Secondary Logo

Journal Logo

Original Research

Physical and Decision-Making Demands of Australian Football Umpires During Competitive Matches

Elsworthy, Nathan1; Burke, Darren2; Scott, Brendan R.1; Stevens, Christopher J.1; Dascombe, Ben J.1,3

Author Information
Journal of Strength and Conditioning Research: December 2014 - Volume 28 - Issue 12 - p 3502-3507
doi: 10.1519/JSC.0000000000000567
  • Free



Australian football (AF) officials are responsible for enforcing the laws of the game and ensuring that matches are played in a fair and safe manner. The on-field AF officials, commonly known as “field umpires,” are responsible for officiating play on large fields (between 135 and 185 m long, 110–155 m wide), with 36 players on the field at 1 time. Given this large playing area, 3 field umpires are appointed to elite matches and officiate together without the benefit of being interchanged. As such, field umpires typically rotate between 3 main areas of the field, specifically the mid-zone and 2 end-zones to equally share the match demands (Figure 1). The umpire officiating the mid-zone typically undertakes greater work demands, given that this is the largest zone and the ball frequently transits through this region. The timing of rotations is largely determined on the field by the umpires; however, they commonly occur after a goal or during a delay in play (i.e., ball out of bounds, player injury). Previous studies have reported that field umpires cover between 10,000 and 11,500 m using the 3 umpire system, with high-speed running (HSR; >14.4 km·h−1) accounting for ∼3,000 m (5,7). However, little is known about the differences in physical demands between the separate umpiring positions.

Figure 1
Figure 1:
Field coverage areas of the mid-zone (no fill) and end-zone (grey fill) umpire positions in Australian football.

Although past literature has reported on the physical demands AF umpiring (5,7), the other requirements of officials (i.e., decision-making requirements) are not well understood. The most important role of umpires is to determine whether the laws of the game have been breached by a player, and if so, apply the appropriate course of action by way of a free kick (8). Such decisions are made in a dynamic multiperson environment and during periods of intense physical work. The ability of umpires to detect law infringements from video-based methods has concluded that elite AF umpires can correctly identify illegal acts of play on 92% of occasions (13). Although video-based assessments provide valid examination of decision-making performance, they do not replicate the physical, physiological, and psychological aspects of an actual match environment.

Specific to match play, soccer referees make up to 140 observable decisions, with ∼45 free kicks directly related to breaches of the rules (8). Other studies have demonstrated that soccer referees correctly adjudicate between 64 and 86% of free-kick decisions (14,15). However, free-kick decisions only account for a limited number of the total decision-making demands, and therefore when encompassing all decisions, it is likely that the accuracy would be considerably higher. With the limited information available regarding the match demands of AF umpires, this study aims to quantify the current demands of elite AF umpiring. Furthermore, as no previous studies have compared the separate areas of the field (i.e., mid-zone vs. end-zone), this will be examined. These data detail the match running requirements of elite AF umpires, specifically describing the variations between different on-field umpiring positions, which is particularly important when prescribing conditioning training sessions. Lastly, the most important role of AF umpiring (decision making) will be assessed to determine the decision-making performance during match play.


Experimental Approach to the Problem

To analyze the current physical and decision-making demands of AF field umpires during elite match play and to examine the differences in these demands in different areas of the field. Firstly, using time-motion analysis methods, this study quantified the overall physical demands associated with match play, as well as comparing the demands of the mid-zone and end-zone positions, typical of AF umpiring. Secondly, the study also determined the decision-making requirements by assessing the number of free kicks awarded and the accuracy of these decisions during elite AF matches.


Male field umpires (n = 29; age, 32.4 ± 6.1 years [range, 22–41 years]; height, 181.7 ± 5.7 cm; body mass, 73.6 ± 5.4 kg; ∑7 skinfolds, 49.6 ± 7.4 mm) volunteered to participate in this study. Each participant had 6.4 ± 5.4 years experience of umpiring in elite AF competition, and all were currently officiating in the Australian Football League (AFL). Before the commencement of any testing procedures, approval was granted by The University of Newcastle Human Ethics Committee (approval number, H-2012-0045) and participants provided informed written consent. This study and its methods were supported by the AFL. Match data were collected on the 3 field umpires per match across 20 matches, resulting in a total of 58 files for analysis (2 files were excluded because of incomplete data files). Individual AF umpires were examined between 1 and 4 times over the course of the data collection period.


