Comparing the Yo-Yo intermittent and Bronco tests and their associations with match demands among amateur rugby union referees

Abstract

This study compared the Yo-Yo intermittent recovery level 1 (YYIR₁) and Bronco tests among amateur rugby referees and examined whether the results of these tests were associated with actual match demands. Sixty-seven (65 male, 2 female; 30 ± 11 years; 77.4 ± 13.1 kg; 175 ± 8 cm) amateur rugby referees participated. Match demands were assessed using Global Navigation Satellite System technology and heart rate (HR) recordings. Significant correlations (all p < 0.05) were found between the YYIR₁ and Bronco test results (r: −0.88). The YYIR₁ test was significantly correlated with match demands including total (r: 0.56) and high-intensity (r: 0.70) distance covered, maximal speed reached (r: 0.73), and the number of high-intensity accelerations (r: 0.54) and sprints (r: 0.68) completed. Furthermore, YYIR₁ test maximum HR was significantly correlated with maximum HR (r: 0.72), average HR (r: 0.54), and internal load (r: 0.55) during match play. In contrast, the Bronco test was only significantly correlated with match demands including maximal speed reached (r: −0.68) and the number of high-intensity accelerations (r: −0.61). Moreover, Bronco test maximum HR was significantly correlated with maximal (r: 0.84) and average HR (r: 0.73) during match play. The results suggest that while the YYIR₁ and Bronco tests might both be efficient tools to guide training prescription, the YYIR₁ test might be as the more valuable test to assess match fitness in rugby referees.

Keywords

Acceleration,global navigation tracking,heart rate,intermittent exercise,sports officials,sprints

Introduction

Rugby union (“rugby”) is a highly demanding intermittent contact sport, consisting of high-intensity efforts interspersed with low-intensity activity played over two 40-min halves.^1,2 During match play, rugby referees combine high-intensity running actions, with low-intensity activity, covering six to eight kilometres per match, with a corresponding mean heart rate of 80% of their maximum.^2–6 On average, during semi-professional and professional matches, a referee covers 11% of the total match distance in high intensity (> 18. 4 km.h⁻¹) running,^4,6 and it has been reported that there is a small to moderate decrease in high-intensity running during the middle period of a match (i.e. 30–70 min), with no differences reported between 0–10 and 70–80 min.⁴ It is important that rugby referees can maintain their physical performance throughout a match, and their ability to meet the physical (i.e. high-intensity running) and physiological (i.e. heart rate) demands imposed on them during match play is crucial for optimal positioning and decision-making.^2,4,7,8

Therefore, minimum mandatory fitness requirements have been established for rugby referees by governing bodies using field tests, with referees who achieve these standards deemed fit enough to cope with the match demands of officiating. However, how the results of these fitness tests relate to the demands placed on rugby referees during actual match play (i.e. direct validity) is unclear. Indeed, studies with soccer referees have reported poor correlations between field tests measures (e.g. mean heart rate during the high-intensity 150-m interval test) and match demands (e.g. total distance covered and high speed running).⁹ However, these results contrast with one study that reported large significant correlations between heart rate during the high-intensity 150-m interval test and match high-intensity running.¹⁰ Therefore, given the mixed results in the literature and the nonspecific nature of the field tests used to assess the fitness of rugby referees, it is important that the relationships between field test results and actual match demands are examined.

The Yo-Yo intermittent recovery level 1 (YYIR₁)¹¹ is regularly used to provide an indirect indication of the physical capacity of intermittent team sport athletes.^12,13 When the standards for the YYIR₁ test were established, they were used globally to evaluate the capacity to repeatedly perform and recover from intense activity.^11–13 It is within this context that this test is used for rugby referees. For instance, acceptable YYIR₁ test levels have been suggested for elite male and female rugby referees (i.e. levels 18 and 15, respectively),¹³ as they can officiate any rugby match, regardless of the gender of the players (i.e. men or women). However, to date, no empirical research has supported these fitness standards, and applied practitioners have started to use other more expedient test with referees. Indeed, recently, the 1.2 km shuttle run test (1.2_SRT),¹⁴ also known as the Bronco test, is gaining popularity in applied rugby contexts.^15,16 It has been used to assess rugby referees’ fitness levels and has been shown to strongly correlate with the YYIR₁ (r = −0.87) among elite youth rugby players.¹⁵ Both tests (i.e. YYIR₁ and Bronco) are used for rugby referees as they incorporate repeated high-intensity sprints as well as periods of acceleration, deceleration and change in direction, which are characteristics of intermittent activities, such as refereeing.¹⁷ Additionally, the YYIR₁ is regularly used to assess aerobic power (i.e. oxygen uptake during dynamic exercise), providing an indirect indication of the physical capacity of intermittent team sport athletes,^12,13 and the Bronco test can be used to determine maximal aerobic running speed.^15,16 The information provided from these tests enables practitioners to monitor individual development and can be used for training prescription and to evaluate the impact of training interventions.^18,19 However, the relationship between the results of these field tests and the actual match demands placed on rugby referees has not yet been investigated. This information is important because direct validity is considered an essential pre-requisite of an accurate sport-specific field test.²⁰

To extend previous research, this study aimed to: (1) directly compare the YYIR₁ and Bronco tests, by correlating the final results of (e.g. distance, completion time), and physiological responses during (e.g. heart rate [HR] and internal load), the two tests; and (2) verify the direct validity of the YYIR₁ and Bronco tests by examining how the results of these tests relate to the physical (e.g. high-intensity distance) and physiological (e.g. internal load) demands placed on rugby referees during actual match play.

Methods

Participants

Overall, 67 referees (65 male, 2 female; age: 30 ± 11 years [range: 18–57 years]; body mass: 77.4 ± 13.1 kg; height: 175 ± 8 cm), with at least two years’ officiating experience (7 ± 5 years), participated in the study. Referees could not participate if they had not been refereeing in the previous six months or more due to injury. This study received institutional ethical approval and informed consent was obtained from each referee.

Field test analysis: YYIR₁ test data was collected from 48 referees (46 male, 2 female; age: 27 ± 8 years; body mass: 76.6 ± 13.9 kg, height: 175 ± 8 cm; experience: 6 ± 4 years), and Bronco test data was collected from 35 referees (34 male, 1 female; age: 29 ± 11 years; body mass: 75.7 ± 12.5 kg; height: 174 ± 9 cm; experience: 7 ± 6 years). Of these referees, 32 completed both field tests (31 male, 1 female; age: 27 ± 8 years; body mass: 74.3 ± 11.9 kg; height: 174 ± 9 cm; experience: 6 ± 4 years).

Match demands analysis: Match demand data was collected from 16 referees (15 male, 1 female; age: 24 ± 5 years; body mass: 76.6 ± 10.3 kg; height: 177 ± 8 cm; experience: 6 ± 4 years) over 27 matches, ranging from two to six matches per referee (1.7 ± 1.5 matches per referee). While all 16 of these referees completed the YYIR₁ test, only 11 performed the Bronco test (11 male, age: 23 ± 5 years; body mass: 79.3 ± 7.3 kg; height: 180 ± 7 cm; experience: 6 ± 3 years).

