ArticlePDF AvailableLiterature Review

Abstract and Figures

Background: To date, athletic performance has been extensively assessed in youth soccer players through laboratory and field testing. Only recently has running performance via time-motion analysis been assessed during match-play. Match running data are often useful in a practical context to aid game understanding and decision-making regarding training content and prescriptions. A plethora of previous reviews have collated and appraised the literature on time-motion analysis in professional senior players, but none have solely examined youth players. Objective: The aim of the present systematic review was to provide a critical appraisal and summary of the original research articles that have evaluated match running performance in young male soccer players. Methods: Following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement, literature searches were performed in four databases, namely, PubMed, ISI Web of Science, SPORTDiscus and SciELO, using the descriptors “soccer”, “football”, “young”, “youth”, “junior”, “physical performance”, “running performance”, “match running performance”, “movement patterns”, “time-motion analysis”, “distances covered”, “activity profile”, “work rate”, “match analysis”, and “match performance”. Articles were included only if they were original articles written in the English language, studied populations of male children and/or adolescents (≤ 20 years of age), were published/ahead of print on or before December 31, 2017, and showed at least one outcome measure regarding match running performance, such as total distance covered, peak game speed or indicators of activities performed at established speed thresholds. Results: A total of 5801 records were found. After removal of duplicates and the application of the exclusion and inclusion criteria, a total of 50 articles were included (n = 2615 participants). Their outcome measures were extracted and findings were synthesized. The majority of the reviewed papers covered the European continent (62%) and used global positioning systems (GPS) (64%). Measurement error of the tools used to obtain position data and running metrics was systematically overlooked among the studies. The main aims of studies were to examine differences across playing positions (20%), age groups (26%) and match halves (36%). Consistent findings pointed to the existence of positional role and age effects on match running output (using fixed running speed thresholds), but there was no clear consensus about reductions in activity over the course of match-play. Congested schedules negatively affected players’ running performance. While over 32% of all studies assessed the relationships between match running performance and physical capacity, biochemical markers and body composition, ~70% of these did not account for playing position. Conclusions: This review collated scientific evidence that can aid soccer conditioning professionals in understanding external match loads across youth categories. Coaches working with youth development programs should consider that data derived from a given population may not be relevant for other populations, since game rules, match format and configuration are essentially unstandardized among studies for age-matched players. Despite limited evidence, periodization training emphasizing technical-tactical content can improve match running performance. Occurrence of acute and residual impairments in the running performance of young soccer players is common. Prescription of postmatch recovery strategies, such as cold-water immersion and spa treatment, can potentially help reduce these declines, although additional research is warranted. This review also highlighted areas requiring further investigation, such as the possible influence of environmental and contextual constraints and a more integrative approach combining tactical and technical data.
This content is subject to copyright. Terms and conditions apply.
Vol.:(0123456789)
Sports Medicine
https://doi.org/10.1007/s40279-018-01048-8
SYSTEMATIC REVIEW
Match Running Performance inYoung Soccer Players: ASystematic
Review
LuizHenriquePalucciVieira1,2,4 · ChristopherCarling3 · FabioAugustoBarbieri4 · RodrigoAquino1,5 ·
PauloRobertoPereiraSantiago1,2
© Springer Nature Switzerland AG 2019
Extended author information available on the last page of the article
Abstract
Background To date, athletic performance has been extensively assessed in youth soccer players through laboratory and field
testing. Only recently has running performance via time–motion analysis been assessed during match play. Match running
data are often useful in a practical context to aid game understanding and decision making regarding training content and
prescriptions. A plethora of previous reviews have collated and appraised the literature on time–motion analysis in profes-
sional senior players, but none have solely examined youth players.
Objective The aim of the present systematic review was to provide a critical appraisal and summary of the original research
articles that have evaluated match running performance in young male soccer players.
Methods Following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement, lit-
erature searches were performed in four databases: PubMed, ISI Web of Science, SPORTDiscus and SciELO. We used the
following descriptors: soccer, football, young, youth, junior, physical performance, running performance, match running
performance, movement patterns, time–motion analysis, distances covered, activity profile, work rate, match analysis, and
match performance. Articles were included only if they were original articles written in the English language, studied popu-
lations of male children and/or adolescents (aged 20years), were published/ahead of print on or before 31 December 2017
and showed at least one outcome measure regarding match running performance, such as total distance covered, peak game
speed or indicators of activities performed at established speed thresholds.
Results A total of 5801 records were found. After duplicates were removed and exclusion and inclusion criteria applied, 50
articles were included (n = 2615 participants). Their outcome measures were extracted and findings were synthesized. The
majority of the reviewed papers covered the European continent (62%) and used global positioning systems (GPS) (64%).
Measurement error of the tools used to obtain position data and running metrics was systematically overlooked among the
studies. The main aims of studies were to examine differences across playing positions (20%), age groups (26%) and match
halves (36%). Consistent findings pointed to the existence of positional role and age effects on match running output (using
fixed running speed thresholds), but there was no clear consensus about reductions in activity over the course of match play.
Congested schedules negatively affected players’ running performance. While over 32% of all studies assessed the relation-
ships between match running performance and physical capacity, biochemical markers and body composition, ~ 70% of these
did not account for playing position.
Conclusions This review collated scientific evidence that can aid soccer conditioning professionals in understanding external
match loads across youth categories. Coaches working with youth development programs should consider that data derived
from a given population may not be relevant for other populations, since game rules, match format and configuration are
essentially unstandardized among studies for age-matched players. Despite limited evidence, periodization training emphasiz-
ing technical-tactical content can improve match running performance. Occurrence of acute and residual impairments in the
running performance of young soccer players is common. Prescription of postmatch recovery strategies, such as cold water
immersion and spa treatment, can potentially help reduce these declines, although additional research is warranted. This
review also highlighted areas requiring further investigation, such as the possible influence of environmental and contextual
constraints and a more integrative approach combining tactical and technical data.
L.H.Palucci Vieira et al.
Key Points
The number of studies on match running performance
has exponentially increased over the current decade in
youth soccer populations, providing information that
could aid the development of physical conditioning
programs and their prescription across different age
categories.
While the current empirical evidence provides a general
understanding of external match loads across differ-
ent age groups, disparities in experimental approaches
across studies in age-matched players exist, whilst meas-
urement error and the potential impact of situational
variables are also frequently overlooked.
Match running activity can decrease over the course of
games and during fixture congestion (i.e., signs of acute
and residual fatigue, respectively), yet information is
limited on the effects of recovery interventions (e.g.,
cold water immersion and spa treatment) to temper per-
formance declines during intensive schedules.
1 Introduction
Professional soccer clubs invest significant amounts of
money to nurture elite players [1, 2]. The monitoring of
match running performance using time–motion analysis
(TMA) is now considered a fundamental part of contem-
porary youth development processes [3]. This is reflected
by a notable shift in the body of knowledge over the current
decade compared with 10years ago, when a review showed
a lack of information on match play running performance in
youth soccer players [4]. While several further reviews have
collated and appraised the TMA literature, none have solely
examined younger populations (e.g., Mohr etal. [5], Sar-
mento and colleagues [6, 10], Lago-Peñas [7], Carling etal.
[8], Reilly etal. [9], Taylor etal. [11]) despite a plethora of
original investigations comparing performance across dif-
ferent age groups [1220].
In general, running data are useful in a practical context
to aid game understanding and decision making in relation
to individual and collective physical training content and
prescriptions [4, 2124]. This information can also help dis-
tinguish player performances across different competitive
standards [6] and improve understanding of the potential
effects of contextual factors such as match location, qual-
ity of opponents, and match status [25]. Regarding youth
players, TMA data can also help to clarify the demands nec-
essary when moving up into older age brackets, especially
when talented youth players (e.g., U18–U20) are promoted
to the senior squad. The data can help determine at which
age(s) young players demonstrate match running outputs that
are sufficient to meet the demands of professional standards.
Insights into athletic and game evolution can be gained that,
in turn, enable the tailoring of age-specific training programs
[3] and improvement of long-term training interventions
[18] and help avoid replication of methods used in senior
players, since very young soccer players should not be con-
sidered small adults [24, 26].
Yet, to our knowledge, critical appraisals of study design
and the information derived from TMA of match running
performance at the youth level are currently lacking [3]. For
example, it is necessary to investigate the potential discrep-
ancies among studies in the cutoffs used for age-band defi-
nitions (12months [14, 15, 18] or 24months [20, 27, 28])
and running speed thresholds (e.g., high-intensity running:
13.1–16 [18], 15.1–18 [29], 15–36 [30], and > 19.8km/h
[31]) for age-matched players. Furthermore, in contrast with
senior players [3, 7, 20], the potential impact of contextual
factors, also known as situational variables (e.g., influence of
match location or result), has not been examined in younger
players. Finally, investigations of the possible effects of
match format (e.g., small-sided games [13, 32] or full-sized
pitches [16, 18, 24]) and decrements in performance (e.g.,
half-times [3, 16, 17] and specific game periods [13, 15, 33])
would be beneficial to aid in the understanding of the charac-
teristics specific to youth soccer match play. Therefore, the
purpose of the present systematic review was to provide a
critical appraisal and summary of original research articles
that have investigated match running performance in young
male soccer players.
2 Methods
2.1 Search Strategy
This systematic review was conducted according to the Pre-
ferred Reporting Items for Systematic Reviews and Meta-
Analyses (PRISMA) statement [34]. Permission was granted
by the Institutional Human Research Ethics Committee
(School of Physical Education and Sport of Ribeirão Preto,
University of São Paulo, Brazil; protocol number CAAE
61884716.9.0000.5659). The searches for relevant content
related to the running performance of young soccer players
during match play were performed on 31 December 2017,
using four electronic databases: PubMed/NCBI (National
Center for Biotechnology Information, US National Library
of Medicine), Institute for Scientific Information (ISI) Web
of Knowledge, SciELO (Scientific Electronic Library
Online), and SPORTDiscus via EBSCOhost. In each data-
base, the following descriptors were used: [soccer OR
Time-Motion Analysis in Youth Soccer
football] AND [young OR youth OR junior] AND [physi-
cal performance OR running performance OR match run-
ning performance OR movement patterns OR time–motion
analysis OR distances covered OR activity profile OR work
rate OR match analysis OR match performance]. Additional
searches were performed on Google Scholar when the full
texts were not available in these databases and for articles
found on ResearchGate™ [35]. Dedicated computer soft-
ware (EndNote X7, Thomson Reuters©, New York, NY,
USA) was used for reference management, facilitating dedu-
plication and screening steps.
