Differences in strength and speed demands between 4v4 and 8v8 SSG

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rjsp20
Download by: [University of Nebraska, Lincoln] Date: 09 June 2016, At: 05:12
Journal of Sports Sciences
ISSN: 0264-0414 (Print) 1466-447X (Online) Journal homepage: http://www.tandfonline.com/loi/rjsp20
Differences in strength and speed demands
between 4v4 and 8v8 small-sided football games
António Natal Campos Rebelo, Pedro Silva, Vincenzo Rago, Daniel Barreira &
Peter Krustrup
To cite this article: António Natal Campos Rebelo, Pedro Silva, Vincenzo Rago, Daniel Barreira
& Peter Krustrup (2016): Differences in strength and speed demands between 4v4 and 8v8
small-sided football games, Journal of Sports Sciences, DOI: 10.1080/02640414.2016.1194527
To link to this article: http://dx.doi.org/10.1080/02640414.2016.1194527
Published online: 09 Jun 2016.
Submit your article to this journal
View related articles
View Crossmark data

Differences in strength and speed demands between 4v4 and 8v8 small-sided
football games
António Natal Campos Rebeloa
, Pedro Silvaa,b
, Vincenzo Ragoa
, Daniel Barreiraa
and Peter Krustrupc,d
a
Centre of Research, Education, Innovation and Intervention in Sport, Faculty of Sports, University of Porto, Porto, Portugal; b
FC Zenit, St.
Petersburg, Russia; c
Department of Nutrition, Exercise and Sports, Section of Human Physiology, University of Copenhagen, Copenhagen,
Denmark; d
Sport and Health Sciences, College of Life and Environmental Sciences, University of Exeter, Exeter, UK
ABSTRACT
The aims of this study were (i) to characterise the acceleration demands of two different formats of
small-sided game (SSG), i.e., 4v4 + goalkeepers (4v4 + GK) and 8v8 + goalkeepers (8v8 + GK); (ii) to
analyse the correlation between performance in power-based tests and acceleration-based physical
loading during the two different SSG formats and (iii) to analyse the neuromuscular-induced fatigue.
Eighteen adult male footballers participated in the study (20.7 ± 1.0 years, 178 ± 5 cm and
71.4 ± 2.1 kg). Baseline measurements were obtained from countermovement jumps, 15 s repeated
jumps and 5 and 15 m sprints. A total of 36 min was analysed for each SSG (4v4 + GK: two sets of
3 × 6 min, and 8v8 + GK: 2 × 18 min). Heart rate, blood lactate, perceived exertion and movement
pattern (GPS) were analysed. Distances covered by very-high-intensity activities and very-high-speed
running were lower in 4v4 + GK than in 8v8 + GK (effect sizes (ES) = −0.69 ± 0.67 and −1.04 ± 0.67,
respectively; very likely), while accelerations and decelerations were higher in 4v4 + GK than in 8v8 + GK
(ES = 1.13–1.52; almost certainly). Blood lactate concentrations were higher (ES = 1.40 ± 0.58; almost
certainly) and players perceived themselves to be more tired (ES = 0.80–2.31; almost certainly) after 4v4
+ GK than after 8v8 + GK. Sprint ability in 5 and 15 m tests decreased (ES = 0.87 ± 0.58 and 0.89 ± 0.58,
respectively; almost certainly) only after 4v4 + GK. This SSG format appeared more demanding in
relation to repetitions and fatigue development of muscle power-based actions than 8v8 + GK. It may
therefore be logical to use the former type of SSG to target development of power-related football
actions.
ARTICLE HISTORY
Accepted 16 May 2016
KEYWORDS
Time-motion; acceleration
demands; fatigue; technical
skills
Introduction
Small-sided games (SSGs) have been extensively used in training
footballers worldwide and have shown very good efficacy in
improving player performance (Hill-Haas, Dawson, Impellizzeri,
& Coutts, 2011). As an example, it has been shown that the
technical performance (Owen, Wong del, McKenna, & Dellal,
2011) and physical performance (Chaouachi et al., 2014; Dellal,
Varliette, Owen, Chirico, & Pialoux, 2012) of footballers can be
enhanced using SSG-based football training programmes.
In the last two decades, extensive research has been pub-
lished on physical and physiological response during SSGs in
football (for refs, see Halouani, Chtourou, Gabbett, Chaouachi,
& Chamari, 2014). It was found that the time-motion charac-
teristics of SSGs could vary greatly depending on certain
structural (e.g., pitch size, number of players, type and number
of goals) and rule (e.g., number of ball touches) constraints.
For example, it was observed that higher maximum speeds are
reached during SSGs played on bigger pitches (Casamichana &
Castellano, 2010). Furthermore, heart rate (HR) and lactate
concentrations were shown to be sensitive to structural and
rule changes in SSGs.
With regard to time-motion characteristics, running speed
has been the most extensively studied variable. However,
recent studies have shown that, in order to better characterise
the physiological loading during SSGs, acceleration analysis is
of major importance. In fact, the physiological strain induced
by acceleration and deceleration movements has a pro-
nounced impact on both energetic and neuromuscular sys-
tems (Gaudino, Alberti, & Iaia, 2014a; Gaudino et al., 2014b).
Thus, when seeking to analyse the physical demands of dif-
ferent SSG formats, acceleration characterisation is strongly
recommended. Although the value of GPS technology in
determining accelerations in the field appears questionable
(Buchheit et al., 2012), acceptable values of validity and relia-
bility in the measurement of constant velocity would be
expected when higher sample rates are used (Varley,
Faiweather, & Aughey, 2013).
It has already been observed that the physical fitness of
football players can be correlated with physical performance
during official matches (Manzi, Impellizzeri, & Castagna, 2014;
Rebelo, Brito, Seabra, Oliveira, & Krustrup, 2014). However,
little (or nothing) is known about the correlation between
the physical fitness of players and their physical and technical
performance in SSGs. This is important information for coa-
ches as the physical loading of the training session should be
individually designed to improve performance and avoid
excess of fatigue and overtraining. Intriguingly, there is limited
scientific data on the effects of SSGs versus other training
CONTACT António Natal Campos Rebelo anatal@fade.up.pt Faculdade de Desporto, Universidade do Porto, R. Dr. Plácido Costa, 91, Porto 4200 – 450, Portugal
JOURNAL OF SPORTS SCIENCES, 2016
http://dx.doi.org/10.1080/02640414.2016.1194527
© 2016 Informa UK Limited, trading as Taylor & Francis Group
Downloadedby[UniversityofNebraska,Lincoln]at05:1209June2016

