Match running performance fluctuations in elite soccer

This article was downloaded by: [University of Sunderland] On: 01 July 2013, At: 02:23 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Sports Sciences Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rjsp20

Match running performance fluctuations in elite soccer: Indicative of fatigue, pacing or situational influences? a

Paul S. Bradley & Timothy D. Noakes

b

a

Department of Sport & Exercise Sciences , University of Sunderland , Sunderland SR1 3SD , UK b

Department of Human Biology , University of Cape Town , South Africa Published online: 01 Jul 2013.

To cite this article: Paul S. Bradley & Timothy D. Noakes (2013): Match running performance fluctuations in elite soccer: Indicative of fatigue, pacing or situational influences?, Journal of Sports Sciences, DOI:10.1080/02640414.2013.796062 To link to this article: http://dx.doi.org/10.1080/02640414.2013.796062

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Journal of Sports Sciences, 2013 http://dx.doi.org/10.1080/02640414.2013.796062

Match running performance fluctuations in elite soccer: Indicative of fatigue, pacing or situational influences?

PAUL S. BRADLEY1 & TIMOTHY D. NOAKES2 1

Department of Sport & Exercise Sciences, University of Sunderland, Sunderland SR1 3SD, UK, and 2Department of Human Biology, University of Cape Town, South Africa

Downloaded by [University of Sunderland] at 02:23 01 July 2013

(Accepted 11 April 2013)

Abstract The aims of this study were to: (1) quantify match running performance in 5-min periods to determine if players fatigue or modulate high-intensity running according to a pacing strategy, and (2) examine factors impacting high-intensity running such as score line, match importance and the introduction of substitutes. All players were analysed using a computerised tracking system. Maintaining ‘high’ levels of activity in the first half resulted in a 12% reduction (P < 0.01) in the second half for high-intensity running (effect size [ES]: 0.8), while no changes were observed in ‘moderate’ and ‘low’ groups (ES: 0.0–0.2). The ‘high’ group covered less (P < 0.01) high-intensity running in the initial 10-min of the second versus first half (ES: 0.6–0.7), but this was not observed in ‘moderate’ and ‘low’ groups (ES: 0.2–0.4). After the most intense periods, players demonstrated an 8% drop in high-intensity running (P < 0.05) compared to the match average (ES: 0.2) and this persisted for 5-min before recovering. Players covered similar high-intensity running distances in matches with differing score lines but position-specific trends indicated central defenders covered 17% less (P < 0.01) and attackers 15% more high-intensity running during matches that were heavily won versus lost (ES: 0.9). High-intensity running distances were comparable in matches of differing importance, but between-half trends indicated that only declines (P < 0.01) occurred in the second half of critical matches (ES: 0.2). Substitutes covered 15% more (P < 0.01) high-intensity running versus the same time period when completing a full match (ES: 0.5). The data demonstrate that high-intensity running in the second half is impacted by the activity of the first half and is reduced for 5-min after intense periods. High-intensity running is also influenced by score line and substitutions but not match importance. More research is warranted to establish if fluctuations in match running performance are primarily a consequence of fatigue, pacing or tactical and situational influences. Keywords: football, pacing, fatigue, score, substitutions, match importance

Introduction The activity pattern of soccer is intermittent, with players switching between brief bouts of high-intensity running and longer periods of low-intensity exercise (Rampinini, Coutts, Castagna, Sassi, & Impellizzeri, 2007). During elite matches players cover a total distance of 10–13 km and 1–3 km of high-intensity running (Bangsbo, Norregaard, & Thorso, 1991; Bradley et al., 2009; Mohr, Krustrup, & Bangsbo, 2003). This results in an average intensity of ~70% of maximal oxygen uptake and elicits blood lactate concentrations of 3–6 mmol ∙ l−1 (Mohr, Krustrup, & Bangsbo, 2005). However, expressing match intensity as an average disguises the unique physiological stress induced during intense periods (Glaister, 2005). During such periods, heart rate can exceed 95% of maximal heart rate and peak blood lactate

concentrations can reach 8–12 mmol ∙ l−1 (Ali & Farrelly 1991; Bangsbo, 1994). Such demands would certainly threaten homeostasis (Edwards & Noakes, 2009). Research demonstrates that high-intensity running declines from the first to the second half of a match (Di Salvo, Gregson, Atkinson, Tordoff & Drust, 2009; Mohr et al., 2003), although some observe minimal changes (Di Salvo et al., 2007). Down regulation of running performance in the second half could be attributed to fatigue, as studies have reported depleted muscle glycogen stores at the end of a match (Bendiksen et al., 2012; Krustrup et al., 2006). Findings demonstrate that high-intensity running is also reduced temporarily after the most intense period of the match (Bradley et al., 2010, 2011; Di Mascio & Bradley, 2013; Mohr et al., 2003). This could be due to a decline in muscle

Correspondence: Dr Paul S. Bradley, Department of Sport & Exercise Sciences, University of Sunderland, Darwin Building, Chester Road, Sunderland SR1 3SD, UK. Email: [email protected] © 2013 Taylor & Francis


2

P. S. Bradley & T. D. Noakes

creatine phosphate, intramuscular acidosis or the accumulation of potassium in the muscle interstitium (Mohr et al., 2005), but temporary drops also seem to be related to the time the ball is out of play and the opportunity to engage in match activities (Carling & Dupont, 2011). Alternatively, some suggest that reductions in match running performance could be due to players employing conscious or subconscious pacing strategies to enable successful completion of the match (Drust, Atkinson, & Reilly, 2007; Edwards & Noakes, 2009). Although this is an attractive hypothesis, limited data exist to support or reject such a statement. Carling and Bloomfield (2010) observed that teams coped with an early player dismissal by sparing low-intensity activity in an attempt to preserve essential high-intensity running, which may suggest pacing or modified tactics. If players pace their efforts then covering ‘low’ to ‘moderate’ distances in the first half would enable them to have the available capacity to maintain match running performances in the second half. However, this has only been established in elite players across 45-min periods (Rampinini et al., 2007), which results in a substantial loss of information when attempting to elucidate factors associated with pacing or fatigue. Others have examined match-running performance across 5-min periods but did not demarcate between positions or investigate the influence of first half activities on second half performances (Weston, Drust & Gregson, 2011a). Moreover, this study did not quantify match running performance in stoppage time periods, which could provide additional information as some suggest that an ‘end spurt’ in the latter stages of the match could be expected if pacing occurs (Aughey, 2010). This would be characterised by a ‘reversed J-shape’ pacing strategy (Abbiss & Laursen, 2008), whereby players are aware of the duration remaining until the end of the match. Although this is a common scenario with stoppage time announcements, no evidence exists of a documented ‘end-spurt’ of activity in soccer. Research using global positioning system (GPS) technology observed reductions in high-intensity running towards the end of an Australian football match but players failed to produce an ‘end spurt’ of activity (Aughey, 2010). However, this study used a small sample and the GPS technology had limited resolution to precisely measure high-intensity running. No studies have investigated whether and to what extent pacing strategies occur for players in other football codes such as soccer. Thus, an in-depth examination of high-intensity running in 5-min periods of matches (plus peak and stoppage time periods) specific to position and the influence of first half activity on second half match running performances may provide new insight into pacing or fatigue in soccer. Although detailed analyses of match running performance could provide information on pacing or