2.2 Selection Criteria
2.2.1 Inclusion Criteria
We opted to include papers if they filled all of the following
criteria: (1) original article; (2) abstract available for screen-
ing; (3) samples of male children and/or adolescents; (4)
published in the English language; (5) published/ahead of
print up to and including 31 December 2017; (6) in a scien-
tific indexed peer-reviewed scientific journal (thus, abstracts
published in conference proceedings, books, theses, disserta-
tions, reviews, systematic reviews, and meta-analyses were
not considered); (7) included at least one outcome measure
regarding the following dependent variables of match run-
ning performance: total distance covered, mean speed or
distance covered per time, peak game speed, activities per-
formed at established speed thresholds (e.g., expressed as
distance covered, distance covered per minute, percentage
of total distance covered) or movement category (e.g., sub-
jective estimates of percentage of time in walking, jogging;
low, medium, and high intensities). No restrictions regard-
ing the date of publication were imposed, other than those
described in item 5.
2.2.2 Exclusion Criteria
Exclusion criteria included (1) goalkeepers as partici-
pants; (2) female participants; (3) samples presenting a
mean age > 20years; (4) matches performed on pitches
with reduced dimensions (i.e., small-sided games, except
as defined by the local soccer governing body according
to information presented in the text); (5) games played as
training/practice sessions; (6) laboratory-based and/or field
tests measuring running performance; (7) use of running
protocols to simulate soccer match play demands; (8) studies
investigating other football codes (American football, Aus-
tralian Rules football, Gaelic football, rugby, indoor soccer)
rather than soccer; (9) unrelated samples (e.g., referees); and
(10) soccer players competing and/or originally described as
senior professionals or semiprofessionals. We also excluded
articles that did not contain one of the descriptors cited in
the search strategy (see Sect.2.1) in the title, abstract and/
or keywords.
2.3 Methodological Quality Assessment
The methodological quality was assessed in line with two
previous review articles related to sport physical perfor-
mance [35] and soccer match running data collection [36].
All of the included studies were appraised using the answers
to nine questions (Q1–9), designed with minor adaptations
from the aforementioned systematic review papers (Table1).
For this purpose, a three-point scale was adopted (where
“yes” = 2 points; “maybe” = 1 point; “no” = 0), except for
Q4 [36]. Strict rules that were applied to Q2, Q3 and Q8 are
also described in a footnote to Table1. Next, a summation
of the attributed points from all the questions was performed
for each study; the possible quality rating varied from 0 to
18 points. Finally, the obtained values were converted into
percentages (minimum 0% to maximum 100%). The studies
were deemed to have an appropriate level of quality when
they scored > 75% [35]. Methodological quality was not
evaluated for the purpose of including/excluding studies.
2.4 Data Extraction
In each search in the aforementioned databases, two evalu-
ators (LV, RA) independently examined the article title,
abstract and keywords in the first stage of screening accord-
ing to the established inclusion and exclusion criteria. Inter-
rater agreement was evaluated by Cohen’s kappa coefficient
(k). If any disagreements occurred, a senior researcher (CC)
examined the situation on a case-by-case and determined the
inclusion or exclusion of a given article using his greater
experience in the field. The agreement rate was k = 0.97. We
examined the texts to identify the terminologies employed in
reference to the method used and for running performance
variable(s) definition. Demographic details of the included
studies were then extracted, including sample size, age/
age group of the participants and the geographical location
where the study was conducted. Methodological descriptions
included match type, format (pitch size, number of players
a-side, whether a rolling substitute policy was adopted) and
configuration (duration and number of periods), measure-
ment techniques/equipment, acquisition frequency used to
obtain running performance data and speed threshold lim-
its. Finally, general results regarding match running perfor-
mance were extracted and the main findings were organized
and described in Sect.3.4. When outcome measures were
presented as figures (e.g., column graphics), a specific rou-
tine that was custom written in the MATLAB® environment
(The MathWorks Inc., Natick, USA) using the “ginput.m”
function was employed to perform a more accurate extrac-
tion of the data.
L.H.Palucci Vieira et al.
3 Results
3.1 Search Results
The search process obtained 5801 records. Figure1 presents
the number of articles found in each electronic database and
a flow chart of the literature search, including all the steps
performed. Following the removal of duplicates, 2102 titles
remained in the reference manager. Following the exami-
nation of titles, abstracts and keywords of all these manu-
scripts, 73 academic studies were eligible and retained for
additional (i.e., full-text) analysis; 34 articles were excluded
at this stage. Upon further inspection of the full text of the
eligible articles and their respective bibliographical refer-
ences, a total of 50 articles [1220, 23, 24, 2633, 3767]
fulfilled all of the inclusion criteria and none of the exclu-
sion criteria and were included in the current systematic
review (i.e., qualitative analysis).
3.2 Methodological Quality
The methodological quality scores attributed to the included
studies can be found in Table2. Scores for the articles ranged
from a minimum of 44% [66] to the maximal possible score
(100%) (two studies [20, 47]). We identified a mean ± stand-
ard deviation quality score for the 50 selected articles of
79 ± 13%. Several papers (N = 20), accounting for 40% of
the total literature, were classified as 80–90% [13, 14, 16,
17, 19, 24, 26, 29, 30, 33, 38, 41, 42, 48, 49, 5254, 57, 60];
in addition, six papers (12%) received very high ratings of
between 90 and 100% [15, 18, 20, 46, 47, 51]. A total of 72%
of the papers (N = 36 publications [1220, 24, 26, 2830, 32,
33, 38, 4043, 4649, 5154, 5658, 6063]) reached an
appropriate quality score, being classified as > 75% [34], and
this was not the case for the remaining articles (28%; N = 14
publications [23, 27, 31, 37, 39, 44, 45, 50, 55, 59, 6467]).
3.3 Research Paradigm
3.3.1 General Information
Table3 describes in detail the demographic and methodo-
logical characteristics of the included papers. Running per-
formance was investigated in match play in a total of 2615
young players. This represents a mean of 52 players per
study. The sample sizes ranged from a minimum of six [37]
to a maximum of 380 [48] participants. The earliest arti-
cles were published in 2001 [37, 59]. Figure2 shows the
yearly distribution frequency of publications since 2000.
A gradually increasing trend has occurred over the current
decade (2010–2017; approximately five articles published
per year) compared with the previous decade (2001–2009;
approximately one article published per year). The mean
age for the youngest group identified was 7.9years [20],
and the oldest group included 20-year-old players [57]. The
majority of evaluations were from the European continent
(62% of the total): England (14%) [13, 15, 19, 26, 33, 44,
47], Italy (10%) [20, 24, 37, 57, 66], Denmark [38, 54, 67],
Portugal [43, 62, 65], Poland [23, 55, 64], Spain [56, 61],
San Marino [39, 40], Norway [31, 59], Turkey [29], Croatia
[50] and Austria [49]. Other investigations were conducted
in Asia, particularly in Qatar (18%) [12, 14, 17, 18, 41, 42,
Table 1 Methodological quality assessment scoring system
Adapted from Bishop etal. [35], Castellano etal. [36], with permission
Strict rules applied to Q2 (no information = 0 point; only age/age group was informed = 1 point; maturity offset also measured = 2 points); Q3
(0–1 item described = 0 point; 2–3 items described = 1 point; 4–5 items described = 2 points); and Q8 (description of mean, standard deviation
and null hypothesis significance test [p-value] = 1 point; also included effect size/magnitude-based inferences = 2 points)
Question Answer Score
Q1 Study objective(s) is/are clearly set out Yes = 2; Maybe = 1; No = 0 0–2
Q2 Demographic data are presented (including assignment of age/age group, maturity status measured) Yes = 2; Maybe = 1; No = 0 0–2
Q3 Game rules (including five items: match duration, field size, players a-side, match type, whether roll-
ing substitute policy was adopted) are described Yes = 2; Maybe = 1; No = 0 0–2
Q4 The reliability/validity of the time–motion system/equipment is not stated, mentioned (i.e., a citation
of previous studies) or measured under local conditions where data collections took place Measured = 2; Mentioned
= 1; Not stated = 0 0–2
Q5 Dependent variables defined Yes = 2; Maybe = 1; No = 0 0–2
Q6 The duration of players recordings/inclusion criteria (an entire half time, a whole match, a certain
percentage, etc.) is clearly indicated Yes = 2; Maybe = 1; No = 0 0–2
Q7 Statistics are appropriate Yes = 2; Maybe = 1; No = 0 0–2
Q8 Results are detailed (mean and standard deviation, percent change/difference, effect size/mechanistic
magnitude-based inference) Yes = 2; Maybe = 1; No = 0 0–2
Q9 Conclusions are insightful (clear, practical applications, and future directions) Yes = 2; Maybe = 1; No = 0 0–2
Total 0–18
Time-Motion Analysis in Youth Soccer
48, 51, 60]. Studies from Oceania (10%), including New
Zealand [15] and Australia [30, 52, 53, 58], were also identi-
fied. The remaining publications were from North America
(USA [53]) and South America (14%), most frequently Bra-
zil (10%) [27, 28, 32, 45, 46]; two records from Bolivia [30,
58] were also found (Table3).