methods (e.g., high-intensity interval training) on fatigue
(Faude, Steffen, Kellmann, & Meyer, 2014). Moreover, analysis
of fatigue induced by different SSGs is extremely scarce (Katis
& Kellis, 2009).
Given the above, the aims of this study were (i) to char-
acterise the acceleration and speed demands of two different
SSG formats (4v4 + goalkeepers (GK) and 8v8 + GK); (ii) to
analyse the effects of each type of SSG on neuromuscular
performance and (iii) to analyse the accumulation of fatigue
during the SSGs.
Methods
Participants
Eighteen adult male footballers were invited to participate in
the study. Their age, height and weight were 20.7 ± 1.0 years,
178 ± 5 cm and 71.4 ± 2.1 kg. The participants were college
students representing different local football clubs. All the
players were interviewed to determine how long they had
been playing football and how many hours a week they
trained. The players had been registered with local football
clubs for 6.0 ± 2.5 years, the last 2‒3 years at semi-professional
level. They had four training sessions and one official match
per week, resulting in a total training time of 7.1 ± 0.7 h per
week. When questioned about familiarity with SSGs, the parti-
cipants reported that they had frequently been exposed to
SSG formats similar to those used in the study. All the players
were informed about the protocol and signed an informed
consent form before the investigation. The Scientific Board of
the Faculty of Sport at the University of Porto, Portugal,
approved the design of the study.
Experimental design
The participants were evaluated during two different SSGs:
4v4 + GK and 8v8 + GK. In order to analyse SSG-induced
fatigue, baseline measurements were obtained from physical
tests performed 1 week before the SSG assessments. On two
separate days, the participants completed tests comprising
countermovement jumps (CMJs; Digitime 1000, Digitest,
Finland), 5 and 15 m sprints (Speed Trap II; Brower Timing
System, Utah, USA) and quadriceps strength tests (Tempo
Technologies, Globus Ergometer). HR, blood lactate, time-
motion, technical actions and rate of perceived exertion
(RPE) were obtained during SSGs.
Following baseline testing, the participants were ran-
domly allocated to four teams of four players. At baseline,
teams were balanced in terms of the players’ jump and
sprint performance (CMJ: 0.43 ± 0.03, 0.38 ± 0.04,
0.42 ± 0.05 and 0.42 ± 0.06 m; 5 m sprint: 1.12 ± 0.06,
1.14 ± 0.06, 1.14 ± 0.06 and 1.10 ± 0.06 s; 15 m sprint:
6.09 ± 0.20, 6.13 ± 0.20, 6.14 ± 0.30 and 6.15 ± 0.28 s;
P > 0.05). All games were played outdoors between 9 and
11 am in dry conditions. Ambient temperature (16–20°C)
and humidity (60–75%) were monitored. All SSGs were
played on a third-generation artificial pitch developed espe-
cially for football consisting of long and widely spread fibres
of polypropylene filled with rubber granules. The two SSG
sessions took place in a random order with an interval of
72 h. A total of 36 min was analysed for each SSG (4v4 + GK:
two sets of 3 × 6 min and 8v8 + GK: 2 × 18 min). The games
included a 3 min rest break between repetitions and a 5 min
rest break between the two sets of 4v4 + GK. The pitch size
was adapted to each SSG format (4v4 + GK: length 47.72 m,
width 29.54 m; 8v8 + GK: length 85.90 m, width 53.18 m).
The goals were 7 × 2 m (width × height). Minor rule mod-
ifications were applied, such as no offside, restart of the
game after a goal by the goalkeeper and kick-in awarded
to the opposing side to that of the player who last touched
the ball. Games were played without referees or sideline
encouragement. The scores of the games were prospectively
recorded by one of the researchers. Six extra footballs were
always available near the goals and at the side of the pitch
to facilitate a quick restart when the ball left the playing
area. The chief investigator was always available to immedi-
ately replace the ball when it was kicked out. All players
were encouraged to drink water before the start of the
game, during the breaks and after the last game prior to
the sprint, jump and strength tests.
Post-game measurements were obtained immediately after
each SSG session.
Testing
Sprint times were determined by a flat sprint test performed in
a straight 20 m line. The times were measured by three pairs
of photoelectric cells positioned at the starting point and at 5
and 15 m. The participants were instructed to run as fast as
possible from a standing start 30 cm behind the starting line.
Baseline and post-game sprint tests were performed outdoors
on a tartan track located within 20 m of the pitch.
For the CMJ test, the participant stood upright, bent the
knees to the squat position and jumped vertically as high as
possible keeping hands on hips and landing with straight
knees on the mat. The flight time was used to calculate the
change in the height of the body’s centre of gravity (Bosco,
Luhtanen, & Komi, 1983).
Maximal voluntary isometric torque of the quadriceps with
knees positioned at 90° of flexion was measured using an
isometric loading cell. After a warm-up set of five submaximal
repetitions of knee extension at the stated angle, the partici-
pants completed two maximal repetitions separated by 60 s of
rest. The participants received verbal encouragement and the
best result of the two was used in further analysis.
Heart rate measurements
HR was recorded at 5 s intervals by short-range radio teleme-
try (Polar Team System, Kempele, Finland). To reduce HR
recording errors, the players were asked to check their HR
monitors before each game. The monitors were attached to
the participants using an adjustable elastic chest strap. Data
were transferred to a computer using the Polar Precision
Performance 4.03 software program (Polar Team System).
Mean HR values during the games were expressed in absolute
values (beats per min).
2 A. N. C. REBELO ET AL.

