How do you Organise Analytics in Sports?

30 May 2018

On Monday 7 May, we were at the “Analytics in Sports” conference at the Johan Cruijff ArenA in Amsterdam. After a bit of a kickabout on the pitch, we had an extremely interesting and inspiring day. On the day there were 39 speakers covering three different tracks: Sports Teams, Federations and Leagues; Brands, Fans and Engagements; and Sports Technology, Technical and Data Science. We’re going to take you through what we learned!


Johan Cruijff ArenA

The Johan Cruijff ArenA which hosted the event also gave two talks. In the first talk, Henk van Raan (the stadium’s CIO) talked about how the ArenA came about, how it was built in 1993 when sustainability wasn’t quite such an important topic as it is today. They have set the bar high for the next few years (f.e. carbon neutral) and agreed with the municipalities that they are going to be a pioneer in sustainability for South East Amsterdam, and Amsterdam more widely. That was the hook into the rest of the objectives which, as well as infrastructure, security and facility management, also include customer journey and fan experience too. The Johan Cruijff ArenA doesn’t just want to become a Smart Stadium, it wants to be part, the leading part in fact, of Smart City South East. They have unlocked €50 million to improve the Fan Experience in the years ahead. This is what the second talk from ArenA, by Sander van Stiphout and Max Reckers, looked at more closely. They showed the audience some dashboards with insights into subjects from ticketing to turf quality. These could be displayed in real time. The specific examples that Max Reckers highlighted were particularly interesting. It was about what data you can use for insights and can quickly take decisions based on them.




Rick Cost is the Physical Training Coordinator at Feyenoord. Rick made a fascinating address introducing us to the role played by data & analytics in the structure and set-up of training at Feyenoord. The greatest challenge in this is finding the ideal combination between science (chiefly data science) and practical application. The result being that Feyenoord’s players can physically cope with playing their best football for a full 90 minutes.

Rick is operating ‘on the edge of chaos’. This means his methods are not based on static, prolonged and repetitive exercises. His training also consists of short, dynamic and match-focused exercises. The ultimate aim of his methodology is to make players faster, more nimble and more flexible. These aims are fixed within the very DNA of the club. It comprises five specific aims that every Feyenoord player – and therefore his training too – must satisfy. With the help of data & analytics, Rick developed a model that allows him to then complete and measure every training session. The model has two parameters:

  • Intensity of the match, based on the ‘worst case minute’
  • Volume of the match, based on the workload.

Armed with the outcomes from this methodology, Rick is able to divide training sessions into periods, collect feedback on training intensity, train specific positions for a match and set up specific exercises for a match. The next step for the model is to use internal parameters which can determine the correlation between the mental and physical condition of a player.


Sports Technology, Technical and Data Science

Data-savvy sport-lovers like us were most drawn to the Sport Technology, Technical and Data Science component. So of course this is the area in which we attended the most talks. There were some smaller talks. Johan spoke about their public APIs to encourage data-sharing; Game On about real-time bird’s-eye video for trainers; and Mylabs about how they measure times in mass events such as running competitions.

One of the most striking trends in Sport Technology is automated video. With this technology, in the future all clubs, whether amateur or otherwise, with limited resources will be able to produce a professional television broadcast of a match.

Some companies focus mainly on the technology. For example, they place three cameras along the football pitch and stick the pictures together. An algorithm can spot where the ball is and where the bulk of the players are. This makes it possible to film a match in the way we are used to, without the need for cameramen and a director. Other companies developed player recognition technologies whereby statistics can be displayed automatically at the right moment. After a goal is scored or a substitution is made, for example.

These technologies are bound to mean that in the not-so-distant future virtually every amateur match will be made available online. And then what comes next? Perhaps automated match commentary as we are already used to it with FIFA…



Also at the Sport Technology track, SciSports told us about their story and their ambitions. SciSports are making pretty good progress, they have received plenty of investments recently and once again the room was jam-packed.

The SciSports story started when they set out to rebuild Football Manager with real data and the founder wrote his thesis at FC Twente on how data can help with scouting. Since then, they have grown strongly and have a purely data-based scoring system to rank players: Neymar is currently the best player in the world and De Ligt is the best under 19. So therefore it’s no wonder that De Ligt won the ‘Johan Cruijff Talent of the Year’ Prize on Tuesday.