fatigue, this would only demonstrate outcome behaviour and not necessarily the motive. Studies have investigated common factors that dictate physical performance, such as match location, outcome and standard (Castellano, Blanco-Villasenor, & Alvarez, 2011; Lago, Casais, Dominguez, & Sampaio, 2010), but not score line or importance, which could present motives that modulate match running performance. Furthermore, if pacing or fatigue is evident in soccer then substitutes introduced in the latter stages of a match would be expected to cover more high-intensity running than the equivalent time period when completing the full match. Research has reported such a trend (Carling, Espié, Le Gall, Bloomfield, & Jullien, 2010; Mohr et al., 2003) but either used a separate groups analysis with a small sample or limited differentiation of player position and time. Thus, it is clear that additional analyses on score line, match importance and the introduction of substitutes may provide an understanding of the factors that influence high-intensity running. Therefore, the aims of this study were to: (1) quantify match running performance in 5-min periods to determine if players fatigue or modulate high-intensity running according to pacing and (2) examine factors impacting high-intensity running in elite soccer matches. Method Match analysis With approval from the institutional ethics committee, English FA Premier League matches during successive seasons (2006/07–2008/09) were analysed using a multiple-camera computerised tracking system (Prozone Sports Ltd, Leeds, UK). Players’ movements were captured during matches by cameras positioned at roof level and analysed using proprietary software. The systems reliability and validity has been quantified to verify the capture process and data accuracy (Bradley et al., 2009; Di Salvo, Collins, McNeill, & Cardinale, 2006). Match running performance variables Players activities were coded into the following categories and speed thresholds: standing (0–0.6 km · h−1), walking (0.7–7.1 km · h−1), jogging (7.2–14.3 km · h−1), running (14.4–19.7 km · h−1), high-speed running (19.8–25.1 km · h−1) and sprinting (>25.1 km · h−1). The speeds for each category have been previously employed (Bradley et al., 2009). The absolute distances covered by players (m) were converted to a relative analysis of the distance covered per unit of time (m ∙ min−1). This enabled comparisons between various periods of the match including stoppage times. Total distance

Match Running Performance Fluctuations in Soccer represented the summation of distances in all categories. High-intensity running consisted of the combined distance in running, high-speed running and sprinting (≥14.4 km · h−1). Peak high-intensity running distance in a 5-min period represented the most intense period of the match (Mohr et al., 2003). Comparisons were then made between the average 5-min period (minus the peak) versus the subsequent 5-min periods up to a maximum of 15-min to quantify the time course of recovery.


Part 1. Criteria for evaluating discrete periods of match running performance Data were collected in pre-defined 5-min periods (plus stoppage times), separated into three levels of activity and classified as ‘high’, ‘moderate’ and ‘low’ based on the total distance covered in the first half (Rampinini et al., 2007). Data were sorted using percentiles to produce each level (‘low’: ≤30th percentile; ‘moderate’: 35–65th percentile and ‘high’: ≥70th percentile). To improve the analytical approach, boundaries were drawn between percentile thresholds to ensured demarcation between groups (e.g. minimising the chance of players in the lowend of ‘high’ covering a very similar total distance to players at the high-end of ‘moderate’). From a total sample of 186 players, 17 were removed to create these boundaries. Matches were played between teams of a similar end of season league placing. Teams were ranked into top, middle and bottom and only matches played between teams within each rank were analysed. Moreover, equal distribution of home and away fixtures was ensured. Based on these criteria, the match running performances of 169 players (single observations) in three levels and five positions were profiled (Table I; Part 1). Part 2. Criteria for evaluating factors influencing match running performance Data were analysed in pre-defined periods (plus stoppage times) to examine the influence of score line, importance and substitutions on match running

3

performance. Owing to limited availability of matches for the above, we were unable to present data in 5-min periods. Although some players were included in each subsection of Part 2, some of the data were comprised of a collection of different players. On account of the subjectivity involved in match selection, recommendations were provided from two UEFA qualified coaches with at least 5 years of elite coaching experience. Data pertaining to score line were collected from matches that were competitive (≤1 goal differential) throughout the entire 90-min compared with those heavily won or lost (score differential ≥3 goals). All data from heavily won or lost matches were single observations versus a median of two competitive matches. Match running performances of 54 players in five positions were profiled (Table I; Part 2) Data relating to match importance were collected from 55 players during critical matches whereby the outcome directly impacted upon Championships/ European places or relegation, with local derby matches also included. This was compared with matches of lower importance within the same seasonal period. Data from critical matches were single observations versus a median of two matches for lower importance. Data were obtained from five playing positions (Table I; Part 2). Data pertaining to substitutions were collected from 65 players completing full matches versus those they were introduced as a substitute. Substitution and full match comparisons were corrected for exactly the same time period and only included single observations within a similar seasonal period. Data were collected from players in various positions (Table I; Part 2). Independent of position, data were also sub-divided into substitutions that occurred early (45- to 65-min: n = 40) or late in the second half (65- to 90-min: n = 25). To further improve the analytical approach adopted, the average match running performance of all outfield players at the corresponding time-point minus the substitute was used to account for match tempo using the procedures described by Weston et al. (2011a,b).

Table I. Sample size (n) differentiated into subsection of the study and positional subsets. Part 1. Discrete Periods Position/Variable

Part 2. Influencing Factors

Low

Moderate

High

Total

Score

Importance

Substitution

Central defenders Full-backs Central midfielders Wide midfielders Attackers

29 10 3 1 13

12 20 11 5 9

1 9 23 18 5

42 39 37 24 27

14 14 12 7 7

14 15 11 7 8

9 9 13 20 14

All players

56

57

56

169

54

55

65

4



Statistical analysis Analyses were conducted using statistical software (SPSS Inc., Chicago, USA). Descriptive statistics were calculated on each variable and z-scores confirmed data normality. Factorial analysis of variance (ANOVA) tests were used to explore the influence of first half activity and the impact of both score line and importance on match running performance. Interactions were examined between measures of physical performance across playing positions and time periods. In the event of a difference occurring, univariate analyses using Bonferroni-corrected pairwise comparisons were employed. Differences between players introduced as substitutes versus the identical period of time when completing the full match or against the mean match running performances of the remaining players were determined using Bonferroni-corrected dependent and independent t-tests. Effect size (ES) was calculated to determine the meaningfulness of the difference (Cohen, 1988). The magnitude of the ES was classified as trivial (0.2–0.6), moderate (>0.6–1.2), large (>1.2–2.0) and very large (>2.0–4.0) based on guidelines from Batterham and Hopkins (2006). Statistical significance was set at P < 0.05. Data are presented as means and standard error of the mean.