Blood lactate
Blood lactate was collected as a physiological indicator of the
contribution of anaerobic glycolysis during exercise. Blood lac-
tate was measured via 5 μl capillary blood samples taken from an
earlobe (Lactate Pro, Arkray Inc., Japan). The analyser was cali-
brated according to precision standards and routinely assessed
by external quality controls. The accuracy and reliability of the
blood lactate analyser has been assessed elsewhere at different
lactate concentrations (coefficient of variation ranged between
2.8% and 5.0%) and between test strips (intraclass correlation
coefficient, r = 0.999) (Baldari et al., 2009). The participants were
asked to temporarily leave the pitch (±30 s) to sit on a chair
positioned 2 m outside the sideline for collection of the blood
samples. One researcher carefully cleaned, disinfected and dried
the participant’s earlobe before blood collection in order to avoid
any possible interference as a result of sweat and dirt. The
participant’s skin was then punctured with a lancet and the first
drop of blood placed straight onto the strip. One blood sample
was taken from each participant in each SSG repetition in a
random order. All samples were collected 1‒2 min after game
activity and analysed within a few seconds of collection
(Taoutaou, Granier, Mercier, Ahmaidi, & Prefaut, 1996). All sam-
ples were collected during SSGs and analysed within a few
seconds of collection. An outfield substitute replaced the parti-
cipant from whom the blood sample was being collected.
Time-motion analysis
Movement pattern during the games was measured using unob-
trusive portable global positioning system (GPS) units (GPSports
SPI Elite, Canberra, Australia). Based on signals from at least three
Earth-orbiting satellites, the GPS receiver recorded the players’
positional data (x- and y-) with a time resolution of 15 Hz
(obtained through interpolation of a 5 Hz signal). The system
used the GPS Doppler data and distances were calculated from
changes in position and subject to the manufacturer’s proprie-
tary algorithm integrated to reduce measurement error. The data
from each receiver were treated and extracted using proprietary
software (GPSports team AMS v1.2.1.12, Canberra, Australia). The
software calculated the total distance covered during a game,
the average speed and maximum speed achieved during the
game, and the time spent in five locomotor categories defined in
advance in accordance with Buchheit, Mendez-Villanueva,
Simpson, and Bourdon (2010). For data analysis purposes, the
following locomotor categories were chosen: low-intensity run-
ning (LIR) (0–13 km ∙ h−1
), high-intensity running (13.1‒16 km ∙
h−1
), very-high-intensity running (16.1–19 km ∙ h−1
), very-high-
intensity activities (VHIA) (16.1‒19 km ∙ h−1
) and very-high-speed
running (VHSR) (>19.1 km ∙ h−1
). The time spent in each locomo-
tor category during the games was presented as a percentage of
total playing time. The GPS software calculated the frequency
and duration of high-intensity running and VHSR in relation to
the specific speed categories.
Accelerations and decelerations
Acceleration and deceleration distances were captured using
GTX3 accelerometers (100 Hz, ActiGraph, version 4.4.0). The
raw data of the accelerometers and the coordinates of the GPS
system in the x- (longitudinal) and y- (lateral) directions were
extracted and matched temporally for each game (using satel-
lite time, recorded in both the GPS and the accelerometers).
The speed of each player was computed (using GPS coordi-
nates) to assess whether the player was accelerating or decel-
erating. Acceleration and deceleration data were then
calculated in m∙s−2
using the accelerometers’ g forces in the
longitudinal (gx) and lateral (gy) directions through the follow-
ing formula –
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
gx2 Â gy2
p
. Finally, all acceleration distances
were measured according to type (acceleration or decelera-
tion) and category (2–3 m∙s−2
and >3 m∙s−2
). All computations
were made using MATLAB routines (R2011a, MathWorks, USA).
Rating of perceived exertion
Perceived exertion, physical effort and fatigue state were
recorded using a visual analogue scale (VAS) questionnaire
(Andersson, Ekblom, & Krustrup, 2008). The questionnaire
comprised five questions worded to analyse the players’
perceptions of playing each of the two SSG formats. The
questions were as follows: “How do you feel after today’s
training session?” (not tired–very tired; VAS1); “How hard phy-
sically was today’s SSG?” (not hard–very hard; VAS2); “In terms
of strength, how demanding was today’s SSG?” (not very
demanding–very demanding; VAS3); “In terms of endurance,
how demanding was today’s SSG?” (not very demanding–very
demanding; VAS4); “In terms of sprinting, how demanding
was today’s SSG?” (not very demanding–very demanding;
VAS5). The VAS was anchored by two items with a 100 mm
horizontal line connecting them and scored from 0 (not tired/
not hard/not very demanding) to 100 (very tired/very hard/very
demanding), but the participants were unaware of the
numbers.
Technical actions
The SSGs were videoed (DCR-HC53E, Sony, Tokyo, Japan), and
defensive and offensive technical skills were measured:
tackles, ball interceptions, defensive duels, passes, offensive
duels, runs with the ball, shots and goals. Games were filmed
with cameras positioned on a platform positioned at the stand
on the side of the pitch at the halfway line, at a height of 5 m
and about 5 m behind the sideline. Two experienced obser-
vers analysed all SSG videos. Data quality was ensured by
assessing intra- and inter-observer reliability. Specifically,
Cohen’s kappa index (Cohen, 1960) was calculated from the
records of two experienced observers. Results yielded values
of 0.88‒0.98, well above the value of 0.75 established as being
indicative of high data quality.
Statistics
Differences between 4v4 and 8v8 in (i) technical, physical and
physiological demands and (ii) post tests and baseline were
analysed using magnitude-based inferences (Hopkins, Marshall,
Batterham, & Hanin, 2009). Between-treatment effect sizes (ES)
with 90% confidence intervals (CIs) were calculated using pooled
standard deviations. Threshold values for Cohen’s ES were >0.2
JOURNAL OF SPORTS SCIENCES 3