3.3.2 Study Objectives
The main study objectives identified were primarily to char-
acterize general game demands (22%) [12, 14, 15, 23, 24, 26,
27, 29, 43, 48, 64] and to compare the running performance
between playing positions (20%) [12, 14, 18, 19, 23, 27, 48,
62, 65, 66], age groups (26%) [1218, 20, 26, 27, 60, 66, 67]
and match halves/periods (36%) [12, 14, 16, 20, 23, 24, 29,
31, 32, 3740, 45, 48, 56, 57, 64]. Further studies also exam-
ined the influence of biological maturity [38, 51, 60], play-
ing standards [33, 38, 48, 54] and retained versus released
players [15, 19, 26] and compared match running perfor-
mances between game formats [37, 54, 61] and between
specific training regimens [63]. Approximately one-third of
all the studies evaluated relationships between the variables
of match running performance and (1) anthropometric meas-
ures (e.g., height, body weight, and skinfolds) [50, 60], (2)
physiological markers [45, 57] (e.g., creatine kinase [CK],
lactate dehydrogenase [LDH], cortisol, interleukin [IL]-6,
and testosterone levels) and, more frequently, (3) indica-
tors of physical capacity provided through laboratory-based
methods (e.g., maximal oxygen consumption [
̇
V
O
2max
]
and speed attained at
̇
V
O
2max
[v
̇
V
O
2max
] obtained through
an incremental treadmill protocol) and field testing (e.g.,
YoYo Intermittent Recovery Test Level 1 (YoYo IR1) [68],
running-based anaerobic sprint test (RAST) [69], Hoff test
[70] and Vam-Eval test [71]) [14, 17, 18, 20, 28, 29, 39,
40, 43, 46, 47, 65]. Researchers also investigated the effects
of match congestion [49, 51, 52], moderate [53] and high
Fig. 1 Flow chart of literature search including all steps performed according to the PRISMA statement. aFailed to meet inclusion criterion 7
(n = 7), and fulfillment of exclusion criteria 3 (n = 3), 4 (n = 1), 5 (n = 3), 6 (n = 17) and 7 (n = 3)
L.H.Palucci Vieira et al.
Table 2 Methodological quality outcomes attributed to studies on match running performance in youth soccer players
Study Year Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Total (∑) Quality score (%)
Al Haddad etal. [17] 2015 2 1 2 1 2 2 2 2 2 16 89
Andrzejewski etal. [55] 2011 1 1 0 0 2 2 2 1 1 10 56
Andrzejewski etal. [23] 2009 2 1 1 0 1 2 0 1 2 10 56
Aquino etal. [45] 2016 2 2 1 0 2 0 2 1 2 12 67
Aquino etal. [32] 2016 2 1 1 1 2 2 2 1 2 14 78
Aquino etal. [28] 2018 2 1 1 1 2 2 2 1 2 14 78
Arruda etal. [49] 2015 2 1 1 1 2 2 2 2 2 15 83
Aslan etal. [29] 2012 2 1 2 0 2 2 2 2 2 15 83
Atan etal. [16] 2016 2 1 2 1 2 1 2 2 2 15 83
Aughey etal. [30] 2013 2 2 2 0 2 1 2 2 2 15 83
Bravo-Sánchez etal. [61] 2017 2 2 2 0 2 0 2 2 2 14 78
Bellistri etal. [20] 2017 2 2 2 2 2 2 2 2 2 18 100
Brito etal. [62] 2017 2 1 2 0 2 1 2 2 2 14 78
Buchheit etal. [18] 2010 2 2 2 1 2 2 2 2 2 17 94
Buchheit etal. [12] 2010 2 1 2 1 2 2 1 1 2 14 78
Buchheit etal. [51] 2011 2 2 2 1 2 2 2 2 2 17 94
Buchheit etal. [42] 2013 2 2 2 1 2 1 2 2 2 16 89
Buchheit and Mendez-Villanueva [60] 2014 2 2 2 1 2 2 2 1 2 16 89
Buchheit etal. [58] 2015 2 1 2 0 2 2 2 2 1 14 78
Capranica etal. [37] 2001 2 1 1 2 2 0 1 1 1 11 61
Castagna etal. [24] 2003 2 1 2 2 2 1 2 1 2 15 83
Castagna etal. [39] 2009 2 1 1 1 2 0 2 1 2 12 67
Castagna etal. [40] 2010 2 1 1 2 2 0 2 2 2 14 78
Dello Iacono etal. [63] 2017 2 1 0 1 2 2 2 2 2 14 78
Doncaster etal. [47] 2016 2 2 2 2 2 2 2 2 2 18 100
Fernandes-da-Silva etal. [46]2016222122222 17 94
Garvican etal. [53] 2014 2 1 1 1 2 2 2 2 2 15 83
Goto etal. [26] 2015 2 2 2 1 2 2 2 2 1 16 89
Goto etal. [15] 2015 2 1 2 2 2 2 2 2 2 17 94
Harley etal. [13] 2010 2 1 2 1 2 2 2 2 2 16 89
Helgerud etal. [59] 2001 2 1 0 2 1 0 2 1 2 11 61
Hunter etal. [44] 2015 2 1 0 0 2 2 2 2 2 13 72
Kęsicki and Lewicki [64] 2017111022002950
Mendez-Villanueva etal. [41] 2011 2 1 2 1 2 2 2 2 2 16 89
Mendez-Villanueva etal. [14] 2013 2 1 2 1 2 2 2 2 2 16 89
Pereira Da Silva etal. [27] 2007 2 1 1 2 2 0 1 1 1 11 61
Time-Motion Analysis in Youth Soccer
altitude [30, 58], specific pitch surface [62] and opponent
quality [48] on running performance. Seasonal changes in
physical capacity (i.e., performance derived from fitness
tests) associated with those in match running performance
were examined [42]. Also examined were interventional pre-
and postmatch strategies, including prematch supplementa-
tion with caffeine [31] and postmatch recovery using cold
water immersion [52] or spa treatments (combined sauna,
cold water immersion and jacuzzi) [51]. Five remaining arti-
cles (10% of the total) investigated the effects of training on
running performance during match play using longitudinal
experimental approaches [19, 32, 42, 55, 59]. Figure3 pre-
sents the various research topics addressed in the studies on
match running performance.
3.3.3 Match Type andConguration
In total, 29 studies (58% of the total) analyzed performance
in official competitions [1320, 23, 24, 2628, 33, 3740,
43, 44, 4851, 61, 63, 6567], eight (16%) were friendly
matches [12, 29, 30, 41, 42, 53, 58, 60] and 11 (20%) were
simulated matches [31, 32, 4547, 52, 54, 56, 57, 62, 64].
Two studies did not clearly specify the match type [55, 59].
Among the studies, 50% exclusively used a game format
with 11 players a-side [1217, 24, 27, 29, 30, 37, 3943,
46, 47, 50, 51, 54, 57, 60, 62, 65]. Some of these compared
age groups and used the 11-a-side format regardless of age
(e.g., U11 to U16 [15], U12 to U16 [13] and U13 to U18 [12,
14, 42, 51]). In contrast, Goto etal. [26] employed 6 versus
6 in the U9–U10 age groups, and Bellistri etal. [20] used 5
versus 5 and 7 versus 7 for U8 and U10 players, respectively.
Additionally, Saward etal. [19] examined U9 to U18 soc-
cer players, adopting 11 versus 11 in U12 to U18s, whereas
the authors adjusted the number of players per side for the
younger players of U9–U10 (5 vs. 5 and 7 vs. 7) and U11s
(7 vs. 7 and 11 vs. 11). The remaining articles (44%) did
not provide sufficient information to fully characterize the
number of players per side [23, 28, 31, 32, 38, 44, 45, 48,
49, 52, 53, 55, 56, 59, 63, 64, 66, 67]. Some studies (8%)
(England [13, 33], Italy [20], New Zealand [16]) adopted
a rolling substitute policy, in which players were allowed
to return to the field after being replaced. One study [53]
used both the official and the interchangeable substitution
methods. Additional information regarding game configura-
tion (field size, duration and number of playing periods) is
presented in Table3.
3.3.4 Speed Thresholds
Table4 presents speed thresholds adopted (i.e., superior
and inferior speed limits) across studies. A total of 34
studies employed arbitrary fixed speed thresholds (68%
of the total) with unit measures in km/h or m/s [12, 15,
Table 2 (continued)
Study Year Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Total (∑) Quality score (%)
Pettersen etal. [31] 2014 2 1 1 0 2 1 2 2 2 13 72
Rago etal. [65] 2017 2 1 1 0 2 0 2 1 2 11 61
Randers etal. [56] 2010 2 1 1 2 2 2 1 1 2 14 78
Randers etal. [54] 2014 2 1 2 1 2 1 2 2 2 15 83
Rebelo etal. [43] 2014 2 1 1 1 2 2 2 1 2 14 78
Izzo and Varde’i [66] 2017210021002844
Romagnoli etal. [57] 2016 2 1 2 2 2 2 2 1 2 16 89
Rowsell etal. [52] 2011 2 1 1 1 2 2 2 2 2 15 83
Saward etal. [19] 2016 2 1 1 1 2 2 2 2 2 15 83
Sporis etal. [50] 2017 2 1 1 0 2 0 2 1 2 11 61
Stroyer etal. [38] 2004 2 2 1 2 2 2 1 1 2 15 83
Varley etal. [48] 2017 2 1 1 1 2 2 2 2 2 15 83
Vigh-Larsen etal. [67] 2018 2 1 0 1 2 2 2 1 2 13 72
Waldron and Murphy [33] 2013 2 1 2 1 2 2 2 1 2 15 83
Mean ± standard deviation 1.96 ± 0.20 1.24 ± 0.43 1.36 ± 0.69 0.94 ± 0.71 1.96 ± 0.20 1.48 ± 0.79 1.78 ± 0.55 1.52 ± 0.58 1.90 ± 0.30 14 ± 2 79 ± 13
L.H.Palucci Vieira et al.
Table 3 Demographic characteristics and methodologies employed in studies measuring match running performance in young soccer players
Study NAge (years)aDuration (min) Field size (m)bMatch type Rolling
substitute
policy
Players a-side Location Technology used and acqui-
sition frequency
Bellistri etal. [20]n = 27 U8
U10 3 × 15 45 × 25
60 × 40 Official Yes 5 vs. 5
7 vs. 7 Italy GPS–K-GPS, K-Sport, Italy
(10Hz)
Goto etal. [26]n = 34 U9
U10 4 × 15
or (2 × 20 + 2 × 15) 44.8 × 26 Official 6 vs. 6 England GPS–GPsports, Australia
(1Hz)
Bravo-Sánchez etal. [61]n = 154 10.7 ± 0.8 2 × 25 40 × 60 Official 7 vs. 7 or 8 vs. 8 Spain GPS–GPsports, Australia
(5Hz)
Capranica etal. [37]n = 6 11 – 100 × 65 Official 11 vs. 11 Italy Observational analysis–
video records
Castagna etal. [24]n = 11 11.8 ± 0.6 2 × 30 100 × 65 Official 11 vs. 11 Italy Triangular surveying
method
Stroyer etal. [38]n = 26 12.1 ± 0.7
12.6 ± 0.6
14 ± 0.2
2 × 30
2 × 30
2 × 35
Official Denmark Observational analysis–
video records
Pereira Da Silva etal. [27]n = 75 U15
U17
U20
2 × 30
2 × 40
2 × 45
Official 11 vs. 11 Brazil Observational analysis–
video records
Randers etal. [54]n = 41 U13 20 105 × 68 Simulated 11 vs. 11 Denmark GPS–MinimaxX, Catapult,
Australia (10Hz)
Andrzejewski etal. [55]n = 19 13.5 ± 0.4 Poland Visual tracking system–soft-
ware BANAL
Andrzejewski etal. [23]n = 10 13.5 ± 0.4 2 × 25 100 × 68 Official Poland Kinematic method–software
BANAL
Fernandes-da-Silva etal.
