielmacher: The use of AI technology personalised the monitoring to a greater extent. Does that offer the opportunity to react more individually and could reoccurring injury problems be reduced eventually?
Tal Brown: Yes, we believe that the most effective intervention is 100 per cent personalised and must be tailored to each athlete’s performance history, medical profile and any additional attributes. A key part of this is the athlete’s injury history that is taken into consideration with every risk and recommendation, so while we cannot guarantee to eliminate recurring injuries, we are seeing that Zone7’s technology is helping older and more injury-prone players to increase their availability and stay longer at a peak physical condition.
Spielmacher: To what extent can training practices be aligned to data, which give a picture of the players’ health and fitness conditions?
Tal Brown: In training, workload and the effort invested by players can be controlled. This control has always been used to accomplish both tactical goals (e.g. learning/improving in a specific football strategy) and fitness related goals (being better prepared physically for the upcoming game). Rehabilitation is also a good example where the health and physical condition of the player determine the level of effort and the progression of their effort to ensure a safe return from injury into maximum form.
Zone7’s technology allows the technical staff of a football team to break down the player’s current physical condition into fine-grained parameters to understand what kind of medical, biomechanical and workload parameters (sprints, overall distances, accelerations, decelerations, etc.) is currently contributing to risk. The second step is to use Zone7 to define optimal ranges for each workload parameter for each player daily. This approach allows players to train, improve performance levels, but avoid aggravating their current risk factors that will likely lead to injury.
Spielmacher: The AI-powered system of Zone7 works with comparative data. How many matching data sets can be used to develop predictions for a single player for example?
Tal Brown: First, the term prediction should be well-defined: Zone7 does not predict the exact moment physical mechanism of an injury. Instead, Zone7 calculates the Injury Risk Forecast over a period of several days, typically over the days proceeding a match.
Overall, Zone7 has analyzed over 5 million games and training sessions and thousands of injuries. This data is used to create a multi-layered model that is customised for an individual player but can also rely on data from similar players in similar leagues.
Spielmacher: You are able to reduce injury days out by 75, injury rates by 70 percent. Could you give an example of a player or club profiting from that recently?
Tal Brown: Yes, one of our earliest adopters is Getafe CF, currently #5 in La Liga. This team has the 17th highest budget in La Liga, yet in the past two seasons since coming up from the second league have been over-performing dramatically. In addition, Getafe have one of the lowest injury rates in the league.
Interestingly, Getafe has a relatively old squad and yet their injury rates with Zone7 have reduced by 65 per cent. Individual examples should be taken lightly when examining AI products, but such example could be Mathieu Flamini, who, at the age of 35, came into Getafe after playing for top tier European clubs, where he sustained 13 injuries over the previous seasons. This season in Getafe, he has sustained no injuries at Getafe and has had nearly 100 per cent availability.
Spielmacher: Your system works via video or wearables. The latter aren’t common in football, so will video analysis be the main factor there? And could such an analysis prevent players from continuing with a concussion or a hamstring problem rather quickly to prevent further damage?
Tal Brown: Zone7 requires performance data to compute the player workload, injury risk and optimal training levels. The data can be collected through wearables (using GPS, accelerometer, etc.) or through video systems that are becoming more accurate in this respect. Already today, many of our customers provide us with game-day data from video/broadcasting systems instead of GPS.
Until today, Zone7 has helped coaches make unbiased decisions with regards to training load and match load. As more real-time data becomes available (through video as well as the next generation of wearables) then our role can potentially grow into minute-by-minute insights to coaches. Ultimately, the coach is responsible for the team’s results and overall performance and he is the expert for that job. We strive to make health/performance data be available in an easily understood manner so unbiased decisions can be made.
Spielmacher: Could the use of AI technology not only prevent injuries but also improve performance levels in the long term?
Tal Brown: Yes, absolutely. Our strategy is to use AI to help athletes reach their goals. In the professional environment, one key goal is to reduce injuries hence we “teach” our AI platform to provide insights that help achieve this goal. Over time, other goals like reaching specific performance levels can also be a part of the same process. A good example for this is long distance running, where many runners have a specific goal in mind for a race – for example running a marathon in under 3 hours. AI can be used to define the day-to-day training leading to such an event to ensure minimal injury risk while pushing the runner forward to reach his/her goal.
Spielmacher: Do you believe that data driven injury prevention and health care in sports is an essential step towards long term success? Can clubs afford to not take advantage of it at all?
Tal Brown: The club’s ultimate goal is to win more. For some clubs, this means a championship every year, and for others, it means simply over-performing vs their budget. Winning translates to growth in both direct revenues and indirect revenues that further fuel winning in the next year. From a pure data perspective, reduced injury rates are strongly correlated with winning because having the best (and most cohesive) squad available for more days means the team has higher chances of winning more games.
I believe that the medical and performance staff in clubs are going through a major revolution. Today tools are available to visualise and create meaning from data, and so decisions can be more objective through repeatable scientific (and validated) processes/computations. As with other aspects in medicine, coaching and professional sports in general, the expert will always be the one making the decisions, but now there is more science that can support these experts with objective decision-making tools. This will undoubtedly lead to fewer injuries, more availability and more success.
Spielmacher: Is your system rather made for professional sportspersons or equally useful and necessary for semi-pro and other levels?
Tal Brown: Our mission is to help athletes reach their goals while remaining injury free. This is applicable to professional athletes as well as semi-pros and lifestyle athletes. However, Zone7 creates value from data, and so we rely on data to deliver. In professional environments, the data needed (e.g. performance, medical records) is available more consistently and accurately than in other markets. However, over time we are bringing our technology to other environments both in the professional sports (like baseball and basketball) and semi-professional: we are working with hundreds of amateur runners preparing for half and full marathons.
Spielmacher: Has the awareness regarding the need for optimised supervision of the athletes’ health grown in recent years?
Tal Brown: Yes, while supervising the athlete’s health will always remain an internal process driven by medical professionals inside clubs, we see more awareness and more openness to evaluate tools to support the internal decision process. Traditionally these tools have been designed to collect data and visualise data, but are now expanding to tools (like Zone7) that can unveil the valuable meaning within the data or provide an objective and validated calculation for risk and optimised performance levels.
Spielmacher: How can you ensure to keep the sensible data concerning athletes’ health safe? Can data breaches, which could have an impact on players’ developments, be ruled out?
Tal Brown: We use security measures that are equivalent to healthcare standards to ensure the data is safe. These include end to end encryption, advanced authentication methods and of course strict control on data access both for our employees and our users.
Also, as dictated by GDPR, our customers own their data, so they have complete control on what is available and for how long.
While breaches cannot be 100 per cent avoided, we invest heavily to ensure the data security. It’s important to note that already today, many products used by teams (e.g. AMS, wearables) manage sensitive data in the cloud successfully so this is already a common practice in many environments.
Spielmacher: From your personal opinion, will AI only enrich the sports ecosystem or actually take some of its precious unpredictability, too?
Tal Brown: I personally do not think that AI will make the sport more predictable. The human brain, operating in high-speed, reacting to a thousand changing factors in a game that are detected through eyes and ears is and will always remain unpredictable. However, I think the ultimate sport experience will be better because we can eliminate (or at least reduce) some factors that negatively impact the fantastic “drama” that everyone loves to follow. One obvious example is the ability to keep the best players on the stage.
Thanks so much for the interview, Tal. Whoever wants to reduce injury rates or days out to not only save money but strengthen the players’ and athletes’ health sould definitely considers using the very contemporary solutions of Zone7, based on AI and supported by an enormous amount of data.