Results Between half match running performance Reductions of 4–7% (P < 0.01) were observed for total distance in the second half for players maintaining ‘moderate’ and ‘high’ levels of activity in the first half (ES: 0.9–1.2), while this did not differ for the ‘low’ group (ES: 0.2). Maintaining ‘high’ levels of activity in the first half resulted in 12% less (P < 0.01) high-intensity running in the second half (ES: 0.8), but no changes were evident for ‘low’ to ‘moderate’ groups (ES: 0.0–0.2). No differences were observed for sprinting. After collapsing categories, total distance and high-intensity running were greater (P < 0.01) in the first versus the second half (ES: 0.2– 0.4) but sprinting remained unchanged (ES: 0.1). Part 1. Discrete periods of match running performance The ‘high’ group covered less (P < 0.01) total distance in the initial 10-min of the second half (ES: 0.9–1.0), in addition to other periods (70- and 85-min) versus the same first half period (ES: 0.8–0.9; Figure 1(a)). The ‘low’ and ‘moderate’ groups only covered less (P < 0.01) total distance in the initial 5- and 10-min periods of the second compared with the first half (ES: 0.7– 0.8). The ‘low’, ‘moderate’ and ‘high’ groups demonstrated declines (P < 0.01) in total distance between

the first versus last 5-min and stoppage time of the first half (ES: 1.0–1.3). Although, similar trends were evident between the first versus last 5-min and stoppage time of the second half in the ‘low’ group (ES: 0.5– 0.8), the ‘high’ group only illustrated differences (P < 0.05) between the first versus last 5-min but not stoppage time (ES: 0.5–0.8), with no differences observed for the ‘moderate’ group (ES: 0.4–0.6). The ‘high’ group covered less (P < 0.05) high-intensity running in the initial 10-min period of the second versus the first half (ES: 0.6–0.7), but this was not observed for ‘moderate’ and ‘low’ groups (ES: 0.2–0.4; Figure 1 (b)). The ‘low’, ‘moderate’ and ‘high’ groups demonstrated declines (P < 0.01) in high-intensity running between the first versus last 5-min of the first half (ES: 0.8–0.9), with differences (P < 0.01) only evident in stoppage time for the ‘low’ group (ES: 0.8). The ‘low’, ‘moderate’ and ‘high’ groups displayed similar highintensity running between the first versus last 5-min of the second half (ES: 0.2–0.5) and stoppage time period (ES: 0.0–0.3). No differences were observed for sprinting (Figure 1(c)). After collapsing categories, total distance covered was greater (P < 0.01) in the initial 10-min of the first versus the second half (ES: 0.5–0.6), with the only other exception occurring at the end of the half (40- versus 85-min; ES: 0.4). Players covered more (P < 0.01) total distance in the first versus final 5-min period of each half (ES: 0.6–1.0) and in first but not second half stoppage time (ES: 0.2–0.9). High-intensity running was greater (P < 0.01) during the initial 10-min of the first versus second half only (ES: 0.3–0.4), with the most comparable periods occurring in stoppage time. High-intensity running was greater (P < 0.01) in the first versus final 5-min and stoppage of the first half (ES: 0.5–0.6) but not in the second half (ES: 0.1–0.3). Central defenders and central/wide midfielders reduced (P < 0.05) their total distance in the initial 10-min of the second versus first half (ES: 0.7–0.9), while full-backs were only lower for the second 5-min period (ES: 0.7; Figure 1(d)). Total distance covered declined (P < 0.01) between the first versus last 5-min period of both halves for full-backs and attackers (ES: 0.8–1.0), while central defenders, central/wide midfielders only exhibited first half reductions (ES: 1.2–1.4). Only midfielders demonstrated reductions (P < 0.01) in high-intensity running in the initial 5-min of the second versus first half (ES: 0.7; Figure 1 (e)). Attackers and central/wide midfielders demonstrated declines (P < 0.01) in high-intensity running between the first versus last 5-min period of the first half (ES: 0.8–1.2) but this was not evident in full-backs and central defenders (ES: 0.4–0.5). Although, full-backs reduced (P < 0.05) their distances in first half stoppage time compared

Total Distance (m · min−1)

80

100

120

140

160

80

100

120

140

160

**

* * * * * * * *

Δ Δ Δ Δ

Δ

Δ Δ

**

Time (min)

+

*

*

** Δ Δ Δ Δ **

D

Time (min)

+

Δ Δ Δ Δ

Δ Δ Δ Δ

Δ Δ Δ Δ

A

* * * *

Δ

Δ

+

+

Δ Δ Δ Δ

Δ Δ

HIR Distance (m · min−1) HIR Distance (m · min−1) 10

20

30

40

50

10

20

30

40

50

st

'High' 1 half

ATT

* * *

FB

**

Time (min)

+

Δ ΔΔ Δ Δ Δ

Δ Δ

E

CB

Time (min)

+

Δ Δ Δ Δ

Δ Δ

Δ Δ

B

st

'Mod' 1 half

CM

+

+

st

WM

0

1

2

3

4

5

6

0

1

2

3

4

5

6

'Low' 1 half

Time (min)

+

F

Time (min)

+

C

+

+

Figure 1. (a) Total distance, (b) high-intensity running and (c) sprinting in 5-min periods of elite matches when players are separated based on ‘high’, ‘moderate’ and ‘low’ activity in the first half. (d) Total distance, (e) high-intensity running and (f) sprinting in 5-min periods of elite matches when players are separated into playing positions. Lower than same first half period: (*P < 0.05; **P < 0.01). Lower than first 5-min period of the half (ΔP < 0.05; ΔΔP < 0.01). Attacker (ATT), Full-backs (FB), Central defender (CB), Central midfielder (CM), Wide midfielder (WM), High-intensity running (HIR), Stoppage time (+). Data are presented as means and standard error of the mean.

