🇬🇧Football Data Team

How Palmeiras Uses Data to Find a New Endrick

🇧🇷 – para ler este artigo em português clique aqui: https://gustavosantana.online/equipe-de-dados-no-futebol/

Data-driven decisions have become increasingly important in companies across various sectors, but what about in football? To what extent can accurate data influence critical decisions such as the recruitment of new players, the definition of tactical schemes in important matches, or even recommendations for the medical support of players?

The Sociedade Esportiva Palmeiras, a Brazilian football club, founded its own data team, the “Palmeiras Data Science Center,” comprised of data scientists, data engineers, and mathematicians. Other Brazilian football teams also use data in their analyses, but let’s understand how Palmeiras is ahead of this, like no other national club.

Each match produces a vast amount of data, far beyond the data we are already accustomed to, such as shots on goal, fouls, or the number of cards. It is possible to quantify how many times the goalkeeper saved with his right hand, how many times a player controlled with his left leg and shot with his right leg, and specialized companies gather some data and sell it to interested clubs. Instead of simply buying this data generically, Palmeiras is internalizing it, capturing its own data at a lower cost, and using its own data team to evolve, refine, and customize the analyses according to the club’s needs. This brings another level of detail to the team’s decision-making process.

Data extraction is done with artificial intelligence, monitoring the players in the match and capturing it through images, transformed and made available on demand. This helps coach Abel Ferreira position the team tactically, understand more about the opponent, and now they are evolving into the health sector, predicting potential injuries. The generated data provides recommendations to assist doctors in decision-making.

Data-driven recommendations are also aimed at the recruitment market. If the coach needs a player with specific characteristics, instead of sending a scout to understand professionals with that capability, the data center runs an analysis with a model to consult various leagues worldwide. Based on characteristics and data, it is possible to make a more accurate search using analysis and prediction.

For example, if the goal is to recruit a defender, exhaustive data collection is refined to verify situations that allow for more precise analysis. In addition to checking the number of correct passes, the context in which they occurred is analyzed. Out of 15 correct passes made by the defender, how many were really under pressure? Or, in the recruitment of a forward, the analysis can focus on the number of accurate shots on goal from outside the box. The data provides the expected goal, calculates where the shot came from, how many players were in front of the shot, what the goalkeeper’s angle was at the time, and the probability of scoring considering all factors, refining the quality, intelligence, and choice of the forward at the time of recruitment.

Palmeiras is also creating “patterns” and studying data generated by Endrick during his journey at the club to try to generate or discover potential talents like him. And for those who follow the club, this innovation adds another layer of excitement and anticipation to their experience.

_f