The sample size justification was resource constraint based²¹ in that we were constrained by the referees’ availability to voluntary attend both tests and matches. All referees were amateur and actively officiating in South Africa at either the residence “Koshuis” tournament at Stellenbosch University, or the Western Province Rugby Union Super League competition, played across the Western Cape Province. The “Koshuis” tournament includes ∼200 matches over seven months, divided into first-, second-, third-, and fourth- league matches, played on weeknights. Except for one “private-residence” team, all teams comprised Stellenbosch University students and could be described as recreational, with some semi-professional players. The total match duration was 60 min for leagues one to three, and 50 min for the fourth league.²² The Western Province Super League is organised by the Western Province Rugby Union (i.e. one of the fourteen provincial unions in South Africa), and is played by amateur clubs, sub-divided into A, B, C, and Under 20 teams. It follows a single round format where each team plays every other team in their league once a season (April to September), and total match duration was 80 min for league A, 70 min for leagues B and Under 20, and 60 min for league C.²³

Study protocol

An observational research design was used. During the field tests and rugby matches, referees wore an elasticated vest with an augmented concurrent multi-GNSS receiver unit (Apex, 10 Hz, STATSports, Belfast, UK), located between their shoulder blades. The validity and reliability of this GNSS unit has been reported previously.²⁴ The GNSS unit was switched on ∼15 min before, and turned off immediately after, each field test and match. HR data was collected during the field tests and matches through a HR sensor (Polar T31, Polar Electro, Kempele, Finland) worn around each referee's chest, which transmitted real-time data via Bluetooth to the GNSS unit. All measurements were obtained at the end of the competitive season, and field testing was performed by the same assessors.

Testing procedures

Both field tests were performed on a rugby field on dry and firm conditions, and a minimum of 48 h and a maximum of six days rest between tests. As the field tests were scheduled by the research team based on the referee's availability, it was not possible to counterbalance the order in which the referees completed the two field tests (i.e. 28 referees performed the YYIR₁ test first, while four referees performed the Bronco test first). All referees were familiar with both testing protocols as they complete them regularly throughout the season. The referees were requested not to undertake vigorous physical activity and to maintain normal dietary intake and sleeping patterns in the preceding 24 h, and avoid consuming a heavy meal and caffeinated beverages in the preceding two hours. Finally, referees were allowed to do their own warm-up for 10 min before each test, with activation exercises and one high-intensity running shuttle (i.e. 20-m and back) advised before the YYIR₁ test, and one practice repetition (i.e. 20-m-40-m-60-m) recommended before the Bronco test

The YYIR₁ test was performed according to standard procedures,^11,12 comprising repeated 2 × 20-m shuttles at a progressively increasing speed controlled by audio signals with an active 10-s rest between each shuttle, which consisted of jogging 2 × 5-m. When the referees failed twice to reach the finish line before the audio signal, the test was ended and the final distance (m) and level were recorded.^11,12 The test lasted ∼20 min.

For the Bronco test, cones were placed at the 0-m, 20-m, 40-m, and 60-m lines and marked using a 50-m measuring tape (Richer 50-meter steel). Referees were asked to run from the 0-m line to the 20-m line and back, then run from the 0-m line to the 40-m line and back, and finally run from the 0-m line to the 60-m line and back. Completion of these 20-m-40-m-60-m shuttles was considered one repetition, with referees completing five repetitions to achieve a distance of 1200 m as quickly as possible.^14,16,25 A hand-held stop watch (Casio HS-80TW-1EF) was used to record completion time in seconds.¹⁶ The test lasted ∼6 min.

In line with previous research,¹⁰ the physical demands assessed during both field tests and actual match play included: total, relative (m.min⁻¹), and high-intensity (> 18.4 km.h⁻¹) distance covered (m), maximal speed (km.h⁻¹), and number of high-intensity (> 2.79 m.s⁻²) accelerations (n) and sprints (20 km.h⁻¹ or greater sustained for at least one second) (n)^6,26 and the frequency (n) of repeated high-intensity efforts (RHIE) bouts was analysed. A RHIE bout was defined as a minimum of three high-intensity efforts (i.e. sprints or high-intensity accelerations) with less than 21 s of recovery between efforts.^6,26,27 The single longest period of a RHIE bout from each match was identified and analysed as the “Worst Case Scenario” (WCS).^6,26,28,29 A bout duration was defined as the time the referee first performed a high-intensity activity (i.e. sprint or acceleration) and repeated a minimum of two other efforts with less than 21 s between those efforts.^6,26 For each WCS, the total duration (s), total distance (m), and total distance relative to the bout duration (m.min⁻¹) was analysed. Additionally, high metabolic load (HML) distance (m) (i.e. distance covered > 18.1 km.h⁻¹ and accelerating and decelerating over 2 m.s⁻²), and HML efforts (n) (i.e. number of separate efforts undertaken in producing HML distance) were analysed.³⁰ Accelerations and decelerations were calculated from a single derivation of the speed during a period of 0.5 s.³¹ These speed zones were selected as they have previously been used in the analysis of rugby referees,^3,4,6 and are in accordance with training programmes guidelines for rugby referees.¹³

HR data was assessed during both field tests and actual match play, with maximal HR (HR_max) determined as the highest of either: (1) HR_max estimated thorough the formula: 208 – (0.7 × age),³² or (2) HR_max values obtained during the field tests^33,34 or match.⁴ HR data was classified based on the percentage of total time spent in six HR zones⁴: zone 1 (< 60% HR_max), zone 2 (61–70% HR_max), zone 3 (71–80% HR_max), zone 4 (81–90% HR_max), zone 5 (91–95% HR_max), and zone 6 (> 96% HR_max). Internal load was calculated by multiplying the time spent (in mins) in the five HR zones by a coefficient assigned to each zone (i.e. 50–60% HR_max = 1, 60–70% HR_max = 2, 70–80% HR_max = 3, 80–90% HR_max = 4, and 90–100% HR_max = 5).³⁵

Statistical analysis

The sample size in all analysis was comparable to previous research,⁹ and deemed adequate (software G*Power 3.1.9.7; Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany) assuming statistical power of 80% and α error probability of 0.05.³⁶ After collection, data from each GNSS unit was downloaded to analysis software (STATSports Apex software, v. 3.0.02011), and then exported to statistical analysis software (IBM SPSS v. 26.0; IBM Corp., Armonk, NY, USA). Kolmogorov-Smirnov tests revealed that data was normally distributed. Next, means, standard deviations, and 95% confidence intervals were calculated. Subsequently, a series of Pearson's product-moment tests were conducted to assess the correlations between: (1) the YYIR₁ final level and distance versus the Bronco test final time, (2) HR and internal load from the YYIR₁ test versus HR and internal load from the Bronco test, and (3) the aforementioned field test data and physical (i.e. total, relative and high-intensity distance, maximal speed, high-intensity accelerations, sprints, RHIE, WCS distance, WCS duration, and WCS relative distance, and HML efforts and distance) and physiological (i.e. HR and internal load) demands during actual match play. Given the different number of matches officiated by each referee, the average match demands were calculated for each referee and used throughout the third step. Correlation coefficients (r) of 0.1–0.3, 0.3–0.5, and > 0.5 were interpreted as small, medium, and large, respectively.³⁷

