See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/261800968

Evaluation of the Match Performances of

Substitution Players in Elite Soccer

Article in International journal of sports physiology and performance May 2014

DOI: 10.1123/IJSPP.2013-0304 Source: PubMed

CITATIONS READS

9 592

3 authors:

Paul S Bradley Carlos Lago Peas

Liverpool John Moores University University of Vigo
59 PUBLICATIONS 1,207 CITATIONS 97 PUBLICATIONS 1,794 CITATIONS

SEE PROFILE SEE PROFILE

Ezequiel Rey
University of Vigo
37 PUBLICATIONS 384 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Exercise-Related Sensations Contribute to Decrease Power Output during Repeated Cycle Sprints with
Limited Influence on Neural Drive View project

Are Current Physical Match Performance Metrics in Elite Football Fit for Purpose or is the Adoption of an
Integrated Approach Needed: Moving Towards a Paradigm Shift or Simply Tinkering with Tradition? View
project

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

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
International Journal of Sports Physiology and Performance, 2014, 9, 415-424
http://dx.doi.org/10.1123/IJSPP.2013-0304
2014 Human Kinetics, Inc.
www.IJSPP-Journal.com
ORIGINAL INVESTIGATION

Evaluation of the Match Performances

of Substitution Players in Elite Soccer
Paul S. Bradley, Carlos Lago-Peas, and Ezequiel Rey

Purpose: To evaluate match performances of substitute players using different research designs. Methods:
English Premier League matches were analyzed using a multiple-camera system. Two research designs were
adopted: an independent-measures analysis comparing the match-performance characteristics of players
completing the entire match (n = 810) vs substitutes (n = 286) and the players they replaced (n = 286) and a
repeated-measures analysis comparing the same players completing full matches vs those in which they were
introduced as a substitute (n = 94). Results: Most substitutions (P < .05) occurred at halftime and between
the 60- to 85-min vs all first-half periods and the remaining second-half periods (effect size [ES]: 0.851.21).
These substitutions become more (P < .01) offensive (eg, more attacking positions were introduced) in relation
to the positions introduced as the half progressed (ES: 0.931.37). Independent-measures analysis indicated
that high-intensity running was greater (P < .01) in substitutes compared with players who either completed
the entire match or were replaced (ES: 0.280.67), but no differences were evident for pass-completion rates
(ES: 0.010.02). Repeated-measures analysis highlighted that players covered more (P < .01) high-intensity
running when they were introduced as substitutes compared with the equivalent period of the second- but not
the first-half period (ES: 0.210.47). Both research designs indicated that attackers covered more (P < .05)
high-intensity running than peers or their own performances when completing the entire match (ES: 0.45
0.71). Conclusions: Substitutes cover greater high-intensity-running distance; this was particularly evident
in attackers, but pass-completion rates did not differ for any position. This information could be beneficial to
coaches regarding optimizing the match running performances of their players, but much more work needs to
be undertaken to investigate the overall impact of substitutes (physical, technical indicators, and contribution
to key moments of matches).

Keywords: positions, high-intensity running, passing

Research demonstrates that physical performance
declines from the first to the second half of elite soccer
match play,1,2 but some observe minimal differences.3,4
Reductions in match running performance in the second
half or temporarily after the most intense period could be
attributed to fatigue,59 pacing strategies,1013 or contex-
tual variables14 or be related to the time the ball is out of
play and the opportunity to engage in match activities.15
Although each factor has the potential to affect the physi-
cal performance of elite players, match-induced fatigue
seems evident as physical capacity markedly declines
after matches in comparison with baseline measures.1618
However, those studies were not conducted using elite
players, so caution is warranted. Thus, fatigue could result
in an inability to repeatedly cover distances during critical
Bradley is with the Dept of Sport and Exercise Sciences, Uni-
versity of Sunderland, Sunderland, UK. Lago-Peas and Rey
are with the Faculty of Education and Sports Sciences, Univer-
sity of Vigo, Pontevedra, Spain. Correspondence concerning
this article should be addressed to Paul S. Bradley. E-mail:
paul.s.bradley@sunderland.ac.uk
situations and may also reduce technical capabilities
that are related to match outcome,19 while others have
not found a relationship between physical fatigue and
declines in skill-related performance.15
Coaches typically attempt to use well-timed substi-
tutions to reduce fatigue across the team or in an effort to
modify tactics.20 However, despite the perceived benefit
of such practices, limited research has been published to
support such assertions. For this practice to be deemed
advantageous from a work-rate perspective, substitutes
would cover equitable or ideally greater distance in
high-intensity running than the replaced players or
other team members during the equivalent time period.
Research has reported such trends21 but generally used
either independent measures, small samples, or lim-
ited differentiation of player position across different
time periods. Moreover, the studies did not investigate
whether and to what extent technical indicators differ
between substitute players and those completing the
full match. This is especially important as soccer perfor-
mance is a consequence of technical skills, and passing
is considered an important indicator.19,22 There is also
scant evidence on the timing and type of substitutions

