Inlemal'ional Joumal of Perfonllance

2014,14,801-8.

Analysis in Sport

Comparison of tactical behaviour and performance of
youth soccer players in 3v3 and 5v5 small-sided games
Daniel CasteHio1, Julio Garganta1, Rodrigo Santos2, and srael Teoldo2
] Faculty of Sport, University of Porto, Portugal
2 Centre of Research and Studies in Soccer, Universidade Federal de Viyosa, Brazil

Abstract

Soccer is a predominantly tactical sport and tactical skills are particularly
important for enhancing performance, since the actions are
unpredictable, thus forcing players to constantly adapt to the situations.
Our aim is to compare the tactical behaviour and performance of soccer
players in 3v3 and 5v5 small-sided games (SSGs). The sample comprised
10 U-ll youth soccer players who performed 1,563 tactical actions in both
31>3and 5v5 SSGs. We used the System of tactical assessment in Soccer
(FUT-SAT) to assess players' tactical behaviour and performance.
Descriptive analyses were performed Shapiro- Wilk normality test was
conducted T-test for repeated measures and Wilcoxon test were used to
compare data between both situations (P<0.05). Significant differences
were found in all categories of variables except in the category Tactical
Performance Index (TPI). Players performed considerably more
frequently the principles of Penetration and Delay in 3v3 and Offensive
Unity and Balance in 5v5 SSGs. We concluded that while players' tactical
performance has not changed, 3v3 and 5v5 demand different tactical
behaviour from the players. These data can contribute to practice since
coaches should take into account the changes in tactical behaviour when
considering the number of players in soccer SSGs during training
sessions.
Keywords:
games.

soccer, tactical behaviour, tactical performance,

small-sided

1. Introduction
Soccer can be understood as a predominantly tactical sport in which most of the actions
during the game are performed without possession of the ball (Garganta, 1997). Also,
players with limited technical skills can play the game if they possess reasonable
tactical comprehension (Oslin et aI., 1998). Therefore, the lack of tactical knowledge
and intelligence is a significant reason for the poor execution of technical skills
(Teodorescu, 1977). Accordingly, tactical skills are particularly important for enhancing
sports performance, since the frequency, order and complexity of the actions are
801

unpredictable, forcing players to constantly adapt to match situations through organized
means (Garganta, 1997; McPherson, 1994).
Aiming to develop players' technical and tactical skills, coaches and managers often
make use of small-sided games (SSGs), since they are expected to make training
sessions more objective, therefore encouraging the commitment of players, besides
stimulating actions related to the tactical component that need to be constantly practised
and developed (Reilly, 2005). Fitness coaches also resort to SSGs, since researches
indicate the induction to different kinds of effects for the development of physical
capacity through this sort of drill (Casamichana et aI., 2012; Silva et aI., 2011).
Within the context of small-sided games (SSGs), 3v3 turns out to be the minimal
structure among all possible numeric settings, allowing the occurrence of the tactical
principles inherent to the game of Soccer (Grehaigne, 1992, 2007). Small-sided games
with more players participating also offer the possibility of increasing the effective playspace both in width and in depth. According to Platt et al. (2001) significant differences
regarding physical constraints can already be observed in 5v5 in comparison to 3v3.
onetheless, there is an apparent lack of researches comparing tactical behaviour and
performance between such distinct situations.
Due to the importance of tactics for players' performances, an increasing number of
researches over this subject have been reported (Dellal et aI., 2012; Jones and Drust,
2007; Silva et al., 2011). Such growth can be explained by the interest of identifying
individual and collective action patterns that are likely to be linked to teams' increased
efficiency. Therefore, in order to obtain objective data with respect to the behaviour and
performance of players and teams, researchers have resorted to several methods to
examine register and assess actions performed within the context of team games.
Among such methods, the observational approach is an objective, reliable and valid way
to assess players' and teams' performances (Grehaigne et al., 2001). In addition, this
methodology can provide coaches and researchers with relevant information that might
be useful for planning training sessions, through cost-effective procedures (pers et al.,
2002).
Thus, this study aims to compare the tactical behaviour and performance of soccer
players in 3v3 and 5v5 small-sided games (SSGs).