[46]n = 33 13.5–15.9 2 × 35 90 × 45 Simulated 11 vs. 11 Brazil GPS–K-GPS, K-Sport, Italy
(10Hz)
Doncaster etal. [47]n = 17 12–14 2 × 20 90 × 50 Simulated 11 vs. 11 England GPS–Catapult, Australia
(10Hz)
Buchheit etal. [18]
Mendez-Villanueva etal.
[14]
Buchheit etal. [51]
Al Haddad etal. [17]
[except U18]
n = 99
n = 103
n = 33
n = 180
U13
U14
U15
U16
U17
U18
2 × 35
2 × 35
2 × 40
2 × 40
2 × 40
2 × 45
100 × 70 Official 11 vs. 11 Qatar GPS–GPsports, Australia
(1Hz)
Buchheit etal. [12]
Buchheit etal. [42]n = 99
n = 124 U13
U14
U15
U16
U17
U18
2 × 35
2 × 35
2 × 40
2 × 40
2 × 40
2 × 45
100 × 70 Friendly 11 vs. 11 Qatar GPS–GPsports, Australia
(1Hz)
Castagna etal. [39]n = 21 14.1 ± 0.2 2 × 30 Official 11 vs. 11 San Marino GPS–GPsports, Australia
(1Hz)
Time-Motion Analysis in Youth Soccer
Table 3 (continued)
Study NAge (years)aDuration (min) Field size (m)bMatch type Rolling
substitute
policy
Players a-side Location Technology used and acqui-
sition frequency
Castagna etal. [40]n = 18 14.4 ± 0.1 2 × 30 Official 11 vs. 11 San Marino GPS–GPsports, Australia
(1Hz)
Harley etal. [13]n = 112 U12
U13
U14
U15
U16
3 × 25 or (2 × 25 + 2 × 12.5)
2 × 40 77 × 60
77 × 60
99 × 65
99 × 65
99 × 65
Official Yes 11 vs. 11 England NdGPS–MinimaxX, Cata-
pult, Australia (5Hz)
Rowsell etal. [52]n = 13 15.9 ± 0.6 90 Simulated Australia GPS–GPsports, Australia
(1Hz)
Mendez-Villanueva etal.
[41]n = 14 16.7 ± 0.7 2 × 40–45 100 × 70 Friendly 11 vs. 11 Qatar GPS–GPsports, Australia
(1Hz)
Brito etal. [62]n = 66 U14 30 100 × 64 Simulated No 11 vs. 11 Portugal GPS–Qstarz, Taiwan
(10Hz)
Waldron and Murphy [33]n = 31 U14 2 × 25 + 2 × 12.5 99 × 65 Official Ye s England GPS–GPsports, Australia
(5Hz)
Buchheit and Mendez-
Villanueva [60]n = 36 U15 2 × 40 100 × 70 Friendly 11 vs. 11 Qatar GPS–GPsports, Australia
(1Hz)
Rebelo etal. [43]n = 30 U17 80 Official 11 vs. 11 Portugal Observational analysis–
video records
Goto etal. [15]n = 81 U11
U12
U13
U14
U15
U16
(2 × 20 + 2 × 15) or 3 × 25
3 × 25
3 × 25
2 × 40
2 × 40
2 × 40
78.7 × 54.1
78.7 × 54.1
88 × 64.2
100.8 × 68.2
100.8 × 68.2
100.8 × 68.2
Official 11 vs. 11 England GPS–GPsports, Australia
(1Hz)
Saward etal. [19]n = 263 U9
U10
U11
U12
U13
U14
U15
U16
U17
U18
60–80
60–80
60–80
75–80
75–80
80–90
80–90
80–90
80–90
80–90
Official 5 vs. 5–7 vs. 7
5 vs. 5–7 vs. 7
7 vs. 7–11 vs. 11
11 vs. 11
11 vs. 11
11 vs. 11
11 vs. 11
11 vs. 11
11 vs. 11
11 vs. 11
England GPS–GPsports, Australia (1
and 5Hz)
Aquino etal. [45]
Aquino etal. [32]n = 18
n = 10 U16
15.4 ± 0.2 2 × 30 70 × 50 Simulated Brazil Videogrammetry method
Automatic tracking–
DVIDEOW, Brazil
(30Hz)
L.H.Palucci Vieira et al.
Table 3 (continued)
Study NAge (years)aDuration (min) Field size (m)bMatch type Rolling
substitute
policy
Players a-side Location Technology used and acqui-
sition frequency
Atan etal. [16]n = 85 U13
U14
U15
2 × 30
2 × 35
2 × 40
100 × 60 Official Yes 11 vs. 11 New Zealand GPS–GPsports, Australia
(5Hz)
Arruda etal. [49]n = 10 U15 2 × 25–30 Official Austria GPS–GPSports, Australia
(15Hz)
Aquino etal. [28]n = 18 U15
U17 2 × 30
2 × 40 105 × 68 Official Brazil Videogrammetry method
Manual tracking–
DVIDEOW, Brazil
(30Hz)
Varley etal. [48]n = 380 U17 2 × 45 Official Qatar Videogrammetry method–
Prozone Sports Ltd
Aslan etal. [29]n = 32 17.6 ± 0.6 2 × 45 105 × 68 Friendly 11 vs. 11 Turkey Videogrammetry method
Manual tracking–Mathball
Match Analysis System,
Algoritma Company,
Turkey
Helgerud etal. [59]n = 19 18.1 ± 0.8 – – Norway Videogrammetry method
Manual tracking–Wacom
Digitizer SD-421-E digital
board (Wacom Co., Ltd,
Japan) and a marking pen
(Arntzen Engineering,
Norway)
Sporis etal. [50]n = 37 18.4 ± 0.1 2 × 45 Official 11 vs. 11 Croatia System 3D tille sport
analyzer
Vigh-Larsen etal. [67]n = 30 U17
U19 Official Denmark Local positioning system
ZXY Tracking System
(20Hz)–radio, Chyron-
Hego Corp., USA
Aughey etal. [30]
Buchheit etal. [58]n = 20
n = 13 U17–U18 2 × 45 105 × 68 Friendly 11 vs. 11 Australia/Bolivia GPS–MinimaxX, Catapult,
Australia (10Hz)
Pettersen etal. [31]n = 22 17.6 ± 1.1 90 Simulated – Norway ZXY Sport Tracking System
(20Hz)–radio, Norway
Hunter etal. [44]n = 12 U18 Official England GPS–Catapult Innovations,
Australia (5Hz)
Rago etal. [65]n = 29 U19 Official 11 vs. 11 Portugal GPS–Qstarz, Taiwan (5Hz)
Dello Iacono etal. [63]n = 24 U19 Official NYL and UEFA GPS–GPsports, Australia
(15Hz)
Time-Motion Analysis in Youth Soccer
Table 3 (continued)
Study NAge (years)aDuration (min) Field size (m)bMatch type Rolling
substitute
policy
Players a-side Location Technology used and acqui-
sition frequency
Kęsicki and Lewicki [64]n = 18 18.1 ± 1.2 90 Simulated Poland GPS–OptimEye, Catapult,
Australia (10Hz)
Garvican etal. [53]n = 20 U20 2 × 45
3 × 25–30 Friendly No
Yes Australia/USA GPS–MinimaxX, Catapult,
Australia (10Hz)
Izzo and Varde’i [66] U20 Official Italy GPS–K-GPS, K-Sport, Italy
(20Hz)
Randers etal. [56]n = 20 19.3 ± 1.2 2 × 47.5 105 × 68 Simulated Spain GPS (MinimaxX, Catapult/
GPsports, Australia) (1
and 5Hz); video-based
time-motion and semi-
automatic video tracking
(Amisco Pro®, France)
Romagnoli etal. [57]n = 22 17–20 2 × 45 Simulated No 11 vs. 11 Italy Semi-automatic video track-
ing–Wisport, Italy
GPS global positioning system, NYL National Youth League, UEFA Union of European Football Associations
a Mean, range or age category
b Field size = length × width
L.H.Palucci Vieira et al.
18, 23, 24, 2833, 39, 40, 4246, 4857, 6062, 6567],
whereas 14 (28%) used individualized speed thresholds
[1217, 19, 20, 26, 29, 41, 42, 44, 47]. Nine studies
(18%) employed both methods [12, 15, 17, 29, 41, 42,
44, 58, 63]. Three types of individualization techniques
were reported: (1) speed thresholds derived according
to individual physical capacity using data from fitness
testing protocols (20%)—these were generally based on
maximal linear sprint speed, lactate concentrations, and
v
̇
V
O2max tests [12, 14, 15, 17, 29, 41, 42, 44, 47, 58];
(2) thresholds proposed according to mean values for
physical capacity (e.g., maximal linear sprint speed test)
for each age group (12%) [13, 15, 16, 19, 20, 26]; and
(3) individual sprinting threshold using a percentage of
individual peak game speed relative to an arbitrary fixed
threshold (25.2km/h) [63]. Four remaining studies [27,
37, 38, 43] were conducted using operator judgment of
the intensities reached during running displacements (i.e.,
video-based time-motion [VTM]) performed by players in
a given movement category. The number of speed thresh-
olds used ranged from a minimum of one to a maximum
of eight (ST1–ST8). There were 31 distinct arbitrary fixed
speed thresholds, 26 distinct individualized speed thresh-
olds by mean age-band physical capacity, and 11 distinct
individualized speed thresholds by individual physical
capacity employed to characterize the youth players’ run-
ning performance (see Table4).