Free-kick decisions were identified post hoc by 3 experienced AF umpire coaches. Each AFL umpire coach had previously umpired at least 300 elite AFL matches and had been employed as an elite AFL umpire coach for between 7 and 12 years. Within 3 days after each match, coaches viewed each individual match using television broadcast footage and 3 additional camera angles (full field from behind the goal, wide-angle side view from halfway, and close-up angle vision). Only free-kick situations with complete agreement between coaches were used for analysis as per Mallo et al. (14). In AF, free kicks are “awarded to or against a player, as the case may be when a field umpire considers any circumstances set out in Law 15” (1). Although free kicks only represent a small portion of the overall decision-making demands, they are easily measurable and require more complex cognitive processing and rule interpretation. Other decisions such as marks, out of bounds, and scoring decisions were excluded from analysis because they represent general play scenarios that occur on a very frequent basis or are adjudicated by the boundary or goal umpires.

Free kicks were assessed as either correct (i.e., field umpire awarded a free kick to a player who was infringed), missed (i.e., the umpire failed to award a free kick to a player who was infringed and a free kick should have been awarded to a player), or unwarranted (i.e., umpire awarded a free kick to a player who was not infringed and no free kick should have been awarded in these circumstances). Missed and unwarranted free kicks were further classified as incorrect decisions. The nature of the infringement, the area it occurred on the playing field (mid-zone or end-zone), time and quarter of the infringement, and the players and umpires involved were also recorded. Free-kick accuracy was calculated as the number of correct decisions divided by the total number of free kicks awarded by each umpire during each quarter of a match. To examine intra-observer reliability of the AF umpire coaches, video footage of 50 free-kick scenarios separate from the data collection period was shown to each umpire coach for assessment. This assessment demonstrated a moderate level of intra-observer reliability (intraclass correlation coefficient [ICC] = 0.91; coefficient of variation [CV] = 10.2%) (9,14).

Time-motion data were collected using MinimaxX global positioning system (GPS) devices (5 Hz; Catapult Innovations, Melbourne, Australia), which were fitted into an undergarment, worn underneath their normal umpiring uniform. Each device was turned on at least 20 minutes before the commencement of a match in an open area to ensure that a satellite lock was established. Data were downloaded and analyzed after each match using the Catapult Sprint 5.0 software (Catapult Innovations). Data were divided into the 4 quarters, according to the start time and duration obtained from official timekeepers. Nonplaying data (i.e., quarter breaks) were not included in analysis.

Movement data were calculated as the distance covered and time spent in 6 locomotor categories: (stand, <0.7 km·h−1; walk, 0.7–7 km·h−1; jog, 7–14.4 km·h−1; run, 14.4–20 km·h−1; fast run, 20–23 km·h−1; and sprint, ≥23 km·h−1). The distance covered in low-intensity activity (<14.4 km·h−1) and HSR (≥14.4 km·h−1) were also collated in addition to the total distance covered and relative distance maintained during each quarter. These zones are consistent with previously published time-motion analysis studies, and thus allow for direct comparison with the demands of umpires and players in AF (4,7). Previously, 5-Hz MinimaxX GPS units have demonstrated acceptable validity (SEE, 3.8 ± 0.6%) and reliability (CV, 3.6%) for measuring the total distance during simulated team sport exercise (10). However, a limitation of GPS use is the decreased reliability when assessing HSR in 2 change of direction courses (typical error, 2.6 m; CV, 7.9–9.2%) (10). As such, care needs to be taken when interpreting these results.

Using the postmatch video analysis, the time periods throughout the match that each umpire spent in the mid- or end-zone were determined. The position of each umpire and elapsed time of each rotation between the mid-zone and end-zone umpires were recorded. Rotations could occur at any time throughout a match as per the discretion of the officiating umpires. Movement data were then manually divided to determine the various physical performance measures for each rotation.