Total Distance (m · min−1)

5 10 15 20 25 30 35 40 45

5 10 15 20 25 30 35 40 45

5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45

50 55 60 65 70 75 80 85 90 50 55 60 65 70 75 80 85 90

50 55 60 65 70 75 80 85 90 50 55 60 65 70 75 80 85 90

5 10 15 20 25 30 35 40 45

5 10 15 20 25 30 35 40 45

Sprint Distance (m · min−1) Sprint Distance (m · min−1)

50 55 60 65 70 75 80 85 90

50 55 60 65 70 75 80 85 90


Match Running Performance Fluctuations in Soccer 5

6


with the first 5-min (ES: 0.8). Sprinting was highly variable and thus no differences were found for any position across periods (Figure 1(f)). Players were tracked in the 15-min after the most intense period of the match to observe the time course of recovery. After the most intense period of the match, players demonstrated an 8% drop in high-intensity running (P < 0.01) compared with the match average, but this only persisted for 5min before recovering to mean values (ES: 0.2; Figure 2).

High-intensity running and sprinting were similar in matches that were competitive, heavily won or lost (Table II). Matches that were competitive or won produced decrements (P < 0.01) in high-intensity running in the second half but when losing highintensity running was maintained (ES: 0.1–0.3). Central defenders covered 10–17% less high-intensity running (P < 0.01) during matches that were heavily won versus lost or competitive (ES: 0.6–0.9). Attackers covered 15% and 54% more high-intensity running and sprinting in matches won (P < 0.01) versus defeated (ES: 0.9–1.4). Match running performance was unchanged between matches of differing importance (Table III). Between half trends indicated that total distance declined (P < 0.01) in the second half of both critical and less important matches (ES: 0.3–0.4) but highintensity running only declined in critical matches (ES: 0.2). Central defenders and full-backs demonstrated reductions (P < 0.05) in second half highintensity running in critically important matches only (ES: 0.5–0.6). Players introduced as substitutes covered more (P < 0.01) total and high-intensity running distance

50

HIR Distance (m · min−1)


Part 2: Factors influencing match running performance

40

30

*

20

10

0 Peak 5-min Post 5-min Post 10-min Post 15-min Average 5-min

Figure 2. The most intense 5-min period of a match and the time course of recovery in the subsequent 15-min. *Lower (P < 0.05) than average 5-min period minus the peak value. High-intensity running (HIR). Data are presented as means and standard error of the mean.

compared with the equivalent time period when completing the full match (ES: 0.5–0.6), although sprinting did not differ (ES: 0.2; Table IV). Total distance and high-intensity running were greater (P < 0.05) in central midfielders when introduced as substitutes compared with the exact time period during full matches (ES: 0.7–0.9). Sprinting was only higher (P < 0.05) in central defenders and full-backs when entering the match as substitutes (ES: 0.6–0.7). Data separated into early and late substitutions, produced similar relative increases for high-intensity running (14–16%; ES: 0.5) and total distance covered (7–8%; ES: 0.6) compared with the equivalent time period when completing the full match (Table IV). To account for match tempo, we compared the mean player match running performances with that of the substitutes. Data trends indicated that substitutes also covered more (P < 0.01) distance in total (ES: 1.0) and at highintensity (ES: 1.1) compared to the mean for all remaining players at the same time period, although sprinting did not differ (ES: 0.5). Discussion The present study revealed that total distance and high-intensity running were markedly lower in the second half for players in a ‘high’ activity group compared with ‘moderate’ and ‘low’ groups. This finding agrees with research separating elite players and referees into similar categories (D’Ottavio & Castagna, 2001; Rampinini et al., 2007; Weston, Castagna, Impellizzeri, Rampinini, & Abt, 2007) but this is the first study to investigate this trend in 5-min periods of matches. Interestingly, high-intensity running was lower in the initial 10-min of the second versus first half for the ‘high’ group but not for ‘moderate’ and ‘low’ groups. The differences denoted immediately after the half time interval could indicate a pacing strategy but given the complexity of soccer match-play and the limitations of the methods used (time-motion analysis), other factors cannot be discounted. For instance, the decline in high-intensity running may indicate fatigue occurs across the course of the second half due to the ‘high’ demands of the first half. This decline was evident across most equivalent periods of the match but the lack of statistical significance can be explained by the stringent Bonferroni correction applied due to the large array of pairwise comparisons (small-moderate effect sizes for all). A multitude of mechanisms have been proposed to explain fatigue development in soccer, but researchers have failed to identify its precise cause (Mohr et al., 2005). This is unsurprising given the complexities of match running performance, which is influenced by a myriad of factors (Drust et al., 2007). Fatigue is not causally linked to

105.3 ± 1.8 21.1 ± 0.9 1.38 ± 0.12

116.9 ± 1.9 31.0 ± 1.5 3.08 ± 0.34

123.7 ± 1.5 33.9 ± 1.6 2.35 ± 0.25

126.9 ± 3.2 36.6 ± 2.6 3.49 ± 0.26

106.3 ± 2.8 24.2 ± 1.6 2.99 ± 0.50

115.3 ± 1.5 29.0 ± 1.1 2.52 ± 0.16

Full-backs (n = 14) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

Central midfielders (n = 12) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

