See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/51070066

The effect of playing formation on high-intensity running and technical

profiles in English FA Premier League soccer matches

Article in Journal of Sports Sciences · May 2011

DOI: 10.1080/02640414.2011.561868 · Source: PubMed

CITATIONS READS

406 6,316

10 authors, including:

Christopher Carling David T Archer

Fédération Française de Football University of Sunderland
159 PUBLICATIONS 12,134 CITATIONS 27 PUBLICATIONS 2,409 CITATIONS

SEE PROFILE SEE PROFILE

Jenny Roberts
University of Sunderland
5 PUBLICATIONS 671 CITATIONS

SEE PROFILE

All content following this page was uploaded by Paul S Bradley on 20 July 2015.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [University of Sunderland]
On: 5 May 2011
Access details: Access Details: [subscription number 930572433]
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.informaworld.com/smpp/title~content=t713721847

The effect of playing formation on high-intensity running and technical

profiles in English FA Premier League soccer matches
Paul S. Bradleya; Chris Carlingbc; Dave Archera; Jenny Robertsa; Andrew Doddsa; Michele Di Mascioa;
Darren Pauld; Antonio Gomez Diaze; Dan Peartf; Peter Krustrupg
a
Department of Sports and Exercise, Soccer Research Group, University of Sunderland, Sunderland,
UK b LOSC Lille Metropole Football Club, Centre de Formation, Lille, France c Institute of Coaching
and Performance, University of Central Lancashire, Preston, UK d Department of Sports Science,
Southampton Solent University, Southampton, UK e Medical Department, Ipswich Football Club,
Ipswich, UK f Department of Sport, Health and Exercise Sciences, University of Hull, Hull, UK g
Department of Exercise and Sport Sciences, University of Copenhagen, Copenhagen, Denmark

First published on: 21 April 2011

To cite this Article Bradley, Paul S. , Carling, Chris , Archer, Dave , Roberts, Jenny , Dodds, Andrew , Di Mascio, Michele ,
Paul, Darren , Gomez Diaz, Antonio , Peart, Dan and Krustrup, Peter(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: 8,
821 — 830, First published on: 21 April 2011 (iFirst)
To link to this Article: DOI: 10.1080/02640414.2011.561868
URL: http://dx.doi.org/10.1080/02640414.2011.561868

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or 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, May 2011; 29(8): 821–830

The effect of playing formation on high-intensity running and technical

profiles in English FA Premier League soccer matches

PAUL S. BRADLEY1, CHRIS CARLING2,3, DAVE ARCHER1, JENNY ROBERTS1,

ANDREW DODDS1, MICHELE DI MASCIO1, DARREN PAUL4, ANTONIO GOMEZ DIAZ5,
DAN PEART6, & PETER KRUSTRUP7
1
Department of Sports and Exercise, Soccer Research Group, University of Sunderland, Sunderland, UK, 2LOSC Lille
Métropole Football Club, Centre de Formation, Lille, France, 3Institute of Coaching and Performance, University of Central
Lancashire, Preston, UK, 4Department of Sports Science, Southampton Solent University, Southampton, UK, 5Medical
Department, Ipswich Football Club, Ipswich, UK, 6Department of Sport, Health and Exercise Sciences, University of Hull,
Hull, UK, and 7Department of Exercise and Sport Sciences, University of Copenhagen, Copenhagen, Denmark
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

(Accepted 7 February 2011)

Abstract
The aim of this study was to examine the effect of playing formation on high-intensity running and technical performance
during elite soccer matches. Twenty English FA Premier League games were analysed using a multiple-camera
computerized tracking system (n ¼ 153 players). Overall ball possession did not differ (P 4 0.05) between 4–4–2, 4–3–3
and 4–5–1 formations (50%, s ¼ 7 vs. 49%, s ¼ 8 vs. 44%, s ¼ 6). No differences were observed in high-intensity running
between 4–4–2, 4–3–3 and 4–5–1 formations. Compared with 4–4–2 and 4–3–3 formations, players in a 4–5–1 formation
performed less very high-intensity running when their team was in possession (312 m, s ¼ 196 vs. 433 m, s ¼ 261 vs. 410 m,
s ¼ 270; P 5 0.05) but more when their team was not in possession (547 m, s ¼ 217 vs. 461 m, s ¼ 156 vs. 459 m, s ¼ 169;
P 5 0.05). Attackers in a 4–3–3 performed *30% more (P 5 0.05) high-intensity running than attackers in 4–4–2 and 4–5–
1 formations. However, the fraction of successful passes was highest in a 4–4–2 (P 5 0.05) compared with 4–3–3 and 4–5–1
formations. The results suggest that playing formation does not influence the overall activity profiles of players, except for
attackers, but impacts on very high-intensity running activity with and without ball possession and some technical elements
of performance.

Keywords: Football, playing system, motion analysis, physical performance, passes

Thomas, 1976), with some authors suggesting that

Introduction
the physical, technical, and tactical demands of
The physiological demands of soccer can be indi- contemporary soccer are increasing (Andersson,
rectly quantified through time–motion analysis Ekblom, & Krustrup, 2008; Carling et al., 2008).
(Bangsbo, Norregaard, & Thorsoe, 1991; Mohr, To accommodate for this, recent studies have
Krustrup, & Bangsbo, 2003), providing a valuable determined player activity profiles in different
method to evaluate match performance (Rampinini, professional leagues (English: Bradley, Di Mascio,
Impellizzeri, Castagna, Coutts, & Wisloff, 2008). Peart, Olsen, & Sheldon, 2010a; Italian and Danish:
This information can subsequently be used to Mohr et al., 2003; Swedish: Andersson et al., 2008;
develop and optimize soccer-specific physical pre- Spanish: Di Salvo et al., 2007). Variations in
paration programmes (Bradley et al., 2009; Carling, performance across playing positions (Di Salvo,
Bloomfield, Nelsen, & Reilly, 2008). Interest has Gregson, Atkinson, Tordoff, & Drust, 2009) and
grown in this form of analysis over the last four the occurrence of fatigue in match-play (Krustrup
decades (Bangsbo et al., 1991; Bloomfield, Polman, et al., 2006b; Mohr, Krustrup, & Bangsbo, 2005;
& O’Donoghue, 2007; Ekblom, 1986; Reilly & Reilly, Drust, & Clarke, 2008) have also been

