Is Thomas Tuchel wrong? Evaluation of hexagonal shaped drills based on machine learning and position data

(Liegt Thomas Tuchel falsch? Bewertung von sechseckigen angeordneten Übungen auf der Grundlage von maschinellem Lernen und Positionsdaten)

As a variation of small-sided games (SSGs), Thomas Tuchel's hexagonal shaped possession drills attracted a lot of attention in coaches` soccer education. Changing certain variables (e.g. pitch size) is one traditional approach to provide an optimal stimulus for a specific training goal. Therefore, the presented field study investigated whether hexagonal shaped drills show changes in tactical key performance indicators (KPIs) using positional data in a controlled experimental setting. Data were collected using player tracking systems (1 Hz) in 5 versus 5 SSGs in both the full-size pitch and pitch with cut-off corners condition. At match-related level, trial duration and outcome were examined. At player-related level, the tactical KPIs effective playing space (EPS), length-per-width ratio, space control and overplayed defenders were analyzed. The results show significant differences in length-per-width ratio (attacking team), the EPS and space control (30m-zone). However, it could not be confirmed that hexagonal shaped drills are played more vertically and faster toward the goal than on a normal pitch, as there were no changes in trial duration, outcome or overplayed defenders. Nevertheless, an experimental positional data analysis paradigm is a useful approach to investigate tactical principles in high-level professional soccer.
© Copyright 2024 International Journal of Sports Science & Coaching. SAGE Publications. Alle Rechte vorbehalten.

Bibliographische Detailangaben
Schlagworte:
Notationen:Spielsportarten
Tagging:Kleinfeld maschinelles Lernen
Veröffentlicht in:International Journal of Sports Science & Coaching
Sprache:Englisch
Veröffentlicht: 2024
Online-Zugang:https://doi.org/10.1177/17479541241255596
Jahrgang:19
Heft:4
Seiten:1697-1705
Dokumentenarten:Artikel
Level:hoch