Statistical Analyses

All data are presented as mean (±SD). Data were initially analyzed for normality and homogeneity of variances using the Shapiro-Wilk statistic and Levene's test, respectively. One-way repeated-measures analysis of variance (ANOVA) was used to determine the differences in physical demands between the 4 quarters. For significant main effects, a Bonferroni's post hoc comparison test was used. Furthermore, a two-way mixed ANOVA (2 [position] × 4 [quarter] design) was used to examine the effect of umpire positioning (i.e., mid-zone and end-zone) on the physical demands across each quarter. Independent t-tests were used to identify comparisons between groups. Furthermore, 95% confidence intervals (CI) were used to report mean differences between the 2 groups. Decision-making accuracy was analyzed by comparing the free-kick decisions made within each quarter of a match. As this data were nonparametric, a Friedman's ANOVA was used. Chi-squared goodness of fit tests (16) were applied to examine the decision-making performance between the mid-zone and end-zone umpire positions. Statistical significance was set at p ≤ 0.05, and statistical tests and procedures were performed using SPSS Statistics (version 19; IBM Corporation, Somers, NY, USA).


The mean values for the physical and free-kick decision-making demands are presented in Table 1. Significant reductions in the distance covered were observed (F[3,168] = 6.116; p = 0.001;

= 0.098), with post hoc analysis identifying that the distance covered in the third (p = 0.006) and fourth (p = 0.001) quarters were significantly less than that of the first quarter. A significant main effect was also observed for the HSR distance (F[3,168] = 4.271; p = 0.006;

= 0.071), with the fourth quarter being significantly lower compared with the first (p = 0.003). Further reductions were observed across various speed thresholds that are collectively shown in Table 1.

Table 1
Table 1:
Physical and decision-making performance measures of Australian football field umpires during match play (mean ±SD).*

Significant differences exist between the mid-zone and end-zone positions in regard to HSR (p = 0.001; 95% CI, 641–895 m) and relative distance (p = 0.001; 95% CI, 47–55 m·min−1) covered. There was no significant difference between the 2 groups for the total distance covered (Figure 2). On average, a rotation between the umpires occurred every 2:38 ± 1:24 minutes:seconds, with no significant difference in rotation length across quarters (p = 0.530).

Figure 2
Figure 2:
The total distance (A), high-speed running distance (B), relative distance (C), and free-kick decision-making accuracy (D) between mid-zone (□) and end-zone ([Black Square]) positions (mean ±SD). *Significantly different between groups; a: significantly different from first quarter; b: significantly different from second quarter; c: significantly different from third quarter. D) No SD because this represents DMA across all decisions (n = 884). DMA, decision-making accuracy.

During the observed matches, umpires awarded a total of 779 free kicks (correct, 742; unwarranted, 37), with a further 105 free kicks categorized as missed during these matches (n = 884 decisions). Subsequent analysis determined that in each match, field umpires had an overall free-kick accuracy of 84 ± 6%, awarding a collective 44 ± 8 free kicks per match. Free-kick accuracy (p = 0.110) and number of free kicks awarded (p = 0.340) were consistent across each quarter of a match.

Chi-square analysis identified a nonuniform distribution in the number of free kicks, with more free kicks awarded in the mid-zone area compared with the end-zone (p = 0.025; χ2 (1) = 5.040). Accuracy between mid-zone and end-zone areas was uniformly distributed. Examination of incorrect free kicks highlighted no significant trend to over or under penalizing across a match (p = 0.731).


This study aimed to examine the physical demands of AF umpires and quantify their free-kick decision-making performance during match play. The results present novel data that examine differences in the physical demands and decision-making requirements in different regions of the field. Together, these data have important applications in the development of AF umpires, and collectively all team sport officials.

Although there is little difference in the total distance covered by umpires in this study, compared with Coutts and Reaburn (5), umpires performed a greater proportion of HSR. Further differences in physical demands exist compared with a subelite population group using similar methodology (7). This may reflect the differing standards of competition examined, with the superior ability of elite-level umpires to read play and predict where to position themselves (2). Although this may not be true for AF umpires as suggested by Larkin et al. (13) using video-based play prediction methods. Similar to various AF research (4,6,7), within-match variations exist, demonstrated by reductions in the total distance and HSR distance toward the end of matches. However, given the role of umpires, their physical demands are likely to be influenced by the running speed and ball movement of players. The significantly faster movement reported for the beginning of a match and the subsequent decline in these demands (4) demonstrate that the physical intensity of umpires may be reduced in response to the demands of play (19).