3.3.5 Technology
In reference to the technologies utilized to quantify match
running performance, 25 studies employed global positioning
systems (GPS) exclusively (64%), with sampling frequencies
preset at 1Hz (28%) [12, 14, 15, 1719, 26, 3942, 51, 52,
60], 5Hz (14%) [13, 16, 19, 33, 44, 61, 65], 10Hz (16%) [20,
30, 46, 47, 53, 54, 62, 64], 15Hz [49, 63] and 20Hz [66].
Among the remaining studies, four employed VTM [27, 37,
38, 43], and eight used video tracking (i.e., videogrammetry)
approaches (16%), which were performed either manually [28,
29, 59], semi-automatically [48, 57] or automatically (30Hz)
[32, 45], or the method was not stated [55]. One investigation
compared 1Hz GPS, 5Hz GPS, VTM techniques and semiau-
tomatic video-tracking methods [56]. Two studies did not pro-
vide sufficient information about their methodologies [23, 50].
Castagna etal. [24] adopted a triangular surveying method
(for more information see Carling etal. [4] and Ohashi etal.
[72]), and two studies utilized a three-dimensional local radio
processing system (20Hz) [31, 67].
3.3.6 Terminology
Nine distinct nomenclatures were used to the report the
methodology used to obtain the running performance vari-
ables, ordered from the most to least frequent, as follows:
(1) match analysis (36%) [12, 15, 16, 19, 24, 26, 27, 29, 31,
Fig. 2 Yearly distribution frequency and cumulative sum of the number of publications included in the current systematic review addressing
match running performance in young soccer players
Time-Motion Analysis in Youth Soccer
37, 41, 43, 44, 46, 51, 57, 60, 61]; (2) TMA (22%) [12, 14,
17, 18, 41, 42, 44, 51, 56, 62, 65]; (3) performance analysis
[28, 32, 39, 52]; (4) motion analysis [13, 15, 16]; (5) kin-
ematic analysis [23, 55]; (6) physical analysis [45]; (7) video
analysis [59]; (8) movement analysis [33]; and (9) external
load data collection [63]. To define and group dependent
variables related to running performance, we verified 20
different descriptions, ordered from the most to least fre-
quent, as follows: (1) match running performance (26%)
[14, 1620, 28, 41, 42, 51, 52, 58, 60]; (2) match activities
(24%) [15, 16, 24, 26, 31, 33, 3840, 43, 47, 50]; (3) activity
profile (22%) [24, 30, 31, 39, 40, 46, 53, 54, 56, 58, 67]; (4)
distances covered (10%) [13, 27, 57, 59, 64]; (5) physical
performance [32, 40, 47, 48]; (6) movement patterns [27, 29,
32, 65]; (7) activity pattern [38, 41, 54]; (8) player’s move-
ments [37, 44, 48]; (9) physical loads [23, 40]; (10) physical
match performance [43, 46]; (11) running performance [53];
(12) running activity [62]; (13) motor performances [23];
(14) displacement patterns [45]; (15) running measures [49];
(16) match intensity [58]; (17) match play intensity [14];
(18) running activity [41]; (19) physical variables [61]; (20)
time–motion variables [63]; and unknown [55].
3.4 Match Running Performance
3.4.1 Playing Standard
For age-matched players (U13 and U14 categories), match
running performance was greater in elite players (e.g.,
total distance covered, distance covered per minute, high-
intensity running, peak game speed) [33, 54]. Conversely,
these players performed fewer standing or low-intensity
Fig. 3 Research topics of studies on match running performance in young soccer players
L.H.Palucci Vieira et al.
Table 4 Speed thresholds used in studies to assess match running performance in youth soccer players
References ST
Measure ST1 ST2 ST3 ST4 ST5 ST6 ST7 ST8
Fixed
Castagna etal. [24] km/h <8.0 8.1–13 13.1–18 >18 >13 –
Castagna etal. [39] km/h 0–0.4 0.4–3 3–8 8–13 13–18 >18 ST5 + ST6
Buchheit etal. [18]
Brito etal. [62]km/h <13 13.1–16 16.1–19 >19.1 ST3 + ST4 – –
Castagna etal. [40] km/h 0–0.4 0.4–3 3–8 8–13 13–18 >18
Aslan etal. [29] km/h 0–6 6.1–8 8.1–12 12.1–15 15.1–18 >18.1
Waldron and Murphy [33] km/h <6 6.1–13 13.1–19 >19.1
Rebelo etal. [43] km/h 0–0.4 0.4–3 3–8 8–13 13–18 >18 RB
Goto etal. [15] m/s 0–1.5 1.6–3.0 3.1–4.5 4.6–6.0 >6.0
Hunter etal. [44] km/h <14.99 15–17.99 18–24.99 25–35
Buchheit etal. [58] km/h >14.4 – – –
Aquino etal. [45]
Aquino etal. [32]km/h ≤0.4 0.4–3 3.1–8 8.1–13 13.1–18 >18 ST5 + ST6
Fernandes-da-Silva etal.
[46]km/h 13–18 >18 ST1 + ST2 – –
Varley etal. [48] m/s <4 >4 ≥5.5 ≥7 – –
Arruda etal. [49] km/h >18 – – –
Sporis etal. [50] km/h 0.4–3 3–8 8–13 13–18 >18
Andrzejewski etal. [23] m/s <3.5 ≥3.5 – – –
Andrzejewski etal. [55] m/s ≥5 – – –
Buchheit etal. [51] km/h <13 13.1–16 16.1–19 >19.1 ST3 + ST4 – –
Rowsell etal. [52] km/h >15 – – –
Aughey etal. [30]
Garvican etal. [53]m/s 0.01–4.16 4.17–10 – – –
Pettersen etal. [31] km/h >19.8 >25.2 – – –
Randers etal. [54] km/h 0–0.2 0.2–4 4–8 8–12 12–16 16–20 >20 –
Aquino etal. [28] km/h 0–0.4 0.41–3 3.01–8 8–13 13.01–16 16.01–19 >19.01 ST6 + ST7
Randers etal. [56] km/h >7 9–13 16–22 >22 – –
Izzo and Varde’i [66] km/h 0–10 10–14 14–16 16–21 21–24 >24
Romagnoli etal. [57] km/h <15 >15 >20
Vigh-Larsen etal. [67] km/h 19.8–25.2 ≥25.2 – – –
Buchheit etal. [12]
Buchheit etal. [42]
Dello Iacono etal. [63]
km/h >19 – – –
Bravo-Sánchez etal. [61] km/h ≥13 ≥15.7 ≥20
Time-Motion Analysis in Youth Soccer
Table 4 (continued)
References ST
Measure ST1 ST2 ST3 ST4 ST5 ST6 ST7 ST8
Buchheit and Mendez-Vil-
lanueva [60]km/h >16 >19 – – -
Rago etal. [65] km/h 16–19 19–22 >22 ≥16 –
Individualized
Age-band
Bellistri etal. [20] km/h U8
U10 <6.3
<6.7 6.4–8.4
6.8–9.6 8.5–11.5
9.7–13.2 11.6–17.3
13.3–18.2 >18.2
>17.3 ST3 + ST4 +ST5 – –
Goto etal. [26] m/s U9
U10 0–1.0
0–1.0 1.1–2.0
1.1–2.1 2.1–3.1
2.2–3.1 3.2–4.1
3.2–4.2 >4.1
>4.2 – –
Saward etal. [19] m/s U9
U10
U11
U12
U13
U14
U15 U16
U17, U18
<2.74
<2.84
<2.91
<3.05
<3.19
<3.34
<3.49
<3.51
<3.59
≥2.74
≥2.84
≥2.91
≥3.05
≥3.19
≥3.34
≥3.49
≥3.51
≥3.59
≥4.56
≥4.73
≥4.85
≥5.08
≥5.31
≥5.6
≥5.81
≥5.85
≥5.98
Goto etal. [15] m/s U11
U12
U13 U14
U15, U16
0–1.1
0–1.1
0–1.1
0–1.2
0–1.2
1.2–2.1
1.2–2.2
1.2–2.2
1.3–2.3
1.3–2.4
2.2–3.2
2.3–3.2
2.3–3.3
2.4–3.5
2.5–3.7
3.3–4.2
3.3–4.3
3.4–4.4
3.6–4.6
3.8–4.9
>4.2
>4.3
>4.4
>4.6
>4.9
Atan etal. [16] km/h U13
U14
U15
0–0.4
0–0.4
0–0.4
0.4–4
0.4–4.5
0.4–5
4–8
4.5–8.5
5–9
8–13
8.5–13.5
9–14
13–18
13.5–18.5
14–19
>18
>18.5
>19
Harley etal. [13] m/s U12
U13 U14
U15
U16
>3.04
>3.16
>3.18
>3.56
>3.66
>4.18
>4.34
>4.37
>4.89
>5.04
≥5.32
≥5.53
≥5.56
≥6.22
≥6.41
Individual physical capacity
Buchheit etal. [58] >80% vs. YoYoIR1
Dello Iacono etal. [63] – (25.2/Vpeak) × 100
Doncaster etal. [47] >50% MLV >70% MLV >90% MLV
Goto etal. [15] m/s 0–20%MS5m 21–40%MS5m 41–60%MS5m 61–80%MS5m 80–MS5m
Mendez-Villanueva etal.
[14]km/h <60%MAS 61–80% MAS 81–100% MAS 101%MAS–30%ASR >31% ASR
L.H.Palucci Vieira et al.