Based on this data, they can help players and clubs to find the right match. They recommended that Memphis Depay join Lyon. They were able to show that when Weghorst left Emmen on a free transfer he was always in the right place and scoring the right goals and that he was therefore a good buy. They were also able to see that Jaap Stam was unfairly dismissed because Ferguson saw that he was making fewer tackles, whereas actually he was actually making his tackles more efficient. What’s more, SciSports is also a long way into the development of BallJames. A system with fourteen cameras around the pitch that makes a 3D display of the game. Based on this positional and event data, they can follow the players, referees and the ball. They can then turn this into statistics for each player or run simulations on different tactics. At both the player level and the team level. This could eventually even serve for VR or hologram purposes.



Jerome Durussel, a Data Scientist at Catapult Sports, gave an interesting presentation on wearables. These little boxes have a variety of sensors to pass on data about an athlete’s performance. These wearables allow analysts to track the performance and strain of a body and compare them with other measurements.
One of their many applications is to see whether an athlete has fully recovered from an injury. From the outside, the athlete may seem to be putting their body under full strain again so they must be fit. But often the data begs to differ. It allows you to see abnormalities compared to previous measurements. For example, these can be the result of an ever-so-slightly different running pattern to (consciously or subconsciously) protect the damaged muscles. In such a situation, the data helps to avoid any further injuries being incurred.

The challenge in this area is to obtain a good benchmark for each person. After all, when is someone in their “normal” state? At the start of the season they may be “too fit”, meaning the benchmark ends up too high. Challenges that still remain unresolved, yet when you consider the pace of development, there will surely soon be a solution to this too.



One of the day’s most interesting presentations came from Xsens. Xsens has developed an advanced suit that can map movements in 3D. It means you can determine the position, speed and direction of various body parts with extreme accuracy.

The suit is actually made up on a collection of sensors that are constantly giving out signals about their current position in relation to other sensors on the body. Because these sensors are stuck onto various body parts (e.g. forearm, upper arm, lower back, hamstring etc.), it can put together a simulation of the posture and movements of the test subject. The simulation can, for example, be used to calculate the forces on joints.

The possible applications for this technology are endless. The system is being used by car manufacturers to work out the strain on muscles when getting into a car. And occupational health and safety organisations can keep an eye on the posture of people suffering from back problems. But Xsens can provide interesting insights in the world of sport too. How much extra jumping power does an athlete need to exert in order to make that extra turn? Or how can a rower’s posture be improved?
Xsens is therefore giving sportspeople a whole mountain of data and new insights. And if there’s one thing that became clear during this conference, it’s that the world of sport is playing major catch-up in the use and acceptance of data. So we will certainly be hearing more from Xsens!

Xsens Presentation



In his brisk Australian accent, the CEO and founder, Luke McCoy, takes us through the concept behind his platform. His aim: to provide automated production of matches for clubs that don’t have the budget themselves.
This product is a response to clubs’ need to broadcast matches for their – often local – fans and, if possible, also to make some money from it through advertising.

These matches aren’t enormously appealing for advertisers, as the viewer numbers are often low. What’s more, there is also a lot of variation in production, information displays, advertisements, etc. The added value therefore comes when you get large numbers by putting all those little clubs together and having a standard template to show advertisements. And that’s exactly how LIGR is trying to link together advertisers and clubs.

LIGR is a production platform that puts the information and advertising layers onto the pictures and provides a standard template to make the whole product look more professional. They buy the match information – goals, red cards, etc. – themselves and show an advertisement for one of the advertisers at key moments in the match. The clubs only need to buy a decent camera system, ideally an automated one. Luke mentions Pixellot, amongst others, as a supplier that makes “filming without a cameraman” possible.

Then Mr McCoy demonstrated his product. We watched an amateur football match, neatly displaying the player information and stand before kick-off. When there’s a goal, then an animation by sponsor Veolia hurtles across the full width of the screen. Not very subtle, but it does look slick. Especially when you consider that the system calculates the timing of the information layers automatically. There is no more human intervention in the process, which makes it a scalable overall concept.

LIGR’s to-do list includes adding more live information. For example, above the player’s head you could see how many points he has scored during this match so far or whether he has any yellow cards. All in all it the Australian gave an impressive talk on the automated creation of content and potentially a good idea for your local korfball club.

Put simply, it was a really inspiring day!

Simon, Kjeld, Hugo, Willem, Joost and Jeroen

Latest news

In 2022, we are going to have a great year of celebration! Many weddings expected

21 February 2022

Plan your calendar free and make sure you have plenty of party clothes in your closet,... read more

Can you escape our “Power BI Escaperoom”?

7 April 2021

In these boring lockdown-times we are all desperately looking for ways to still interact with our... read more

Building a book recommender from scratch

1 April 2021

Almost every day we go online we encounter recommender systems; if you are listening to your... read more

Subscribe to our newsletter

Never miss anything in the field of advanced analytics, data science and its application within organizations!