With the help of data science, Almere City FC is able to map players that are overlooked by the competition. In this way they generate a lasting advantage over other clubs.
Young. Ambitious. Stubborn. These are the core values of Almere City FC. Young, because the club has only existed under this name since 2010. Ambitious, because the plan to be promoted to the highest division within a few years is not under the chairs or benches. Stubborn, as shown by the club's mentality to do things differently from the competition. These core values are reflected throughout the organization, so of course also among the staff. Teun Jacobs, technical manager at Almere City FC, is an excellent example. As technical manager he is responsible for the entire technical football affairs within the club. His original and thoughtful vision can be found in the composition of the first team up to the youth academy. At the age of 35, he is also a young technical manager bursting with ambition.
Partly because of these core values, a collaboration between Almere City FC and Vantage AI has been established. Director John Bes and Teun Jacobs saw opportunities to create a scouting platform based on data and thus distinguish themselves from the competition. “We wanted to make more use of data in detecting valuable players for our first team,” Jacobs said. The goal was to creatively extract insights from the data that exposes inefficiencies in the football market. In this way, a competitive advantage can be achieved in attracting players.
This project fits seamlessly with the vision of Vantage AI. The basis of data science is to look for hidden insights and not let yourself be guided by conventional wisdom. Player scouting is essential to the quality and performance of a football club. Both parties share the opinion that a competitive advantage can only be gained in this area through a different approach than the competition. Each transfer period, clubs in the smaller leagues fish in a large pond for the same players. However, the use of data science can cause the club to fish in a smaller and quieter pond with at least as much quality.
Jacobs knows that this sounds nice in theory, but is not always easy in practice: “During my time at the KNVB, data science projects were also started that did not always have the desired result”. At Vantage AI we also know that the pitfalls and challenges must be properly identified. For example, the presence of sufficient and structured data is essential to make a project successful. We found that a reliable third party provides a consistent data stream, containing all registered actions per match. In addition, we realized that within Vantage AI there had to be knowledge related to the football industry for this project to be successful. There are two main reasons for this. First of all, the football industry differs from other sectors in the way it operates. When Vantage AI consultants do not speak the same language as domain experts (trainers, managers, etc.), the insights and expectations between both parties cannot be optimally exchanged. As a result, the potential value of the project could never be achieved. Second, football is a very complex sport. Scoring is low compared to other sports, it is continuous, it varies over time, it is highly strategic and very subjective. If two people were to analyze the same match, they could come up with different opinions. The key to success is to ask the right questions with the right language. That is one of the things that surprised Jacobs positively. "I experienced the consultants' short lines of communication and expertise as very pleasant, with the essential football link also present."
After the pitfalls were identified, we came up with the solution to gain a competitive advantage in scouting. Many clubs nowadays look at statistics such as assists, duels won or passes that have arrived well. But is every pass or duel in a match equally important? At first glance, a player who wins 8 out of 10 matches seems like a good footballer in that respect. But if you know that these 8 duels won didn't have much added value and the two lost duels came against goals, do you still share the same opinion? The individual actions therefore contain more information than many clubs now use for assessing players. Giving a value to each action could thus generate a huge competitive advantage.
The above value is assigned with the help of an algorithm that is now supporting the Almere City FC scouting process. The algorithm actually acts as a scout for every player on every match. Via a website, which is automatically updated every week, those involved have access to a list of players that the algorithm has rated as potentially interesting purchases. This list is an amalgamation of the preferences that Almere City FC wants to see reflected in players and the results of the algorithm based on the data. In addition, there is access to various more specific insights per player that have been discovered by the algorithm, such as excelling in tackles or the right passing. This substantiates the algorithm's decision for this player. In addition, video images can also easily be searched by the video scouts, because the most interesting actions are shown.
This data science solution delivers the first positive results by bringing out undervalued talent. Interesting players are found who then move on to the next step in the scouting process. “We had the algorithm predicted for all positions and interesting options emerged,” Jacobs says. From here, live scouting can be used more specifically and effectively. Consider, for example, not attending or not attending a particular match with interesting players and reconsidering a doubtful case among scouts. Thanks to these positive findings, a follow-up is on the horizon. Jacobs: “We would like to continue this project and continue to benefit from it!