Table 4 (continued)
References ST
Measure ST1 ST2 ST3 ST4 ST5 ST6 ST7 ST8
Hunter etal. [44] km/h <RCT
<79% MAS
<49% MSS
<79%MAS
RCT–v ̇
V
O2max
80–99% MAS
50–59% MSS
80–99% MAS
v̇
V
O2max –29%ASR
100–139%MAS
60–79% MSS
100%MAS–29%ASR
30%ASR–MSS
10%MAS–35
80–100%MSS
30%ASR–MSS
Aslan etal. [29] m/s <FBL2FBL2–4 >FBL4 – –
Buchheit etal. [42]
Buchheit etal. [12]km/h >61%MSS – – –
VTM
Capranica etal. [37] Running Walking Inactivity Jumping
Stroyer etal. [38] Standing Walking LIR HIR
Pereira Da Silva etal. [27] – Jogging Walking Sprint LM WB RB JB SPR-b
Rebelo etal. [43] Standing Walking Jogging MIR HIR SPR RB –
ASR anaerobic speed reserve, obtained from the difference between MSS and MAS value, FBL running velocity correspondent to fixed blood lactate (FBL) < 2 (FBL2), 2–4 (FBL2–4)
and > 4mmol·L−1 (FBL2–4) obtained from an incremental field test [29], HIR high-intensity running, JB jogging with the ball, LIR low-intensity running, LM lateral movement, MAS maximal
aerobic speed, provided from the Vam-Eval incremental field test [71] or an incremental treadmill test [44], MIR medium-intensity running, MLV maximal linear velocity in a 20-m sprint test,
MS5m “flying” 5-m sprint speed reached during the 10-m sprint test, MSS maximal sprinting speed in a 40-m sprint test, RB running backward, RCT speed corresponding to respiratory com-
pensation threshold, SPR sprinting, SPR-b sprint with the ball, ST speed threshold, v
̇
V
O2max velocity corresponding to 95% of the maximal oxygen consumption, Vpeak peak game speed, VTM
video-based time-motion, vYoYoIR1 final velocity reached in the YoYo Intermittent recovery test level 1 [68], WB walking backward
Time-Motion Analysis in Youth Soccer
displacements [33, 38] than did nonelite or recreational
peers. Studies also highlighted differences in match running
performance, favoring academy players who were retained
(e.g., greater distance covered per minute and low- to mod-
erate-intensity running) compared with those released in
some age groups [15, 19, 26]. In addition, a greater sprint
distance was covered by top- and middle-ranked teams than
by bottom-ranked U17 peers [48]. Table5 summarizes the
general match running performance results extracted from
the reviewed papers. Indicators of running load included
total distance covered, distance covered per minute, peak
game speed reached and the distribution of distances covered
according to speed thresholds.
3.4.2 Match Halves
Contrasting results regarding comparisons between halves
were observed. Several studies (~ 47%) reported reductions
in second-half measures of running performance (e.g., total
distance covered, high-intensity running, repeated sprint
sequences) [12, 14, 16, 29, 39, 40, 57]. In contrast, numer-
ous articles (40%) reported no changes (e.g., total distance
covered, time spent in high-intensity running, high-intensity
running distance) [16, 20, 23, 24, 37, 43], and two (~ 13%)
identified an increase (e.g., average speed, high-intensity
running, peak game speed and number of sprints) [32, 45]
in match running performance during the second half.
3.4.3 Age Group Comparisons
Concerning age-related performance, the values for some
parameters (e.g., peak game speed, total distance covered,
repeated sprint sequences, high-intensity running, high-
intensity activities) were greater in the older than in the
younger groups in cross-sectional studies [1315, 17, 18,
20, 67] that used fixed speed thresholds. In addition, older
players performed more high-intensity actions than younger
peers in the same U15 age group [60]. Age-related differ-
ences varied from slight [18] to large [15]. On the other
hand, when individualized speed thresholds were applied or
distances covered were adjusted by effective playing time,
the differences were less evident [12, 13, 15, 16, 18]. While
studies rarely identified greater running outputs in younger
than in older players in absolute terms (e.g., distance covered
per minute [27]), this finding was more frequent when indi-
vidualized speed bands were employed (e.g., repeated sprint
sequences and peak game speed relative to maximal sprint-
ing speed in field tests, distance covered above maximal aer-
obic speed) [12, 14, 16, 17]. In addition, players commenc-
ing puberty spent more time in standing/walking and lower
jogging movements than did mature players [38]. There
was evidence that more mature players achieved greater
peak speeds and performed more high-intensity actions and
repeated high-intensity actions in match-play than did less
mature peers in the U15 category [60].
3.4.4 Between‑Position Dierences
Studies were in accordance that match running performance
measures (e.g., total distance covered, peak game speed, fre-
quency of sprints, sprinting distance) were position depend-
ent [12, 14, 1719, 23, 27, 29, 41, 42, 48, 62, 65, 66]. Centre
backs reported the lowest values for total distance covered
[14] and high-intensity activities [18, 23, 48, 62]; midfield-
ers and second forwards covered the highest total distance
covered; wide midfielders and forwards demonstrated the
highest peak game speeds and frequency of high-intensity
activities [17, 18, 66].
3.4.5 Association withPhysical andPhysiological Factors
Match running performance was moderately to strongly cor-
related with postmatch physiological markers (CK, LDH,
cortisol, IL-6) in two studies [45, 57]. Positive relationships
were also revealed on several occasions between match run-
ning performance (e.g., total distance covered, sprinting,
high-intensity running, high-intensity activities) and physi-
cal capacity, as determined by the following tests: YoYo
IR-1, YoYo IR-2, multistage fitness, Carminatti, 20-m shut-
tle run, Zig Zag, Hoff, RAST and 40-m sprint [17, 20, 28,
29, 39, 40, 43, 46, 47]. On the other hand, analysis across
positional roles showed significant relationships between
match running performance and physical capacity only in
strikers and second strikers (e.g., very high-intensity match
activities vs. Vam-Eval test [71]). Otherwise, nonsignifi-
cant trivial correlations were identified for fullbacks, center
backs, midfielders and wide midfielders [18] (see an exam-
ple of contrasting results in Rago etal. [65]). Similarly, low
explanation power was revealed for several of the aforemen-
tioned tests (i.e., Zig-Zag Test, RAST, YoYo IR-1) used to
predict match running performance (peak game speed, total
distance covered, and percentage at velocity 8–13km/h)
(R2 = 17–22%) [28]. Additionally,
̇
V
O2max, whether esti-
mated or directly determined, was not associated with match
running performance in some papers [29, 43], whereas in
other papers, VO2 kinetics were significantly related to total
distance covered and high-intensity running [47]. Mendez-
Villanueva etal. [14] reported no significant relationships
between match running performance (i.e., differences
between first and second half) and physical capacity (i.e.,
maximal aerobic speed) determined using the Vam-Eval test,
irrespective of playing position. There were also contrasting
examples of correlation outcomes between match running
performance and anthropometric measures. Nonsignificant
(vs. body mass, height, body fat percentage) and weak to
moderate correlations (vs. subscapular and abdominal
L.H.Palucci Vieira et al.
Table 5 Results for measures of match running performance extracted from a literature research of studies on youth soccer playersa
Reference Ag (years)bTD (m) TD (m/min) Vpeak (km/h) ST1 (m) ST2 (m) ST3 (m) ST4 (m) ST5 (m) ST6 (m) ST7 (m) ST8 (m)
Bellistri etal. [20] U8
U10 2229
3541 50
78
1073 (49%)
1249 (36%) 318 (14%)
800 (22%) 440 (20%)
855 (24%) 365 (16%)
556 (16%) 25 (1%)
86 (2%) 836 (39%)
1503 (42%)
Goto etal. [26] U9
U10 4356
4056 77.92
79.63
966 (22%)
865 (21%) 1560 (36%)
1594 (39%) 1189 (27%)
927 (23%) 462 (11%)
485 (12%) 166 (4%)
186 (5%)
Bravo-Sánchez etal. [61] 10.7 85.39 11m/min 5m/min 0.4m/min –
Capranica etal. [37] 11 55% 38% 4% 3% – –
Castagna etal. [24] 11.8 6175 103 1112 (18%) 32 (1%) 217 (4%) 3200 (52%) 986 (16%) 468 (8%) 114 (2%) 9.4%
Stroyer etal. [38] 12.1
12.6
14
9.6%
3.6%
3.1%
63.9%
57.1%
53.8%
19.6%
31.3%
34.0%
6.8%
7.9%
9.0%
Doncaster etal. [47] 12–14 110–116 1207 (26%) 184 (7.9%)
Buchheit etal. [51] 12.8
15.9 5966
7130
24–25
27–28 4866 (82%)
5536 (78%) 607 (10%)
788 (11%) 299 (5%)
420 (6%) 196 (3%)
386 (5%) 1101 (18%)
1593 (22%)
Andrzejewski etal. [23] 13.5 4252 83 23–25 3596 (85%) 656 (15%)
Andrzejewski etal. [55] 13.5 – – 135–196 – – – – –
Atan etal. [16] U13
U14
U15
4516
5385
6600
96.6
95.8
94.3
23.5
25.4
26.5
2205 (5%)
324 (6%)
306 (5%)
747 (17%)
746 (14%)
536 (8%)
491 (33%)
1847 (34%)
2152 (32%)
1413 (31%)
1628 (30%)
2000 (30%)
612 (14%)
835 (16%)
1599 (24%)
4.7 (0.1%)
4.3 (0.08%)
5.5 (0.08%)
Randers etal. [54] U13 2038 102 22.5 18 (1%) 311 (15%) 694 (34%) 527 (26%) 297 (15%) 134 (7%) 54 (3%)
Brito etal. [62] U14 2964 2353 (79%) 318 (11%) 161 (5%) 95 (3%) 271 (9%)
Castagna etal. [39] 14.1 6204 103 508 (8%) 2981 (48%) 1694 (27%) 741(12%) 234 (4%) 975.6 (16%)
Pereira Da Silva etal. [27] U15
U17
U20
7077
8639
9810
118
105
109
48%
44%
46%
33%
35%
32%
4%
6%
6%
5%
5%
6%
4%
4%
4%
2%
2%
3%
3%
3%
3%
0.4%
1%
1%
Buchheit etal. [18] U13
U14
U15
U16
U17
U18
6549
7383
8129
8312
8707
8867
94
105
102
104
109
99
22.3
24.4
26
26.3
26.6
28.3
5370 (82%)
5799 (79%)
6288 (77%)
6480 (78%)
6749 (78%)
6650 (75%)
671 (10%)
821 (11%)
954 (12%)
968 (12%)
991 (11%)
976 (11%)
323 (5%)
446 (6%)
477 (6%)
479 (6%)
519 (6%)
574 (6%)
186 (3%)
318 (4%)
410 (5%)
384 (5%)
449 (5%)
666 (8%)
509 (8%)
763 (10%)
887 (11%)
864 (10%)
967 (11%)
1239 (14%)
Castagna etal. [40] 14.4 6087 101 486 (8%) 3029 (50%) 1630 (27%) 713 (12%) 217 (4%) 930 (15%)
Fernandes-da-Silva etal.
