26 August 2016
Analysis and Scouting: Evaluating a Player’s True Worth
We all know the story of Billy Beane and the Oakland A’s, right? How, using maths, the second poorest team in Major League Baseball went on a record-breaking winning streak and changed the way baseball scouts and general managers operated. The story was turned into a Hollywood blockbuster and the word ‘Moneyball’ became synonymous with statistical analysis in sport. Well, a similar story is underway in football. By combining the roles of head of scouting and head of analysis, Daniel Stenz is changing the game.
CONQA spoke with Stenz on how he is implementing the use of modern technology to map talent.
You’ve probably never heard of him, but Daniel Stenz is one of the most innovative and revolutionary minds in world football. By assuming the combined responsibilities of two leading roles within a team, this 36 year old from the small town of Mayen, Germany, is setting a benchmark that all elite clubs and national teams will soon follow.
Stenz has recently been appointed as the Technical Director at the Chinese Football Club Shandong Luneng Taishan FC, after stints with the Hungarian National Team, the Vancouver Whitecaps in Major League Soccer, as well as FC Union Berlin and FC Köln (Cologne) in Germany. By amalgamating these two positions which were previously considered mutually exclusive, Stenz and his teams gain a unique advantage over the competition.
“Scouting is external evaluation and evaluation based on data analysis is internal scouting,” Stenz explains. “If we combine the two roles we are able to create a system whereby we value the players we have as well as the players that are out there according to the same variables. This way we create consistencies in how we develop internal talent and how we scout external talent.”
The idea first came to Stenz back in 2004 while working as an intern in Cologne. He noticed that there was no central database within the club where a coach could access information regarding a specific player. If a strength and conditioning coach wanted to learn about the medical condition of a player, he would have to speak directly with one of the analysts.
More worryingly, a coach responsible for the tactical and technical development of players, when leaving the club, would take all the information he had learned during his stay at the club with him. What made this archaic structure all the more baffling was that a computer game, Football Manager, had a more sophisticated approach.
In the simulation game, players assume the role of the manager of a club and have access to a centralised database where every piece of information regarding a player is accessible under a single player profile.
As a fan of the game, this provided Stenz with the inspiration that he needed. “I realised that if we implemented a similar structure with a single profile for each player in our team, we would be able to access and store information that we could use to change the way we played on the field, the way we developed youth players and the way we scouted. More importantly, we would be doing all of this under the unified philosophy of the club.”
By using a simple tagging system where Stenz and his team of analysts could categorise information – for example, ‘injury’, or ‘successful pass’ – coaches and trainers could keep tabs on a particular player’s development. Installing an access rights structure ensured that information was only made available to those who would benefit from it. “It was not easy in the beginning as some of the older coaches and scouts struggled to grasp the new technology but once it got going it made our jobs so much more efficient,” Stenz says.
The first step in the process of using data to evaluate the worth of a player requires agreed upon definitions of what you are observing and analysing. These are divided into two categories: hard criteria and soft criteria.
Hard criteria are easily defined events such as goals scored or time on the ball. These stats give the casual observer a general analysis on the game and are universally understood. However, soft criteria require a nuanced understanding of the game and their definitions are hotly debated. For example, a duel is an encounter between two players competing for the ball. As simple as that sounds, there is a degree of subjectivity. Stenz explains:
“I might consider a duel to be an encounter where two players compete for a ball of which neither has possession. You might consider a duel as any battle for the ball including tackles. I might limit it to balls on the ground and you might include aerial encounters. If we both have different definitions, but use the same algorithms, we will have very different results which will yield wholly different player evaluations.”
In order to create cohesion throughout the domestic system in Germany, the Bundesliga assembled an expert group of analysts to compile a list of definitions and establish a database for all players competing in the top two divisions of German Football. Stenz represented Berlin alongside compatriots from Hamburger SV, Cologne, Bayern Munich and Borussia Dortmund.
This crack team wrote a 200 page book detailing the precise definitions of every possible event on a football field and what constituted a success or failure. Is the right back out of position when a counter attack breaks on the opposite flank? Did the holding midfielder cover enough ground when the opposition had the ball? What constitutes a successful through ball in the attacking third?
By establishing universal definitions to answer these questions, Bundesliga managers and scouts were able to evaluate the true worth of a player. But Germany was on a different plane to the rest of Europe, and a lack of consistency proved how disparaging a disagreement in definitions can be.