Correspondence: Dr P. S. Bradley, Department of Sport Sciences, University of Sunderland, Darwin Building, Chester Road, Sunderland SR1 3SD, UK.
E-mail: [email protected]
ISSN 0264-0414 print/ISSN 1466-447X online Ó 2011 Taylor & Francis
DOI: 10.1080/02640414.2011.561868
 822 P. S. Bradley et al.

investigated. To date, no studies have examined the players, in particular their high-intensity running
effect of playing formation on physical and technical profiles; and (2) to analyse technical performance
performance in elite soccer. In recognition of the across playing formations.
significant role that time–motion analysis now plays
as a tool for analysing performance in elite soccer, it
is necessary to examine activity patterns across a Methods
variety of popular formations to attain an expression
Match analysis
of modern-day soccer.
Some researchers have suggested that high-intensity With institutional ethics approval and informed
running distance covered during match-play is a valid consent, 20 English FA Premier League games
measure of physical performance in elite soccer due to during the 2006–07 competitive season were analysed
its strong relationship with training status (Krustrup using a multiple-camera computerized tracking sys-
et al., 2003, 2006a; Krustrup, Mohr, Ellingsgaard, & tem (ProZone Version 3.0, ProZone Sports Ltd.1,
Bangsbo, 2005). Previous research demonstrated that Leeds, UK). All outfield players’ movements were
high-intensity running was a distinguishing character- captured during each game by eight colour cameras
istic between players in different playing positions, (Vicon surveyor dome SVFT-W23, Oxford, UK)
whereby midfielders covered more distance than positioned in each of the stadiums at roof level. The
central defenders, full backs, and attackers (Rampinini data captured were analysed using match analysis
et al., 2008). However, performance in all positions software (ProZone 3 and MatchViewer, ProZone
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

was identified by using the highly recognizable 4–4–2 Sports Ltd.1, Leeds, UK) to produce a dataset on
formation. In addition, research suggests that high- each player’s physical and technical performance
intensity running distance is reduced towards the end during a game. This system has been validated to
of the game and temporarily after intense periods verify the capture process and accuracy of the data
(Mohr et al., 2003). Bradley et al. (2010a) identified (Di Salvo, Collins, McNeill, & Cardinale, 2006).
temporary decrements in high-intensity running dur-
ing domestic and international match-play, for differ-
Game selection criteria
ent playing positions and with and without ball
possession. However, these studies failed to examine Originally, 70 games were included for analysis.
variations in high-intensity running according to Physical and technical performance data were col-
playing formation. lected from teams playing standard formations (4–4–
Performance in soccer is a consequence of techni- 2, 4–5–1, and 4–3–3). Each formation played against
cal and tactical skills of an individual (Bangsbo, the most commonly used 4–4–2 formation due to
1994). Although there is evidence to suggest that insufficient numbers in other categories. Two UEFA
physical performance deteriorates due to match- qualified coaches observed each of the games to verify
related fatigue, few studies have examined decre- that formation was consistent throughout the game.
ments in technical performance. In a recent study, Furthermore, care was taken by the coaches to exclude
Rampinini et al. (2008) observed a decrease during any games that involved dynamic formation transi-
match-play in several measures of technical skills, tions, as offensive formations such as a 4–3–3 can be
including involvements with the ball, short passes, switched to a more defensive 4–5–1 formation during
and successful short passes. Furthermore, Helgerud a game. The number of games for 4–4–2 (n ¼ 7), 4–3–
and colleagues (Helgerud, Engen, Wisloff, & Hoff, 3 (n ¼ 7) and 4–5–1 (n ¼ 6) formations was equally
2001) reported that 8 weeks of aerobic interval distributed. The games selected were played between
training led to both improved aerobic fitness and an teams of a similar placing in the league table. Teams
increase in the number of successful passes. Limited were ranked into top, middle, and bottom and only
information is available about the interaction of high- games played between teams within each rank were
intensity running patterns and technical performance analysed. Ranking was based on the end-of-season
in various formations. Positional roles and formations position. Games were played at the same time of day to
have evolved over the last few decades and physical reduce the possible influence of circadian variations
performance has been analysed using limited forma- on performance (Reilly & Brooks, 1986). Equal
tional changes (4–3–3: Reilly & Thomas, 1976; 4–4– distribution of games according to home and away
2: Bradley et al., 2010a). An in-depth examination of fixtures was ensured and the differential in score was
high-intensity running and technical performance no greater than two goals. The game outcome was
across formation could provide insight into position- similar for each of the respective formations to
specific changes during the game and provide a tool minimize any influence on ball possession (Lago &
for optimal training preparation. Therefore, the aims Martin, 2007). Thus, based on these stringent game
of the present study were: (1) to examine the effect of selection criteria, the performance of 10 teams in a
formation on the physical performance of elite soccer total of 20 matches was analysed. Data were collected
 Effect of playing formation on soccer performance 823