415
416Bradley, Lago-Peas, and Rey
made in matches, and it would be of interest to quantify
this to determine if substitutions are mainly made for
physical or tactical reasons.
Research using elite soccer populations highlights
substantial match-related differences between players
when analyzed according to 5 positional subsets,1,3,2327
although most researchers have not investigated match-
performance characteristics of substitutes in the 5 most
common positions. Studies have either collapsed all posi-
tions together for analysis or have only measured match
performance of 2 positions (midfielders and attackers).9,21
Thus, quantifying physical and technical indicators
of substitutes in a variety of positions across multiple
time periods could provide coaches with additional
information on the effectiveness of position-specific
introductions.
Thus, the aims of this study were (1) to evaluate the
match performances of a large sample of substitute play-
ers versus those replaced or completing the entire match
using independent-measures analysis, (2) to compare
players introduced as substitutes compared with previ-
ous performances across identical time periods of the
first and second half when completing a full match using
repeated-measures analysis, and (3) to quantify physical
and technical indicators across various playing positions.
Methods
Players
Player performances in official competition were
analyzed from English FA Premier League matches.
Match-performance characteristics of players complet-
ing the entire match were compared against substitute
performances using different research designs. Ethical
approval was obtained from the University of Sunder-
land, and Prozone Sports Ltd granted permission to
publish. To ensure player and team confidentiality, all
data were anonymized. We, the authors, have no con-
flicts of interest that are directly relevant to the content
of this article.
Study Design
English FA Premier League matches were analyzed
using a multiple-camera computerized tracking system
(Prozone Sports Ltd, Leeds, UK). Players activities
were captured during matches by cameras positioned at
roof level and analyzed using proprietary software. The
systems reliability and validity have been quantified
to verify the capture process and data accuracy.28,29 To
evaluate substitutions in elite soccer, different approaches
were adopted:
An independent-measures analysis whereby match-
performance characteristics of players completing
the entire match (n = 810) were compared with
substitutes (n = 286) and the players they replaced
(n = 286) (eg, substitute: 15 min vs replaced: 75 min
vs entire match: 90 min).
A repeated-measures analysis comparing players
match performances when completing full matches
versus those in which they were introduced as sub-
stitutes (n = 94). Substitution and full-match com-
parisons were corrected for the equivalent period of
the match (eg, substitute: 15 min vs final 15 min of
second half vs first 15 min of first half).
Due to the large sample used in each section of the
study, the data were separated into 5 playing positions
(attackers, central defenders, fullbacks, and central and
wide midfielders). Please note that the data from play-
ers completing the entire match (n = 810) have been
previously published,22 but all other data have not been
reported previously.
Match-Performance Variables
Players activities were coded into the following catego-
ries and speed thresholds: standing (00.6 km/h), walk-
ing (0.77.1 km/h), jogging (7.214.3 km/h), running
(14.419.7 km/h), high-speed running (19.825.1 km/h),
and sprinting (>25.1 km/h). The speeds for each category
have been employed previously.2 Absolute distances (m)
were converted to relative distance covered per unit of
time (m/min). Total distance represented the summation
of distances in all categories. High-intensity running
consisted of the combined distance in high-speed running
and sprinting (19.8 km/h) and subsequently separated
into 2 subsets based on the team possession status (with
or without the ball). Maximal running speed was defined
as the peak speed reached during a match. In addition,
pass-completion rate was included in the independent-
measures analysis.
Statistical Analysis
All analyses were conducted using statistical soft-
ware (SPSS Inc., V20, Chicago, IL, USA). One-way
independent-measures analysis of variance (ANOVA)
tests were used to compare players completing the
entire match versus substitute and replaced players.
One-way repeated-measures ANOVAs were also per-
formed to examine differences in match-performance
characteristics for players in the first and second half
of full matches versus the equivalent time period when
introduced as substitutes. In addition, 2-way ANOVAs
were used to explore differences between substitutes
and examine the interaction between measures of
performance across playing positions. Bonferroni-
corrected comparisons were employed to verify local-
ized effects. Finally, differences between 5-minute
periods for substitute players were determined using
1-way ANOVAs with Tukey post hoc tests used to
verify isolated trends. Effect size (ES) was quantified
to indicate the meaningfulness of the difference.30 The
ES was classified as trivial (<0.2), small (>0.20.6),
moderate (>0.61.2), large (>1.22.0), and very large
(>2.04.0) based on guidelines from Batterham and
Hopkins.31 Alpha was set at P < .05.