Next, a single two-way repeated measure analysis of variance (ANOVA) with follow-up paired samples t-tests was used to determine the differences between the six HR zones for the YYIR₁ and Bronco tests. Effect sizes were calculated as partial eta squared (η_p²), and values of ≥ 0^.01, ≥ 0.06, and ≥ 0.14 were interpreted as small, medium, and large, respectively.³⁸ Additionally, differences in HR and internal load between the YYIR₁ and Bronco tests were assessed via a dependent t-test, with Cohen's d effect sizes of 0.2, 0.5, and 0.8 interpreted as small, medium, and large, respectively.³⁸

Finally, since the YYIR₁ test final distance was significantly correlated with the time taken to complete the Bronco test, a separate linear regression equation was generated to estimate the YYIR₁ test final distance. The YYIR₁ test final distance was entered as the dependent variable, and the time taken to complete the Bronco test was entered as independent variable. The adequacy of the results generated by the equation was investigated through Bland-Altman's graphical analysis between true and estimated values, and concomitant analysis of result significance through: (1) simple Student's t-test between “0” and the difference in results between true and estimated values of YYIR₁ test final distance, and (2) linear regression analysis of agreement values.³⁹ The bias (mean difference between true and estimated values), accuracy (overall distance between true and estimated values), and precision (statistical variance) of the equation were calculated by the scaled mean error (SME), the root mean square error (RMSE), and the coefficient of variance (CV), respectively⁴⁰:

SME = MEAN

where SME is the bias and MEAN is the mean of all differences between the true and estimated values,

R M S E = \sqrt{\frac{\sum_{i = 1}^{n} {(X_{o b s, i} – X_{m o d e l, i})}^{2}}{n}}

where RMSE is the accuracy, X_obs,i are the observational true values and X_model,i are the estimated values obtained by the equation, and

CV = (S D_{diff}) / (MEA N_{diff}), X 100

where CV is the coefficient of variation (precision indicator), SD_diff is the standard deviation from the difference between true and estimated values, and MEAN_diff is the average difference obtained from true and estimated values. The limits of agreement (LoA) between methods were also evaluated (random error = SD of the differences × 1.96) and represents the 95% likely range for the difference between the true values and those estimated by Equation 1.³⁹

The reliability between true and estimated values (YYIR₁ test final distance) generated by Equation 1 was also analysed using an absolute-agreement, two-way mixed model, intraclass correlation coefficient (ICC), and values of < 0.5, 0.5–0.75, 0.75–0.9, and > 0.90 were interpreted as poor, moderate, good, and excellent reliability, respectively.⁴¹ An α of 0.05 was used in all analyses.

Results

Field tests results

Compared to the Bronco test, HR_max was significantly higher in the YYIR₁ test both in terms of bpm⁻¹, t₍₂₆₎ = 5.16, p < 0.001, d = 0.74, and percentage, t₍₂₇₎ = 5.35, p < 0.001, d = 1.27. Additionally, internal load was significantly greater in the YYIR₁ test compared to the Bronco test, t₍₂₆₎ = 6.72, p < 0.001, d = 1.89. Average HR did not differ significantly (p = 0.764) (Table 1).

Table 1.

Mean ± SD _(CI95%) results, HR, and internal load data during the YYIR₁ (n = 48) and Bronco (n = 35) tests.

	YYIR₁	Bronco
YYIR₁ final level	16.4 ± 1.4 _{(16.1–16.8)}	-
YYIR₁ final distance (m)	1243 ± 427 _{(1119–1367)}	-
Bronco result (s)	-	326 ± 34 _(314–337)
HR_max (bpm⁻¹)	197 ± 9 _(194–200)	192 ± 6 _(189–194)*
HR_max (%)	98.9 ± 1.9 _{(98.2–99.2)}	96.2 ± 2.4 _{(95.3–97.2)}*
Average HR (%max)	89.6 ± 3.5 _{(88.2–91.0)}	89.2 ± 4.7 _{(87.3–91.1)}
Internal load (au)	40.1 ± 14.9 _{(34.2–46.0)}	19.2 ± 4.7 _{(17.4–21.1)} *

YYIR₁ = Yo-Yo intermittent recovery test level 1, HR = Heart Rate, *Significant difference between tests at the p < 0.05 level.

Relationships between YYIR₁ and Bronco test data

There was a significant negative correlation between total distance covered in the YYIR₁ test and the time taken to complete the Bronco test (p < 0.001), and a significant positive correlation between average HR in the YYIR₁ and Bronco tests (p = 0.034). Finally, there was a significant negative correlation between internal load during the YYIR₁ test and the time taken to complete the Bronco test (p < 0.001) (Table 2).

Table 2.

Correlations coefficients between performance, HR, and internal load from the YYIR₁ and Bronco tests (n = 32).

	Bronco final time	Bronco avg. HR	Bronco IL
YYIR₁ total distance (m)	−0.88**	−0.12	−0.34
YYIR₁ avg. HR (%max)	0.27	0.67*	0.08
YYIR₁ IL (au)	−0.74**	−0.10	−0.12

YYIR₁ = Yo-Yo intermittent recovery test level 1, HR = Heart Rate, IL = Internal Load. *correlation is significant at the p < 0.05 level. **correlation is significant at the p < 0.01 level.

Comparisons between HR zone data during the YYIR₁ and Bronco tests

There was no significant main effect for Test, F_{(1, 307)} = 0.44, p = 0.509, η_p² = 0.00. However, there was a significant main effect for HR Zone, F_{(5, 307)} = 40.90, p < 0.001, η_p² = 0.40, and a significant interaction effect, F_{(5, 307)} = 3.24, p = 0.007, η_p² = 0.05. Follow up t-tests revealed that referees spent significantly less time in HR zone 5 (91–95 HR_max, p = 0.040), and more time in HR zone 6 (> 96% HR_max, p = 0.049), during the YYIR₁ test compared to the Bronco test Furthermore, during the YYIR₁ test, referees spent longer in HR zone 6 (> 96% HR_max) compared to HR zones 1 (< 60% HR_max, p < 0.001), 2 (61–70% HR_max, p < 0.001), 3 (71–80% HR_max, p < 0.001), and 4 (80–90% HR_max, p = 0.021). Finally, during the Bronco test, referees spent longer in HR zones 4 (81–90% HR_max), 5 (91–95% HR_max) and 6 (> 96% HR_max) compared to HR zones 1 (< 60% HR_max, ps < 0.001), 2 (61–70% HR_max, ps < 0.002), and 3 (71–80% HR_max, ps < 0.018), and longer in HR zone 3 (71–80% HR_max) compared to HR zones 1 (< 60% HR_max, p = 0.003) and 2 (61–70% HR_max, p = 0.008) (Figure 1).