2. Methods
2.1. Sample
The sample comprised 1,563 tactical actions (768 actions in 3v3 and 795 in 5v5 SSGs)
performed by 10 U-Il youth soccer players from a Portuguese club. The actions in
which players performed throw-ins, free kicks, corner-kicks, as well as those in which
they did not perform any tactical principle, were not considered for assessment.

802

2.2. Procedure
2.2.1. Data collection
Data were collected with the penmSSlOn of club's representatives. All partIcIpants
played during eight minutes in both situations (3v3 and 5v5). Playing area was adapted
according to the number of players involved, and in 3v3 field size was 36m long by
27m wide, while in 5v5 it was 60m long by 45m wide. In 3v3, participants were
grouped in teams of three players plus a goalkeeper (GK +3 vs. 3+GK), while in 5v5 the
distribution consisted of five players for each teams plus a goalkeeper (GK +5 vs.
5+GK). Actions performed by goalkeepers were not analyzed. Prior to the start of each
test session, players were informed over the objectives of such tests and were given 30
seconds in order to familiarize with test procedures. All players wore numbered vests in
order be clearly identified during video analysis.
2.2.2. Ethical procedures
This study had the approval of the Ethics Committee of the University of Porto
(CEFADE 15/2013) and meets the standards of the Declaration of Helsinki for
researches with human beings (2008).
2.3. Instrument
We used the System of tactical assessment in Soccer (FUT-SAT) (Teoldo et at., 2011)
which enables the assessment of tactical actions performed by players with and without
ball possession. Such assessment is based on ten core tactical principles of the Soccer
game, with five offensive - (i) Penetration; (ii) Offensive Coverage; (iii) Width and
Length; (iv) Depth Mobility; (v) Offensive Unity - and five defensive principles - (vi)
Delay; (vii) Defensive Coverage; (viii) Balance; (ix) Concentration; (x) Defensive
Unity, (Teoldo et at., 2009b; Worthington, 1974). FUT-SAT comprises two MacroCategories, seven categories and 76 variables (as described in Chart 1) that are
organized according to the type of information dealt with by the system. The MacroCategory "Observation" involves three categories and 24 variables: the category
"Tactical Principles" includes ten variables; the category "Place of Action in the Playing
Field" encompasses four variables; and the category "Action Outcome" contains ten
variables.

803

Chart 1. Definitions, categories and sub-categories of variables assessed by FUT-SAT.
SubDefinitions
Categories
Variables
Categories
tovement of player with the ball tov"ards the goal line.

Penetration

Offensive supports ta the player with the ball.

Offensive Caverage
Offensive

lavement 'Ofplayers between the last defender and gaalline.

Depth Mability

. lovement of players to extend and use the effective play-space.

Width and Length
Offensive

M'Ovement 'Ofthe last line 'Ofdefenders t'Owards the 'Offensive
midfield, in 'Order t'Osupp'Ort 'Offensive acti'Ons 'Ofthe teammates.
Actians t'Oslaw dawn the appanent's atl'empt t'Om'Ove f'Orward with
dleball

'nity

Delay

Tactical
Principles

Defensive

Defensive Coverage

Pasitianing 'Of'Off-ball defenders behind the "delay" player,
providing defensive suppart'.

Balance

P'Ositi'Oning'Of'Off-ball defenders in reacti'On t'Om'Ovements 'Of
atl'ackers, trying t'Oachieve the numerical stability 'Orsuperi'Ority in
the 'Oppasitian relatianship,

Concentration

P'Ositi'Oning'Of'Off-ball defenders t'Ooccupy vital spaces and pr'Otect
the scaring area,

Defensive

nity

Positioning of off-ball defenders to reduce the effect'ive play-space
'Ofdle appanents.

Offensive
Midfield

Offensive

ctians

Offensive actions perfonned in the offensive midfield.

Defensive.

ction

Defensive
Midfield

Offensive Actians

Offensive acti'Ons perfanned

in the defensive midfield.