Results
Time and Type of Substitution
Match data separated into 5-minute periods indicated
that more (P < .05) substitutions occurred at halftime
and between the 60- to 85-minute marks versus all first-
half periods and the remaining second-half periods (ES:
0.851.21; Figure 1). More (P < .01) offensive substitu-
tions involving attackers and wide and central midfielders
were made between the 60- to 90-minute marks compared
with defensive substitutes such as central defenders and
fullbacks (ES: 0.931.37).
Independent-Measures Analysis
Table 1 displays match-performance characteristics
of players using an independent-measures analysis.
Substitutes covered greater (P < .05) total distance
(ES: 0.330.67), jogging (ES: 0.230.41), running
(ES: 0.410.77), high-speed running (ES: 0.290.74),
sprinting (ES: 0.210.49), and high-intensity running
(ES: 0.280.67) than players who either completed the
entire match or were replaced. Maximal running speeds
were higher (P < .01) in players completing the entire
match and those replaced compared with substitutes (ES:
035068). No differences were found between players
for pass-completion rates (ES: 0.010.02). No differences
were observed for the total and high-intensity running dis-
tances covered by substitutes across 5-minute periods of
matches (0.230.36; Figure 2), but there was a tendency
for this to increase as the match progressed.
Figure 1 Time periods when substitutes were introduced into the match (independent-measures analysis only; n = 286).
Player Substitution and Match Performance 417
Repeated-Measures Analysis
Table 2 displays match-performance characteristics of
players using a repeated-measures analysis. Players
covered greater (P < .05) total distance (ES: 0.130.82),
jogging (ES: 0.250.74), running (ES: 0.030.59),
high-speed running (ES: 0.160.46), and high-intensity
running (ES: 0.210.47) when they were introduced as
substitutes compared with the equivalent period of the
second-half but not the first-half period when tracked
from the start of the match. Similar maximal running
speeds were observed for the same players across the first
and second halves of full matches (ES: 0.01).
Positional Variation
Using an independent-measures analysis, substitutes
in all positions except fullback covered more (P <
.05) high-intensity running than players who either
completed the entire match or were replaced (ES: 0.33
0.87), while substitute fullbacks and central midfielders
sprinted more (P < .05) than players in the same position
completing the entire match (ES: 0.420.43; Table 1).
Although no differences were observed between posi-
tions for pass completions, the maximal running speeds
attained during the entire match were (P < .05) greater
in all playing positions compared with substitutes (ES:
0.750.99).
Repeated-measures analysis indicated that central
defenders, attackers, and central and wide midfielders
covered more (P < .05) total distance when they were
introduced as substitutes compared with the equivalent
 Table 1 Independent-Measures Analysis of Match Performance of Substitute Players Versus Those Who Completed the Entire Match
or Were Replaced, Mean SD

418
Independent of Position Central Defenders Fullbacks
Entire match Replaced Substitute Entire match Replaced Substitute Entire match Replaced Substitute
Indicator (n = 810) (n = 286) (n = 286) (n = 208) (n = 26) (n = 20) (n = 177) (n = 26) (n = 26)
Distances covered (m/min)
total 112.3 10.3* 116.2 10.6# 120.1 14.5 102.8 5.8 103.0 5.6 108.2 13.7 112.1 6.6 113.5 7.2 115.7 8.9
standing 0.3 0.1 0.3 0.1 0.2 0.1 0.3 0.1 0.3 0.1 0.3 0.2 0.3 0.1 0.3 0.1 0.3 0.1
walking 39.6 3.1# 39.3 3.4# 37.9 4.1 40.6 2.3# 39.8 3.6 38.1 2.9 39.3 2.3 39.3 2.7 37.9 3.3
jogging 44.6 6.7# 45.7 7.6# 47.7 9.3 41.5 4.2 40.8 3.6 43.8 8.0 43.7 4.5 44.9 4.7 46.2 6.2
running 17.9 4.5* 19.5 4.6# 21.8 6.4 13.8 2.3#+ 15.1 2.5 17.8 7.1 18.1 2.7 18.5 2.7 20.5 3.8
high-speed running 7.1 2.3* 8.2 2.3# 9.1 3.6 4.8 1.2# 5.2 1.2 6.3 2.5 7.5 1.8 7.4 1.8 7.3 2.9
sprinting 2.6 1.3#+ 3.0 1.3 3.4 2.3 1.6 0.7 1.7 1.1 1.8 1.0 2.9 1.1 3.0 1.1 3.4 1.6
HIR 9.8 3.2* 11.3 3.2# 12.4 5.3 6.5 1.6# 6.9 2.2 8.1 3.4 10.5 2.4 10.5 2.4 10.7 3.3
HIR with possession of the ball 4.1 2.7* 5.6 2.6 6.1 3.6 1.1 0.6 1.2 0.6 1.4 1.6 4.1 1.7 3.8 1.5 4.2 2.5
HIR without possession of the ball 5.6 1.9# 5.7 2.9 6.2 3.4 5.3 1.3# 5.7 1.9 6.7 2.4 6.3 1.5 6.6 1.6 6.5 2.1
Maximal running speed (m/s) 9.0 0.3# 8.9 0.3# 8.6 0.6 8.9 0.3* 8.6 0.4 8.4 0.7 9.1 0.3# 8.9 0.3 8.8 0.3
Successful passes (%) 77.5 11.9 76.9 12.2 77.3 23.7 79.0 13.4 73.3 17.6 73.3 26.3 74.3 12.2 75.2 10.7 71.6 21.9
Central Midfielders Wide Midfielders Attackers
Entire match Replaced Substitute Entire match Replaced Substitute Entire match Replaced Substitute
(n = 188) (n = 68) (n = 66) (n = 108) (n = 79) (n = 86) (n = 129) (n = 87) (n = 88)
Distances covered (m/min)
total 120.0 6.9# 121.1 7.7 125.3 12.3 121.5 8.4 120.8 8.8# 125.2 10.9 108.2 10.5* 113.0 9.4 115.5 15.3
standing 0.2 0.1 0.2 0.1 0.2 0.1 0.2 0.1 0.2 0.1 0.2 0.1 0.3 0.1 0.3 0.1 0.3 0.1
walking 38.1 2.5# 37.2 2.9 36.4 3.9 38.2 3.3 38.7 3.4 37.1 3.6 42.1 2.8# 41.3 3.1# 39.7 4.6
jogging 50.2 5.2# 51.1 6.2 51.9 7.7 48.7 6.3 47.5 6.4# 50.8 7.8 38.7 7.7* 41.8 7.1 42.8 9.3
running 21.4 3.5# 22.1 4.1# 24.7 6.0 21.5 4.2 20.9 4.4# 23.2 5.0 16.1 4.2#+ 18.0 4.3 19.8 6.8
high-speed running 7.7 1.9# 8.1 2.2 9.1 3.4 9.3 1.9 9.5 2.1 10.1 2.7 7.6 1.9# 8.2 1.8 9.2 4.3
sprinting 2.3 0.9# 2.3 1.0 2.8 1.7 3.5 1.2 3.9 1.1 3.8 2.1 3.2 1.2 3.2 1.1 3.7 2.9
HIR 9.9 2.7# 10.4 3.0 11.9 4.4 12.7 2.7 13.4 2.7 13.9 4.4 10.8 2.7# 11.6 2.4 12.9 6.5
HIR with possession of the ball 3.8 1.9 4.2 1.9 4.7 3.1 6.7 2.1 7.4 2.1 7.5 2.9 6.8 1.8 7.1 1.8 7.4 3.6
HIR without possession of the ball 6.1 2.0# 6.2 2.1 7.1 3.4 6.0 1.9 5.9 1.9 6.3 2.8 4.1 1.6# 4.4 1.6 5.5 4.1
Maximal running speed (m/s) 8.9 0.3# 8.9 0.3 8.4 0.6 9.1 0.3# 9.1 0.3# 8.6 0.6 9.1 0.3# 9.1 0.3# 8.7 0.6
Successful passes (%) 81.2 9.2 79.4 10.3 76.7 25.8 77.5 11.3 77.3 11.1 79.3 20.3 73.9 12.4 76.1 12.7 77.0 25.9
Abbreviations: HIR, high-intensity running.
*Different from replaced and substitute players, P < .01. +Different from replaced players, P < .05. Different from substitute players, P < .05. #Different from substitute players, P < .01.
 Player Substitution and Match Performance 419

