Parameterized Search

Finding the box-to-box with data

We are looking for players who fit the box-to-box profile to understand how to choose specific metrics that outline a specific profile.

Is that recognizable box-to-box style still present from the early 2000s, when the double pivot of the 4-2-3-1 was the most commonly used system? It is likely that this specific role has been gradually fading, but not necessarily in favor of something concrete and demanded by the game. Now, midfielders, in addition to being good in defensive and offensive areas, aerial duels, or finishing, do more things and are not clearly seen with that label. We can think of Jude Bellingham, who would undoubtedly be the partner of a central midfielder in a 4-2-3-1 formation of a 2005 team.

Certainly, a great example of this would be Federico Valverde, a paradigmatic box-to-box player, who also does more things but is undoubtedly decisive in recovering balls, carrying the ball, and shooting when approaching the box. Finding such a profile may not be easy, but if one has a specific role in mind among those emerging in football, it can be molded through the selection of metrics and their volume and accuracy in each of them.

For this, it is important to decide which parameters seem most relevant to identify a specific profile. If we are clear that we are looking for a dynamic midfielder who is successful in defensive duels and generates shots and ball progressions in the opponent's half as key aspects, we can add and complement with different metrics until obtaining a more specific profile. For example, we choose metrics and volumes that ensure they are above average in each of them:

Yes, not all midfielders with these characteristics and volumes are box-to-box, and those who do not excel in some of them also remain as such. However, the data enables a flexible framework of analysis and decision-making that brings us closer to the profile. Using these metrics, we have found, for example, two cases where you probably agree with the label. Roberto Gagliardini and Christian Norgaard represent something quite similar to the dynamic midfielder we might have in mind, and they perform notably different tasks between a defensive and a more offensive phase.

Heat maps are an interesting graphical support tool to contextualize and consolidate the sought-after profiles. In addition to Gagliardini or Norgaard, we find Adrien Rabiot and Baptiste Santamaría. Do they seem similar as players? The truth is that they don't. The first fits much more with the classic box-to-box profile, while the second is a high-paced player but more creative in passing.

Getting used to using data to profile players, refining metrics, finding similarities to go to the video with prior work done is much more useful. This search has been conducted with a part of everything we have at our disposal, indicating that data works and can be of extraordinary value in the hands of great professionals with extensive football and data knowledge.

Founded in 2017 as a consultancy, Driblab has driven innovation through data in all aspects of professional football. Thanks to a transversal model, its database collects and models statistics in all directions. From converting matches and videos into bespoke data for training academies to developing cutting-edge technology, helping clubs, federations and representative agencies in talent scouting and transfer markets. Driblab’s smart data is used by clubs all over the world, with success stories such as Dinamo Zagreb, Real Betis and Girondins Bordeaux among others. Here you can find out more about how we work and what we offer.

Melde dich noch heute für unseren Newsletter an

Wir veröffentlichen jede Woche einen Newsletter mit verschiedenen Analysen, neuen Kennzahlen, neuen Talenten und statistischen Berichten.

STARTE JETZT

Kontaktieren Sie uns, fordern Sie eine Demo an und seien Sie Ihren Mitbewerbern einen Schritt voraus