(small), <0.6 (moderate) and >1.2 (large) (Cohen, 1992).
Probabilities were calculated to assess whether true effects
obtained represented substantial changes (Batterham &
Hopkins, 2006). The smallest standardised change for each vari-
able was considered to be 0.2 multiplied by the between-subject
standard deviation value, based on Cohen’s ES principle.
Quantitative probabilities of higher or lower differences were
evaluated qualitatively as <1% = almost certainly not,
1–5% = very unlikely, 5–25% = unlikely, 25–75% = possibly,
75–95% = likely, 95–99% = very likely, >99% = almost certainly.
If the probabilities of the effect being higher or lower than the
smallest worthwhile difference were simultaneously >5%, the
effect was deemed unclear. Otherwise, the effect was clear and
reported as the magnitude of the observed value. The tendency
of physical and technical parameters during the different stages
of each format (4v4 + GK or 8v8 + GK) were analysed through
linear regression slopes (describing tendency).
Pearson correlation coefficients (r, with 90% confidence inter-
vals) were calculated to verify the association between variables.
To interpret the magnitude of the correlation coefficients, the
following criteria were used: <0.10 trivial, 0.10 < r < 0.30 small,
0.30 < r < 0.50 moderate, 0.50 < r < 0.70 large, 0.70 < r < 0.90 very
large and 0.90 < r < 1 almost perfect (Hopkins, Marshall,
Batterham, & Hanin, 2009). Practical inferences of the correlation
coefficients were also considered. The default threshold of r was
considered 0.1 based on Cohen’s assumption of the smallest
clinically important correlation. Data analysis was performed in
SPSS statistical software (version 23, IBM SPSS Statistics, Chicago,
IL, USA) and a modified statistical Excel spread sheet (Hopkins,
2007).
Results
Time-motion analysis, physiological demands and
perceived exertion
Between-game differences (i.e., 4v4 + GK vs. 8v8 + GK) in distance
covered by VHIA and VHSR were very likely (4/0/96; 0.9/0/99,
respectively) moderate (ES = −0.69 ± 0.67 and −1.04 ± 0.67;
Table 1). Very likely (96/0/4) moderate (ES = 0.67 ± 0.67) differ-
ences were observed in accelerations of 2‒3 m∙s−2
, while decel-
erations of 2‒3 m∙s−2
, >3 m∙s−2
and total decelerations were
almost certainly (99.9/0/0.1; 99.9/0/0.1; 99.5/0/0.5) moderately
to largely (ES = 1.52 ± 0.67, 1.13 ± 0.67 and 1.44 ± 0.67) different.
Blood lactate differences between games were almost cer-
tainly (99.9, 0, 0) large (ES = 1.40 ± 0.58).
Differences in RPE (VAS1, VAS2, VAS3 and VAS4) during
games were almost certainly (99.9/0/0; 99.6/0/0.3; 99.7/0/0.3
Table 1. Overall comparison of 4v4 + GK versus 8v8 + GK reported as distance covered per minute.
4v4 + GK 8v8 + GK 4v4 + GK vs 8v8 + GK
Mean ± SD Mean ± SD SMD (90% CL) Uncertainty in true differences
Locomotor activities
LIR (m) 87.11 ± 4.43 84.18 ± 12.61 0.29 (−0.37; 0.96) Unlikely (15.8, 83.4, 0.8)
HIR (m) 11.21 ± 1.78 11.88 ± 3.70 −0.22 (−0.89; 0.44) Unclear (11.8, 34.8, 53.4)
VHIR (m) 4.96 ± 0.82 5.45 ± 2.12 −0.29 (−0.96; 0.37) Unclear (12.9, 16.8, 70.3)
VHIA (m) 8.04 ± 1.80 9.96 ± 3.33 −0.69 (−1.36; −0.02) Very likely (4, 0, 96)
VHSR (m) 2.92 ± 1.00 4.35 ± 1.59 −1.04 (−1.71; 0.36) Very likely (0.9, 0, 99)
TD (m) 107.07 ± 5.47 107.01 ± 18.68 0.00 (−0.66; 0.67) Almost certainly not (0.8, 98.5, 0.7)
Acc 2‒3 m∙s−2
(m) 4.81 ± 0.44 4.17 ± 1.21 0.67 (0.005; 1.34) Very likely (96, 0, 4)
Acc >3 m∙s−2
(m) 2.36 ± 0.37 2.00 ± 0.84 0.53 (−0.13; 1.20) Unclear (93.4, 0, 6.6)
Acc tot (m) 28.81 ± 5.36 29.91 ± 1.81 −0.26 (−0.93; 0.40) Possibly (4.8, 37.5, 57.7)
Dec 2‒3 m∙s−2
(m) 5.28 ± 0.61 4.09 ± 0.87 1.52 (0.85; 2.20) Almost certainly (99.9, 0, 0.1)
Dec >3 m∙s−2
(m) 2.52 ± 0.41 1.88 ± 0.66 1.13 (0.46; 1.80) Almost certainly (99.9, 0, 0.1)
Dec tot (m) 32.63 ± 1.93 28.96 ± 2.88 1.44 (0.77; 2.11) Almost certainly (99.5, 0, 0.5)
Physiological variables
BLa (mmol · l−1
) 6.56 ± 1.23 4.59 ± 1.48 1.40 (0.82; 1.99) Almost certainly (99.9, 0, 0)
HR (beats ∙ min−1
) 170.2 ± 2.0 169.6 ± 3.5 0.08 (−0.49; 0.66) Likely (91.7, 0, 8.3)
Subjective ratings
VAS1 (au) 238.5 ± 43.5 177.3 ± 47.2 2.31 (0.73; 1.89) Almost certainly (99.9, 0, 0)
VAS2 (au) 242.3 ± 59.9 175.5 ± 50.6 1.17 (0.59; 1.75) Almost certainly (99.6, 0, 0.3)
VAS3 (au) 228.4 ± 54.2 162.0 ± 60.2 1.12 (0.54; 1.71) Almost certainly (99.7, 0, 0.3)
VAS4 (au) 245.9 ± 67.7 194.6 ± 56.5 0.80 (0.22; 1.38) Almost certainly (98.6, 0.1, 1.3)
VAS5 (au) 197.3 ± 61.3 196.9 ± 53.7 0.00 (−0.57; 0.58) Almost certainly not (0, 100, 0)
Technical actions
Tackles (n) 4.2 ± 2.9 2.9 ± 2.23 0.77 (0.17; 1.37) Very likely (98.2, 0, 1,7)
Ball interceptions (n) 6.3 ± 1.9 3.3 ± 1.53 1.70 (1.10; 2.30) Almost certainly (100, 0, 0)
Defensive duels (n) 23.3 ± 7.7 12.2 ± 9.21 1.26 (0.66; 1.86) Almost certainly (100, 0, 0)
Passes (n) 43.2 ± 12.5 22.4 ± 6.8 2.00 (1.40; 2.60) Almost certainly (99.9, 0.1, 0)
Offensive duels (n) 7.7 ± 5.7 4.2 ± 4.4 0.67 (0.07; 1.27) Almost certainly (99.4, 0, 0.5)
Runs with the ball (n) 8.5 ± 4.4 3.6 ± 2.1 1.39 (0.79; 1.99) Almost certainly (100, 0, 0)
Shots (n) 5.1 ± 3.0 1.5 ± 1.7 1.45 (0.85; 2.05) Almost certainly (100, 0, 0)
Goals (n) 1.9 ± 1.3 0.2 ± 0.4 1.77 (1.17; 2.37) Almost certainly (100, 0, 0)
Threshold values for effect sizes were <0.2 (trivial), 0.2‒0.6 (small), 0.6‒1.2 (moderate), >1.2 (large).
Quantitative probabilities of higher or lower differences were evaluated qualitatively as: <1% = almost certainly not, 1–5% = very unlikely, 5–25% = unlikely,
25–75% = possibly, 75–95% = likely, 95–99% = very likely, >99% almost certainly.
Acc tot: total distance covered by accelerations; Acc 2‒3: distance covered by accelerations of 2‒3 m∙s−2
; Acc >3: distance covered by accelerations greater than
3 m∙s−2
; BLa: blood lactate; Dec tot: total distance covered by decelerations; Dec 2‒3: distance covered by decelerations of 2‒3 m s−2
; Dec >3: distance covered by
decelerations greater than 3 m s−2
; HR: heart rate; HIR: high-intensity running; LIR: low-intensity running; TD: total distance covered; VAS: visual analogue scale;
VAS1: “How do you feel after today’s training session?” (not tired–very tired); VAS2: “How hard physically was today’s SSG?” (not hard–very hard); VAS3: “In terms
of strength, how demanding was today’s SSG?” (not very demanding–very demanding); VAS4: “In terms of endurance, how demanding was today’s SSG?” (not
very demanding–very demanding); VAS5: “In terms of sprinting, how demanding was today’s SSG?” (not very demanding–very demanding); VHIR: very-high-
intensity running; VHIA: very-high-intensity activities; VHSR: very-high-speed running.