Figure 2 (a) Total distance covered and (b) high-intensity-running distance of substitute players in 5-minute periods of match
play in relation to the minute in which substitutions occurred. Broken lines denote mean values over an entire match.

period of the second-half but not the first-half period
when tracked from the start of the match (ES: 0.681.18).
Attackers covered more (P < .01) high-intensity-running
distance (ES: 0.71) than the equivalent period of the
second-half but not the first-half period when tracked
from the start of the match.
Discussion
The current study revealed that (1) most substitutions
occurred at halftime and the latter part of the second
half, with substitutions becoming more offensive as
the half progressed (eg, more attacking positions were
introduced); (2) independent-measures analysis indicates
that substitutes covered greater high-intensity-running
distance than players who either completed the entire
match or were replaced, but no differences were evident
for pass-completion rate; (3) repeated-measures analysis
highlighted that the same players covered more high-
intensity running when they were introduced as substitutes
compared with the equivalent period of the second-half
but not the first half-period when tracked from the start of
the match; and (4) in both types of analysis high-intensity
running was greater for attacking substitutes.
Although preliminary research has been con-
ducted on the physical performance of substitutions in
 Table 2 Repeated-Measures Analysis of Match Performance of Players Introduced as Substitutes Compared With the Equivalent Period

420
of the Second Half and the First Half When Tracked From the Start of the Match, Mean SD
Independent of position (n = 94) Central Defenders (n = 11) Fullbacks (n = 12)
Indicators First half Second half Substitute First half Second half Substitute First half Second half Substitute
Distances covered (m/min)
total 119.8 12.6 105.9 16.2* 118.1 13.6 110.2 7.2 98.1 7.3+ 110.4 13.9 117.5 10.3 109.5 17.6 121.1 14.2
standing 0.2 0.1 0.4 0.2* 0.3 0.1 0.2 0.1 0.5 0.2 0.3 0.2 0.3 0.1 0.4 0.2 0.3 0.1
walking 39.1 4.4 38.6 4.6 38.4 3.9 39.9 1.8 35.6 5.2 36.5 4.5 38.2 2.4 39.8 4.0 36.9 3.3
jogging 48.4 8.5 39.6 9.1* 46.2 8.7 47.8 6.6 38.8 4.1+ 47.5 9.2 46.3 6.4 41.1 9.8 48.2 6.6
running 20.7 6.0 17.5 5.6* 20.9 5.9 15.8 3.4 16.2 6.6 18.9 6.8 21.2 3.7 17.5 4.7 22.0 6.3
high-speed running 8.3 2.9 7.3 3.2# 8.8 3.2 5.0 1.6 5.8 2.8 5.6 2.1 8.4 2.9 7.6 3.3 9.3 4.2
sprinting 3.0 1.9 2.7 1.6 3.4 2.0 1.2 0.9 1.1 0.9 1.7 1.1 3.1 2.2 3.0 1.7 4.3 1.9
HIR 11.3 3.9 10.1 4.1# 12.2 4.7 6.3 2.3 6.9 3.0 7.4 2.7 11.6 3.9 10.7 3.9 13.6 5.9
HIR with possession of the ball 5.7 3.4 4.8 3.1# 6.1 3.3 1.1 0.7 1.4 1.1 1.3 1.1 5.7 3.1 4.5 2.8 6.2 4.0
HIR without possession of the ball 5.7 2.8 5.3 2.8 6.0 2.8 5.2 2.5 5.4 2.2 6.1 2.4 5.7 1.5 6.2 2.1 7.4 2.8
Maximal running speed (m/s) 8.5 0.5 8.5 0.6 8.6 0.6 8.3 0.5 8.1 0.8 8.2 0.7 8.6 0.3 8.8 0.4 8.8 0.3
Central Midfielders (n = 18) Wide Midfielders (n = 22) Attackers (n = 31)
First half Second half Substitute First half Second half Substitute First half Second half Substitute
Distances covered (m/min)
total 126.8 9.0 115.4 10.2+ 124.5 12.0 127.9 10.1 114.8 11.3+ 122.2 10.2 114.5 13.4 95.3 16.6* 113.0 13.2
standing 0.2 0.1 0.3 0.2 0.2 0.1 0.2 0.1 0.3 0.1 0.2 0.1 0.3 0.2 0.4 0.2 0.3 0.1
walking 38.7 1.9 37.6 3.1 36.2 2.7 37.2 3.5 38.8 2.8 38.1 3.1 40.7 6.3 39.7 5.7 41.0 3.7
jogging 52.9 5.1 47.4 5.8 51.4 7.2 52.6 7.9 43.1 5.9 47.5 5.9 43.9 9.4 32.3 8.1* 40.9 9.5
running 24.5 5.3 20.4 5.3 24.1 6.1 24.2 4.5 19.8 3.5 21.6 4.4 17.6 6.1 13.5 5.1+ 19.1 5.7
high-speed running 8.4 2.7 7.4 3.0 9.7 3.4 10.2 2.4 9.1 4.1 10.2 2.4 7.9 2.6 6.5 2.3 8.3 2.7
sprinting 2.1 1.2 2.1 1.7 2.7 1.5 3.4 1.5 3.7 1.8 4.4 2.3 3.9 2.1 2.8 1.1 3.2 1.9
HIR 10.5 3.6 9.6 4.2 12.5 4.6 13.5 3.4 12.8 4.7 14.6 4.3 11.9 3.5 9.3 2.9+ 11.6 3.6
HIR with possession of the ball 4.4 2.3 3.5 2.4 5.4 3.4 7.6 2.1 7.3 2.6 7.8 2.4 6.9 3.8 5.2 2.7 7.1 2.2
HIR without possession of the ball 6.1 2.4 6.1 2.9 7.1 2.2 6.2 2.5 5.4 3.1 6.7 3.1 5.4 3.5 4.3 2.9 4.5 2.6
Maximal running speed (m/s) 8.4 0.5 8.3 0.7 8.5 0.5 8.6 0.6 8.6 0.4 8.7 0.5 8.6 0.5 8.6 0.4 8.6 0.6
Abbreviations: HIR, high-intensity running.
*Different from first half and substitute performance, P < .01. #Different from substitute performance, P < .01. +Different from first half and substitute performance, P < .05. Different from substitute
performance, P < .05. Different from first half, P < .05.

