Cloud competition and sports analytics | Dekiutte Netherlands


Sport Faces Big Data Dilemmas

There are significant risks, as well as rewards, in collecting detailed data on athletes’ performances

Professional athletes are accustomed to being in the public spotlight. But how intense should that spotlight be?

If every aspect of your professional performance was analysed under a microscope by a large number of people, you might feel pretty uncomfortable. But that is now what happens in sport: broadcasters provide football fans, for example, with detailed data on everything from how fast a player runs to the number of tackles they make to the passes they compete. If the statistics show an ongoing decline, the athlete could be subject to public criticism or humiliation, and their personal brand could be damaged.
A combination of high definition video footage, wearable sensors and artificial intelligence (AI) has made it feasible to analyse trends in athletes’ performance in unparalleled detail. While this kind of high-tech scrutiny can make sport more competitive and engaging, it needs to be accompanied with measures to protect personal privacy and data security.
With performance data proliferating fast, the sport ecosystem should begin to address this challenge now. The multiple commercial drivers for collecting and sharing detailed data need to be balanced with a regulatory and ethical requirement to ensure that individuals give their explicit and informed consent.  
As a key pillar of data protection legislation, transparency is crucial. Athletes need to fully understand how their performance data could be employed. Will the data be used as the basis of betting odds, to select a team, to decide a bonus or even negotiate a future contract? 

Context can be king

Although detailed performance statistics can make a sport more enjoyable to watch, an athlete can reasonably argue that they often lack important context – the number of tackles a defender makes in a particular football or hockey game might hinge on the tactics employed by the opposing team, for example. This kind of context could be important, particularly if the data is being used commercially. 
In the UK, a group of 400 former professional footballers are reported to be backing a planned lawsuit against video game developers, gambling firms and data processing companies over the use and commercialisation of performance data and statistics relating to them. These claimants are likely to argue that they haven’t had the opportunity to change data they feel misrepresents them – a requirement of both the UK’s and EU’s General Data Protection Regulation (GDPR). 
The GDPR also affords stringent protections for sensitive data, such as health-related information. In some cases, the data collected by wearable sensors could fall into that category as it may give indications about an individual’s underlying health. Data related to heart rate or lung capacity, for example, could be of interest to health insurers or even potential employers.  

Even the amateurs need protection

Of course, privacy concerns don’t just apply to professional sports people. The falling cost of wearables and video production is making it feasible to analyse the performance of amateur athletes in fine detail. While some amateurs may understand the risks and embrace this kind of analysis, children may need special protections. Should a football or hockey club be allowed to collect detailed data on 14-year-old players, for example? 
Even if there are safeguards in place to prevent performance data from being shared publicly, there is always the risk of security breaches that could compromise an athlete’s privacy. Europeans also need to be aware that, in many cases, their performance data is being collected by companies based somewhere else in the world, typically, the U.S. or perhaps China. Although all service providers operating and/or targeting data subjects in the EU need to comply with the GDPR, a third-party government may seek to override the provisions in that legislation.  

Clarity and control required

In light of such risks, each sport needs to figure out who will control performance data and its usage – is it the league, the team or the individual player themselves? It may be that different models apply to data collected in different ways. Data collected by wearables may be under the control of the player, for example, whereas data derived from video footage may be controlled by the league and licensed through broadcast rights.   
Once ‘controllership’ is clear, this party should be responsible for managing the usage and security of the data in a way that maximises the long-term benefits to all the stakeholders involved. It may be, for example, that individual players should be allowed to sell their data rights in the same way that they sell their image rights today. 

Anonymity versus opacity

When it comes to commercialisation, automated anonymization and aggregation tools will help. Software can enable third parties to draw insights from a dataset without actually seeing the data. For example, such systems could enable a gambling company to see how the overall movement of a football team is changing during a match without seeing data relating to individual players. Similarly, such software could enable video game developers to derive scores for dribbling, heading, passing and other technical skills without actually seeing the underlying data.  
Indeed, rapid advances in AI are making it easier for a wide range of stakeholders to extract valuable insights from performance data, without the need for manual analysis. But where AI is perceived to be opaque, flawed or unfair, there will surely be a backlash against so-called black box systems. A new proposal for EU regulation on AI is designed to introduce a number of safeguards to ensure that such systems are both transparent and non-discriminatory.   
 But crafting future-proof regulation is tough. One of the biggest challenges for the sports ecosystem will be anticipating how technological advances could change the usage and value of performance data over time. For example, collectors may want to own the original dataset relating to a landmark sports performance. Meanwhile, the gambling industry may develop new forms of betting that draw on real-time biometric data, while game developers may use performance data to create virtual worlds in which consumers compete against real athletes.  
The first step is to discuss the opportunities and the related risks with all the key stakeholders. Such discussions would be the precursor to developing robust agreements that are in the best long-term interests of the sport. In our view, the athlete should be at the center of every decision and conversation.  
Without transparency and trust, sport will struggle. 

Get in touch

If you would like to discuss the risks and rewards big data has on the world of sports please get in touch.

Did you find this useful?