Defensive Actians

Defensive acti'Ons perfanned

in the defensive midfield.

Place 'OfAction

Defensive actions perfonned in the offensive midfield.

When a player shoots at gaal, and (a) scares a g'Oal, (b) the
gaalkeeper makes a save, (c) the ball tauches 'One'Ofthe gaalpasts
'Orthe crassbar.
Wlletl team pi ayers execute passes to each other and keep up with
the ball,

Shoot at goal
Keep passes sian 'Ofthe ball
Offensive

Earn a foul, win a comer or
throw-in

Wlletl dle match is stopped due to a foul, comer or throw-in; the
team that' wns al't'acking KEEPS passessian 'Ofthe ball.

Commit a foul, give away a
carner 'Orthrow in

'When dle match is stopped due to a foul, comer or throw-in; the
possession of I'lie ball C
'GES ta the team that was in defence.

Loss afball

When the attacking team lases the ball passessian.

passessian

Action Outcames

Defensive

Regain the ball passessian

When the defensive players regain the ball p'Ossessi'On.

Eam a faul, win a camer 'Or
throw-in

'When the match is stopped due to a foul, comer or throw-in and
dle passessian afdle ball CHANGES ta the team that was in
defence,

Canumt a faul, give away a
camer 'Orthrow in

When the match is st'Opped due t'Oa loul, carner 'OrthroW-ill; the
team that was attacking KEEPS passession afthe ball.

Ball pas e ian 'Ofthe
'Opponent

\Vhen the defensive players da nat' regain the ball possessian.

Take a shot at own goal

v"hen dle defensive team takes a shot at their own goal, and (a)
takes a goal, (b) the goalkeeper makes a save, (c) the ball tauches
one oftlie goalposts or the crossbar.

The field test within this system is conducted in a playing field with dimensions ranging
between 27 and 36 metres long by 13.5 and 27 metres wide (test specifications for 5v5
SSGs are mentioned in the next paragraph). Field dimensions were calculated based on
soccer field dimensions established by the International Football Association Board, and
on the ratio calculation of the use (by the players) of playing space. For all the abovementioned dimensions, the centre of play has a 5-meter radius, taking into account the
constant values found between the lower and higher values. During the test, participants

804

are asked to play according to the official rules of Soccer, except for the offside rule
(Teoldo et al., 2009a).
For this study, we used field dimensions of 36 metres long by 27 metres wide and a
time of seven minutes and twenty seconds (without stoppages) for the test in both 3v3
and 5v5 SSGs. Such option is justified by the pilot study ofTeoldo et al. (201 Ob), which
did not find significant differences in the use oflarger amounts of time.
For 5v5 SSGs, FUT-SAT was conducted in a field of60 metres long by 45 metres wide.
These dimensions were conceived based on the ratio of the maximum dimensions
established for 3v3 SSGs. Taking these dimensions into account, the value found for the
radius of the centre of play was 6 metres.
2.4. Material
To record field tests, a digital video camera (Panasonic NV-DS35EG) was used. Video
footage was introduced in digital format in a laptop (DELL Inspiron 1545, Intel T6500
processor) via USB cable (IEEE 1394) and converted to ".avi" format.
For image processing and video analysis, the software Utilius V. and Soccer Analyser®
were used. The first was developed for the analysis, categorization and registry of the
actions assessed. The latter, specially designed for FUT-SAT, is used for the insertion
of spatial references in the video, enabling the accurate assessment of players'
positioning and movement.
Insert Chart 1
2.5. Statistical Analysis
Descriptive analyses were performed. Within the categories Tactical Principles, Place of
Action in the Game Field and Action Outcome, analyses of absolute and relative
frequency were performed. For the variables within the categories TPI, Percentage of
Errors, Tactical Actions and P ARP, we conducted analyses of means and standard
deviations.
For statistical analysis of frequencies in the categories Tactical Principles, Place of
Action in the Game Field and Action Outcome, Chi-square (l) test was conducted, with
significance level of P<0.05. For variables in the categories TPI, Percentage of Errors,
Tactical Actions and PARP, Shapiro-Wilk normality test was performed to verify data
distribution. For variables with normal distribution, comparisons were made through
Student's t test (P<0.05). For non-parametric data, we used Wilcoxon test (P<0.05). We
performed data analyses through the software PASW (Predictive Analytics Software for
Windows~, version 18.0.
2.5.1.. Reliability analysis
We performed test-retest reliability for the observations, respecting a 20-day interval for
reanalysis, thus avoiding task familiarity issues (Robinson and Q'Donoghue, 2007). For
reliability calculation we used Cohen's Kappa test. Analyses were verified through the
reassessment of220 tactical actions, or 14.07% of the overall sample, a value which is
greater than the percentage (10%) suggested by literature (Tabachnick and Fidell,
805