and 98.6/0.1/1.3) moderate to large (ES = 2.31 ± 1.58;
1.17 ± 0.58; 1.12 ± 0.58 and 0.80 ± 0.58).
Technical-action differences between games were very
likely or almost certainly moderate to large (ES = 0.67–2.00).
Changes in physical performance throughout game
repetitions
Total distance and distance covered by LIR and by accelera-
tions and decelerations tend to decrease throughout repeti-
tions of 4v4 + GK, as can be depicted from the negative slopes
represented in Figure 1. In 8v8 + GK, total distance
and distance covered in LIR also showed a tendency to
decrease (more markedly than in 4v4 + GK) between the
two repetitions.
Differences in physical capacity between baseline and
post SSG
Differences in repeated jump ability (15″ RJ) and in sprint
times (5 and 15 m) between baseline and after 4v4 + GK
were almost certainly (100/0/0; 99.1/0/0.9 and 99.8/0/0.1,
respectively; Table 2) small to moderate (ES = −0.38 ± 0.58,
0.87 ± 0.58 and 0.89 ± 0.58, respectively). Repeated jump
ability after 8v8 + GK was almost certainly (100/0/0) moder-
ately (ES = 0.78 ± 0.58) different from baseline measurements.
Relationship between neuromuscular testing and
performance in 4v4 + GK and 8v8 + GK
Positive moderate correlations between CMJ performance
and distance covered in VHIA and VHSR in 4v4 + GK
(r = 0.51 ± 0.33 and r = 0.55 ± 0.32, respectively; Table 3) and
between repeated jump ability and VHSR (r = 0.54 ± 0.32, CI
(90%)) were very likely (95.2/3.9/0.8; 96.9/2.6/0.5 and 96.5/2.9/
0.6). A positive moderate correlation between 15 m sprint per-
formance and distance covered by accelerations >3 m∙s−2
was
very likely (r = 0.52 ± 0.33; CI (90%); 95.7/3.6/0.7).
Discussion
In this study, we aimed to characterise acceleration and VHSR
demands as well as the neuromuscular fatigue induced by two
commonly used SSG formats in football training, i.e., 4v4 + GK
and 8v8 + GK, and to analyse the link between the neuromus-
cular performance of players and acceleration- and high-speed-
based physical loading during these SSGs. 4v4 + GK appeared
more demanding in relation to repetitions and fatigue develop-
ment of muscle-based actions than 8v8 + GK based on higher
anaerobic energy turnover, perceived exertion and number of
technical actions. Moreover, jump ability and repeated jump
performance in specific physical tests correlated with high-inten-
sity running in 4v4 + GK, while 15 m sprint performance corre-
lated with distance covered by accelerations during 8v8 + GK.
20
a) b)
19
18
17
16
15
14
13
12
8
7
6
5
4
0’-6’
40
45
50
60
55
30’-36’ 18’-36’0’-18’
SLOPE = –0.12
SLOPE = –0.12
SLOPE = –0.36
SLOPE = –0.39
LIR (m)
TD (m)
LIR (m)
TD (m)
Acc tot (m)
Dec tot (m) SLOPE = –2.38
SLOPE = –3.43
Time playing Time playing
Distancecoveredperminute(m)
Distancecoveredperminute(m)
24’-30’18’-24’12’-18’6’-12’
Figure 1. Changes in physical performance throughout 4v4 + GK (a) and 8v8 + GK (b).
Acc tot: total distance covered by accelerations; Dec tot: total distance covered by decelerations; LIR: low-intensity running; TD: total distance covered.
Table 2. Comparison of neuromuscular performance between baseline values and post SSGs.
Baseline Post 4v4 + GK Post 8v8 + GK
Mean + SD Mean + SD SMD ± 90% CI
Uncertainty in
true differences Mean ± SD SMD ± 90% CI
Uncertainty in
true differences
CMJ (cm) 42.88 ± 11.9 38.84 ± 5.2 0.45 (−0.13; 1.03) Likely (95.9, 0, 4.1) 39.46 ± 5.69 0.37 (−0.20; 0.95) Likely (93, 0, 7)
15″ RJ (num) 17.37 ± 2.6 18.31 ± 2.15 −0.38 (−0.96; 0.19) Almost certainly
(100, 0, 0)
19.28 ± 2.11 −0.78 (−1.36; −0.20) Almost certainly
(100, 0, 0)
5 m sprint (m∙s) 4.8 ± 0.28 4.08 ± 1.073 0.87 (0.29; 1.46) Almost certainly
(99.1, 0, 0.9)
4.76 ± 0.2 0.09 (−0.48; 0.67) Unclear
(62.3, 3.5, 34.2)
15 m sprint (m∙s) 6.14 ± 0.27 5.92 ± 0.224 0.89 (0.31; 1.47) Almost certainly
(99.8, 0, 0.1)
6.09 ± 0.201 0.21 (−0.36; 0.79) Unclear
(85.2, 0, 14.7)
Data are presented as means ± SD.
Threshold values for effect sizes were <0.2 (trivial), 0.2–0.6 (small), 0.6‒1.2 (moderate), >1.2 (large). Quantitative probabilities of higher or lower differences were
evaluated qualitatively as <1% = almost certainly not, 1–5% = very unlikely, 5–25% = unlikely, 25–75% = possibly, 75–95% = likely, 95–99% = very likely, >99%
almost certainly.
CMJ: countermovement jump; RJ: repeated jumps.

Table3.RelationshipbetweenphysicalperformanceduringSSGsandneuromuscularperformancepostSSGs.
HIRVHIRVHIAVHSRTDAcc2–3m∙s−2
Acc>3m∙s−2
AcctotDec2–3m∙s−2
Dec>3m∙s−2
Dectot
4v4+GK(N=16)
CMJ0.29±0.400.43±0.360.51±0.330.55±0.320.48±0.34−0.063±0.440.097±0.44−0.30±0.44−0.17±0.430.44±0.37−0.60±0.30
UnclearLikelymoderateVerylikely
moderate
Verylikely
moderate
LikelymoderateUnclearUnclearLikelymoderateUnclearLikelymoderateVerylikelylarge
15″RJ−0.22±0.410.28±0.400.35±0.410.54±0.320.38±0.38−0.01±0.440.07±0.440.23±0.42−0.31±0.410.21±0.43−0.27±0.42
UnclearUnclearLikelymoderateVerylikely
moderate
LikelymoderateUnclearUnclearUnclearUnclearUnclearUnclear
MIVC−0.48±0.340.01±0.430.03±0.430.05±0.43−0.32±0.40−0.36±0.39−0.08±0.440.42±0.38−0.40±0.38−0.28±0.410.19±0.43
LikelymoderateUnclearUnclearUnclearUnclearLikelymoderateUnclearLikelymoderateUnclearUnclear
5msprint0.18±0.420.15±0.420.37±0.380.46±0.350.007±0.44−0.58±0.310.04±0.44−0.19±0.43−0.57±0.320.05±0.44−0.45±0.37
UnclearUnclearLikelymoderateLikelymoderateUnclearVerylikelylargeUnclearUnclearVerylikely
moderate
UnclearLikelymoderate
15msprint0.17±0.420.46±0.350.43±0.360.43±0.360.10±0.44−0.29±0.41−0.03±0.44−0.18±0.43−0.26±0.420.03±0.44−0.40±0.38
UnclearLikelymoderateLikelymoderateLikelymoderateUnclearLikelysmallUnclearUnclearUnclearUnclearLikelymoderate
8v8+GK(N=16)
CMJ0.08±0.420.27±0.400.26±0.400.18±0.420.22±0.42−0.43±0.370.13±0.44−0.34±0.40−0.39±0.39−0.05±0.440.11±0.44
UnclearUnclearUnclearUnclearUnclearLikelymoderateUnclearUnclearLikelymoderateUnclearUnclear
15″RJ−0.17±0.42−0.12±0.42−0.06±0.430.01±0.43−0.21±0.430.38±0.390.03±0.440.32±0.400.39±0.390.23±0.42−0.28±0.41
UnclearUnclearUnclearUnclearUnclearLikelymoderateUnclearUnclearLikelymoderateUnclear
MIVC−0.44±0.36−0.25±0.40−0.28±0.40−0.28±0.40−0.41±0.380.27±0.420.12±0.440.51±0.340.14±0.440.15±0.43−0.18±0.44
LikelymoderateUnclearUnclearUnclearLikelymoderateUnclearUnclearVerylikely
moderate
UnclearUnclearUnclear
5msprint0.13±0.42−0.01±0.430.16±0.420.36±0.380.15±0.43−0.30±0.410.14±0.44−0.23±0.420.003±0.440.18±0.43−0.03±0.44
UnclearUnclearUnclearLikelymoderateUnclearUnclearUnclearUnclearUnclearUnclearUnclear
15msprint0.21±0.410.27±0.400.32±0.390.30±0.400.28±0.41−0.01±0.440.52±0.340.06±0.440.043±0.440.37±0.390.48±0.36
UnclearUnclearUnclearUnclearUnclearUnclearVerylikely
moderate
UnclearUnclearLikelymoderate
Dataarepresentedasmagnitudeofcorrelation±90%CI.
Acctot:totalaccelerations;Acc2‒3:accelerationsof2‒3m∙s−2
;Acc>3:accelerationsgreaterthan3m∙s−2
;CMJ:countermovementjump;CV:coefficientofvariation;Dectot:totaldecelerations;Dec2‒3;decelerationsof
2‒3m∙s−2
;Dec>3:decelerationsgreaterthan3m∙s−2;
HIR:high-intensityrunning;LIR:low-intensityrunning;MIVC:maximalisometricvoluntarycontraction;RJ:repeatedjumps;TD:totaldistancecovered;VHIR:very-high-
intensityrunning;VHIA:very-high-intensityactivities;VHSR:very-high-speedrunning.

