Match analysis of soccer refereeing using spatiotemporal data: A case study

Bruno Gonçalves, Diogo Coutinho, Bruno Travassos, João Brito, Pedro Figueiredo

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)


This case study explored how spatiotemporal data can develop key metrics to evaluate and understand elite soccer referees’ performance during one elite soccer match. The dynamic position of players from both teams, the ball and three elite referees allowed to capture the following performance metrics: (i) assistant referees: alignment with the second last defender; (ii) referee: referee diagonal movement—a position density was computed and a principal component analysis was carried to identify the directions of greatest variability; and (iii) referee: assessing the distance from the referee to the ball. All computations were processed when the ball was in-play and separated by 1st and 2nd halves. The first metric showed an alignment lower than 1 m between the assistant referee and the second last defender. The second metric showed that in the 1st half, the referee position ellipsis area was 548 m2, which increased during the 2nd half (671 m2 ). The third metric showed an increase in the distance from the referee to the ball and >80% of the distance between 5–30 m during the 2nd half. The findings may be used as a starting point to elaborate normative behavior models from the referee’s movement performance in soccer.

Original languageEnglish
Article number2541
Issue number7
Publication statusPublished - Apr 1 2021
Externally publishedYes


  • Performance analysis
  • Positional data
  • Refereeing
  • Tactical positioning
  • Tracking systems

ASJC Scopus subject areas

  • Analytical Chemistry
  • Information Systems
  • Atomic and Molecular Physics, and Optics
  • Biochemistry
  • Instrumentation
  • Electrical and Electronic Engineering


Dive into the research topics of 'Match analysis of soccer refereeing using spatiotemporal data: A case study'. Together they form a unique fingerprint.

Cite this