2012). Intra-observer reliability displayed Kappa value of 0.91 (SE=O.OI). This value is
classified as "Almost Perfect" (0.81 - 1.00) by literature (Landis and Koch, 1977).

3. Results
Results are initially presented according to the Macro Categories that compose the test.
Therefore, results regarding variables in the categories Tactical Principles, Place of
Action in the Game Field and Action Outcome, and later, TPI, Percentage of Errors,
Tactical Actions and PARP are presented. After these, TPI values of each player in 3v3
and 5v5 SSGs are displayed.
3.1. Macro-Category Observation
Table 1 presents frequencies, percentages and percentage vanatIOn of the variables
within the categories Tactical Principles, Place of Action in the Game Field and Action
Outcome in 3v3 and 5v5 SSGs. Among the 24 variables presented in Table 1, six of
them displayed significant differences.
After analysing the variables within the category Tactical Principles (Table 1), we
observed the incidence of four statistically significant differences. Among these, we
observed that frequencies of the principles of Penetration and Delay were higher in 3v3,
whereas frequencies related to the principles of Offensive Unity and Balance were
higher in 5v5.
The lowest percentage variations found regard the principles of Width and Length and
Depth Mobility, in offensive phase, and Concentration and Defensive Unity, in
defensi ve phase.
No significant differences were found for the variables in the category Place of Action
in the Game Field. It is possible to see a distribution of actions in the game field that
was similar in both SSGs, what can be verified through the low values in the column
that refers to percentage variation.

806

Table 1. Absolute and relative frequencies and percentage variation of the variables
within the categories Tactical Principles, Place of Action in the Game Field and Action
Outcome in 3v3 and 5v5 SSGs.
Categories

3v3

and Variables

n

TACTICAL PRINCIPLES
Offensive
Penetration*
Offensive Coverage
Widtll and Length
Depth Mobility
Offensive Unity*
Defensive
Delay*
Defensive Coverage
Balance*
Co ncentratio n
Defensive Unity
PLACE OF ACTION IN THE GAME FIELD
Offensive
Offensive Midfield
Defensive Midfield
Defensive
Offensive Midfield
Defensive Midfield
ACTION OUTCOME
Offensive
Shoot at goal*
Keep possession of the ball
Earn a foul, win a comer or throw-in
Conunit a foul, give away a comer or tllfow-in
Loss of ball possession
Defensive
Regain ball possession
Earn a foul, win a comer or tllfow-in
Conunit a foul, give away a comer or tllfow-in
Ball possession of tIle opponent
Take a shot at own goal*
TOTAL