from players completing an entire match, allowing variation of the present match analysis system. The
performance in 153 players in three playing positions inter-observer coefficients of variation for total
to be profiled. Match performance was analysed for a distance covered, walking, running, high-speed run-
total of 71 defenders, 48 midfielders, and 34 attackers. ning, and high-intensity running were 52%, with
To ensure team and player confidentiality, all data the exception of sprinting, for which it was 3.5%
were desensitized before analysis. (Bradley et al., 2009). Furthermore, Bradley and
colleagues (Bradley, O’Donoghue, Wooster, &
Tordoff, 2007) observed good inter- and intra-
Match activities
observer agreement for the number and type of
Players’ activities were coded into the following recorded technical events and the player and second
categories and speed thresholds: standing player involved in the event (k 4 0.9).
(0–0.6 km  h71), walking (0.7–7.1 km  h71),
jogging (7.2–14.3 km  h71), running (14.4–
Statistical analysis
19.7 km  h71), high-speed running (19.8–
25.1 km  h71), and sprinting (425.1 km  h71). All statistical analyses were conducted using SPSS for
The speeds for each category are similar to those Windows v.14.0 (SPSS Inc., Chicago, IL). Descrip-
employed in previous time–motion studies (Di Salvo tive statistics were calculated on each variable using
et al., 2009; Mohr et al., 2003). z-scores to confirm the assumptions of normality.
Three-way analysis of variance was used to explore
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

differences between formations and examine the

High-intensity distances
interaction between measures of physical and tech-
High-intensity running consisted of running, high- nical performance across playing positions and match
speed running, and sprinting (14.4 km  h71). halves (3 6 3 6 2 design: formation, positions, half).
Very high-intensity running consisted of high-speed In the event of a significant difference, univariate
running and sprinting (19.8 km  h71). Very high- analyses using Bonferroni-corrected pair-wise com-
intensity running with ball possession was defined as parisons were employed. Statistical significance was
the very high-intensity running distance covered set at P 5 0.05. Data are presented as means and
when the player’s own team was in possession of standard deviations (s) unless otherwise stated.
the ball. Very high-intensity running without ball
possession was defined as the very high-intensity
running distance covered when the opposition team Results
was in possession. Very high-intensity running when
Match distances
the ball was out of play consisted of the very high-
intensity distance covered during natural breaks in The total distance covered during a match was
the game (i.e. corners and free kicks). Peak distance similar across formations (Table I). Greater dis-
covered in high-intensity running in a 5-min period tances were covered in walking (P 5 0.05) in a 4–5–
represented the 5 min that contained the most high- 1 than a 4–4–2 formation. The amount of distance
intensity running in a game, and was specific for each covered in jogging was higher (P 5 0.01) for both
player profiled (Mohr et al., 2003). the 4–4–2 and 4–3–3 formations than for the 4–5–1
formation (4290 m, s ¼ 620 and 4304 m, s ¼ 665 vs.
4121 m, s ¼ 662). The analysis of distance covered
Additional match analysis measures
across match halves showed no differences
Mean recovery time was defined as the time that (P 4 0.05) between formations (Table I).
elapsed between discrete very high-intensity bouts
(19.8 km  h71). Match analysis also included the
High-intensity running profile
concomitant coding of technical game events (num-
ber of passes and percentage completion rate, passes The distance covered in high- and very high-
received, touches per possession, dribbles, crosses, intensity running was similar (P 4 0.05) in 4–4–2,
final third entries, possession won and lost, and total 4–3–3 and 4–5–1 formations (Table I). The distance
ball possession) with the operation definitions in covered in very high-intensity running with ball
accordance with those previously employed (Anders- possession in 4–3–3 and 4–4–2 formations was 32–
son et al., 2008). 39% higher (P 5 0.01) than in a 4–5–1 (Figure 1).
In contrast, *19% more distance was covered
(P 5 0.01) at very high-intensity without possession
Inter- and intra-observer reliability
in a 4–5–1 versus 4–4–2 and 4–3–3 formations
Reliability studies have previously been conducted to (Figure 1). There were no differences (P 4 0.05)
determine the inter- and intra-observer coefficient of between playing formation and distances covered at
 824 P. S. Bradley et al.

high intensities, according to possession across game intensity bouts or for the mean recovery times
halves. No differences (P 4 0.05) were found between these actions (Table 1).
between formations for the total number of high-
Decrements in peak high-intensity running
Table I. Physical performance according to team formation. The decline in peak high-intensity running immedi-
Formation
ately after the most intense 5-min period was similar
(P 4 0.05) across formations: 4–4–2 (229 vs. 117 m
4–4–2 4–3–3 4–5–1 or 49% decline), 4–3–3 (227 vs. 113 m or 50%
Variables (n ¼ 58) (n ¼ 49) (n ¼ 46)
decline) and 4–5–1 (234 vs. 122 m or 48% decline).
Distance (m) However, performance was reduced (P 5 0.05) in
Total 10697 + 945 10786 + 1041 10613 + 1104 the 5-min period immediately after the most intense
1st half 5371 + 482 5457 + 534 5347 + 557 5-min period compared with the mean high-intensity
2nd half 5327 + 514 5329 + 539 5266 + 641 distance for 5-min game periods in 4–4–2 and 4–3–3
VHI running 956 + 302 924 + 316 901 + 305
formations (Figure 2).
1st half 475 + 158 478 + 160 444 + 175
2nd half 482 + 172 447 + 178 457 + 160
HI running 2633 + 671 2649 + 706 2585 + 734
Playing positions
1st half 1330 + 353 1375 + 368 1311 + 402
2nd half 1304 + 369 1275 + 367 1274 + 386 Defenders playing in a 4–4–2 formation covered
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