Figure 1.

Percentage of time spent in each HR zone during matches, as well as the YYIR₁ and bronco tests. Notes: Differences between the six HR zones for the YYIR₁ and Bronco tests = *Significant difference between tests at the p < 0.05 level; ^a Significantly different from < 60% HR_max, ^b Significantly different from 61–70% HR_max, ^c Significantly different from 71–80% HR_max, and ^d Significantly different from 81–90% HR_max.

Relationships between field tests results and match demands

The average descriptive match demand data from 16 referees for between one and six matches per referee, with a total of 27 matches (mean 1.7 ± 1.5 matches per referee) are presented in Table 3. There were significant positive correlations between YYIR₁ test final distance and match demands including: total (p = 0.023) and high-intensity (p = 0.002) distance, maximal speed (p = 0.001), number of high-intensity accelerations (p = 0.022) and sprints (p = 0.004), and number of repeated high-intensity efforts (p = 0.019), WCS duration (p = 0.035), distance covered during the WCS (p = 0.047), number of HML efforts (p = 0.002) and HML distance (p = 0.003). Additionally, there were significant negative correlations between the final time taken to complete the Bronco test and match demands including: maximal speed (p = 0.022), number of high-intensity accelerations (p = 0.048), and relative distance covered during the WCS (p = 0.030) (Table 4). Finally, additional correlation analyses were conducted with a sample of 11 referees who undertook both tests. There were significant positive correlations between YYIR₁ test final distance and match demands including total (p = 0.037) and high-intensity (p = 0.006) distance, maximal speed (p = 0.003), number of high-intensity accelerations (p = 0.036) and sprints (p = 0.007), and number of repeated high-intensity efforts (p = 0.046), distance covered during the WCS (p = 0.034), relative distance covered during the WCS (p = 0.031), number of HML efforts (p = 0.005), and HML distance (p = 0.007). Additionally, there were significant negative correlations between the final time taken to complete the Bronco test and match demands including maximal speed (p = 0.022), number of high-intensity accelerations (p = 0.048), and relative distance covered during the WCS (p = 0.030) (Table 5).

Table 3.

Mean ± SD _(CI95%) match (i.e. physical and physiological) demand data per match.

Total Distance (m)	4828.8 ± 867.0 _{(4328.2–5329.4)}
Relative Distance (m.min⁻¹)	75.8 ± 5.0 _{(72.9–78.7)}
High-intensity distance (>18.4 km.h⁻¹)	561.3 ± 265.5 _{(408.0–714.6)}
Maximal speed (km.h⁻¹)	27.4 ± 2.1 _{(26.2–28.6)}
High-intensity accelerations (n)	42 ± 15 _(34–51)
Sprints (n)	21 ± 11_(15–28)
RHIE (n)	6 ± 5 _(3–9)
WCS duration (s)	38 ± 12 _(31–45)
WCS distance (m)	88.7 ± 34.2 _{(69.0–108.5)}
WCS relative distance (m.min⁻¹)	140.2 ± 39.0 _{(117.7–162.7)}
HML efforts (n)	110 ± 25 _(96–125)
HML distance (m)	1173.0 ± 368.3 _{(960.3–1385.6)}
HR_max (bpm⁻¹)	196 ± 9 _(191–200)
Average HR (%max)	82 ± 4_(80–84)
Internal load (au)	233 ± 48 _(206–259)

RHIE = Repeated High-Intensity Efforts; WCS = Worst Case Scenario; HML = High Metabolic Load, HR = Heart Rate.

Table 4.

Correlation coefficients _(CI95%) between YYIR₁ (n = 16) and bronco (n = 11) test results and physical and physiological demands during actual match play.

	YYIR₁ final distance	Bronco final time
Total distance (m)	0.56* _{(0.54–0.86)}	−0.42 _{(−0.84 to 0.39)}
Relative distance (m.min⁻¹)	0.28 _{(−0.19 to 0.68)}	−0.44 _{(−0.77 to 0.30)}
High-intensity distance (>18.4 km.h⁻¹)	0.70^** _{(0.32–0.88)}	−0.52 _{(−0.87 to 0.63)}
Maximal speed (km.h⁻¹)	0.73** _{(0.32–0.92)}	−0.68^* _{(−0.96 to −0.08)}
High-intensity accelerations (n)	0.54* _{(0.13–0.83)}	−0.61^* _{(−0.87 to −0.09)}
Sprints (n)	0.68* _{(0.34–0.86)}	−0.55 _{(−0.86 to 0.02)}
RHIE (n)	0.62* _{(0.25–0.84)}	−0.40 _{(−0.79 to 0.31)}
WCS duration (s)	0.56* _{(0.04–0.87)}	−0.44 _{(−0.87 to 0.43)}
WCS distance (m)	0.54* _{(−0.04 to 0.84)}	−0.53_{(−0.94 to 0.57)}
WCS relative distance (m.min⁻¹)	0.10 _{(−0.58 to 0.77)}	−0.71^* _{(−0.94 to 0.61)}
HML efforts (n)	0.70^** _{(0.33–0.91)}	−0.56 _{(−0.91 to 0.36)}
HML distance (m)	0.69^** _{(0.36–0.89)}	−0.55 _{(−0.89 to 0.21)}
Match HR_max (bpm⁻¹)	−0.21 _{(−0.56 to 0.18)}	0.27 _{(−0.24 to 0.78)}
Match average HR (%max)	−0.06_{(−0.48 to 0.54)}	0.31 _{(−0.45 to 0.81)}
Match internal load (au)	0.36 _{(−0.19 to 0.76)}	0.01 _{(−0.45 to 0.52)}

RHIE = Repeated High-Intensity Efforts; WCS = Worst Case Scenario; HML = High Metabolic Load, HR = Heart Rate. *correlation is significant at the p < 0.05 level. **correlation is significant at the p < 0.01 level.

Table 5.

Correlation coefficients _(CI95%) between YYIR₁ and Bronco test results and physical and physiological demands during actual match play (n = 11).