Wide midfielders (n = 7) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

Attackers (n = 7) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

All players (n = 54) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

First

Central defenders (n = 14) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

Position/Variable

110.6 ± 1.5Δ 27.2 ± 1.0Δ 2.71 ± 0.16

103.0 ± 2.8= 24.3 ± 1.7 3.89 ± 0.42

123.9 ± 2.2 35.7 ± 1.9 3.51 ± 0.39

117.6 ± 1.8Δ 31.3 ± 1.2= 2.41 ± 0.23

112.2 ± 2.1= 29.0 ± 1.4 2.99 ± 0.32

100.1 ± 1.6Δ 19.2 ± 1.1 1.69 ± 0.15

Second

Competitive

112.9 ± 1.4 28.1 ± 1.0 2.61 ± 0.15

104.6 ± 2.7 24.3 ± 1.6 3.44 ± 0.42

125.4 ± 2.2 36.1 ± 1.9 3.50 ± 0.23

120.6 ± 1.6 32.6 ± 1.3 2.39 ± 0.21

114.5 ± 1.8 30.0 ± 1.4 3.02 ± 0.30

102.6 ± 1.6 20.1 ± 0.9 1.54 ± 0.12

Total

115.6 ± 1.6 29.4 ± 1.23 2.71 ± 0.20

106.7 ± 2.1 25.0 ± 1.2 3.73 ± 0.35x

127.3 ± 3.2 37.5 ± 2.6 3.49 ± 0.27

125.7 ± 1.8 36.1 ± 2.0 2.71 ± 0.39

117.7 ± 1.9 32.0 ± 1.6 3.51 ± 0.30

103.2 ± 1.8 19.1 ± 1.1 1.01 ± 0.15º

First

109.5 ± 1.5Δ 26.6 ± 1.1Δ 2.69 ± 0.18

106.3 ± 2.9x 26.2 ± 2.0 4.20 ± 0.46x

120.6 ± 2.7Δ 35.4 ± 1.8 3.37 ± 0.30

116.9 ± 1.7Δ 31.0 ± 1.5Δ 2.44 ± 0.24

111.6 ± 2.1Δ 28.2 ± 1.4= 3.00 ± 0.35

97.2 ± 1.6Δ 17.0 ± 0.9º= 1.49 ± 0.13

Second

Heavy Win

112.5 ± 1.5 28.0 ± 1.1 2.69 ± 0.17

106.5 ± 2.4 25.6 ± 1.5x 3.96 ± 0.38α

124.0 ± 2.9 36.5 ± 2.1 3.44 ± 0.12

121.3 ± 1.6 33.5 ± 1.6 2.57 ± 0.25

114.6 ± 1.9 30.1 ± 1.4 3.26 ± 0.28

100.2 ± 1.6 18.0 ± 0.9~ 1.21 ± 0.10º

Total

115.2 ± 1.6 29.5 ± 1.2 2.45 ± 0.15

107.4 ± 2.8 23.8 ± 2.0 2.60 ± 0.49

121.8 ± 3.9 34.2 ± 2.8 2.70 ± 0.30

123.1 ± 1.7 34.8 ± 1.4 2.42 ± 0.24

119.4 ± 3.0 33.4 ± 2.1 3.06 ± 0.34

105.0 ± 2.3 21.5 ± 1.5 1.67 ± 0.18

First

111.3 ± 1.5Δ 28.4 ± 1.1 2.74 ± 0.16

100.4 ± 1.5= 20.6 ± 1.0+ 2.53 ± 0.28+

121.9 ± 4.4 34.4 ± 3.5 2.53 ± 0.39

120.1 ± 1.5 33.1 ± 1.1 2.83 ± 0.29

113.5 ± 2.0 31.9 ± 1.6 3.69 ± 0.31

101.5 ± 1.9 21.9 ± 1.27 1.94 ± 0.25

Second

Heavy Loss

113.2 ± 1.5 28.9 ± 1.1 2.60 ± 0.14

103.9 ± 2.0 22.2 ± 1.4 2.57 ± 0.37

121.8 ± 3.9 34.2 ± 2.9 2.60 ± 0.35

121.6 ± 1.4 33.9 ± 1.1 2.64 ± 0.20

116.4 ± 2.1 32.7 ± 1.6 3.38 ± 0.28

103.2 ± 2.0 21.7 ± 1.3 1.80 ± 0.19

Total

Table II. Match running performance in competitive matches (1 ≤ goal differential) versus those heavily won or lost (≥3 goal differential). αHigher (P < 0.01) in win than loss. xHigher (P < 0.05) in win than loss. ºLower (P < 0.05) in win compared to loss and competitive. ~Lower (P < 0.01) in win compared to loss and competitive. +Lower (P < 0.05) in loss compared to win and competitive. Δ Lower (P < 0.01) in second half. =Lower (P < 0.05) in second half. Total distance covered (TDC), High-intensity running (HIR) and Sprinting (SPR). Data are presented as means and standard error of the mean.


Match Running Performance Fluctuations in Soccer 7

103.9 ± 1.6 20.4 ± 0.7 1.53 ± 0.12 117.8 ± 1.7 31.5 ± 1.2 3.53 ± 0.28 123.8 ± 2.5 33.8 ± 1.6 2.52 ± 0.34 121.9 ± 2.8 32.8 ± 1.6 3.63 ± 0.48 113.9 ± 3.3 27.3 ± 2.3 2.93 ± 0.52 115.4 ± 1.4 28.7 ± 0.9 2.74 ± 0.18


Central midfielders (n = 11) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)




First


Position/Variable

110.8 ± 1.4Δ 27.3 ± 1.0Δ 2.77 ± 0.17

108.4 ± 4.1 27.1 ± 2.8 3.22 ± 0.37

120.7 ± 4.4 34.1 ± 2.7 3.72 ± 0.49

118.6 ± 2.3Δ 30.9 ± 1.4 2.62 ± 0.32

112.3 ± 1.5Δ 29.3 ± 1.2= 3.28 ± 0.34

99.6 ± 1.2Δ 19.0 ± 0.6= 1.59 ± 0.14

Second

Critical Importance

113.1 ± 1.4 28.0 ± 0.9 2.75 ± 0.16

111.1 ± 3.5 27.2 ± 2.5 3.08 ± 0.43

121.3 ± 3.4 33.4 ± 2.0 3.67 ± 0.41

121.1 ± 2.3 32.3 ± 1.4 2.57 ± 0.28

115.0 ± 1.5 30.4 ± 1.2 3.40 ± 0.29

101.7 ± 1.3 19.7 ± 0.6 1.56 ± 0.10

Total

113.9 ± 1.4 27.9 ± 0.9 2.53 ± 0.16

111.5 ± 3.3 26.9 ± 1.9 3.06 ± 0.20

119.5 ± 4.3 31.7 ± 3.0 3.49 ± 0.46

123.8 ± 2.7 33.7 ± 1.9 2.59 ± 0.39

114.7 ± 1.3 29.5 ± 0.9 2.84 ± 0.26

103.7 ± 1.6 20.1 ± 0.8 1.36 ± 0.12

First

110.4 ± 1.4Δ 26.9 ± 0.9 2.68 ± 0.18

108.6 ± 3.4 27.8 ± 2.1 3.36 ± 0.36

115.6 ± 3.5 30.0 ± 2.3 3.10 ± 0.50

117.2 ± 2.8Δ 29.2 ± 1.6Δ 2.25 ± 0.26

113.7 ± 1.7 30.3 ± 1.2 3.43 ± 0.36

99.8 ± 1.6Δ 19.3 ± 1.3 1.60 ± 0.20

Second

Lower Importance

112.1 ± 1.3 27.3 ± 0.9 2.60 ± 0.15

110.0 ± 3.3 27.4 ± 2.0 3.21 ± 0.26

117.5 ± 3.7 30.8 ± 2.6 3.29 ± 0.45

120.5 ± 2.7 31.4 ± 1.6 2.41 ± 0.25

114.2 ± 1.3 29.9 ± 0.9 3.14 ± 0.28

101.7 ± 1.5 19.7 ± 1.0 1.48 ± 0.13

Total

Table III. Match running performance in critical versus less important matches. ΔLower (P < 0.01) in second half. =Lower (P < 0.05) in second half. Total distance covered (TDC), High-intensity running (HIR) and Sprinting (SPR). Data are presented as means and standard error of the mean.