Jogging 4290 + 620* 4304 + 665* 4121 + 662

more (P 5 0.01) total distance than defenders in a
1st half 2178 + 337 2206 + 359 2114 + 337
2nd half 2112 + 321 2099 + 334 2008 + 370 4–3–3 or 4–5–1 formation (Table II). Attackers in a
Walking 3774 + 307 3832 + 279 3907 + 257# 4–3–3 formation covered 28–32% more high-inten-
1st half 1864 + 176 1876 + 155 1923 + 139 sity running (P 5 0.05) and 22–32% more very high-
2nd half 1911 + 160 1956 + 142 1984 + 149 intensity running compared with attackers in 4–5–1
Number HI 122 + 37 120 + 39 116 + 40
and 4–4–2 formations (Table II). In contrast,
actions
1st half 60 + 19 60 + 20 57 + 23 defenders in a 4–4–2 ran *11% more distance in
2nd half 62 + 22 60 + 21 58 + 22 high-intensity running (P 5 0.01) than those in a 4–
Recovery 50 + 18 54 + 25 57 + 24 5–1 formation. For very high-intensity running,
time HI (s) defenders in a 4–4–2 ran more distance (P 5 0.05)
1st half 49 + 18 51 + 20 56 + 30
than defenders in a 4–5–1 and 4–3–3 formation
2nd half 52 + 23 56 + 33 57 + 25
(Table II). Defenders in a 4–3–3 formation covered a
Note: Data represent means and standard deviations. HI ¼ high- greater (P 5 0.01) total distance and distance in
intensity, VHI ¼ very high-intensity. *Greater distance covered in high- and very high-intensity running in the first
4–4–2 and 4–3–3 than 4–5–1 (P 5 0.01). #Greater distance compared with the second half (Table II). Attackers
covered in 4–5–1 than 4–4–2.
in a 4–5–1 formation covered less distance

Figure 1. Very high-intensity running distance with possession of Figure 2. Peak high-intensity running distance in 5-min periods for
the ball, without possession of the ball, and ball out of play for various formations. High-intensity running in the most intense 5-
various playing formations. Data represent means and standard min period during the game (peak 5-min), the following 5-min
deviations. *Distance covered with possession of the ball was lower period (next 5-min), and the game mean 5-min period (minus the
with the 4–5–1 compared with 4–4–2 and 4–3–3 formations peak value) for 4–4–2, 4–3–3 and 4–5–1 formations. Data
(P 5 0.01). DDistance covered without possession of the ball was represent means and standard deviations. *Different from the
higher with the 4–5–1 compared with 4–4–2 and 4–3–3 formations mean 5-min period (P 5 0.01). DDifferent from the mean 5-min
(P 5 0.01). period (P 5 0.05).
 Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

Table II. Physical performance across playing position within formations.

Defenders Midfielders Attackers

Variables
4–4–2 (n ¼ 30) 4–3–3 (n ¼ 22) 4–5–1 (n ¼ 19) 4–4–2 (n ¼ 18) 4–3–3 (n ¼ 14) 4–5–1 (n ¼ 16) 4–4–2 (n ¼ 10) 4–3–3 (n ¼ 13) 4–5–1 (n ¼ 11)

Distance (m)
Total 10452 + 755* 10073 + 852 10123 + 875 11505 + 783 11586 + 494 11606 + 722 9982 + 769 11130 + 999 10012 + 946
1st half 5221 + 392 5123 + 456# 5052 + 519 5823 + 343 5913 + 287 5819 + 324 5009 + 355 5530 + 487 5167 + 447
2nd half 5231 + 445 4950 + 430 5071 + 423 5681 + 457 5673 + 261 5787 + 492 4974 + 473 5600 + 533 4842 + 542
VHI running 862 + 309{ 751 + 273 748 + 293 1118 + 262 985 + 299 1103 + 259 950 + 236 1155 + 231¤ 870 + 227
1st half 437 + 179 406 + 146$ 355 + 159 537 + 135 497 + 170 535 + 168 474 + 94 578 + 114 464 + 146#
2nd half 424 + 176 344 + 160 393 + 161 581 + 137 488 + 145 568 + 137 475 + 149 576 + 137 406 + 102
HI running 2454 + 6328 2218 + 625 2207 + 691 3146 + 550 3013 + 538 3207 + 555 2250 + 454 2988 + 614D 2333 + 458
1st half 1249 + 352 1183 + 343$ 1093 + 382 1583 + 289 1568 + 303 1616 + 306 1114 + 180 1491 + 331 1243 + 284$
2nd half 1205 + 361 1035 + 323 1114 + 350 1563 + 278 1445 + 255 1591 + 321 1135 + 306 1497 + 302 1090 + 216
Number HI actions 111 + 37 98 + 36 93 + 33 146 + 31 134 + 28 147 + 32 110 + 24 142 + 34þ 108 + 25
1st half 56 + 21 51 + 21 46 + 18 70 + 16 67 + 17 72 + 25 53 + 12 67 + 18 57 + 18
2nd half 55 + 22 47 + 19 47 + 20 76 + 20 67 + 13 75 + 21 57 + 15 75 + 18 52 + 13
Recovery time HI (s) 56 + 21 67 + 29 70 + 28 39 + 8 44 + 10 41 + 11 53 + 11 42 + 13 55 + 15
1st half 54 + 22 60 + 23 69 + 30 40 + 8 43 + 14 42 + 14 52 + 9 44 + 15 52 + 17
2nd half 58 + 27 74 + 41 71 + 39 38 + 10 43 + 11 39 + 11 53 + 16 40 + 12 58 + 17

Note: Data represent means and standard deviations. HI ¼ high-intensity, VHI ¼ very high-intensity. *Defenders playing in a 4–4–2 formation ran a greater total distance than defenders in a 4–3–3 or
4–5–1 formation (both P 5 0.01). {Defenders in a 4–4–2 formation ran greater distances than defenders in a 4–5–1 (P 5 0.01) or 4–3–3 (P 5 0.05) formation. 8Defenders in a 4–4–2 formation
covered a greater distance in high-intensity running (P 5 0.001) than those in a 4–5–1 formation. DAttackers in a 4–3–3 formation covered greater distances in high-intensity running than those in a 4–
5–1 (P 5 0.05) or 4–4–2 (P 5 0.01) formation. ¤Attackers in a 4–3–3 formation covered greater distances than attackers in a 4–5–1 formation (P 5 0.01). þAttackers in a 4–4–3 formation undertook
more high-intensity actions than those in a 4–5–1 (P 5 0.05) or 4–4–2 (P 5 0.01) formation. #Difference between halves (P 5 0.05). $Difference between halves (P 5 0.01).
Effect of playing formation on soccer performance
825
 826 P. S. Bradley et al.