5v5

Percentage
Variation**

n

0/0

7.03
13.41
16.93
3.65
7.55

27
117
134
30

3.40
14.72
16.86

82

10.31

62

7.80

23
68
58

183

12.11
2.47
5.47
7.55
23.83

194

2.89
8.55
7.30
24.40

203
210

25.06
25.93

217
188

26.79
23.21

6.90
-10.48

156
241

19.26
29.75

138
267

17.04
32.96

-11.54
10.79

52
273
21
19

6.42
33.70
2.59
2.35
5.93

30
285

3.70
35.19
3.09
3.09
4.94

-42.31
4.40
]9.05
31.58
-16.67

6.17
2.10
2.84
31.36
6.54

40
25
25
285
30
795

4.94
3.09
3.09
35.19
3.70

-20.00
47.06
8.70
]2.20
-43.40

54
103
130

28
58

93
19
42

58

48
50
17
23
254
53
768

%

25
25
40

3.77

-50.00
13.59
3.08
7.14
41.38
-33.33
21.05
61.90

0.00
6.01

*Statistically significant differences (P<O.05). Tactical Principles: Penetration (P=0.03), Offensive
Unity (P=0.043), Delay (P=O.013), Balance (P=0.013). Action Outcome: Shoot at goal (P=0.015), Take
a shot at own goal (P=0.012). ** Values of percent.age variation were obt.ained from t.he comparison of
3v3 to 5v5 SSGs.

In the category Action Outcome, we found statistically significant differences regarding
the frequency of the variables Shoot at Goal and Take a Shot at Own Goal, in 3v3
SSGs.
The lowest percentage variations found refer to the variable Keep Possession of the
Ball, in offensive phase, and Commit a Foul, Give Away a Corner or Throw-in, in
defensive phase.

807

0\

.,-.t

In

N

0\ <'1 N
###BOT_TEXT### In

0'\ N
("'")
\D: \D:

'I""""'C

~o\...t

0 """
(')
-tj -tj -tj -tj -tj

-

l/i

-tj -tj -tj

000

00000

•....•
000
00("')-:
,...-c ("""I

..-< 'n..-< ~ """.
In

0

00

("')

000<'1-

I:-

""". 00 """. ~
'f""""l
N.,.....( l,n

1:-.
N

-tj -tj -tj -tj -tj
00000
N

~

!"-I

00 (')

l/i

Nl/iOo\N

N

'D

or)

<'1

•.••-.:t:0\~-.:t:
NOON

"""