soccer,9,10,21 this is the first study to quantify the time
and type of substitutions made in matches, and our data
highlight some noteworthy trends. In the current study,
most coaches introduced substitutes at halftime and in
the latter stages of the second half, with these introduc-
tions becoming more offensive as the half progressed.
This links with the general premise that substitutions are
made to replace fatigued or injured players in addition to
those producing suboptimal performances.20 Additional
factors influencing substitution decisions include the
score line, experience for nonestablished players, and to
create match interruptions for tactical benefit.32 Due to
the type of players introduced into the latter stages of the
match, it seems that tactical factors are the primary reason
for substitutions. For instance, more offensive players
introduced later in the match would possibly enhance
the teams chances of creating scoring opportunities. It
is noteworthy that alongside attackers, other positions
such as central and wide midfielders were introduced
regularly during last 30 minutes of the match. The latter
2 positions are arguably the most physically demanding,
and this change could also be due to the manifestation of
fatigue in players starting the match,1,3,2325,27,33 but the
lack of information pertaining to the reasoning behind
each substitution indicates that additional research
employing such an approach is warranted. However, the
reader must also be aware that although offensive posi-
tions were more commonly introduced as substitutes in
the later stages of matches, it is important to clarify that
this type of introduction is not always made with offensive
strategies in mind. For instance, these offensive players
may have been introduced as part of a defensive strategy
to hold a score line or even to waste time.
Coaches use substitutions to reduce fatigue across the
team or in an effort to modify tactics.20 For this practice
to be deemed advantageous from a work-rate perspective,
substitutes would cover equitable or ideally greater high-
intensity-running distance than the replaced players or
other team members during the equivalent time period.
This is especially important as high-intensity running
could be a useful indicator of physical performance in
soccer.16,34 Research demonstrates that high-intensity
running was 25% greater during the final 15 minutes of
matches in substitutes versus players starting the match.9
Carling et al21 observed that substitute midfielders cov-
ered 34% more high-intensity running in the second half
than the midfielders they replaced. These findings are in
agreement with the current study, as players entering the
match as substitutes covered 10% or 27% more high-
intensity running than those replaced or completing the
entire match, respectively. The reader should be aware
that previous research used small sample sizes with lim-
ited differentiation of position, thus making comparisons
across studies difficult. To comprehensively examine
the match running performances of substitutes, a large
sample of players representing all playing positions
is needed, given the considerable variability observed
from match to match.27 Thus, our data seem to provide
a more valid expression of full- versus partial-match
Player Substitution and Match Performance 421
running-performance comparisons across all playing
positions than achieved in previous studies, although the
current study does have some limitations.
The independent-measures analysis performed in
this study examined the match performances of sub-
stitutes compared with those replaced or completing
the entire match, and it might be that a more sensitive
repeated-measures analysis would improve our under-
standing of the work rate of substitutes versus their
habitual performances when completing an entire match.
Although we corrected the distances covered per minute
of the match, it could be argued that the trends observed
could be due to the differences in the time periods in
which match running performance was monitored in each
sample (eg, substitute: 15 min vs replaced: 75 min vs
entire match: 90 min). Although the data are interesting
from a comparison perspective, more research should be
conducted that compares the substitutes who remained
on the pitch over the same time period and over the 20
minutes before substitution in the player replaced. This
was not possible in the independent-measures analysis, as
data were only available in 45-minute periods. Thus, the
research design and the subsequent interpretation of the
data may be influenced by pacing strategies, the tempo,
and the intensity of selected periods of the match10,35,36
and could indicate that comparisons using independent
analyses are not robust. Our repeated-measures analysis
improved the analytical approach to this area and high-
lighted that players covered 21% more high-intensity
running when they were introduced as substitutes
compared with the equivalent period of the second-half
but not the first-half period when tracked from the start
of the match. Carling et al21 found that high-intensity
running in midfielders during their first 10 minutes as
substitutes was similar to their running performance
during the opening 10 minutes when starting matches.
This finding is noteworthy, as substitute players seem to
pace their efforts in a similar fashion to the opening of a
match, despite their teammates and opponents reduc-
ing their physical performances during the same period
of the match. These data support the notion that players
completing full matches either lower their match running
performances due to fatigue in the second half or use a
more effective pacing strategy compared with being intro-
duced as substitutes.1013 However, we did not account
for contextual variables14 or the time the ball is out of
play and the opportunity to engage in match activities,15
which could be confounding variables.
The large differences observed between various
playing positions for the energetic and physical-perfor-
mance characteristics of elite players is one of the most
robust findings in timemotion analysis studies.1,3,2326,33
Given the higher work rate and inferior physical capac-
ity of selected positions,4,37 an insight into the physi-
cal performance of substitute players in offensive and
defensive positions would definitely contribute to the