	YYIR₁ final distance	Bronco final time
Total distance (m)	0.63* _{(0.02–0.92)}	−0.40 _{(−0.83 to 0.31)}
Relative distance (m.min⁻¹)	0.33 _{(−0.44 to 0.78)}	−0.44 _{(−0.77 to 0.22)}
High-intensity distance (>18.4 km.h⁻¹)	0.77^** _{(0.35–0.92)}	−0.52 _{(−0.86 to 0.05)}
Maximal speed (km.h⁻¹)	0.80** _{(0.22–0.97)}	−0.68^* _{(−0.96 to −0.08)}
High-intensity accelerations (n)	0.63* _{(0.20–0.86)}	−0.61^* _{(−0.87 to −0.10)}
Sprints (n)	0.76* _{(0.39–0.92)}	−0.55 _{(−0.84 to −0.02)}
RHIE (n)	0.67* _{(0.17–0.94)}	−0.40 _{(−0.81 to 0.31)}
WCS duration (s)	0.62 _{(−0.21 to 0.92)}	−0.44 _{(−0.86 to 0.44)}
WCS distance (m)	0.70* _{(−0.32 to 0.96)}	−0.53 _{(−0.94 to 0.57)}
WCS relative distance (m.min⁻¹)	0.71^* _{(−0.40 to 0.97)}	−0.71^* _{(−0.96 to 0.59)}
HML efforts (n)	0.77^** _{(0.12–0.96)}	−0.56 _{(−0.91 to 0.31)}
HML distance (m)	0.75^** _{(0.26–0.94)}	−0.55 _{(−0.88 to 0.20)}
Match HR_max (bpm⁻¹)	−0.31 _{(−0.75 to 0.09)}	0.26 _{(−0.26 to 0.80)}
Match average HR (%max)	−0.38 _{(−0.76 to 0.21)}	0.31 _{(−0.47 to 0.80)}
Match internal load (au)	0.26 _{(−0.48 to 0.75)}	0.01 _{(−0.45 to 0.52)}

Relationships between field test physiological demands and match demands

There was a significant negative correlation between YYIR₁ test HR_max and relative distance covered in match play (p = 0.038), and significant positive correlations between YYIR₁ test HR_max and HR_max (p = 0.004), average HR (p = 0.046), and internal load (p = 0.043) during match play. The average HR in the YYIR₁ test was significantly and positively correlated with the number of high-intensity accelerations per match (p = 0.013), and significantly and negatively correlated with the relative distance covered during the WCS (p = 0.010). Significant positive correlations were found between the YYIR₁ test internal load and high-intensity distance (p = 0.023), maximal speed (p = 0.022), number of high-intensity accelerations (p = 0.008) and sprints (p = 0.018), RHIE (p = 0.004), total (p = 0.016) and relative (p = 0.042) distance covered during the WCS, and number of HML efforts (p = 0.011) and distance (p = 0.014) during match play. Finally, there were significant positive correlations between the Bronco test HR_max and HR_max (p = 0.001) and average HR (p = 0.010) during match play, and there was a significant positive correlation between average HR during the Bronco test and average HR during match play (p = 0.025) (Table 6).

Table 6.

Correlation coefficients _(CI95%) between physiological response data during the YYIR₁ (n = 16) and Bronco (n = 11) tests and physical and physiological demands during actual match play.

	YYIR₁ HR_max	Bronco HR_max	YYIR₁ avg.HR	Bronco avg.HR	YYIR₁ IL	Bronco test IL
Total distance (m)	0.03 _{(−0.37 to 0.38)}	−0.24 _{(−0.77 to 0.17)}	0.09 _{(−0.55 to 0.35)}	−0.27 _{(−0.88 to 0.42)}	0.41 _{(−0.18 to 0.88)}	−0.16 _{(−0.69 to 0.67)}
Relative distance (m.min⁻¹)	−0.54* _{(−0.88 to −0.16)}	−0.26 _{(−0.68 to 0.19)}	0.00 _{(−0.68 to 0.25)}	−0.17 _{(−0.73 to 0.34)}	0.16 _{(−0.42 to 0.64)}	−0.11 _{(−0.64 to 0.57)}
High-intensity distance (>18.4 km.h⁻¹)	0.05 _{(−0.53 to 0.42)}	−0.14 _{(−0.69 to 0.35)}	0.18 _{(−0.41 to 0.44)}	−0.31 _{(−0.92 to 0.42)}	0.58* _{(0.12–0.86)}	−0.36 _{(−0.79 to 0.54)}
Maximal speed (km.h⁻¹)	0.18 _{(−0.45 to 0.62)}	0.17 _{(−0.52 to 0.86)}	0.12 _{(−0.31 to 0.40)}	−0.14 _{(−0.80 to 0.76)}	0.58* _{(0.12–0.88)}	−0.45 _{(−0.81 to 0.58)}
High-intensity accelerations (n)	0.15 _{(−0.56 to 0.67)}	0.16 _{(−0.63 to 0.67)}	0.60* _{(−0.09 to 0.88)}	0.12 _{(−0.81 to 0.66)}	0.66** _{(0.20–0.89)}	−0.01 _{(−0.69 to 0.67)}
Sprints (n)	0.09 _{(−0.44 to 0.44)}	−0.09 _{(−0.69 to 0.39)}	0.21_{(−0.38 to 0.46)}	−0.19 _{(−0.93 to 0.49)}	0.60** _{(0.11–0.91)}	−0.25 _{(−0.74 to 0.60)}
RHIE (n)	0.31 _{(−0.16 to 0.57)}	0.17 _{(−0.51 to 0.70)}	0.36 _{(0.24–0.71)}	−0.16 _{(−0.92 to 0.53)}	0.74** _{(0.39–0.94)}	−0.42 _{(−0.86 to 0.47)}
WCS duration (s)	0.15 _{(−0.31 to 0.51)}	0.09 _{(−0.45 to 0.47)}	0.35 _{(−0.36 to 0.61)}	−0.06 _{(−0.84 to 0.62)}	0.55 _{(−0.22 to 0.90)}	−0.18 _{(−0.79 to 0.72)}
WCS distance (m)	0.12 _{(−0.41 to 0.51)}	0.01 _{(−0.59 to 0.55)}	−0.01 _{(−0.31 to 0.41)}	−0.22 _{(−0.84 to 0.55)}	0.65* _{(0.20–0.91)}	−0.41 _{(−0.85 to 0.61)}
WCS relative distance (m.min⁻¹)	0.04 _{(−0.38 to 0.53)}	−0.12 _{(−0.84 to 0.50)}	−0.66^** _{(−0.90 to 0.69)}	−0.30 _{(−0.83 to 0.57)}	0.57* _{(0.12–0.86)}	−0.63 _{(−0.94 to 0.45)}
HML efforts (n)	0.15 _{(−0.32 to 0.48)}	−0.15 _{(−0.71 to 0.33)}	0.47 _{(−0.18 to 0.78)}	−0.28 _{(−0.88 to 0.47)}	0.64* _{(0.21–0.90)}	−0.34 _{(−0.76 to 0.57)}
HML distance (m)	0.05 _{(−0.39 to 0.39)}	−0.15 _{(−0.68 to 0.29)}	0.29 _{(−0.32 to 0.59)}	−0.24 _{(−0.92 to 0.46)}	0.62* _{(0.08–0.92)}	−0.26_{(−0.81 to 0.61)}
Match HR_max (bpm⁻¹)	0.72** _{(0.22–0.91)}	0.84** _{(0.30–0.96)}	0.23 _{(0.04–0.80)}	0.42 _{(−0.61 to 0.82)}	−0.18 _{(−0.69 to 0.27)}	−0.27_{(−0.79 to 0.12)}
Match average HR (%max)	0.54*_{(−0.31 to 0.85)}	0.74** _{(0.14–0.93)}	0.40 _{(−0.03 to 0.75)}	0.67* _{(0.01–0.92)}	0.15 _{(−0.41 to 0.63)}	0.39 _{(−0.16 to 0.76)}
Match IL (au)	0.55* _{(−0.09 to 0.80)}	0.24 _{(−0.64 to 0.78)}	0.33 _{(−0.29 to 0.62)}	0.15 _{(−0.75 to 0.87)}	0.37 _{(−0.20 to 0.74)}	0.08 _{(−0.46 to 0.71)}