Demands of 4v4 + GK and 8v8 + GK
In this study, it was observed that distance covered by VHIA
and high-speed running was greater in 8v8 + GK than in
4v4 + GK. Similar results were reported in other studies
investigating SSGs that found an increase in distance cov-
ered in higher speed categories with a number of players.
Hill-Haas, Rowsell, Dawson, and Coutts (2009b) examined
the effect on physiological responses of three game formats
(2v2, 4v4 and 6v6) with a constant ratio of player number to
pitch size. The largest game format (6v6) correlated with a
greater range of distances covered at speeds >18 km · h–1
.
Castellano, Casamichana, and Dellal (2013) also found sig-
nificant differences for the high-intensity speed category
depending on the number of players involved in different
game formats, with greater distances covered in this speed
category in 7v7 than in 3v3. In fact, it appears that a con-
current increase in player number and relative pitch area per
player in SSGs induces lower exercise intensity (Hill-Haas
et al., 2011).
For a more detailed analysis of SSG demands, we need to
take into account an essential element of football, namely
accelerations and decelerations, as a massive metabolic load
is imposed on players not only during high-speed running but
every time acceleration is elevated, even when speed is low
(Casamichana, Castellano, & Dellal, 2013; Gaudino et al., 2014a;
Varley & Aughey, 2013). Moreover, the high-intensity demands
of SSGs on elite footballers are underestimated by running
speed alone, particularly in “small” SSGs (Gaudino et al.,
2014b).
Previous studies showed an increase in acceleration/
deceleration demands as the number of players in the SSG
decreased. Gaudino et al. (2014a) analysed three SSG for-
mats (5v5, 7v7 and 10v10) and found that both the number
of moderate (2–3 m·s−2
) accelerations and decelerations, and
the total number of changes in velocity, were greater as the
pitch dimensions decreased. Predicted energy cost, average
metabolic power and distance covered in all metabolic
power categories were higher for large pitches than for
small pitches. Our study confirmed these results, with a
greater distance covered by accelerations and decelerations
performed in 4v4 + GK than in 8v8 + GK, which could mean
a higher metabolic and mechanical loading of the neuro-
muscular system in the smaller format. Thus, this informa-
tion should be taken into consideration when designing
weekly training schedules for footballers. As eccentric exer-
cises impose a high mechanical strain (Clarkson & Hubal,
2002), requiring long recovery times, coaches should not
schedule SSGs with a low number of players too close to
official matches (either before or after) in order to allow
adequate recovery. However, since it is the internal response
to training (e.g., HR, blood lactate concentration and RPE)
and not the external training load that determines each
player’s adaptation to a training stimulus, it is recommended
that each player’s internal load should be monitored to
assess how they are coping with different SSG training
designs (Hill-Haas et al., 2011).
We found higher values of blood lactate in 4v4 + GK than
in 8v8 + GK as a result of a higher anaerobic energy turnover.
Other researchers reported similar findings, pointing out that
as the number of players in the SSG teams decreased, the
overall physiological and perceptual responses increased. In
fact, it appears that a concurrent decrease in player number
and relative pitch area per player in SSGs elicits higher phy-
siological loading. Little and Williams (2006) investigated the
effect on [La–
] and RPE of six SSG football formats and pitch
area. The results showed an increase in [La–
]
(5.8 − 9.6 mmol·L−1
) when the number of players and pitch
area decreased. Other studies have also shown that SSG for-
mats with fewer players elicit greater RPE than larger formats
(Aroso, Rebelo, & Gomes-Pereira, 2004; Hill-Haas, Coutts,
Dawson, & Rowsell, 2010; Hill-Haas, Coutts, Rowsell, &
Dawson, 2008; Hill-Haas et al., 2009b; Impellizzeri et al., 2006;
Rampinini et al., 2007). Additionally, we found higher percep-
tion scores after 4v4 + GK in relation to physical, strength and
endurance demands. Altogether, these results suggest that,
compared with 8v8, 4x4 calls for more strength-demanding
actions requiring muscular eccentric work and relying on
anaerobic metabolism. Interestingly, we did not find any dif-
ference in HR between 4v4 + GK and 8v8 + GK. It could be
hypothesised that both games highly tax the cardiorespiratory
system.
As previously reported (Aslan, 2013; Clemente, Wong del,
Martins, & Mendes, 2014; Jones & Drust, 2007; Katis & Kellis,
2009), another argument for the use of a low number of
players in SSGs is the high frequency of technical actions.
We found a higher number of technical actions in 4v4 + GK
than in 8v8 + GK (tackles +45%; ball interceptions +91%;
defensive duels +91%; passes +93%; offensive duels +83%;
runs with the ball +136%; shots +240%; and goals +850%).
Reducing the number of players in the game increases the
number of repetitions of football skills and the number of
decision-making actions, and this provides a logical argument
for the use of this strategy both in teaching football and in
elite training environments. Thus, this information should be
taken into consideration when designing weekly training sche-
dules for footballers.
Fatigue development during and after 4v4 + GK and
8v8 + GK
Total distance and LIR tended to decrease throughout repeti-
tions of 4v4 + GK and 8v8 + GK. Moreover, the distance
covered by accelerations and decelerations showed an iden-
tical trend throughout 4v4 + GK repetitions. These results
indicate that players experienced fatigue throughout repeti-
tions of SSG, particularly during 4v4.
Dellal, Drust, and Lago-Penas (2012) found that the total
distance covered by high-intensity activities of elite footballers
decreased as a result of the number of exercise periods in 2v2,
3v3 and 4v4, and that the changes associated with the exer-
cise period seem to be related to the number of players. The
authors suggested that a reduction in the number of players
in an SSG induces a decrease in the number and duration of
the recovery periods between efforts. This information should
not be neglected when using SSG to improve the physical
fitness of footballers; if the aim of the training session is to