existing knowledge base in this area. This approach is
particularly novel, as previous studies have failed to dif-
ferentiate substitutes into the 5 most common positions
422Bradley, Lago-Peas, and Rey
and either collapsed all positions together for analysis
or only measured match performance of midfielders and
attackers.9,21 We found in both the independent- and the
repeated-measures analyses that more total distance was
covered by substitutes in most positions versus peers
completing the entire match and the same players for
the equivalent second- but not first-half period, although
total distance is not considered an important indicator
of physical performance compared with high-intensity
running, given that it includes nonessential low-intensity
activities such as walking and jogging.
Attacking substitutes were the only position to
consistently cover more high-intensity running in the
independent and repeated measures analysis. Carling et
al21 demarcated between midfielders and attackers but
observed that only the latter failed to improve their match
running performance during the first 10 minutes of being
introduced as substitutes compared with the equivalent
full time period. In contrast, the current study found this
trend for attackers, thus further supporting the assertion
that coaches replacing players still need to consider the
contribution that a particular substitute can make based
on situational, technical, and positional factors. Bradley
et al2 observed that English Premier League attackers per-
formed less high-intensity running than their counterparts
in the Spanish and Italian top tiers. It could be that the
fitness levels of attackers are not sufficient to meet the
demands of elite European leagues, and this point has also
been supported by low physical capacity in this specific
position.37 Thus, given the general trend for attackers
found in both independent- and repeated-measures analy-
ses, the use of timing substitutions of attacking players
seems to be beneficial from a work-rate perspective and
could be optimized by introducing the players later in the
second half, especially given that substitutes have lower
measures of leg power, intermittent-exercise capacity,
and sprint speed than players who start matches.38 It
must be acknowledged that this study was conducted on
Australian Football players and not soccer players, so
the reader should interpret with caution. The latter was
evidenced in the independent-measures analysis of the
current study, as substitutes had lower maximal running
speeds than those completing the entire match. Given the
complexities of soccer, the match-performance trends
of the players in the current study could also reflect the
actions of the opposition and teammates. Thus, the lower
work rate in the second-half period for players completing
the entire match could be related to fatigue and the fact
that the higher intensity from substitutes may provide a
tactical advantage that would be greater with a delayed
substitution. However, to fully support or refute such
an assertion, future research should possibly examine
not just the relative high-intensity-running distances
but also the number of efforts in this category for sub-
stitutes and their teammates. This would enable one to
conclude whether players completing the entire match
are having an impact on substitutes. It could be possible
that substitutes are covering more relative distance late
in the game, but these attacking runs are less effective
as the fatigued players on their team cannot respond to
their runs (eg, substitute attacker makes a surging break
but teammates are too fatigued to respond). If this is the
case, then it might be more prudent to substitute earlier
rather than later. Moreover, the primary reason behind
substitutions is for tactical considerations,20 and although
we found no differences between pass-completion rates
of substitutes versus players completing the entire match
or those they replaced, it seems that the true impact of
substitutions cannot be determined in the current study.
Soccer performance is a consequence of technical skills,
and passing is considered an important indicator.22
However, the impact a substitute has on a match could
be very specific based on the tactical philosophy of the
coach, contextual factors, and obviously the position of
the player. For instance, an attacking substitute introduced
in the latter stages of a match may display worse match
running performance and pass-completion rates than
the player replaced, but he could score goals or create
numerous opportunities for the team and thus this would
be classified as a successful introduction. Moreover, our
research group has recently published data investigating
substitutes work-rate profiles in relation to fatigue and/
or pacing using a small sample and found some conflict-
ing data trends compared with the current study.10 Thus,
more research is definitely needed, particularly to fully
quantify the overall impact of substitutions.
Practical Applications
Coaches replacing players still need to consider the
contribution that a particular substitute can make based
on situational, technical, and positional factors. It seems
that, from a work-rate perspective, substitution seems to
be effective, but from a technical and outcome perspective
this still needs to be established. This information could
be beneficial to coaches regarding optimizing the match
running performances of their players.
Conclusions
The data demonstrate that most substitutions occur at
halftime, with substitutions becoming more offensive as
the half progresses. Independent-measures analysis indi-
cated that high-intensity running was greater in substitutes
compared with players who either completed the entire
match or were replaced, but no differences were evident
for pass-completion rates. Repeated-measures analysis
highlighted that players covered more high-intensity run-
ning when they were introduced as substitutes compared
with the equivalent period of the second-half but not the
first-half period. Position-specific trends for both types
of analysis indicated that attackers covered more high-
intensity running than peers or their own performances
when completing the entire match. Although this informa-
tion could be important for coaches, more work should
still be undertaken on the overall impact of substitutions
from a technical and physical perspective.