RHIE = Repeated High-Intensity Efforts; WCS = Worst Case Scenario; HML = High Metabolic Load; HR = Heart Rate; IL = Internal Load. *correlation is significant at the p < 0.05 level. **correlation is significant at the p < 0.01 level.

Field-test results regression analysis

A linear regression model was applied between the YYIR₁ test final distance and the time taken to complete the Bronco test, which was statistically significant (R² = 0.78, F_{(1, 30)} = 104.46, p < 0.001), and had an explanatory power of 76.9% for YYIR₁ test distance. Hence, the following equation estimates the distance covered in the YYIR₁ test while taking into consideration the time taken in seconds to complete the Bronco test:

Equation 1 : Yo - Y o_{dist} = (– 15 .744 * Bronc o_{time}) + 6323.282

Where Yo-Yo_dist is the estimated distance in the YYIR₁ test (m) and Bronco_time is the time taken to complete the Bronco test (s). Equation 1 was created from the data of 32 rugby referees and has a bias of −0.11 m, an accuracy of 194.2 m, and a precision of 15.1%. Figure 2 shows Bland-Altman's graphical analysis between true and estimated values of the YYIR₁ distance from Equation 1, indicating acceptable agreement limits with all but two values falling within the 95% LoA (i.e. −386.7-m and 386.5-m). The average difference between true and estimated values of distance in the YYIR₁ test was similar to zero (t₃₁ = - 0.003, p = 0.998). By plotting a regression line between values identified in Bland-Altman's diagram, it was found that it was not statistically significant (R² = 0.07, F_(1,31) = 2.15, p = 0.153), and the unstandardized coeficient was similar to zero (β = 0.134), which means that no proportional bias in the data was obtained by Equation 1. Thus, the difference in values resulting from the two measures (true and estimated values) does not increase or decrease in proportion to the average values,⁴² resulting in an accurate estimation of the result from Equation 1. The average ICC was 0.93, with 95% confidence intervals from 0.87 to 0.97 (F_(31,31) = 14.93, p < 0.001). Thus, excellent reliability was found between the true and estimated values generated by Equation 1.

Figure 2.

Bland-Altmann plot analysis of differences and means for YYIR₁ final distance between actual values and estimated values from equation 1. The centre line is the mean difference, while the lines below and above the centre are the respective 95% of limits of agreement (1.96 × SD of Differences): −386.74 to 386.52-m.

Discussion

This study aimed to compare the YYIR₁ and Bronco tests in amateur rugby referees and verify the direct validity of these tests in relation to the physical and physiological demands placed on referees during actual match play. The results revealed a large significant correlation between the YYIR₁ and Bronco test results. Furthermore, significant medium correlations were found between the YYIR₁ test result and physical and physiological demands during actual match play, including total and high-intensity distance covered, maximal speed, number of high-intensity accelerations and sprints, number of RHIEs, WCS duration, distance covered during the WCS, number of HML efforts, and match HML distance. Finally, the time taken to complete the Bronco test was significant medium correlated with physical match demands including maximal speed, number of high-intensity accelerations, and relative distance covered within the WCS.

The large significant correlation found in this study between the distance covered in the YYIR₁ test and time taken to complete the Bronco test is similar to the results found for elite youth rugby players.¹⁵ Although the nature of the two field tests differ (e.g. continuous self-paced Bronco test and an intermittent externally-paced YYIR_1,¹⁵⁾ the average HR reported by rugby referees during both tests was around 90% of their maximum.³⁴ Additionally, the strength of the correlation between the YYIR₁ and Bronco test results might indicate that the tests measure similar characteristics (e.g. ability to perform intermittent high-intensity exercise) and are therefore potentially interchangeable. However, the greatest percentage of time (∼36%) was spent in the 91–95% HR_max zone during the Bronco test, whereas the largest percentage of time (∼35%) was spent in the > 96% HR_max zone during the YYIR₁ test This might be explained by the differences in pacing between the tests, with the externally-paced nature of the YYIR₁ test allowing referees to perform at a greater percentage of their maximum HR. Moreover, compared with the Bronco test, the maximum HR reached during the YYIR₁ test was significantly higher. Thus, in comparison with the Bronco test, the YYIR₁ test might better determine the maximal HR of rugby referees,⁹^,³⁴ information that is critical for training prescription.

A field test cannot be considered sufficiently sport-specific until a direct association has been observed between the field test result and the most relevant aspects of match performance.²⁰ Taking that into consideration, the YYIR₁ test result and internal load was significantly correlated with referees’ physical match demands. Furthermore, the YYIR₁ test maximum HR was significantly correlated with referees’ physiological match demands, supporting the direct validity of the test Collectively, the significant correlations found between the YYIR₁ test result and the high-intensity demands (i.e. high-intensity distance covered, maximal speed, high-intensity accelerations, sprints, RHIE, WCS duration, WCS distance, HML efforts, HML distance) placed on referees in a match, implies that the test is a good tool to evaluate if the referee is fit enough to cope with match demands, since it is the amount of high-intensity exercise that best predicts the development of fatigue.⁴³ Additionally, the significant correlation found between the YYIR₁ test results and number of sprints and RHIEs performed per match is an interesting finding, suggesting that the physical and physiological attributes required to perform the YYIR₁ test are similar to those required to maintain RHIEs and sprinting during a match. Further, the number of HML efforts performed by the referees was significantly correlated with YYIR₁ test result, which is in agreement with a study performed with elite soccer referees,⁴³ suggesting that the YYIR₁ test might provide a valid measure of physical performance in intermittent activities. Finally, the significant correlation found between the YYIR₁ test results and the WCS, might suggest that the better the rugby referee performed in the YYIR₁ test, the better they coped with the most demanding period in a match. Since, the YYIR₁ test elicits maximal aerobic responses and is focused on evaluating the ability to perform and recover from high-intensity exercise, which are essential components during the most intense periods in a match,^11,12 it seems logical to be performed by referees, because if a rugby referee can recover more quickly from maximal or near-maximal efforts, the referee would be most likely to complete more of these efforts during a match.⁴⁴ However, the constant changes of direction in a set distance (i.e. 20-m) does not accurately replicate the movements during rugby officiating and should be considered as a limitation of using the YYIR₁ test with rugby referees. Thus, alternative tests might be needed that incorporate more valid movement patterns while evaluating the ability to perform high-intensity exercise.