Before 2013, the Bundesliga was the only major European league that did not use Opta Sports as their data mining company. “This caused massive problems,” Stenz says. “We had an incident where Opta evaluated that the Borussia Dortmund defender, Neven Subotíc only won 25% of his duels. We had that figure at 80%. The difference was the result of different definitions.”
As a centre back, the amount of duels won by Subotíc greatly influences his evaluated worth. Winning 25% of them is mediocre at best. However, 80% places him at an elite level. Stenz cautions against using data and trusting numbers that you yourself did not have an active hand in calculating. He says, “If you spend time writing algorithms, and you use someone else’s data set, you will get an entirely different outcome. You have to be in control, from defining the definitions through to mining the data and then to evaluating its worth.”
This led Stenz to reassess the way his teams were evaluating and identifying talent. He became the middle man between those actively searching for external prospects and for those honing the skills of players already on the books.
“I see myself as a translator,” Stenz says. “I translate the IT to the coaches and I translate the technical elements to the IT guys who are analysing the data.” By placing himself firmly in the middle of the two traditionally separated spheres, Stenz has bridged a gap that few sports practitioners may have even known existed.
With his feet in both camps, Stenz links the philosophies of the club and the way the manager wants his team to play and the data that is collected. He explains how a previous manager was in need of a midfielder who could switch the ball quickly and with quality once his side had won possession.
Stenz, with both his scouting and analysis hats on, then defined the terms ‘switching’ and ‘quality’. He put a time restraint on the play, set a parameter on the distance the ball needed to travel and set a bench mark for the outcome of the pass.
“That was a 200 page algorithm,” Stenz says. “Once you have that algorithm and overlay that over the data, not just for your league, but the data from England, Spain, France, Italy and the Netherlands, you find a collection of players that fulfil that specific requirement.” Combine that with the specific variables that your centre midfielder must have, say, a certain percent of duels won as well as distance covered per game, and you create a system where you can track talent globally.
This method of scouting not only improved the way external talent was evaluated, but provided Stenz with a quality measurement of the scouts themselves. If a scout submitted a report on a young striker from a foreign league and rated him as a 9 out of 10, Stenz would then use a video analysis and crunch the numbers.
If his evaluation was substantially different to that of his scouts, a conversation with the scout would then need to take place. “I’d have to try and ascertain why my scout’s gut feel was so wrong,” says Stenz. “Human biases often get in the way of accurate scouting and sometimes an average player in a very good team can seem better than he is.”
Data analysis and event tracking can also be used to help prevent injuries. As mentioned before, with centralised player profiles, physiotherapists and strength and conditioning coaches can use deductive reasoning to determine the root cause of the injury. If a player has injured his hamstring, coaches can work out exactly how many kilometres he ran or how many long passes he made before the injury took place.
Stenz reasons that, with the amount of money spent on players today, such as the eye watering £89.3 million Manchester United payed to re-sign Paul Pogba, clubs should be using every tool at their disposal to ensure the longevity of their investment. By measuring every variable, no matter how seemingly inconsequential, coaches and trainers can dilute a player’s performance, cause of injury or potential value to its purest form. This can only happen if every piece of information is recorded.
“You can’t go back in time and measure something again,” Stenz points out. “Rather have a wealth of information you don’t need than cry because you missed something that you need. Sometimes the medical staff or trainers get annoyed with me because I am always measuring the small details. But my job is to measure the data on a very wide bandwidth and to gain as much information as possible.”
But data is simply a tool and all the information in the world won’t guarantee success. As Stenz says, “If I buy my mother a Ferrari, she will still drive it like a Volkswagen.”
At Vancouver, Stenz’s approach saw the club field the youngest team in last year’s MLS season. During his time with the Whitecaps, they won the domestic cup, twice qualified for the playoffs and gave eight home grown players game time. Few other clubs in world football, with the possible exception of AZ Alkmaar in the Netherlands, can rival the Canadians when it comes to nurturing locally produced talent.
Now with Hungary, a nation once regarded as a colossus in world football, Stenz has taken a gigantic leap towards cementing himself as a revolutionary in the sport. At the recent European Championships in France, the Nemzeti Tizenegy (National Eleven) produced their best performance in a major championship for over 40 years by qualifying for the last 16. With fellow German, Benrd Storck, as head coach, a giant of the game is slowly waking up.