Although global changes occur in the movement demands across a match, perhaps more importantly, significant differences in the demands (i.e., relative distance, HSR) within the mid-zone highlight the more physically demanding region for AF umpires. It is proposed that rotating umpires between mid- and end-zone positions during match play permits umpires to alternate periods of high-intensity work and rest, allowing them to maintain a lower physiological intensity and limit fatigue. However, while positioned within the end-zone, umpires are still required to perform their role, particularly if play transitions toward this area. As such, this may limit the potential for complete physical recovery before rotating back into the mid-zone.

Although reductions in the movement demands exist, the maintenance of decision-making accuracy suggests that these activity changes are in response to game speed and structure, rather than physical fatigue. It would be expected that with increasing physical fatigue, umpires are unable to maintain the required intensity to position themselves correctly around play (11,12), and as a result, are more likely to incorrectly adjudicate. As the majority of match play is spent in the middle of the ground, it was expected that the mid-zone umpire would make the highest number of decisions while undertaking the highest relative physical demands. Mental fatigue has been suggested as a mechanism for poorer decision-making performance in soccer referees (14), hence the 3 umpire system may assist in alleviating this reduction in performance late in matches by allowing mid-zone umpires to rotate to recover in the end-zone. Therefore, this may be beneficial for umpires to help maintain decision-making accuracy throughout a match.

Although decision-making performance is stable in this study, previous research has reported conflicting findings across similar studies. Mallo et al. (14) reported that soccer referees demonstrate the greatest decision-making error rate in the final 15 minutes, suggesting that physical and mental fatigue occurs in the final stages of a match. In contrast, Mascarenhas et al. (15) reported poorer decision-making performance of soccer referees during the initial 15 minutes of matches because of more players being around the ball. Although there were no significant changes in decision making within a match in this study, a considerable number of errors (16%) in free-kick judgment are committed by AF umpires. As such, other factors may facilitate errors in the decision-making performance of AF umpires, including obstructed view, distance from play (14), movement speed (17), and changes in central nervous system arousal (3). These factors are yet to be examined in AF umpires and therefore could be a focus of future research.

It is important to note that the free-kick judgments made by umpires represent only a small proportion of the total decision-making demands. With respect to soccer referees, Plessner et al. (18) estimated that there are between 200 and 250 potential decision-making judgments across a match. Alternatively, Helsen and Bultynck (8) determined that soccer referees make, on average, 137 observable decisions per match, of which 45 related to free-kick scenarios. As such, free kicks represent only a portion of the overall decision making, with referees making numerous nonobservable decisions (i.e., when the referee does not interfere with play and it continues on). Specifically, each AF match involves more than 700 disposals (i.e., passing the ball to a team mate), each representing either a legal kick or handball. In these instances, the AF umpire is effectively making a decision to determine whether the method of disposal was legal. Furthermore, body-to-body contact (i.e., tackling, marking contests) between opposing players is only permitted within specific areas of the body, placing further decision-making demands on the umpire. Therefore, it is likely that the total decision-making requirements of AF officiating is above and beyond the free-kick judgments and likely to be upward of 2,000 decisions per match; however, this remains difficult to quantify. Each of these acts of play require some assessment by the umpire, most of which occur without interference (i.e., a “play-on” decision when there is no illegal act of play) or obvious reaction from the umpire. As such, future studies should aim to establish methods to identify the total decision-making demands associated with AF umpiring and team sport officiating in general.

The main findings of this study present current information regarding the match demands of AF umpires whereby reductions in various physical performance measures existed during the fourth quarter. There were reduced physical and decision-making demands in the end-zone position, which umpires use for periods of on-field physical recovery. Furthermore, the study demonstrated that the free-kick decision-making accuracy of AF umpires remains consistent across each quarter of a match. The positional rotations of the umpires allow for physical recovery between bouts within the mid-zone, and thus the match demands are shared, enabling all umpires to complete each match while maintaining their decision-making performance.