(P 5 0.05) in high-intensity and very high-intensity (Table III). However, total ball possession did not
running in the second half (Table II). differ (P 4 0.05) between 4–4–2, 4–3–3 and 4–5–1
In possession, attackers, defenders, and midfielders formations (50%, s ¼ 7 vs. 49%, s ¼ 8 vs. 44%,
in 4–4–2 and 4–3–3 formations performed more s ¼ 6). A 22% reduction (P 5 0.01) was evident in
distance (P 5 0.01) in very high-intensity running the number of passes per player in the second half
than players in a 4–5–1 formation (Figure 3). Out of compared with the first half in a 4–4–2 system. The
possession, attackers in a 4–5–1 and 4–3–3 formation percentage of successful passes was highest in a 4–4–
ran 37–68% more (P 5 0.01) distance at very high 2 formation (P 5 0.05) compared with 4–3–3 and 4–
intensity than attackers in a 4–4–2. Similarly, greater 5–1 formations (80%, s ¼ 11 vs. 73%, s ¼ 13 vs.
distances were covered at very high intensities out of 72%, s ¼ 16). Players in 4–4–2 (P 5 0.01) and 4–3–3
possession by defenders and midfielders in a 4–5–1 (P 5 0.05) formations received more passes than
(P 5 0.01) compared with those in a 4–4–2 and 4–3–3 those in a 4–5–1 formation (Table III).
formation (Figure 3). Finally, analysis of the number
of high-intensity runs across playing positions indi-
Discussion
cated that attackers performed more actions in a 4–4–3
compared with a 4–5–1 (P 5 0.05) or 4–4–2 The present study provides a detailed investigation of
(P 5 0.01) formation. The decline in peak high- the physical and technical performance of elite soccer
intensity running immediately after the most intense
5-min period was more pronounced (P 5 0.05) in
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

Table III. Technical profile according to team formation.

midfielders in a 4–5–1 (58% decline) than a 4–4–2
formation (43% decline). Furthermore, a difference Formation
(P 5 0.05) was observed between performance in the 4–4–2 4–3–3 4–5–1
next 5-min and average 5-min period for midfielders Per player (n ¼ 58) (n ¼ 49) (n ¼ 46)
playing in a 4–5–1 formation.
Passes 32.1 + 11.7 28.8 + 16.7 21.2 + 11.0*
1st half 18.2 + 7.2 14.4 + 8.4 9.9 + 5.4
Technical analysis 2nd half 13.9 + 6.9# 14.4 + 9.1 11.3 + 6.8
% Successful 79.5 + 10.5þ 73.4 + 13.2 71.6 + 15.5
Players in a 4–4–2 and a 4–3–3 performed more passes
(P 5 0.01) passes than those in a 4–5–1 formation 1st half 79.7 + 13.7þ 71.7 + 19.1 71.3 + 15.4
2nd half 79.5 + 12.6þ 73.1 + 16.4 71.5 + 19.8
Passes received 34.3 + 12.2 30.8 + 18.0 24.9 + 11.1D
1st half 19.1 + 7.2 15.2 + 9.5 12.1 + 5.8
2nd half 15.2 + 7.0 15.6 + 10.0 12.8 + 6.9
Touches per 2.5 + 0.4 2.6 + 0.7 2.4 + 0.5
possession
1st half 2.5 + 0.4 2.6 + 0.8 2.3 + 0.7
2nd half 2.5 + 0.6 2.6 + 0.7 2.4 + 0.6
Dribbles 0.3 + 0.7 0.5 + 1.2 0.4 + 0.9
1st half 0.2 + 0.6 0.3 + 0.7 0.1 + 0.3
2nd half 0.1 + 0.3 0.2 + 0.7 0.3 + 0.7
Crosses 1.7 + 2.1 1.6 + 2.1 1.0 + 1.3
1st half 0.8 + 1.2 0.9 + 1.3 0.5 + 0.6
2nd half 0.9 + 1.3 0.7 + 1.2 0.5 + 0.9
Final third 5.9 + 4.0 6.3 + 4.1 5.5 + 3.5
entries
1st half 3.3 + 2.9 2.8 + 2.3 2.7 + 1.9
2nd half 2.6 + 2.2 3.5 + 2.3 2.8 + 2.0
Possessions 22.8 + 11.9 21.7 + 10.4 18.3 + 9.9
won
1st half 11.4 + 6.6 10.9 + 5.7 9.2 + 5.7
2nd half 11.4 + 6.3 10.8 + 5.7 9.1 + 5.1
Possessions 21.8 + 6.0 22.5 + 6.7 20.8 + 7.4
lost
Figure 3. Positional variations in very high-intensity running 1st half 11.1 + 3.9 11.8 + 3.8 10.9 + 4.2
distance with possession of the ball, without possession of the ball, 2nd half 10.7 + 3.7 10.7 + 4.7 9.9 + 4.4
and ball out of play in different playing formations. Data represent
means and standard deviations. *Distance covered with possession Note: Data represent means and standard deviations. *Lower
was lower with the 4–5–1 compared with 4–4–2 and 4–3–3 number of passes in the 4–5–1 than 4–4–2 and 4–3–3 formations
formations (P 5 0.001). .Distance covered without possession (P 5 0.01). DLower number of passes received in the 4–5–1 than
was lower (P 5 0.01) with the 4–4–2 compared with 4–5–1 and 4– 4–4–2 (P 5 0.05) and 4–3–3 (P 5 0.05) formations. #Lower
3–3 formations. DDistance covered without possession was higher number of passes in the second half versus the first half with 4–4–2
with the 4–5–1 compared with 4–4–2 and 4–3–3 formations formation (P 5 0.01). þHigher number of passes in 4–4–2
(P 5 0.01). compared with 4–3–3 and 4–5–1 formations (P 5 0.05).
 Effect of playing formation on soccer performance 827