References
1. Di Salvo V, Gregson W, Atkinson G, Tordoff P, Drust
B. Analysis of high intensity activity in Premier League
soccer. Int J Sports Med. 2009;30:205212. PubMed
doi:10.1055/s-0028-1105950
2. Bradley PS, Sheldon W, Wooster B, Olsen P, Boanas P,
Krustrup P. High-intensity running in English FA Premier
League soccer matches. J Sports Sci. 2009;27:159168.
PubMed doi:10.1080/02640410802512775
3. Di Salvo V, Baron R, Tschan H, Calderon Montero FJ,
Bachl N, Pigozzi F. Performance characteristics accord-
ing to playing position in elite soccer. Int J Sports Med.
2007;28:222227. PubMed doi:10.1055/s-2006-924294
4. Bradley PS, Carling C, Gomez Diaz A, et al. Match
performance and physical capacity of players in the top
three competitive standards of English professional soccer.
Hum Mov Sci. 2031;32(4):808821.
5. Bradley PS, Di Mascio M, Peart D, Olsen P, Shel-
don B. High-intensity activity profiles of elite soccer
players at different performance levels. J Strength
Cond Res. 2010;24:23432351. PubMed doi:10.1519/
JSC.0b013e3181aeb1b3
6. Bendiksen M, Bischoff R, Randers MB, et al. The Copen-
hagen Soccer Test: physiological response and fatigue
development. Med Sci Sports Exerc. 2012;44:15951603.
PubMed doi:10.1249/MSS.0b013e31824cc23b
7. Di Mascio M, Bradley PS. Evaluation of the most intense
high-intensity running period in English FA premier league
soccer matches. J Strength Cond Res. 2013;27:909915.
PubMed doi:10.1519/JSC.0b013e31825ff099
8. Krustrup P, Mohr M, Steensberg A, Bencke J, Kjaer M,
Bangsbo J. Muscle and blood metabolites during a soccer
game: implications for sprint performance. Med Sci Sports
Exerc. 2006;38:11651174. PubMed doi:10.1249/01.
mss.0000222845.89262.cd
9. Mohr M, Krustrup P, Bangsbo J. Match performance of
high-standard soccer players with special reference to
development of fatigue. J Sports Sci. 2003;21:519528.
PubMed doi:10.1080/0264041031000071182
10. Bradley PS, Noakes TD. Match running performance
fluctuations in elite soccer: indicative of fatigue, pacing
or situational variables [Published online ahead of print
2013]. J Sports Sci. doi:10.1080/02640414.2013.796062
11. Carling C, Bloomfield J. The effect of an early dismissal
on player work-rate in a professional soccer match. J Sci
Med Sport. 2010;13:126128. PubMed doi:10.1016/j.
jsams.2008.09.004
12. Drust B, Atkinson G, Reilly T. Future perspectives in the
evaluation of the physiological demands of soccer. Sports
Med. 2007;37:783805. PubMed doi:10.2165/00007256-
200737090-00003
13. Edwards AM, Noakes TD. Dehydration: cause of fatigue or
sign of pacing in elite soccer? Sports Med. 2009;39:113.
PubMed doi:10.2165/00007256-200939010-00001
14. Castellano J, Blanco-Villasenor A, Alvarez D.
Contextual variables and timemotion analysis in
soccer. Int J Sports Med. 2011;32:415421. PubMed
doi:10.1055/s-0031-1271771
Player Substitution and Match Performance 423
15. Carling C, Dupont G. Are declines in physical performance
associated with a reduction in skill-related performance
during professional soccer match-play? J Sports Sci.
2011;29:6371. PubMed doi:10.1080/02640414.2010.5
21945
16. Krustrup P, Mohr M, Amstrup T, et al. The Yo-Yo Intermit-
tent Recovery Test: physiological response, reliability, and
validity. Med Sci Sports Exerc. 2003;35:697705. PubMed
doi:10.1249/01.MSS.0000058441.94520.32
17. Krustrup P, Zebis M, Jensen JM, Mohr M. Game-
induced fatigue patterns in elite female soccer. J Strength
Cond Res. 2010;24:437441. PubMed doi:10.1519/
JSC.0b013e3181c09b79
18. Mohr M, Krustrup P, Nybo L, Nielsen JJ, Bangsbo J.
Muscle temperature and sprint performance during soccer
matchesbeneficial effect of re-warm-up at half-time.
Scand J Med Sci Sports. 2004;14:156162. PubMed
doi:10.1111/j.1600-0838.2004.00349.x
19. Rampinini E, Impellizzeri FM, Castagna C, Coutts AJ,