The Bronco test appears to have ecological validity for rugby referees as it includes shuttles across a variety of distances rather than a single fixed distance, as well as periods of acceleration, deceleration, and changes in direction, thus partially replicating the demands placed on rugby referees during actual match play. Indeed, the referee's maximal speed and number of high-intensity accelerations per match were significantly medium correlated with the Bronco test result. Additionally, maximum and average HR during the Bronco test and actual match play were significantly large correlated, which could provide an indicator of match fitness. However, the self-paced continuous nature of the Bronco test, unlike the intermittent nature of rugby refereeing, and the lack of correlations found between the Bronco test result and high-intensity match demands (e.g. high-intensity distance, number of sprints, and RHIEs), suggests that the Bronco test may not replicate the true metabolic demands placed on rugby referees during matches, which raises questions about the direct validity of the test, since the quantity of high-intensity exercise is a better indicator of match demands.⁴³ Based on that, it was not possible to establish the direct validity of the Bronco test for rugby referees in this study.

When analysing the results obtained by the Bland-Altman's graphical analysis, the findings demonstrate the possibility to use Equation 1 to estimate the YYIR₁ test final distance. From a practical viewpoint, Equation 1 provides an estimation of which YYIR₁ test final distance a referee would achieve considering the time taken to complete the Bronco test Although Equations 1 have excellent reliability,⁴¹ it should be acknowledged that the outcome (i.e. YYIR₁ test final distance) can have a variance of ∼15% compared to the true value. Also, the results generated by the Equation 1 may be considered a contrived variable when compared to the results generated when using the YYIR₁ test directly. Thus, Equation 1 should only be used as an estimation of YYIR₁ test performance and should not replace the actual test result.

This study uniquely adds to the literature regarding the fitness testing of rugby referees. However, some limitations should be noted. First, the correlation analyses were performed with a relatively small sample size, and so, similar studies with a larger sample size and from different levels (e.g. amateur, and elite referees) should be conducted in order to replicate the results of this study and to generate even more accurate estimating equation. Second, this study was conducted with southern hemisphere amateur rugby referees during a range of competitive 15-a-side matches of differing durations (i.e. 50 to 80 min), potentially limiting the generalizability of the findings, since no elite matches or referees were included in the sample. Finally, only observational data was collected during this study, and no laboratory-based tests were performed (e.g. VO_2max), thus preventing a more detailed analysis against a “gold standard” and ventilatory threshold zones.¹² Future research is therefore encouraged to include laboratory-based tests (e.g. VO_2max) and more physiological variables (e.g. blood lactate) to further investigate the relationships between field test results and actual match demands.

In conclusion, this study has shown that the YYIR₁ and Bronco tests are strongly correlated, therefore both tests might be useful in prescribing individualised fitness training for rugby referees. Additionally, this study has shown that the results of the YYIR₁ test are significantly correlated with the physical and physiological demands placed on rugby referees during actual match play and may therefore be considered a valuable tool for assessing the fitness of rugby referees. Moreover, given the relevance of high-intensity running for a rugby referee's physical match performance,⁴³ this study offers empirical support for the direct validity²⁰ of the YYIR₁ test for the physical assessment of rugby referees. However, the direct validity of the Bronco test could not be confirmed by the findings of this study. Overall, the results imply that while the field tests assessed in this study could be useful in providing information on rugby referees’ fitness, they should never be used to predict the referee's overall on-field performance because of the multitude of factors (e.g. positioning, psychological stress) impacting referee decision-making.⁴⁵

Practical applications

The findings of this study highlight the importance of selecting tests that accurately reflect sport-specific requirements. In this regard, the findings indicate that the YYIR₁ test better correlates with actual match demands than the Bronco test Thus, the use of the YYIR₁ test in high-intensity training sessions is recommended.⁴³ Additionally, for the purpose of evaluating if a referee is fit enough to cope with the match demands, the YYIR₁ test should be used over the Bronco test Moreover, this study showed that the YYIR₁ and Bronco tests are strongly correlated. Thus, the Bronco test might be a reasonable alternative to use in referees’ training sessions under time-constraints, as it provides a less disruptive option for practitioners to incorporate into training, and to measure intermittent sport performance.¹⁵ Finally, should a practitioner or referee want to estimate the equivalent YYIR₁ test result from the Bronco test, then the regression Equation 1 outlined in this study could be used as a guide to such comparison. However, since the Bronco test requires 60-m length on a field and is designed for outdoor use, caution must be advised when comparing seasonal changes or using the test to monitor fitness changes when environmental test conditions are not comparable.

Footnotes

Acknowledgements

The authors would like to thank the referees who took part. Further, we would like to thank the Western Province Referees Society, SARefs Academy, and Linston Manuels for allowing access to the referees. Also, the authors would like to thank STATSports for providing the GNSS units. Finally, the authors would like to thank Stephanie Kruger, Tayla Van Heerden, Hannah Bevis, Joshua Schenck, and Cameron Donkin for their help during the data collection. The authors report no conflicts of interest This study was partially supported by a Santander Mobility Award given to the lead author.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Wilbur J Kraak

Ricardo T Sant’Anna

References

Roberts

Trewartha

Higgitt

, et al. The physical demands of elite English rugby union. J Sports Sci 2008; 26: 825–833.

Elsworthy

Blair

Lastella

. On-field movements, heart rate responses and perceived exertion of lead referees in rugby world Cup matches, 2019. J Sci Med Sport 2020; 24: 386–390.

Blair

Elsworthy

Rehrer

, et al. Physical and physiological demands of elite rugby union officials. Int J Sports Physiol Perform 2018; 13: 1199–1207.

Suarez-Arrones

Garcia

Calvo-Lluch

, et al. Running demands and heart rate response in rugby union referees. J Strength Cond Res 2013; 27: 2946–2951.

Bester

Coetzee

Schall

, et al. Physical demands on elite lead rugby union referees. International Journal of Performance Analysis in Sport 2019; 19: 258–273.

Sant ‘Anna

Roberts

Moore

, et al. Rugby union referees’ physical and physiological demands across different competitive levels. J Sports Med Phys Fitness 2021; 61(6): 788–796.

Mallo

Gonzalez Frutos

Juarez

, et al. Effect of positioning on the accuracy of decision making of association football top-class referees and assistant referees during competitive matches. J Sports Sci 2012; 30: 1437–1445.

Sant’Anna

Roberts

Moore

, et al. Referee positioning, but not match demands, score difference, or field location, are associated with breakdown decision-making accuracy in elite rugby sevens referees. International Journal of Performance Analysis in Sport 2021; 21(6): 1127–1139.

Mallo

Navarro

Aranda

, et al. Activity profile of top-class association football referees in relation to fitness-test performance and match standard. J Sports Sci 2009; 27: 9–17.