players within three common playing formations. coaching domain, a 4–5–1 is a more defensive
Players in 4–4–2 and 4–3–3 formations covered system than a 4–4–2 or 4–3–3 due to the reinforce-
greater distances in jogging while players in a 4–5–1 ment of the midfield zones at the expense of a
covered more distance walking. Despite the similar- forward player (Bangsbo & Peitersen, 2000; Bauer,
ity in the overall distance covered in very high- 1993). Therefore, the defensive nature of the 4–5–1
intensity running between various formations, formation may lead to substantially less very high-
players in a 4–5–1 formation performed less very intensity distance being covered by players when
high-intensity running when their team was in attacking and more very high-intensity running
possession, but more when not in possession, distance when defending.
compared with 4–4–2 and 4–3–3 formations. Re- Previous research has shown that the physical
ductions between halves also varied according to efforts of players are highly dependent on their
playing position, with attackers covering *30% positional role in the team (Di Salvo et al., 2007,
more very high-intensity running in a 4–3–3 forma- 2009; Mohr, Krustrup, Andersson, Kirkendall, &
tion than in 4–4–2 and 4–5–1 formations. Passing Bangsbo, 2008) and that physical conditioning
frequency was dependent on team formation, with should therefore be position-specific (Carling et al.,
the number of passes highest in a 4–4–2 compared 2008; Di Salvo & Pigozzi, 1998). This study is the
with 4–3–3 and 4–5–1 formations. first to demonstrate that physical performance across
The present results indicate that general match playing positions is also dependent upon the team
activity profiles do not differ considerably between formation employed. Attackers within a 4–3–3
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

players in three common playing formations. Irre- covered a greater total, high- and very high-intensity
spective of formation, players were shown to cover a running distance than players in 4–4–2 and 4–5–1
similar total and high-intensity running distance. formations. Similarly, defenders in a 4–4–2 covered a
The number of high-intensity runs and recovery greater total and high-intensity running distance than
times between very high-intensity bouts were similar defenders in the 4–3–3 and 4–5–1 formations.
across formations, although more efforts were Further analysis of very high-intensity activity pat-
observed in players in 4–4–2 and 4–3–3 formations. terns showed that players in all positions across both
Altogether, these findings would suggest that team 4–3–3 and 4–4–2 systems ran greater distances when
formation does not impact on the overall physical their team was in ball possession compared with
demands of elite soccer match-play. Indeed, the aim players in a 4–5–1 formation. In contrast, players in
of any team formation is to ensure optimal team all positions within a 4–5–1 generally covered greater
organization so as to best utilize the physical distances at very high intensities when their team was
capacities of players and reduce the efforts required not in ball possession. Again, these discrepancies are
to gain and use ball possession (Doucet, 2007). perhaps linked to the tactical characteristics respec-
However, players in 4–4–2 and 4–3–3 formations tive to the defensive and offensive nature of the three
covered greater distances in jogging while players in a formations (Bangsbo & Peitersen, 2000). These
4–5–1 covered more distance in walking. Combined results would suggest the need for conditioning
with a greater activity in high-intensity running programmes based not only on the individual
(albeit non-significant compared with the 4–5–1), position of players as suggested by Di Salvo and
this finding suggests that players in a 4–4–2 or 4–3–3 Pigozzi (1998), but also on the specific formation
are, to a certain extent, taxed more physically than employed by the team. However, the findings should
those in a 4–5–1 formation. This difference in a 4–5– be interpreted with caution, as the small sample size
1 formation may be due to the increased number of only allowed a basic division into three player
players performing a defensive role and occupying a positions: defender, midfielder, and attacker. Recent
smaller area of play on the field. studies have shown that the physical demands and
The difference observed between formations for bioenergetics across positions vary markedly accord-
very high-intensity running according to ball posses- ing to the precise tactical role of the player (Di Salvo
sion is noteworthy. More distance was covered by et al., 2009; Reilly, 1997). Defenders and midfielders
players in very high-intensity running when their in wide positions (full-backs and wide midfielders)
team was in possession in 4–4–2 and 4–3–3 forma- cover more distance in high-intensity running than
tions compared with a 4–5–1 formation. In contrast, defenders with a central role (Di Salvo et al., 2007).
more very high-intensity running was performed Thus, more research examining the precise physical
when their team was without ball possession in a 4– profiles within different formations of each individual
5–1 formation compared with 4–4–2 and 4–3–3 positional role is necessary, specifically using a
formations. These differences in very high-intensity sample similar to recent large-scale analyses (Bradley
running patterns may reflect the attacking and et al., 2009: Di Salvo et al., 2009).
defensive characteristics inherent to these three The decline in the physical efforts of elite soccer
common playing formations. According to the players during the second half is a consistent finding
 828 P. S. Bradley et al.

(Carling et al., 2008). However, the present results full physical potential during the first 10 min of being
suggest that a decline in total and high-intensity introduced as a substitute compared with the
running distances is not dependent upon the equivalent time frame when starting. This may
formation employed by teams. In contrast, a fall in explain why certain players find it difficult to enter
physical activity was observed in defending and a game, particularly if play is away from their zonal
attacking playing positions within the different area. As a result, coaches should carefully consider
formations. To counteract the effects of fatigue the contribution that a particular player can make
across the team, the identification of players with and should justify introducing players dependent on
the most pronounced falls in work rate is prudent the demands of the game.
(Carling, Espié, Le Gall, Bloomfield, & Jullien, The comparison of several components of techni-
2010; Clarke, Drust, MacLaren, & Reilly, 2008). cal performance between team formations revealed
Therefore, the decline in high-intensity performance minimal differences, suggesting that the technical
between halves reported in attackers in a 4–5–1 demands of match-play are common across forma-
formation is noteworthy. It has been speculated that tions. However, only a basic overview of technical
the 4–5–1 formation places the largest physical performance is provided in the present study, and
burden on attackers, as they are often isolated and future work should categorize passing in relation to
marked by several defenders (Bangsbo & Peitersen, distance or distribution to various areas of the field to
2000; Bauer, 1993). Nevertheless, the greater allow demarcation in the various styles of play within
distance covered in very high-intensity running when the same formation (direct vs. compact vs. counter-
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