8 P. S. Bradley & T. D. Noakes

Match Running Performance Fluctuations in Soccer


Table IV. Match running performance in various playing positions and time periods for substitutes introduced in the second half versus the identical period of time in a full match. *Higher (P < 0.01) than the identical time period in full match. ΔHigher (P < 0.05) than the identical time period in full match. Early (45- to 65min) or late (65- to 90-min) refers to the time period of the second half the substitutes were introduced. Total distance covered (TDC), High-intensity running (HIR) and Sprinting (SPR). Data are presented as means and standard error of the mean.

Position/Variable

Substitution Equivalent Time in Appearance Full Match


108.8 ± 4.4 25.3 ± 2.7 1.85 ± 0.39Δ

99.7 ± 2.5 22.5 ± 2.6 1.11 ± 0.34


118.3 ± 4.7 32.8 ± 3.1 3.85 ± 0.59Δ

105.0 ± 4.8 25.3 ± 1.7 2.86 ± 0.49

Central midfielders (n = 13) 123.3 ± 2.6Δ TDC (m ∙ min−1) HIR (m ∙ min−1) 35.9 ± 2.4Δ SPR (m ∙ min−1) 2.34 ± 0.31

111.7 ± 4.3 30.28 ± 2.3 2.24 ± 0.41


124.4 ± 2.7Δ 37.5 ± 1.8 4.67 ± 0.58

117.9 ± 2.5 33.3 ± 1.8 3.63 ± 0.35


106.0 ± 3.0 26.7 ± 2.3 2.60 ± 0.43

103.2 ± 3.3 25.2 ± 1.4 3.37 ± 0.44


117.2 ± 1.7* 32.5 ± 1.2* 3.25 ± 0.27

109.2 ± 1.7 28.3 ± 1.0 2.84 ± 0.21

Early substitution (n = 40) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

114.1 ± 2.1* 30.9 ± 1.5* 3.13 ± 0.31

106.7 ± 2.0 27.1 ± 1.2 2.68 ± 0.25

Late substitution (n = 25) TDC (m ∙ min−1) HIR (m ∙ min−1) SPR (m ∙ min−1)

122.2 ± 2.8Δ 35.1 ± 2.0Δ 3.46 ± 0.48

113.1 ± 3.1 30.4 ± 1.8 3.10 ± 0.37

a single factor (Edwards & Noakes, 2009); however the sustained decline in match running performance in the second half has been associated with a lowering of muscle glycogen (Krustrup et al., 2006; Saltin, 1973). This could explain the marked reductions in high-intensity running during the second half period for the ‘high’ compared with ‘low’ and ‘moderate’ groups. Reductions in match running performance in the first period of the second half should be expected and could be attributed to the fall in muscle temperature during the half-time break. Research observed a 2°C drop in muscle temperature after the half-time break and this was

9

associated with a more pronounced decline in sprint performance before the second half (Mohr, Krustrup, Nybo, Nielsen, & Bangsbo, 2004). Although additional factors could explain the reduction in running performance at the start of the second half, such as the tempo of the match (Weston et al., 2011b) and the intense nature of the initial period of the first half which could be an inappropriate benchmark to use as a comparison period (Lovell, Barrett, Portas, & Weston, 2013). Alternatively, some suggest that reductions in match running performance could be due to players employing a pacing strategy to enable the completion of a match without any single physiological system failing (Drust et al., 2007). Although limited data exists to fully explain pacing in team sports, a model proposed by Edwards and Noakes (2009) suggests that players may dynamically modulate their high-intensity efforts in an attempt to avoid fatigue. The results of the present study and others (D’Ottavio & Castagna, 2001; Rampinini et al., 2007; Weston et al., 2007) could provide support for this model as players covering the lowest total distances in the first half had the available capacity to be able to maintain match running parameters in the second half. Additional complexities are found on closer inspection of the data in 5-min periods for various activity groups. Unsurprisingly, the ‘high’ and ‘low’ groups were primarily composed of midfielders and central defenders/attackers, respectively. Rampinini et al. (2007) did not report such a trend when observing second half declines in match running performance after ‘high’ first half activity, despite reporting positional variation in match running performance. Midfielders perform more total distance and high-intensity running than attackers and central defenders (Bradley et al., 2009, 2010, 2011; Di Salvo et al., 2009) and this finding could simply reflect position-specific decrements or tactical requirements. Thus, caution is needed before attributing these findings to fatigue or pacing as the tactical role of the player could also dictate the amount of high-intensity running undertaken in the first half and its resultant impact on the second half. Moreover, the analytical approach used in this section of the study (percentiles) could have marginally impacted on the data trends. We separated various groups based on the total distance covered in the first half but a limitation of such an approach is that there was limited separation between some players (e.g. players in the low-end of ‘high’ could have a total distance close to players at the high-end of ‘moderate’). This could have also been an issue with Rampinini et al.’s (2007) analytical approach in which an order function separated similar groups. Although we removed 17 players from the analysis to separate


10


groups, this is still a factor to consider when evaluating the data trends. Data collapsed independent of activity level or position highlights some noteworthy trends. Highintensity running distance was greater in the initial period of the first versus second half but after this period no differences were evident, with the most comparable periods occurring in stoppage time. On the contrary, total distance declined in the first and last part of the second half compared with the same first-half period. If team sports players do pace their efforts then some suggest an ‘end spurt’ in highintensity running in the latter stages of the match could be expected (Aughey, 2010). Although our data failed to support the ‘end spurt’ hypothesis, high-intensity running and sprinting were maintained throughout the second half compared with the same first half period. The maintenance of high-intensity running and sprinting probably occurred at the expense of a more pronounced decline in total distance throughout the second half. Carling and Bloomfield (2010) also observed a similar finding when teams attempted to cope with an early player dismissal. This trend could demonstrate a pacing strategy that spares low-intensity activity such as walking and jogging in an attempt to preserve essential high-intensity running (Drust et al., 2007; Edwards & Noakes, 2009). Studies have generally observed reductions in total distance and high-intensity running as the match progresses (Bradley et al. 2009; Rampinini et al., 2007) and concluded that fatigue mechanisms are responsible. However, these studies have tracked match running performance in 15- and 45-min periods, which results in a substantial loss of specific information. Thus, this study illustrates the importance of analysing data in 5-min periods (including stoppage times) to allow perturbations in match performance to be observed. Although others have observed declines in highintensity running after the most intense period (Di Mascio & Bradley, 2013; Mohr et al., 2003) this is the first study published to date that has quantified the time course of recovery. We observed that highintensity running after the most intense period was below the match average for 5-min before recovering. Krustrup et al. (2006) found sprint performance declined after intense periods in the first and second half but this was not associated with muscle lactate, pH, or glycogen content. It has been suggested that the temporary decline in high-intensity running during matches could be related to the accumulation of potassium in the muscle, which results in electrical disturbances that impact force development (Mohr et al., 2005). Alternatively, Edwards and Noakes (2009) suggest that such high demands would certainly threaten homeostasis and players thus seek out