maintain the level of physical response during each repetition
of the SSG, coaches should pay attention to the number of
repetitions prescribed and players should be given enough
time between repetitions to completely recover.
We observed a decrease in neuromuscular-based perfor-
mance after 4v4 + GK and 8v8 + GK. Repeated jump ability
decreased after 4v4 + GK and 8v8 + GK, while sprint performance
decreased after 4v4 + GK. Results from Katis and Kellis (2009) also
showed a deterioration in 30 m sprint test performance after
repetitions of 3v3 SSG. This observed reduction in neuromuscular
performance after SSG completion could be interpreted as a
result of SSG-induced fatigue. However, reductions in sprint
ability in 5 and 15 m tests were found only after 4v4 + GK. It
should be stressed that, as discussed above, 4v4 + GK induced
higher neuromuscular stress and, based on decreased sprint
ability, this was only observable in this SSG format.
Altogether, the results suggest that 4v4 + GK provides a
greater stimulus for neuromuscular conditioning and technical
improvement than 8v8 + GK, and its use in training elite
footballers is highly recommended.
Relationship between neuromuscular testing and
performance in 4v4 + GK and 8v8 + GK
As far as we know, no previous studies have examined the
relationship between sprint and jump testing and SSG perfor-
mance. In our study, there was a positive correlation between
CMJ and distance covered by VHIA and VHSR during 4v4 + GK,
while repeated jump ability correlated with VHSR in the same
game format. Moreover, 15 m sprint performance positively
correlated with distance covered by accelerations at highest
intensity (Acc >3 m∙s−2
) during 8v8 + GK. These results suggest
that neuromuscular performance, as evaluated by jump ability,
could influence high-intensity activities and power-based
actions in 4v4 + GK and 8v8 + GK.
In recent years, the physiological stress generated in SSG
football has been examined with regard to its potential for
improving aerobic fitness (Hill-Haas, Coutts, Rowsell, &
Dawson, 2009a). However, periodic training interventions
involving SSG with a reduced number of players could be
capable of improving elite-level football players’ neuromuscu-
lar characteristics and this should be investigated in future
intervention studies.
There were some limitations of this study that should be
considered and addressed in future studies. We recommend
the analysis of a large sample of professional soccer players in
a high number of SSG-based training sessions throughout the
season in order to improve the reliability of the preliminary
conclusions of the present study. Moreover, it should be
emphasised that the metrics of speed calculations with GPS
show a great variability and must be interpreted cautiously. A
further limitation of this study might be the different duration
of bouts of 4v4 and 8v8 SSGs analysed as it was already
described that this variable could influence the exercise inten-
sity of SSGs (see Aguiar, Botelho, Lago, Maçãs, & Sampaio,
2012). However, it should be clarified that in this study we
intended to analyse the influence of two examples of soccer
training sessions based on two different types of SSG. As the
performance of SSGs with different numbers of players
requires an adjustment in exercise duration, we decided to
choose a duration for each SSG based on standard soccer
practices. In this way, the intention was to improve the validity
of the study.
Conclusions
In this study, we compared the physical demands of two SSG
formats: 4v4 + GK and 8v8 + GK. It was observed that 4v4 + GK
appears more demanding in relation to repetitions and fatigue
development of muscle power-based actions based on higher
anaerobic energy turnover, perceived exertion and number of
technical actions, whereas during 8v8 + GK players run more
at high intensity. Furthermore, repeated jump performance
evaluated in corresponding tests seems to correlate with dis-
tance covered at higher intensities in 4v4 + GK. It may there-
fore be logical to use 4v4 + GK for targeting the development
of power-related football actions, which should be investi-
gated in future intervention studies.
Disclosure statement
No potential conflict of interest was reported by the authors.
References
Aguiar, M., Botelho, G., Lago, C., Maçãs, V., & Sampaio, J. (2012). A review
on the effect of soccer small-sided games. Journal of Human Kinetics,
33, 103–113. doi:10.2478/v10078-012-0049-x
Andersson, H., Ekblom, B., & Krustrup, P. (2008). Elite football on artificial
turf versus natural grass: Movement patterns, technical standards, and
player impressions. Journal of Sports Sciences, 26(2), 113–122.
doi:10.1080/02640410701422076
Aroso, J., Rebelo, A. N., & Gomes-Pereira, J. (2004). Physiological impact of
selected game-related exercises. Journal of Sports Sciences, 22(6), 521–
566.
Aslan, A. (2013). Cardiovascular responses, perceived exertion and techni-
cal actions during small-sided recreational soccer: Effects of pitch size
and number of players. Journal of Human Kinetics, 38, 95–105.
doi:10.2478/hukin-2013-0049
Batterham, A., & Hopkins, G. (2006). Making meaningful inferences about
magnitudes. International Journal of Sports Physiology and Performance,
1(1), 50–57.
Baldari, C., Bonavolonta, V., Emerenziani, G. P., Gallotta, M. C., Silva, A. J., &
Guidetti, L. (2009). Accuracy, reliability, linearity of accutrend and lac-
tate pro versus EBIO plus analyzer. European Journal of Applied
Physiology, 107(1), 105–111. doi:10.1007/s00421-009-1107-5
Bosco, C., Luhtanen, P., & Komi, P. V. (1983). A simple method for mea-
surement of mechanical power in jumping. European Journal of Applied
Physiology and Occupational Physiology, 50(2), 273–282. doi:10.1007/
BF00422166
Buchheit, M., Al Haddad, H., Simpson, B., Pallazi, D., Bourdon, P., DiSalvo,
V., & Mendez-Villanueva, A. (2012). Monitoring accelerations with GPS
infootball: Time to slow down? International Journal of Sports Physiology
and Performance, 9(3), 442–445. doi:10.1123/IJSPP.2013-0187
Buchheit, M., Mendez-Villanueva, A., Simpson, B. M., & Bourdon, P. C.
(2010). Match running performance and fitness in youth soccer.
International Journal of Sports Medicine, 31(11), 818–825. doi:10.1055/
s-0030-1262838
Casamichana, D., & Castellano, J. (2010). Time-motion, heart rate, percep-
tual and motor behaviour demands in small-sided soccer games.
Journal of Sports Sciences, 28(14), 1615–1623. doi:10.1080/
02640414.2010.521168
Casamichana, D., Castellano, J., & Dellal, A. (2013). Influence of different
training regimes on physical and physiological demands during small-