out of possession for attackers in a 4–5–1 formation attack play). Nevertheless, despite no statistical
may be attributed to the physical capacity of the differences in overall ball possession, the number of
players and variation in task-specific requirements passes made and received was markedly higher for
for each position (Bangsbo & Peitersen, 2000; Mohr players in a 4–4–2 compared with 4–3–3 and 4–5–1
et al., 2003). Attackers may be expected to pressure formations. Notably, defenders in the 4–4–2 per-
the back line and close down opponents when formed a larger number of passes and other technical
unfavourably outnumbered. The results from the actions than defenders in the other formations. The
present study show similarities to the observations of aforementioned physical efforts identified in 4–4–2
Mohr et al. (2003), who observed a decrease in defensive players combined with these greater tech-
sprinting ability in attackers towards the end of a nical requirements demonstrate the high demands
game compared with that of defenders and mid- placed on these players in elite soccer match-play.
fielders. The resultant decrease in high-intensity Caution is needed when interpreting the findings,
running shown during the second half of games as a relatively small number of games were analysed,
could be related to the aforementioned demands with a limited sample of players within certain playing
eliciting the onset of fatigue. Furthermore, it has positions. This is especially relevant given the high
been speculated that the fitness of attackers is not match-to-match variability in high-intensity running
sufficient to meet the demands of elite European (Gregson, Drust, Atkinson, & Di Salvo, 2010).
leagues (Bradley et al., 2009). In support of this, However, the low number of games was due to the
attackers were shown to perform more poorly in the stringent inclusion and exclusion criteria that con-
game-specific Yo-Yo intermittent recovery and en- trolled for time of day, home advantage, score, and
durance tests (Bradley et al., 2010b; Krustrup et al., standard of opposition. Furthermore, the physical
2003, 2006a). Alternatively, such findings could be and technical performance of players in other team
explained by attackers becoming despondent with formations employed in elite soccer (i.e. 3–5–2 or 4–
their lack of involvement, performing more calcu- 2–3–1 playing formations) was not available, due to
lated decisions that reduce their exertional move- the low number of elite teams adopting these
ment or using pacing strategies (Edwards & Noakes, formations. Finally, this study examined performance
2009). Results of the present study tend to confirm in three common formations during games played
this supposition, as performance was not affected in against opponents who all employed a 4–4–2 system.
attackers when playing in 4–4–2 or 4–3–3. On the However, teams will adapt their playing style during a
basis of this finding, teams employing a 4–5–1 game that is extraneous to formation and actually
formation may want to consider the strategic use of based on factors such as injuries, characteristics of
substitute attacking players during the second half of players, and score line (Bloomfield et al., 2007).
matches. However, coaches must be cautious when Team strategies are influenced by score line in soccer
introducing substitutes in the latter stages of a game. and teams employ different playing styles when ahead,
Carling et al. (2010) found that incoming midfield level, and behind (Lago & Martin, 2007). Higher
players covered a greater total and high-intensity ranked teams have been shown to dominate possession
distance compared with their team-mates playing the over their opponents whether ahead, level or behind
full game. In contrast, attackers failed to utilize their (Bloomfield, Polman, Butterly, & O’Donoghue, 2005)
 Effect of playing formation on soccer performance 829

and, therefore, an investigation into the effects on Carling, C., Bloomfield, J., Nelsen, L., & Reilly, T. (2008). The
physical and technical performance when playing role of motion analysis in elite soccer: Contemporary perfor-
mance measurement techniques and work rate data. Sports
against higher ranked teams or teams who employ Medicine, 338, 839–862.
other common formations as well is merited. Re- Carling, C., Espié, V., Le Gall, F., Bloomfield, J., & Jullien, H.
searchers should also be aware of the season in which (2010). Work-rate of substitutes in elite soccer: A
the data were collected (2006–07), since given the preliminary study. Journal of Science and Medicine in Sport,
13, 253–255.
rapidly changing trends in contemporary soccer the
Clarke, N., Drust, B., Maclaren, D., & Reilly, T. (2008).
data could be considered outdated. Fluid provision and metabolic responses to soccer-specific
In summary, the results suggest that playing exercise. European Journal of Applied Physiology, 104, 1069–
formation does not influence the overall physical 1077.
activity profiles of players. Formation did, however, Di Salvo, V., Baron, R., Tschan, H., Calderon Montero, F. J.,
impact on overall high-intensity running perfor- Bachl, N., & Pigozzi, F. (2007). Performance characteristics
according to playing position in elite soccer. International
mance according to whether teams were with or Journal of Sports Medicine, 28, 222–227.
without ball possession and some technical elements Di Salvo, V., Collins, A., McNeill, B., & Cardinale, M. (2006).
of performance. Distances covered at different Validation of ProZone: A new video-based performance
intensities varied substantially for playing positions analysis system. International Journal of Performance Analysis in
across formations. The current findings provide Sport, 6, 108–119.
Di Salvo, V., Gregson, W., Atkinson, G., Tordoff, P., & Drust,
valuable information for managers and fitness B. (2009). Analysis of high intensity activity in Premier
coaches on physical and technical performance
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

