Back to Search
Start Over
Soccer goalkeeper expertise identification based on eye movements.
- Source :
-
PloS one [PLoS One] 2021 May 19; Vol. 16 (5), pp. e0251070. Date of Electronic Publication: 2021 May 19 (Print Publication: 2021). - Publication Year :
- 2021
-
Abstract
- By focusing on high experimental control and realistic presentation, the latest research in expertise assessment of soccer players demonstrates the importance of perceptual skills, especially in decision making. Our work captured omnidirectional in-field scenes displayed through virtual reality glasses to 12 expert players (picked by DFB), 10 regional league intermediate players, and13 novice soccer goalkeepers in order to assess the perceptual skills of athletes in an optimized manner. All scenes were shown from the perspective of the same natural goalkeeper and ended after the return pass to that goalkeeper. Based on the gaze behavior of each player, we classified their expertise with common machine learning techniques. Our results show that eye movements contain highly informative features and thus enable a classification of goalkeepers between three stages of expertise, namely elite youth player, regional league player, and novice, at a high accuracy of 78.2%. This research underscores the importance of eye tracking and machine learning in perceptual expertise research and paves the way for perceptual-cognitive diagnosis as well as future training systems.<br />Competing Interests: The authors have declared that no competing interests exist.
- Subjects :
- Adolescent
Adult
Athletic Performance classification
Cognition physiology
Decision Making
Eye-Tracking Technology
Germany
Humans
Machine Learning
Male
Perception physiology
Virtual Reality
Young Adult
Athletes psychology
Athletic Performance physiology
Athletic Performance psychology
Eye Movements physiology
Soccer physiology
Soccer psychology
Subjects
Details
- Language :
- English
- ISSN :
- 1932-6203
- Volume :
- 16
- Issue :
- 5
- Database :
- MEDLINE
- Journal :
- PloS one
- Publication Type :
- Academic Journal
- Accession number :
- 34010305
- Full Text :
- https://doi.org/10.1371/journal.pone.0251070