Practical Applications

The results of this study provide beneficial information for strength and conditioning coaches to apply current physical demands to AF umpire training. Important information such as comparison between the demands of mid-zone and end-zone AF umpires has never previously been quantified. Therefore, this study highlights how these positions are used for methods of on-field recovery. Furthermore, no previous studies examined the free-kick decision-making aspect of AF umpire performance, and as this is considered their most important role during play, these findings are important to include for future training. However, free kicks only account for a proportion of the total decision-making demands, and therefore, this needs to be considered during match performance assessment.


There was no outside financial support for this project. The authors would like to express their gratitude to the AFL Umpiring Department for their assistance in this project. No conflicts of interest exist for this research.


1. Australian Football League. Laws of Australian Football. Melbourne, Australia: Australian Football League, 2012.
2. Castagna C, Abt G, D'Ottavio S. Activity profile of international-level soccer referees during competitive matches. J Strength Cond Res 18: 486–490, 2004.
3. Chmura J, Nazar K. Parallel changes in the onset of blood lactate accumulation (OBLA) and threshold of psychomotor performance deterioration during incremental exercise after training in athletes. Int J Psychophysiol 75: 287–290, 2010.
4. Coutts AJ, Quinn J, Hocking J, Castagna C, Rampinini E. Match running performance in elite Australian Rules Football. J Sci Med Sport 13: 543–548, 2010.
5. Coutts AJ, Reaburn PR. Time and motion analysis of the AFL field umpire. J Sci Med Sport 3: 132–139, 2000.
6. Duffield R, Coutts AJ, Quinn J. Core temperature responses and match running performance during intermittent-sprint exercise competition in warm conditions. J Strength Cond Res 23: 1238–1244, 2009.
7. Elsworthy N, Dascombe BJ. The match demands of Australian rules football umpires in a state-based competition. Int J Sports Physiol Perform 6: 559–571, 2011.
8. Helsen W, Bultynck JB. Physical and perceptual-cognitive demands of top-class refereeing in association football. J Sports Sci 22: 179–189, 2004.
9. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000.
10. Jennings D, Cormack S, Coutts AJ, Boyd L, Aughey RJ. The validity and reliability of GPS units in team sport specific running patterns. Int J Sports Physiol Perform 5: 328–341, 2010.
11. Krustrup P, Bangsbo J. Physiological demands of top-class soccer refereeing in relation to physical capacity: Effect of intense intermittent exercise training. J Sports Sci 19: 881–891, 2001.
12. Krustrup P, Helsen W, Randers MB, Christensen JF, MacDonald C, Rebelo AN, Bangsbo J. Activity profile and physical demands of football referees and assistant referees in international games. J Sports Sci 27: 1167–1176, 2009.
13. Larkin P, Berry J, Dawson B, Lay B. Perceptual and decision-making skills of Australian Football umpires. Int J Perform Anal Sport 11: 427–437, 2011.
14. Mallo J, Frutos PG, Juarez D, Navarro E. Effect of positioning on the accuracy of decision making of association football top-class referees and assistant referees during competitive matches. J Sports Sci 30: 1437–1445, 2012.
15. Mascarenhas DRD, Button C, O'Hare D, Dicks M. Physical performance and decision making in association football referees: A naturalistic study. Open Sports Sci J 2: 1–9, 2009.
16. Nevill AM, Atkinson G, Hughes MD, Cooper S. Statistical methods for analyzing discrete and categorical data recorded in performance analysis. J Sports Sci 20: 829–844, 2002.
17. Oudejans RRD, Bakker FC, Verheijen R, Gerrits JC, Steinbrückner M, Beek PJ. How position and motion of expert assistant referees in soccer relate to the quality of their offside judgements during actual match play. Int J Sport Psychol 36: 3–21, 2005.
18. Plessner H, Schweizer G, Brand R, O'Hare D. A multiple-cue learning approach as the basis for understanding and improving soccer referees' decision making. Prog Brain Res 174: 151–158, 2009.
19. Weston M, Castagna C, Impellizzeri FM, Rampinini E, Abt G. Analysis of physical match performance in English Premier League soccer referees with particular reference to first half and player work rates. J Sci Med Sport 10: 390–397, 2007.

match analysis; cognition; perception; motor skills; perceptual-cognitive

Copyright © 2014 by the National Strength & Conditioning Association.