Wisloff U. Technical performance during soccer matches
of the Italian Serie A league: effect of fatigue and competi-
tive level. J Sci Med Sport. 2009;12:227233. PubMed
doi:10.1016/j.jsams.2007.10.002
20. Reilly T, Drust B, Clarke N. Muscle fatigue during foot-
ball match-play. Sports Med. 2008;38:357367. PubMed
doi:10.2165/00007256-200838050-00001
21. Carling C, Espie V, Le Gall F, Bloomfield J, Jullien H.
Work-rate of substitutes in elite soccer: a preliminary
study. J Sci Med Sport. 2010;13:253255. PubMed
doi:10.1016/j.jsams.2009.02.012
22. Bradley PS, Lago C, Rey E, Diaz AG. The effect of high
and low percentage ball possession on physical and techni-
cal profiles in English FA Premier League soccer matches
[published online ahead of print 2013]. J Sports Sci. doi:
10.1080/02640414.2013.786185
23. Buchheit M, Mendez-Villanueva A, Simpson BM, Bour-
don PC. Match running performance and fitness in youth
soccer. Int J Sports Med. 2010;31:818825. PubMed
doi:10.1055/s-0030-1262838
24. Carling C. Analysis of physical activity profiles
when running with the ball in a professional soccer
team. J Sports Sci. 2010;28:319326. PubMed
doi:10.1080/02640410903473851
25. Di Salvo V, Baron R, Gonzalez-Haro C, Gormasz C,
Pigozzi F, Bachl N. Sprinting analysis of elite soccer
players during European Champions League and UEFA
Cup matches. J Sports Sci. 2010;28:14891494. PubMed
doi:10.1080/02640414.2010.521166
26. Di Salvo V, Pigozzi F, Gonzalez-Haro C, Laughlin MS, De
Witt JK. Match performance comparison in top English
soccer leagues. Int J Sports Med. 2013;34:526532. PubMed
27. Gregson W, Drust B, Atkinson G, Salvo VD. Match-to-
match variability of high-speed activities in premier league
soccer. Int J Sports Med. 2010;31:237242. PubMed
doi:10.1055/s-0030-1247546
28. Bradley PS, ODonogue P, Wooster B, Tordoff P. The reli-
ability of Prozone Match Viewer: a video-based technical
performance analysis system. Int J Perform Anal Sport.
2007;7:117129.
 424Bradley, Lago-Peas, and Rey
29. Di Salvo V, Collins A, McNeill B, Cardinale M. Valida-
tion of ProZone: a new video-based performance analysis
system. Int J Perform Anal Sport. 2006;6:108119.
30. Cohen J. Statistical Power Analysis for the Behavioural
Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 1988.
31. Batterham AM, Hopkins WG. Making meaningful infer-
ences about magnitudes. Int J Sports Physiol Perform.
2006;1:5057. PubMed
32. Siegle M, Lames M. Game interruptions in elite soccer. J
Sports Sci. 2012;30:619624. PubMed doi:10.1080/0264
0414.2012.667877
33. Bradley PS, Carling C, Archer D, et al. The effect of
playing formation on high-intensity running and technical
profiles in English FA Premier League soccer matches. J
Sports Sci. 2011;29:821830. PubMed doi:10.1080/0264
0414.2011.561868
34. Bangsbo J, Norregaard L, Thorso F. Activity profile of com-
petition soccer. Can J Sport Sci. 1991;16:110116. PubMed
View publication stats
35. Lovell R, Barrett S, Portas M, Weston M. Re-examination
of the post half-time reduction in soccer work-rate. J Sci
Med Sport. 2013;16:250254. PubMed doi:10.1016/j.
jsams.2012.06.004
36. Weston M, Drust B, Gregson W. Intensities of exercise
during match-play in FA Premier League referees and
players. J Sports Sci. 2011;29:527532. PubMed doi:10.
1080/02640414.2010.543914
37. Bradley PS, Mohr M, Bendiksen M, et al. Sub-maximal
and maximal Yo-Yo Intermittent Endurance Test level 2:
heart rate response, reproducibility and application to elite
soccer. Eur J Appl Physiol. 2011;111:969978. PubMed
doi:10.1007/s00421-010-1721-2
38. Young WB, Newton RU, Doyle TL, et al. Physiological and
anthropometric characteristics of starters and non-starters
and playing positions in elite Australian Rules football: a
case study. J Sci Med Sport. 2005;8:333345. PubMed
doi:10.1016/S1440-2440(05)80044-1