extended opportunities to minimise energy expenditure. Thus, this study provides evidence of the need for recovery after the most intense periods and sports scientists should design training aimed at coping with multiple intense actions. However, caution is needed when attributing these temporary drops to fatigue or pacing due to the low magnitude of the effect size statistic and given that they also seem to be related to the time the ball is out of play and the opportunity to engage in match activities (Carling & Dupont, 2011). This was evident in the present study for players in the same match whereby a large stoppage between the 20- to 25-min period resulted in a substantial drop in match running performance as opposed to fatigue or pacing. Furthermore, the temporary drop in high-intensity running may have been underestimated, as our results are based on pre-defined 5-min periods of matches as opposed to rolling 1-min periods. Mohr et al. (2003) observed a 12% decline in high-intensity running after the most intense period using a rolling method, indicating that the drop could be even greater and extend for a longer period (Varley, Elias, & Aughey, 2012). Thus, future studies are advised to use rolling periods that also account for the degree the ball is out of play to fully establish the magnitude and time course of transient fluctuations in high-intensity running. Results demonstrated that players entering the second half as substitutes covered 15% more highintensity running compared with the identical time period when completing a full match. This was particularly evident in midfielders. Research reported a 25% greater distance in high-intensity running during the final 15-min of matches in substitutes versus separate players completing the full match (Mohr et al., 2003). Discrepancies between studies are due to the separate groups analyses and the low sample size in the Mohr et al. (2003) study. It could be argued that the differences are due to the time periods in which match running performance was monitored in each study (second half versus final 15-min). However, this is the first study to separated data into early and late substitutions, and this resulted in very similar relative changes in highintensity running compared with the same full match period. Carling et al. (2010) demarcated between midfielders and attackers but observed that the latter failed to improve their match running performance during the first 10-min of being introduced as substitutes compared with the equivalent full time period. This trend was evident within the present study and further supports the assertion that coaches replacing players still need to consider the contribution that a particular substitute can make based on situational and positional factors. Our data provide a more valid expression of full versus partial match


Match Running Performance Fluctuations in Soccer running performance comparisons across all playing positions and time periods of the second half. This data support the notion that selected playing positions completing full matches either experience fatigue in the second half or use a more effective pacing strategy compared with being introduced as an early or late substitute. This is the first study to demonstrate that players are able to maintain their high-intensity running performances in the second half of heavily defeated matches, but this was not evident in heavily won matches. This seems unsurprising as successful teams cover less distance in total and at high-intensity compared with unsuccessful teams (Di Salvo et al., 2009; Rampinini et al., 2007). Playing against strong opposition has been found to be associated with lower ball possession (Bloomfield, Polman, & O’Donoghue, 2005; Lago, 2009) and it is possible that teams that are heavily defeated have to cover greater high-intensity running distance without the ball in an attempt to close players down and regain possession (Bradley, Lago-Peñas, Rey, & Gomez Diaz, 2013). Team strategies are also influence by score line as teams will employ different playing styles when ahead, level or behind (Lago and Martin, 2007). This is the first study to highlight positional differences in match running performance during various score lines, with attackers covering more and central defenders less high-intensity running in matches heavily won versus lost. Bradley et al. (2011) found attackers in offensive formations covered more high-intensity running than in defensive formations and our findings could be linked to the tactical characteristics inherent within these systems. This highlights the complexities of match running performance and sports scientists must consider the influence of tactical and situational factors before attempting to draw inferences. Caution is also needed when interpreting these findings as a small sample size was used in some positional subsets, and this is especially relevant given the variable nature of some match running parameters (Gregson, Drust, Atkinson, & Di Salvo, 2010). However, this is an unavoidable drawback given the elite nature of the players and the rarity of the data set (e.g. heavily defeated/won matches using a repeated measures design). The results indicated that match running performance did not change considerably between matches of differing importance. Altogether, these findings would suggest that match importance does not impact on the overall physical demands of elite soccer match play. However, the aim of any teams’ tactics is to ensure optimal team organisation in order to best utilise the physical and technical capabilities of its players and thus this trend could simply reflect tactical discipline nullifying the

11

importance of the occasion. However, failure to find an effect for match importance could be due to the analysis of 45-min periods as opposed to highly sensitive 5-min periods that could have indicated that the initial period of the first half was higher in critical versus less important matches as teams attempt to establish their authority through increased tempo, and more research using such an approach is warranted. This study attempted to investigate match running performance fluctuations in elite soccer matches, but the reader should be aware of certain limitations. For instance, studies investigating influencing factors in soccer (Lago et al., 2010), generally use multivariate statistical models on variables common to the entire sample (match outcome, location, standard) and adjusted data for effective playing time (Castellano et al., 2011) but this was not possible within the present study given the lack of commonality in the data set (critical/less importance versus heavily won/lost matches). Finally, a limitation of using distances covered in various speed thresholds to determine fatigue or pacing is that most maximal accelerations do not result in speeds associated with high-intensity running but are metabolically taxing (Varley & Aughey, 2013). Thus, the true energy cost of match running performance cannot be established. In summary, the data demonstrate that highintensity running in the second half is impacted by the activity of the first half and is reduced for 5-min after intense periods. High-intensity running is influenced by score line and substitutions but not match importance. However, the reader must be aware of the present study’s limitations and the challenging nature of using time-motion data to determine if pacing or fatigue occurs in complex sports such as soccer. Thus, more research is warranted to establish if perturbations in match running performance are primarily a consequence of fatigue, pacing or tactical and situational influences.

References Abbiss, C.R., & Laursen, P.B. (2008). Describing and understanding pacing strategies during athletic competition. Sports Medicine, 38, 239–252. Ali, A., & Farrally, M. (1991). Recording soccer players’ heart rates during matches. Journal of Sports Sciences, 9, 183–189. Aughey, R.J. (2010). Australian football player work rate: evidence of fatigue and pacing? International Journal of Sports Physiology and Performance, 5, 394–405. Bangsbo, J. (1994). The physiology of soccer – with special reference to intense intermittent exercise. Acta Physiologica Scandinavica, 151(suppl 619), 1–155. Bangsbo, J., Norregaard, L., & Thorso. F. (1991). Activity profile of competition soccer. Canadian Journal of Sport Sciences, 16, 110–116.