[46]14.5 7159 102.3 1397 (18%) 569 (7%) 1967 (25%)
Waldron and Murphy [33] U14 105–116 26–27 40% 44% 13% 3% 0.3% – –
Goto etal. [15] U11
U12
U13
U14
U15
U16
5800
7700
95
112
2011m/h
2119m/h
2004m/h
1908m/h
1830m/h
1927m/h
2166m/h
2277m/h
2319m/h
2242m/h
2282m/h
2343m/h
1334m/h
1257m/h
1427m/h
1595m/h
1709m/h
1675m/h
349m/h
363m/h
420m/h
515m/h
629m/h
578m/h
29m/h
52m/h
72m/h
118m/h
148m/h
64m/h
Time-Motion Analysis in Youth Soccer
Table 5 (continued)
Reference Ag (years)bTD (m) TD (m/min) Vpeak (km/h) ST1 (m) ST2 (m) ST3 (m) ST4 (m) ST5 (m) ST6 (m) ST7 (m) ST8 (m)
Harley etal. [13] U12
U13
U14
U15
U16
5967
5813
5715
6016
7872
103.7
98.8
106.5
118.7
115.2
1713 (29%)
1756 (30%)
1841 (32%)
1755 (29%)
2481 (32%)
662 (11%)
644 (11%)
748 (13%)
669 (11%)
951 (12%)
174 (3%)
167 (3%)
248 (4%)
194 (3%)
302 (4%)
Arruda etal. [49] 15.1 5485 97–114 27–28 382 (7%)
Aquino etal. [28] 15.2 119 31 0.1% 4.6% 32.6% 28.9% 13.8% 9.4% 10.7% 20.1%
Aquino etal. [32] 15.4 6430 107.33 30–34 0.07–0.10% 6.03–6.71% 39.3–42.7% 27.3–30.2% 14.4–16.8% 7.7–10% 22–27%
Aquino etal. [45] 15.6 6249 104.17 0.08% 6.17% 41.64% 27.98% 15.67% 9.06% 24.73%
Rebelo etal. [43] 15.6 6331 79.14 0 (14%) 2116 (33%) 2025 (32%) 1102 (17%) 529 (8%) 230 (4%) 309 (5%)
Rowsell etal. [52] 15.9 9950 111 1170 (12%)
Mendez-Villanueva etal.
[41]16.7 – 26–31 – – – – – –
Varley etal. [48]cU17 5509
5252
4135 (75%)
4012 (76%) 1375 (25%)
1243 (24%) 458 (8%)
429 (8%) 113 (2%)
109 (2%)
Aughey etal. [30] U17/18 97.26 78% 13% – – –
Vigh-Larsen etal. [67] U17
U19
114.97
117.31
8.2m/min
7.9m/min 1.6m/min
2m/min
Aslan etal. [29]c17.6 5146
4754 114
106
1443 (28%)
1528 (32%) 671 (13%)
596 (13%) 1263 (25%)
1076 (23%) 724 (14%)
630 (13%) 454 (9%)
395 (8%) 591 (11%)
529 (11%)
Pettersen etal. [31] 17.6 9958 111 28 600 (6%) 111 (1%)
Mendez-Villanueva etal.
[14]cU13
U14
U15
U16
U17
U18
4024
3449
4099
3845
4118
3883
4420
4099
4316
4137
4256
4052
2015 (50%)
1741 (50%)
2175 (53%)
2090 (54%)
2194 (53%)
2156 (56%)
2420 (55%)
2250 (55%)
2467 (57%)
2411 (58%)
2505 (59%)
2477 (61%)
792 (20%)
632 (18%)
849 (21%)
717 (19%)
877 (21%)
811 (21%)
963 (22%)
859 (21%)
897 (21%)
830 (20%)
840 (20%)
726 (18%)
613 (15%)
547 (16%)
557 (14%)
519 (13%)
557 (14%)
491 (13%)
575 (13%)
538 (13%)
519 (12%)
472 (11%)
462 (11%)
396 (10%)
349 (9%)
349 (10%)
330 (8%)
321 (8%)
292 (7%)
264 (7%)
311 (7%)
311 (8%)
311 (7%)
283 (7%) 292 (7%)
274 (7%)
255 (6%)
180 (5%)
188 (5%)
198 (5%)
198 (5%)
161 (4%)
151 (3%)
141 (3%)
122 (3%)
141 (3%)
161 (4%)
179 (4%)
L.H.Palucci Vieira et al.
Table 5 (continued)
Reference Ag (years)bTD (m) TD (m/min) Vpeak (km/h) ST1 (m) ST2 (m) ST3 (m) ST4 (m) ST5 (m) ST6 (m) ST7 (m) ST8 (m)
Saward etal. [19] U9
U10
U11
U12
U13
U14
U15
U16
U17
U18
4586
5326
5492
5869
6549
6549
7001
6609
7001
6247
76.43
88.77
91.53
97.82
109.15
109.15
116.68
110.15
116.68
104.12
3620 (79%)
3968 (75%)
4013 (73%)
4118 (70%)
4390 (67%)
4586 (70%)
4888 (70%)
4798 (73%)
4949 (71%)
4330 (69%)
979 (21%)
1326 (25%)
1492 (27%)
1703 (29%)
2126 (32%)
1945 (30%)
2096 (30%)
1809 (27%)
2066 (30%)
1900 (30%)
323.48 (6%)
240.46 (4%)
297.07 (5%)
334.80 (5%)
300.84 (4%)
312.16 (5%)
305.56 (4%)
338.58 (5%)
Al Haddad etal. [17] U13
U14
U15
U16
U17
23.4
25.1
25.6
26.2
26.8
Hunter etal. [44] U18 10296 114.4 83% 9% 7% 1% – – –
Kęsicki and Lewicki [64] 18.1 9213 – 27.2 – – – – – –
Helgerud etal. [59] 18.1 9292 – – – – – – – –
Sporis etal. [50] 18.4 9951 111 5535 (56%) 1603 (16%) 1726 (17%) 684 (7%) 403 (4%)
Garvican etal. [53] 18.8 113.6 86% 16% – – –
Buchheit etal. [58] U17/18 – – 737 – – – – –
Rago etal. [65] U19 8712 495 (6%) 235 (3%) 146 (2%) 876 (10%)
Dello Iacono etal. [63] U19 111.8 4m/min – – – – –
Izzo and Varde’i [66] U20 8464 100 2699 (32%) 2466 (29%) 2079 (25%) 755 (9%) 350 (4%) 115 (1%)
Romagnoli etal. [57] 17–20 11,734 130 9068 (77%) 2665 (23%) 905 (8%)
Randers etal. [56]d19.3 10,830 114 27–36 6230 (58%) 3600 (33%) 2650 (24%) 380 (4%)
ST speed threshold, TD total distance covered, Vpeak peak game speed
a See Table4 to identify the corresponding (fixed) speed threshold reported. In studies where data were presented for different conditions and playing positions, the approximated mean value was
calculated for all conditions/positions. Estimates were also performed regarding the percentage of distance covered at given speed thresholds relative to the total distance covered
b Mean, range or age category
c Results are presented for first (first line) and second (second line) halves
d Results from semi-automatic tracking method
Time-Motion Analysis in Youth Soccer
skinfolds) were found in all player positions when the data
were pooled [50]. In contrast, wingers showed moderate to
large relationships for high-intensity running variables com-
bined with body mass and height [60].
3.4.6 Environmental Constraints
Four studies assessed the possible effects of environment
factors on match running performance in young soccer play-
ers. Locations with moderate (Denver, USA; 1600m) [53]
or high altitude (La Paz, Bolivia; 3600m) [30, 58] had detri-
mental effects on match running performance (e.g., distance
covered per minute, high-intensity activities) compared with
sea level, regardless of where players were born/living. A
single study assessing the effects of pitch surface identified
slightly to moderately greater demands (total distance cov-
ered and very high-intensity running) when competing on
artificial versus natural turf [62].
3.4.7 Congested Match Schedules
Three studies investigated congested match schedules. First,
Buchheit etal. [51] showed that, in post peak height veloc-
ity (PHV) players (15.9 ± 1year), two successive matches
played within 48h resulted in impaired running performance
in the second fixture, whereas this was not the case in pre-
PHV players (12.8 ± 0.6years). Similarly, Rowsell etal.
[52] observed decrements in match running performance
over the course of four matches played in a 4-day period in
players with a mean age of 15.9years. In contrast, Arruda
etal. [49] did not observe differences regarding total dis-
tance covered, distance covered per minute, high-intensity
running efforts and distance, and body-load impacts (which
had higher values in the final than in other matches) in U15
players (mean age 15.1years) across five matches played
over a 3-day competitive period. In contrast, the absolute
frequency of accelerations (> 1.8m/s2) and accelerations
per minute decreased over the course of the competition.
3.4.8 Recovery Methods
Two studies assessed the impact of postmatch recovery strat-
egies on match running performance. Spa treatment (2min
hot shower at 33–43°C + three times [sequence of 2min
sauna at 85–90°C, 2-min jacuzzi/hydromassage at 36 ± 1.5
°C and cold water immersion at ~ 12 ± 1°C]) in post-PHVs
with a mean age of 15.4years [51] or only cold water
immersion (5 × 1min at 10°C) (players with a mean age
of 15.9years) [52] were associated with beneficial effects
on subsequent match running performance output such as
total distance covered, sprinting distance, repeated-sprint
sequences and peak game speed. Additionally, a single pub-
lication evaluated the effects of supplementation on match
running performance in youth soccer. In this sense, caffeine
supplementation (6mg·kg−1) did not enhance running output
in players with a mean age of 17.6years [31].