League soccer. International Journal of Sports Medicine, 30,

requirements across different formations and could 205–212.
be of use for adapting training programmes. Di Salvo, V., & Pigozzi, F. (1998). Physical training of football
players based on their positional rules in the team. Journal of
Sports Medicine and Physical Fitness, 38, 294–297.
Doucet, C. (2007). Football: Entraı̂nement tactique. Paris: Editions
References Amphora.
Edwards, A., & Noakes, T. (2009). Dehydration: Cause of fatigue
Andersson, H., Ekblom, B., & Krustrup, P. (2008). Elite football or sign of pacing in elite soccer. Sports Medicine, 39, 1–13.
on artificial turf versus natural grass: Movement pattern, Ekblom, B. (1986). Applied physiology of soccer. Sports Medicine,
technical standard and player opinion. Journal of Sports Sciences, 3, 50–60.
8, 1–10. Gregson, W., Drust, B., Atkinson, G., & Di Salvo, V. (2010).
Bangsbo, J. (1994). The physiology of soccer: With special Match-to-match variability of high-speed activities in premier
reference to intense intermittent exercise. Acta Physiologica league soccer. International Journal of Sports Medicine, 31, 237–
Scandinavica, 151 (suppl. 619), 1–156. 242.
Bangsbo, J., Norregaard, L., & Thorsoe, F. (1991). Activity profile Helgerud, J., Engen, L. C., Wisloff, U., & Hoff, J. (2001). Aerobic
of competition soccer. Canadian Journal of Sport Sciences, 16, endurance training improves soccer performance. Medicine and
110–116. Science in Sports and Exercise, 33, 1925–1931.
Bangsbo, J., & Peitersen, B. (2000). Soccer systems and strategies. Krustrup, P., Mohr, M., Amstrup, T., Rysgaard, T., Johansen, J.,
Champaign, IL: Human Kinetics. Steensberg, A. et al. (2003). The yo-yo intermittent recovery
Bauer, G. (1993). Soccer techniques, tactics and teamwork. New test: Physiological response, reliability, and validity. Medicine
York: Sterling Publishing Co. and Science in Sports and Exercise, 35, 697–705.
Bloomfield, J., Polman, R., Butterly, R., & O’Donoghue, P. Krustrup, P., Mohr, M., Ellingsgaard, H., & Bangsbo, J. (2005).
(2005). Analysis of age, stature, body mass, BMI and quality of Physical demands of elite female soccer games: Importance of
elite soccer players from 4 European leagues. Journal of Sports training status. Medicine and Science in Sports and Exercise, 37,
Medicine and Physical Fitness, 45, 58–67. 1242–1248.
Bloomfield, J., Polman, R., & O’Donoghue, P. (2007). Physical Krustrup, P., Mohr, M., Nybo, L., Jensen, J.M., Nielsen, J.J., &
demands of different positions in FA Premier League soccer. Bangsbo, J. (2006a). The Yo-Yo IR2 test: Physiological
Journal of Sports Science and Medicine, 6, 63–70. response, reliability and application to elite soccer. Medicine
Bradley, P., Di Mascio, M., Peart, D., Olsen, P., & Sheldon, B. and Science in Sports and Exercise, 38, 1666–1673.
(2010a). High-intensity activity profiles of elite soccer players at Krustrup, P., Mohr, M., Steensberg, A., Bencke, J., Kjær, M., &
different performance levels. Journal of Strength and Conditioning Bangsbo, J. (2006b). Muscle and blood metabolites during a
Research, 24, 2343–2351. soccer game: Implications for sprint performance. Medicine and
Bradley, P. S., Mohr, M., Bendiksen, M., Randers, M. B., Flindt, Science in Sports and Exercise, 38, 1165–1172.
M., Barnes, C. et al. (2010b). Sub-maximal and maximal Yo- Lago, C., & Martin, R. (2007). Determinants of possession of the
Yo intermittent endurance test level 2: Heart rate response, ball in soccer. Journal of Sports Sciences, 25, 969–974.
reproducibility and application to elite soccer. European Journal Mohr, M., Krustrup, P., Andersson, H., Kirkendall, D., &
of Applied Physiology, 28 October, Epub ahead of print. Bangsbo, J. (2008). Match activities of elite women soccer
Bradley, P. S., O’Donoghue, P., Wooster, B., & Tordoff, P. players at different performance levels. Journal of Strength and
(2007). The reliability of Prozone MatchViewer: A video-based Conditioning Research, 22, 341–349.
technical performance analysis system. International Journal of Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match perfor-
Performance Analysis in Sport, 7, 117–129. mance of high standard soccer players with special reference
Bradley, P. S., Sheldon, W., Wooster, B., Olsen, P. D., Boanas, to development of fatigue. Journal of Sports Sciences, 21, 519–
P., & Krustrup, P. (2009). High-intensity running in English 528.
FA Premier League soccer matches. Journal of Sports Sciences, Mohr, M., Krustrup, P., & Bangsbo, J. (2005). Fatigue in soccer:
27, 159–168. A brief review. Journal of Sports Sciences, 23, 593–599.
 830 P. S. Bradley et al.

Rampinini, E., Impellizzeri, F. M., Castagna, C., Coutts, A. J., & Reilly, T., & Brooks, G. A. (1986). Exercise and the circadian
Wisloff, U. (2008). Technical performance during soccer variation in body temperature measures. International Journal of
matches of the Italian Serie A league: Effect of fatigue and Sports Medicine, 7, 358–368.
competitive level. Journal of Science and Medicine in Sport, 12, Reilly, T., Drust, B., & Clarke, N. (2008). Muscle fatigue during
227–233. football match-play. Sports Medicine, 38, 357–367.
Reilly, T. (1997). Energetics of high-intensity exercise (soccer) Reilly, T., & Thomas, V. (1976). A motion analysis of work-rate in
with particular reference to fatigue. Journal of Sports Sciences, 15, different positional roles in professional football match play.
257–263. Journal of Human Movement Studies, 2, 87–89.
Downloaded By: [University of Sunderland] At: 07:31 5 May 2011

View publication stats