10.

Weston

Castagna

Helsen

, et al. Relationships among field-test measures and physical match performance in elite-standard soccer referees. J Sports Sci 2009; 27: 1177–1184.

11.

Bangsbo

Iaia

Krustrup

. The Yo-Yo intermittent recovery test : a useful tool for evaluation of physical performance in intermittent sports. Sports Med 2008; 38: 37–51.

12.

Krustrup

Mohr

Amstrup

, et al. The yo-yo intermittent recovery test: physiological response, reliability, and validity. Med Sci Sports Exerc 2003; 35: 697–705.

13.

Blair

Cronin

Rehrer

, et al. Contextual review of physical requirements of refereeing rugby union at an elite level. Strength & Conditioning Journal 2018; 40: 17–30.

14.

Kelly

Wood

. The correlation between the 30-15 intermittent fitness test and a novel test of running performance. Journal of Australian Strength and Conditioning 2013; 21: 91–94.

15.

Deuchrass

Smith

Elliot

, et al. The 1.2 km shuttle run test: reliability and comparison with the Yo-Yo intermittent recovery level 1 test in young elite rugby union players. Journal of Australian Strength & Conditioning 2019; 27: 14–20.

16.

Kelly

Jackson

Wood

. Typical scores from the 1.2 km shuttle run test to determine maximal aerobic speed. Journal of Australian Strength and Conditioning 2014; 22: 183–185.

17.

Brew DJK

. The reliability of the 1.2 km shuttle run test for intermittent sport athletes. Journal of Australian Strength & Conditioning 2014; 22: 127–131.

18.

Gabbett

. Science of rugby league football: a review. J Sports Sci 2005; 23: 961–976.

19.

Dobbin

Highton

Moss

, et al. Concurrent validity of a rugby-specific Yo-Yo intermittent recovery test (level 1) for assessing match-related running performance. J Strength Cond Res 2018; 35: 1.

20.

Boddington

Lambert

Waldeck

Validity of a 5-meter multiple shuttle run test for assessing fitness of women field hockey players. J Strength Cond Res 2004; 18: 97–100.

21.

Lakens

. Sample Size Justification. PsyArXiv; 2021.

22.

Brown

Starling

Stokes

, et al. High concussion rate in student community rugby union players during the 2018 season: implications for future research directions. Front Hum Neurosci 2019; 13: 423.

23.

Painczyk

Hendricks

Kraak

. Intra and inter-reliability testing of a South African developed computerised notational system among western province club rugby coaches. Int J Sports Sci Coach 2018; 13: 1163–1170.

24.

Beato

Coratella

Stiff

, et al. The validity and between-unit variability of GNSS units (STATSports apex 10 and 18 Hz) for measuring distance and peak speed in team sports. Front Physiol 2018; 9: 1288.

25.

Mayo

Miles

Sims

, et al. The effect of resistance training in a hypoxic chamber on physical performance in elite rugby athletes. High Alt Med Biol 2018; 19: 28–34.

26.

Sant’Anna

Roberts

Moore

, et al. Physical demands of refereeing rugby sevens matches at different competitive levels. J Strength Cond Res 2021; 35(11): 3164–3169.

27.

Austin

Gabbett

Jenkins

. Repeated high-intensity exercise in professional rugby union. J Sports Sci 2011; 29: 1105–1112.

28.

Reardon

Tobin

Tierney

, et al. The worst case scenario: locomotor and collision demands of the longest periods of gameplay in professional rugby union. Plos One 2017; 12(5).

29.

Whitehead

Till

Weaving

, et al. The use of microtechnology to quantify the peak match demands of the football codes: a systematic review. Sports Med 2018; 48(11): 2549–2575.

30.

Cunningham

Shearer

Drawer

, et al. Movement demands of elite under-20 s and senior international rugby union players. PLoS One 2016; 11: e0164990.

31.

Couderc

Gabbett

Piscione

, et al. Repeated high-intensity effort activity in international male rugby sevens. J Strength Cond Res. Epub ahead of print 22 January 2019.

32.

Tanaka

Monahan

Seals

Age-predicted maximal heart rate revisited. J Am Coll Cardiol 2001; 37: 153–156.

33.

Boudet

Garet

Bedu

, et al. Median maximal heart rate for heart rate calibration in different conditions: laboratory, field and competition. Int J Sports Med 2002; 23: 290–297.

34.

Póvoas

Krustrup

Pereira

, et al. Maximal heart rate assessment in recreational football players. A study involving a multiple testing approach. Scand J Med Sci Sports 2019; 29.

35.

Edwards

. High performance training and racing. In: Edwards

(ed) The heart rate monitor book. Sacramento, CA: Fleet Feet Press, 1993, pp.113–123.

36.

Faul

Erdfelder

Lang

A-G

, et al. G*power 3: a flexible statistical power analysis program for the social, behavior, and biomedical sciences. Behav Res Methods Instrum Comput 2007; 39: 175–191.

37.

Hopkins

Marshall

Batterham

, et al. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exercise 2009; 41: 3–12.

38.

Cohen

. Statistical power analysis for the behavioral sciences. Hillsdale, N.J.: Lawrence Erlbaum Associates Publishers, 1988.

39.

Bland

Altman

. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986: 1:307–310.

40.

Walther

Moore

. The concepts of bias, precision and accuracy, and their use in testing the performance of species richness estimators, with a literature review of estimator performance. Ecography 2005; 28: 815–829.

41.

Koo

. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 2016; 15: 155–163.

42.

Ludbrook

. Confidence in altman-bland plots: a critical review of the method of differences. Clin Exp Pharmacol Physiol 2010; 37: 143–149.

43.

Krustrup

Bangsbo

. Physiological demands of top-class soccer refereeing in relation to physical capacity: effect of intense intermittent exercise training. J Sports Sci 2001; 19: 881–891.

44.

Lockie

Jalilvand

Moreno

, et al. Yo-yo intermittent recovery test level 2 and its relationship with other typical soccer field tests in female collegiate soccer players. J Strength Cond Res 2017; 31: 2667–2677.

45.

Rampinini

Bishop

Marcora

, et al. Validity of simple field tests as indicators of match-related physical performance in top-level professional soccer players. Int J Sports Med 2007; 28: 228–235.

Comparing the Yo-Yo intermittent and Bronco tests and their associations with match demands among amateur rugby union referees

Abstract

Keywords

Introduction

Methods

Participants

Study protocol

Testing procedures

Statistical analysis

Results

Field tests results

Relationships between YYIR1 and Bronco test data

Comparisons between HR zone data during the YYIR1 and Bronco tests

Relationships between field tests results and match demands

Relationships between field test physiological demands and match demands

Field-test results regression analysis

Discussion

Practical applications

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

ORCID iD

References

Relationships between YYIR₁ and Bronco test data

Comparisons between HR zone data during the YYIR₁ and Bronco tests