3.4.9 Comparisons Between Formal Games andOther
Game Types
Two studies compared 11 versus 11 (i.e., formal match-play)
with other game formats (i.e., small-sided games). Capranica
etal. [37] observed no differences on TMA variables, includ-
ing running (forward, backwards and with the ball), walking
(forwards, backwards and sideways) and inactivity (no loco-
motion) between 11 versus 11 (field size 100 × 65m) and
7 versus 7 (60 × 40m) in 11-year-old players. In contrast,
Randers etal. [54] reported that 11 versus 11 (105 × 68m)
resulted in a greater total distance covered, peak game speed
and distances in several speed bands (e.g., > 4km/h) than 8
versus 8 (52.5 × 68m) in a sample of U13 players. Official
U12 7-a-side matches were more demanding (i.e., greater
total, high-intensity, very high-intensity and sprint distances)
than 8-a-side, size-matched (40 × 60m) games. An addi-
tional study comparing official U19 matches and game pro-
file-based training sessions (3 × 8-min bouts including 30s
of physical and technical exercises at 50–75–105% vYoY-
oIR1, followed by 30s of active recovery) revealed lower
total and high-speed distances covered per minute and fewer
high-intensity efforts in the former than in training [63].
3.4.10 Longitudinal Interventions
Four articles were longitudinal intervention studies. After
periodization training (22weeks), emphasizing technical-
tactical ability, U16 players showed an increase in running
performance, mainly in peak game speed and high-inten-
sity activities during simulated matches (players from all
positions grouped together) [32]. Buchheit etal. [42] dem-
onstrated that seasonal changes (~ 7months) in running
performance during friendly matches (i.e., repeated sprint
sequences and repeated high-speed running sequences)
were position dependent in U13–U16s. Andrzejewski etal.
[55] divided players (mean age 13.5years) into two groups
(endurance group ≤ average sprint test performance < speed
group) in which “speed-type” players participated in train-
ing sessions of bouts of 9–17m, whereas “endurance-type”
players ran 4–9m (with 40–60 and 25–50s for recovery,
respectively). After a 6-month macrocycle, the former group
performed more sprints and distance 5m/s during match
play but covered a lower total distance than their endurance-
type peers [55]. Finally, Helgerud etal. [59] reported an
increase in the total distance covered and number of sprints
in players with a mean age of 18.1years after an 8-week
training regimen consisting of regular (match play, technical,
tactical, strength and sprint training) plus aerobic interval
L.H.Palucci Vieira et al.
training. The control group, for whom training involved only
regular and additional technical training, did not experience
similar improvements.
4 Discussion
The purpose of the present analysis was to systematically
review the body of knowledge available on match running
performance measures in young male soccer players. A total
of 50 studies provided reference data from 17 different geo-
graphical locations. Most studies examined official match
play, a broad range of chronological ages and, occasionally,
positional roles. Hence, a reasonable amount of knowledge
exists, although a lack of interventions that directly impacted
training and preparation was evident. This discrepancy may
be due to most of the research conducted being relatively
recent (Fig.2). Nevertheless, the available information con-
tributes to the understanding of game requirements and can
inform training content for physical conditioning sessions
[73]. Here, the strengths and limitations of the current lit-
erature are discussed, and recommendations are made for
further investigation.
4.1 Methodological Quality
Overall, the identified research articles generally presented
high methodological quality ratings (79 ± 13%; Table2). The
questions with the highest mean rating, taking into account
all included studies, were Q1 and Q5, suggesting that the lit-
erature aims and the dependent variables that were analyzed
were generally clearly set out. In contrast, Q4 exhibited the
lowest mean quality score. This highlights that essential
characteristics of the time–motion system/equipment used
were either unaccounted for and/or lacking in the text. This
issue is explored in Sect.4.2.5, where measurement tech-
niques are discussed.
4.2 Research Paradigm
4.2.1 Location ofPlayer Populations
An analysis of the geographical location of populations
showed that almost all continents were represented. The
greater portion of studies concerned European countries—
particularly England—and also Qatar, with a recent trend for
an increase in the number of publications in South America.
No scientific studies were identified from African countries,
clearly indicating a need for research in youth populations
in these countries.
4.2.2 Player Categorization
An issue requiring further debate is the definition of criteria
for categorizing youth players in soccer competitions for
research purposes. As such, the majority of the scientific
evidence observed used comparisons based on chronologi-
cal age cutoff points. Studies in England [13, 15, 19, 26]
and New Zealand [16] used 12-month age bands, according
to the Premier League [74] and Auckland Football Federa-
tion regulations [75], respectively. Studies in the São Paulo
State Championship (i.e., the leading state-level tournament
in Brazil [32]) adopted age banding with 24-month cutoff
points [3, 27, 28], similar to that employed in Italian youth
competitions [20], according to Federazione Italiana Giuoco
Calcio standards [76].
It is important to highlight that, in chronologically age-
matched individuals, those demonstrating advanced biologi-
cal maturity status can have an advantage in anthropometric
(weight and height) and physical ability (aerobic resist-
ance, sprint and jump performance) indicators over their
less mature peers [77]. The same point is valid for match
running performance. Indeed, Buchheit and Mendez-Vil-
lanueva [60] reported slightly to moderately greater peak
game speed, high-intensity actions (> 19km/h) and repeated
high-intensity actions (two > 19-km/h runs interspersed by a
maximum of 60s) in more mature than in less mature play-
ers, both belonging to the same U15 group. Goto etal. [78]
reported similar results in U13/U14 players. Recently, Cum-
ming etal. [79] analyzed the effects of a BioBanded soccer
tournament (i.e., games played by youth [U12–U15] play-
ers having 85–90% of predicted adult stature) and reported
that players had a positive perception toward such practice
compared with usual age group competitions. Early and
late maturing players deemed the matches as more and less
physically challenging, respectively. Future studies are nec-
essary to assess match running performance using TMA in
BioBanded players.
4.2.3 Match Type andCongurations
Simulated soccer matches can generate physiological
responses that differ from those in official competition.
However, no studies have directly compared match running
measures among different soccer match play modalities
(e.g., simulated, friendly, preseason tournament vs. official).
Freitas etal. [80] reported that psychophysiological stress
was greater during official competition than during a simu-
lated game, and a lower reduced internal workload (e.g.,
session rating of perceived exertion) was also observed in
the latter. Hence, caution is necessary when using running
data from simulated and friendly matches to determine com-
petition demands and inform physical preparation regimens.
Time-Motion Analysis in Youth Soccer
In general, standardization was lacking across study
designs in age-matched players, particularly with respect to
game configurations. For example, pitch sizes, game dura-
tions, division of playing periods, the number of players per
side, and rolling substitute policy application varied consid-
erably (see Table3) [3]. Variations in the aforementioned
parameters have been shown to influence on-field running
performance in soccer match play [54, 81, 82]. As such,
conditioning practitioners working in youth soccer should
consider that data published for a given population may not
be pertinent for another.
4.2.4 Speed Thresholds
The absence of standardization in speed thresholds (see
Table4) across the literature makes it difficult to compare
findings on running output [44]. Nevertheless, the use of
fixed speed thresholds does provide useful information
regarding player development (e.g., comparisons between
age groups) and the effects of training on match running
performance [15, 32, 42]. In addition, fixed speed thresh-
olds allow direct outcome comparisons between studies
[18]. However, several authors suggest that individualizing
speed thresholds provides a more accurate representation of
match running loads in young soccer players [13, 16, 44].
This practice can also aid individualized assessments and
comparisons between players of differing maturational lev-
els [16] and management of external workloads through the
design of appropriate recovery and periodization schedules
[44].
The majority of studies employed field testing proce-
dures, such as linear sprint speed [1317, 19, 20, 26, 47], as
a means to determine age-specific or individualized speed
thresholds. Yet, recent evidence suggests that maximal
sprinting speed has limitations for establishing game speed
thresholds for several reasons: (1) peak game speed values
might exceed those derived from tests of maximal sprinting
speed [83]; (2) individualized game speed thresholds do not
enhance the dose-response determination to soccer training
[44, 84]; and (3) seasonal changes in match running perfor-
mance do not necessarily match those observed in sprinting
test performance; players with decreased maximal sprinting
speed demonstrated concomitantly increased match running
performance (e.g., number of repeated sprint sequences and
repeated-high speed sequences) [42]. Although two of the
above-cited studies were not specifically performed in youth
[83] and males [84], further investigations are required to
assess the relevance of creating speed thresholds from game
parameters per se to depict match running performance. A
comparison of speed thresholds as a percentage of peak
game speed and distances covered in fixed speed thresholds
relative to total distance covered [3, 63], rather than using
fitness testing performance to determine speed thresholds,
would be useful.
4.2.5 Measurement Techniques
A key study by Randers etal. [56] compared data derived
from GPS-based technologies, VTM and computational
videogrammetry tracking techniques. Large between-sys-
tem differences were present in the determination of the
absolute distances covered, meaning that results between
match analysis systems should be compared with caution
[56]. Indeed, a wide variety of data acquisition methods was
employed across studies (see Table3). GPS is considered the
most time-efficient method to collect and report match data
in contemporary soccer [85]. Yet, most studies (42%) used
low sampling frequency GPS devices (i.e., 1–5Hz), which
exhibit consequential error rates in determining high-speed
activity [86]. Computer-based tracking adopted in investi-
gations in youth players (e.g., Prozone® [48, 87], Mathball
Match Analysis System [29] and DVIDEOW™ [28, 32,
45]) show low absolute error [3, 21, 29, 88]. Indeed, when
compared with GPS and local position measurement, vide-
ogrammetry tracking methods are shown to have the most
constant magnitude of error in computing distances when
running occurs at low- and high-intensity speed thresholds
[85]. However, logistical constraints can favor the use of
GPS, especially when youth matches take place in training
ground facilities [3, 13, 48].
Specific issues affect TMA data collection: (1) for GPS:
the environment (e.g., topography), number of satellites
connected and software/unit updates; and (2) for computer-
based tracking: lighting, background objects, camera posi-
tion and calibration quality. All these issues can interfere
with data signal [85, 86, 8991], and measurement error
should ideally be calculated in the locations where matches
are specifically played, a factor that was systematically over-
looked in the current literature (see Table2).
4.2.6 Terminology Issues
A total of 20 descriptors were adopted among studies
included in the current systematic review to define and group
dependent variables (e.g., total distance covered, mean speed
or distance covered per time, peak game speed, indicators
of activities performed at established speed thresholds or
movement categories). Thus, there is a need to standardize
terminology in this research area. To make it easier for read-
ers to identify evidence related to this research area topic in
the future, we suggest the authors simply use “match run-
ning performance”, as this was the most cited term—used
approximately 26% of the time—in the published articles.