Application of global positioning system technology and
machine learning to characterise the activity profile of elite
Gaelic football referees
Brady, AidanORCID: 0000-0002-9427-5771
(2023)
Application of global positioning system technology and
machine learning to characterise the activity profile of elite
Gaelic football referees.
PhD thesis, Dublin City University.
Gaelic football is officiated by a referee whose decisions can influence the outcome of the game. To ensure optimal positioning for decision making, Gaelic football referees (GFR) must keep up with play at all times which poses a considerable physical challenge. Currently no data are available on the activity profile of GFR during match play.
Study 1 applied a two-stage unsupervised clustering technique to activity data to generate a set of unique movement category velocity thresholds for elite GFR. The activity data were collected during 338 All-Ireland Championship (AIC) and National Football League (NFL) games using 10 Hz global positioning system technology. The newly generated thresholds were more strongly associated with the raw velocity data and resulted in a significant difference in the distribution of the total distance in comparison to arbitrary thresholds.
Study 2 and Study 3 examined the activity profile and temporal changes of elite GFR during 206 AIC and 132 NFL games, respectively, using the newly generated movement categories. During both AIC and NFL games, the relative distance and the distance covered in each movement category, except very high-speed running, was significantly different between halves. The first quarter was the most intense period of the game with a significantly greater relative and high-speed running (HSR) distance compared to any other quarter of match play. There were minimal differences in the activity profile between the three phases of the AIC and the four divisions of the NFL.
Study 4 compared the activity profile of elite GFR between the NFL and AIC, and examined the influence of age, experience, and aerobic fitness on the activity profile during both competitions. Games in the AIC were significantly longer in duration but minimal difference was found in the activity profile between NFL and AIC games. Relative and HSR distance during AIC games was influenced by aerobic fitness level. Age and experience level did not influence the activity profile during NFL or AIC games.
In summary, this research generated a unique set of velocity-based movement categories that are specific to elite GFR. These categories facilitated the detailed characterisation of the activity profile of elite GFR during both NFL and AIC games. It is envisaged that this information will assist in the design of sport specific conditioning programmes for elite GFR.
Irish Research Council, Gaelic Athletic Association
ID Code:
27306
Deposited On:
17 Nov 2022 14:33 by
Niall Moyna
. Last Modified 03 Apr 2023 10:15
Documents
Full text available as:
PDF
- Archive staff only. This file is embargoed until 8 February 2026
- Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader Creative Commons: Attribution-Noncommercial-No Derivative Works 4.0 4MB