12


Batterham, A.M., & Hopkins, W.G. (2006). Making meaningful inferences about magnitudes. International Journal of Sports Physiology and Performance, 1, 50–57. Bendiksen, M., Bischoff, R., Randers, M.B., Mohr, M., Rollo, I., Suetta, C., Bangsbo, J., & Krustrup, P. (2012). The Copenhagen soccer test: physiological response and fatigue development. Medicine and Science in Sport and Exercise, 44, 1595–1603. Bloomfield, J., Polman, R., & O’Donoghue, P. (2005). Effects of score-line on team strategies in FA Premier League soccer. Journal of Sports Sciences, 23, 192–193. Bradley, P.S., Carling, C., Archer, D., Roberts, J., Dodds, A., Di Mascio, M., Paul, D., Diaz, A.G., Peart, D., & Krustrup, P. (2011). The effect of playing formation on high-intensity running and technical profiles in English FA Premier League soccer matches. Journal of Sports Sciences, 29, 821–830. Bradley, P.S., Di Mascio, M., Peart, D., Olsen, P., & Sheldon, B. (2010). High-intensity activity profiles of elite soccer players at different performance levels. Journal of Strength and Conditioning Research, 24, 2343–2351. Bradley, P.S., Lago-Peñas, C., Rey, E., & Gomez Diaz, A. (2013). The effect of high and low percentage ball possession on physical and technical profiles in English FA Premier League soccer matches. Journal of Sports Sciences, Epub ahead of print. Bradley, P.S., Sheldon, W., Wooster, B., Olsen, P., Boanas, P., & Krustrup, P. (2009). High-intensity running in English FA Premier League soccer matches. Journal of Sports Sciences, 27, 159–168. Carling, C., & Bloomfield, J. (2010). The effect of an early dismissal on player work-rate in a professional soccer match. Journal of Science and Medicine in Sport, 13, 126–128. Carling, C., & Dupont, G. (2011). Are declines in physical performance associated with a reduction in skill-related performance during professional soccer match-play? Journal of Sports Sciences, 21, 63–67. Carling, C., Espié, V., Le Gall F., Bloomfield, J., & Jullien, H. (2010). Work-rate of substitutes in elite soccer: A preliminary study. Journal of Science and Medicine in Sport, 13, 253–255. Castellano, J., Blanco-Villasenor, A., & Alvarez, D. (2011). Contextual variables and time-motion analysis in soccer. International Journal of Sports Medicine, 32, 415–421. Cohen, J. (1988). Statistical power analysis for the behavioural sciences (2nd ed.). Hillsdale: Lawrence Erlbaum. D’Ottavio, S., & Castagna, C. (2001). Physiological load imposed on elite soccer referees during actual match play. Journal of Sports Medicine and Physical Fitness, 41, 27–32. Di Mascio, M., & Bradley, P.S. (2013). Evaluation of the most intense period of high-intensity running in English FA Premier League soccer matches. Journal of Strength and Conditioning Research. 27, 909–915. Di Salvo, V., Collins, A., McNeill, B., & Cardinale, M. (2006). Validation of ProZone: A new video-based performance analysis system. International Journal of Performance Analysis in Sport, 6, 108–119. Di Salvo, V., Baron, R., Tschan, H., Calderon Montero, F.J., Bachl, N., & Pigozzi, F. (2007). Performance characteristics according to playing position in elite soccer. International Journal of Sports Medicine, 28, 222–227. Di Salvo, V., Gregson, W., Atkinson, G., Tordoff, P., & Drust, B. (2009). Analysis of high intensity activity in Premier League soccer. International Journal of Sports Medicine, 30, 205–212. Drust, B., Atkinson, G., & Reilly, T. (2007). Future perspectives in the evaluation of the physiological demands of soccer. Sports Medicine, 37, 783–805.

Edwards, A.M., & Noakes, T.D. (2009). Dehydration: Cause of fatigue or sign of pacing in elite soccer? Sports Medicine, 39, 1–13. Glaister, M. (2005). Multiple sprint work: Physiological responses, mechanisms of fatigue and the influence of aerobic fitness. Sports Medicine, 35, 757–777. Gregson, W., Drust, B., Atkinson, G., & Di Salvo, V. (2010). Match-to-match variability of high-speed activities in premier league soccer. International Journal of Sports Medicine, 31, 237–242. Krustrup, P., Mohr, M., Steensberg, A., Bencke, J., Kjaer, M., & Bangsbo, J. (2006). Muscle and blood metabolites during a soccer game: Implications for sprint performance. Medicine and Science in Sports and Exercise, 38, 1165–1174. Lago, C. (2009). The influence of match location, quality of opposition, and match status on possession strategies in professional association football. Journal of Sports Sciences, 27, 1463–1469. Lago, C., & Martin, R. (2007). Determinants of possession of the ball in soccer. Journal of Sports Sciences, 25, 969–974. Lago, C., Casais, L., Dominguez, E., & Sampaio, J. (2010). The effect of situational variables on distance covered at various speeds in elite soccer. European Journal of Sport Science, 10, 103–109. Lovell, R., Barrett, S., Portas, M., & Weston, M. (2013). Reexamination of the post half-time reduction in soccer workrate. Journal of Science and Medicine in Sport, 16, 250–254. Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match performance of high-standard soccer players with special reference to development of fatigue. Journal of Sports Sciences, 21, 519–528. Mohr, M., Krustrup, P., & Bangsbo, J. (2005) Fatigue in soccer: A brief review. Journal of Sports Sciences, 23, 593–599. Mohr, M., Krustrup, P., Nybo, L., Nielsen, J., & Bangsbo, J. (2004). Muscle temperature and sprint performance during soccer matches – beneficial effects of re-warm-up at half time. Scandinavian Journal of Medicine and Science in Sports, 15, 136–143. Rampinini, E., Coutts, A.J., Castagna, C., Sassi, R., & Impellizzeri, F.M. (2007). Variation in top level soccer match performance. International Journal of Sports Medicine, 28, 1018–1024. Saltin, B. (1973). Metabolic fundamentals in exercise. Medicine and Science in Sports, 5, 137–146. Varley, M.C., & Aughey, R.J. (2013). Acceleration profiles in elite Australian soccer. International Journal of Sports Medicine, 34 (1), 34–39. Varley, M.C., Elias, G.P., & Aughey, R.J. (2012). Current matchanalysis techniques’ underestimation of intense periods of highvelocity running. International Journal of Sports Physiology and Performance, 7, 183–185. Weston, M., Castagna, C., Impellizzeri, F.M., Rampinini, E., & Abt, G. (2007). Analysis of physical match performance in English Premier League soccer referees with particular reference to first half and player work rates. Journal of Science and Medicine in Sport, 10, 390–397. Weston, M., Drust, B., & Gregson, W. (2011a). Intensities of exercise during match-play in FA Premier League referees and players. Journal of Sports Sciences, 29, 527– 532. Weston, M., Batterham, A.M., Castagna, C., Portas, M.D., Barnes, C., Harley, J., & Lovell, R. (2011b). Reduction in physical match performance at the start of the second half in elite soccer. International Journal of Sports Physiology and Performance, 6, 174–182.