sided soccer games: Continuous vs. intermittent format. Journal of
Strength and Conditioning Research, 27(3), 690–697. doi:10.1519/
JSC.0b013e31825d99dc
Castellano, J., Casamichana, D., & Dellal, A. (2013). Influence of game format
and number of players on heart rate responses and physical demands in
small-sided soccer games. Journal of Strength and Conditioning Research,
27(5), 1295–1303. doi:10.1519/JSC.0b013e318267a5d1
Chaouachi, A., Chtara, M., Hammami, R., Chtara, H., Turki, O., & Castagna, C.
(2014). Multidirectional sprints and small-sided games training effect
on agility and change of direction abilities in youth soccer. Journal of
JSC.0000000000000505
Clarkson, P. M., & Hubal, M. J. (2002). Exercise-induced muscle damage in
humans. American Journal of Physical Medicine & Rehabilitation, 81
(Suppl. 11), S52–69. doi:10.1097/00002060-200211001-00007
Clemente, F. M., Wong del, P., Martins, F. M., & Mendes, R. S. (2014). Acute
effects of the number of players and scoring method on physiological,
physical, and technical performance in small-sided soccer games. Research
in Sports Medicine, 22(4), 380–397. doi:10.1080/15438627.2014.951761
Cohen, J. (1960). A cofficient of agreement for nominal scales.
Education and Psychological Measurement, 20, 37–46. doi:10.1177/
001316446002000104
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
doi:10.1037/0033-2909.112.1.155
Dellal, A., Drust, B., & Lago-Penas, C. (2012). Variation of activity demands
in small-sided soccer games. International Journal of Sports Medicine, 33
(5), 370–375. doi:10.1055/s-0031-1295476
Dellal, A., Varliette, C., Owen, A., Chirico, E. N., & Pialoux, V. (2012). Small-
sided games versus interval training in amateur soccer players: Effects
on the aerobic capacity and the ability to perform intermittent exer-
cises with changes of direction. Journal of Strength and Conditioning
Research, 26(10), 2712–2720. doi:10.1519/JSC.0b013e31824294c4
Faude, O., Steffen, A., Kellmann, M., & Meyer, T. (2014). The effect of short-
term interval training during the competitive season on physical fitness
and signs of fatigue: A crossover trial in high-level youth football
players. International Journal of Sports Physiology and Performance, 9
(6), 936–944. doi:10.1123/ijspp.2013-0429
Gaudino, P., Alberti, G., & Iaia, F. M. (2014a). Estimated metabolic and
mechanical demands during different small-sided games in elite soccer
players. Human Movement Science, 36, 123–133. doi:10.1016/j.
humov.2014.05.006
Gaudino, P., Iaia, F. M., Alberti, G., Hawkins, R. D., Strudwick, A. J., &
Gregson, W. (2014b). Systematic bias between running speed and
metabolic power data in elite soccer players: Influence of drill type.
International Journal of Sports Medicine, 35(6), 489–493.
Halouani, J., Chtourou, H., Gabbett, T., Chaouachi, A., & Chamari, K. (2014).
Small-sided games in team sports training: A brief review. Journal of
JSC.0000000000000564
Hill-Haas, S. V., Coutts, A. J., Dawson, B. T., & Rowsell, G. J. (2010). Time-
motion characteristics and physiological responses of small-sided
games in elite youth players: The influence of player number and rule
changes. Journal of Strength and Conditioning Research, 24(8), 2149–
2156. doi:10.1519/JSC.0b013e3181af5265
Hill-Haas, S. V., Coutts, A., Rowsell, G., & Dawson, B. (2008). Variability of
acute physiological responses and performance profiles of youth soccer
players in small-sided games. Journal of Science and Medicine in Sport,
11(5), 487–490. doi:10.1016/j.jsams.2007.07.006
Hill-Haas, S. V., Coutts, A. J., Rowsell, G. J., & Dawson, B. T. (2009a). Generic
versus small-sided game training in soccer. International Journal of
Sports Medicine, 30(9), 636–642. doi:10.1055/s-0029-1220730
Hill-Haas, S. V., Dawson, B., Impellizzeri, F. M., & Coutts, A. J. (2011).
Physiology of small-sided games training in football: A systematic
review. Sports Medicine, 41(3), 199–220. doi:10.2165/11539740-
000000000-00000
Hill-Haas, S. V., Rowsell, G. J., Dawson, B. T., & Coutts, A. J. (2009b). Acute
physiological responses and time-motion characteristics of two small-
sided training regimes in youth soccer players. Journal of Strength and
Conditioning Research, 23(1), 111–115. doi:10.1519/JSC.0b013e31818efc1a
Hopkins, W. (2007). A spreadsheet for deriving a confidence interval,
mechanistic inference and clinical inference from a P value.
Sportscience, 11, 16–21.
Hopkins, W., Marshall, S., Batterham, A., & Hanin, J. (2009). Progressive
statistics for studies in sports medicine and exercise science. Medicine
and Science in Sports and Exercise, 41(1), 3–13.
Impellizzeri, F., Marcora, S., Castagna, C., Reilly, T., Sassi, A., Iaia, F., &
Rampinini, E. (2006). Physiological and performance effects of generic
versus specific aerobic training in soccer players. International Journal
of Sports Medicine, 27(6), 483–492. doi:10.1055/s-2005-865839
Jones, S., & Drust, B. (2007). Physiological and technical demands of 4 v 4 and
8 v 8 games in elite youth soccer players. Kinesiology, 39(2), 150–156.
Katis, A., & Kellis, E. (2009). Effects of small-sided games on physical
conditioning and performance in young soccer players. Journal of
Sports Science & Medicine, 8(3), 374–380.
Little, T., & Williams, A. G. (2006). Suitability of soccer training drills for
endurance training. Journal of Strength and Conditioning Research, 20
(2), 316–319. doi:10.1519/r-17094.1
Manzi, V., Impellizzeri, F., & Castagna, C. (2014). Aerobic fitness ecological
validity in elite soccer players: A metabolic power approach. Journal of
JSC.0000000000000239
Owen, A. L., Wong del, P., McKenna, M., & Dellal, A. (2011). Heart rate
responses and technical comparison between small- vs. large-sided
games in elite professional soccer. Journal of Strength and Conditioning
Research, 25(8), 2104–2110. doi:10.1519/JSC.0b013e3181f0a8a3
Rampinini, E., Impellizzeri, F., Castagna, C., Abt, G., Chamari, K., Sassi, A., &
Marcora, S. (2007). Factors influencing physiological responses to small-
sided soccer games. Journal of Sports Sciences, 25(6), 659–666.
doi:10.1080/02640410600811858
Rebelo, A., Brito, J., Seabra, A., Oliveira, J., & Krustrup, P. (2014). Physical
match performance of youth football players in relation to physical
capacity. European Journal of Sports Sciences, 14(Suppl. 1), S148–156.
doi:10.1080/17461391.2012.664171
Taoutaou, Z., Granier, P., Mercier, B., Ahmaidi, S., & Prefaut, C. (1996).
Lactate kinetics during passive and partially active recovery in endur-
ance and sprint athletes. European Journal of Applied Physiology, 73(5),
465–470. doi:10.1007/BF00334425
Varley, M. C., & Aughey, R. J. (2013). Acceleration profiles in elite Australian
soccer. International Journal of Sports Medicine, 34(1), 34–39.
Varley, M. C., Faiweather, J., & Aughey, R. J. (2013). Validity and reliability of
GPS for measuring inatanteneous velocity during acceleration and
constant motion. Journal of Sports Medicine, 30, 121–127.

Differences in strength and speed demands between 4v4 and 8v8 SSG

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Differences in strength and speed demands between 4v4 and 8v8 SSG (20)

More from Fernando Farias (6)

Recently uploaded (20)

Differences in strength and speed demands between 4v4 and 8v8 SSG