-tj -tj -tj -tj -tj

00000
"-<N~I:-O\

MO

'!"'"""I

.....•
N~

<'1<'1-

~oo~
###BOT_TEXT### -0

'f"l

~~~
00 l/i

(')

II

o\orior)
!"-I

,...-c

---N
"""
o
o

S

("""I

Ir, Ir, In """

<"'"l<"'"l'0<"'"l0

•.•..•
N •.•..•
00 N
N •.••N""" N

tii~(t\t\t\
<'100 00 <'1<'1
•.•..•
N •.•...•
-oN
'D N """ •.•..•
N
0

I:- l/i

""" 'f"l N
,.-.t .,-( r't

(') N
N ("l

00 I:-

1:-. '0 N. 1:-. 1:-.

~tii~tii(t1
<'1I:- <'10\ N-o<'1N...t
N..-r..-:

N

(,'l-

co

o
C()

~ r- ,.......(
(''\-0~ 1:-. 1:-. ~ 1:-.
00 <'1 0\ N 'f"l
,.-.t.,-(N--<C"l

("<")

('''I

«!«!~

In

0\ <'1 or)

-tj -tj -tj

J!.-#tii3t~
'Dor)",,"O\N

00 00 or)

-O~o\o\~

00 N

'n """ <'1<'1"""

.....•

~~

0

"""<'1,,,,"

By observing means related to the variables Offensive Phase and Defensive Phase, we
observed higher values in 5v5 SSGs, in respect to the categories TPI and Tactical
Actions, whereas In 3v3 SSGs, we verified higher values within the categories
Percentage of Errors and P ARP.
3.3. Individual TPIs
Table 3 presents the Tactical Performance Indexes (TPI) of each player. Such indexes
are related to players' performances regarding each tactical principle, offensive and
defensive phases and the entire game. The values presented demonstrated that players'
performances were not homogeneous.
By comparing TPls in 3v3 and 5v5 SSGs, it is possible to observe that three players
obtained higher values ofTPI in 3v3 and seven players had higher values in 5v5.
After analysing Offensive Tactical Performance Indexes (OTPIs), we observed that six
players displayed higher values in 5v5 SSGs. Instead, when considering the Defensive
Tactical Performance Indexes (DTPIs), among all ten players assessed, five displayed
higher values in 3v3 and the other five in 5v5 SSGs.
Table 3. Players' individual Tactical Performance
Tactical Principles

~

~

Penetration
Offensive Coverage
Depth Mobility
Width and Length
Offensive Unity
Offensive TPI
Delay
Defensive Coverage
Balance
Concentration
Defensive Unity
Defensive TPI
TPI
Penetration
Offensive Coverage
Depth Mobility
Width and Length
Offensive Unity
Offensive TPI
Delay
Defensive Coverage
Balance
Concentration
Defensive Unity
Defensive TPI
TPI

4
45.00
50.00
40.00
21.43
50.00 62.94
38.18
48.00 55.00
29.29 44.39 55.00 55.22
40.71 28.33 41.25 44.38
40.00 48.00
45.00
40.00 17.50 16.00 17.50
40.77 27.50 22.22 17.50
34.00 32.06 28.44 32.50
38.08 29.71 28.03 32.50
34.54 35.41 42.65 44.82
80.00 62.00 90.00
53.68 56.67 37.06 65.71
73.64
66.67
30.00 52.19 27.50 54.21
46.00 10.00 51.43 85.00
46.83 59.62 44.25 63.66
30.00 27.50 20.00 22.14
35.00 10.00 25.00 20.00
28.33 13.57 56.25 27.50
40.00 18.33 53.33 18.00
46.84 30.22 "'9.13 21.32
39.50 24.88 42.93 22.25
43.21 41.81 43.58 43.21
1
23.57
29.00

2
69.00
36.67
51.00
31.11

3
47.50
67.86

809

ndexes (TP s)
Players
567
64.38 88.00
67.14 60.00
80.00
49.29 45.00
46.00 54.38
59.27 59.48
39.17 55.00
53.33
10.00 40.00
40.00 45.00
32.75 28.33
35.41 37.57
47.76 47.20
70.00
80.00 58.75
60.00
45.00 45.77
72.00 49.38
54.47 53.90
25.00 32.73
53.33
48.75 25.00
25.00 31.25
30.65 42.14
32.80 36.50
43.23 45.31

63.57
36.36
50.00
30.00
58.00
49.57
27.00
10.00
30.00
16.67
35.00
30.41
39.72
30.00
41.54
28.75
61.25
46.00
25.00
40.00
41.11
45.00
36.67
37.56
41.73

8

9

10

70.00
42.00
72.00
38.24
100.00
50.32
27.50

43.33
50.00
45.00
34.29
50.00
39.71
14.44
53.33
30.00
24.00
42.67
32.57
36.14
52.50
74.29
80.00
42.83
80.00
56.25
31.67
50.00
26.67
27.78
21.00
27.56
41.73

52.50
36.36
56.67
37.69
22.50
39.59
27.50
40.00
17.50
21.67
32.50
27.71
33.82
46.00
44.17

25.00
60.00
27.27
28.82
38.48
70.00
62.73
80.00
47.06
80.00
57.43
25.56
36.00
27.50
26.67
31.05
29.75
42.67

40.00
53.24
48.71
40.00
46.67
40.00
46.67
48.70
46.50
47.53

With respect to Tactical Performance ndexes for the entire game (TPls), higher values
with lower variations among the ten players were found in 5v5 SSGs. Moreover, we
observed that players 3, 4, 5 and 6 displayed relatively better results when compared to
the others, with their TPls being ranked within the top five values both in 3v3 and 5v5
SSGs.

4. Discussion
The present study aimed to compare tactical behaviour and performance of soccer
players in 3v3 and 5v5 small-sided games (SSGs). Results within this study allowed us
to observe variations in eight of the 76 variables analysed. Such variations indicate that
both small-sided games (SSGs) demand different tactical behaviours from the players.
Although no differences were found between both SSGs with respect to tactical
performance, players' tactical behaviour displayed significant variations and enabled us
making important interpretations. The study conducted by Silva et al. (2014), which
compared 3v3 and 6v6 SSGs, also found significant differences regarding players'
tactical behaviour in such configurations. However, the authors did not examine tactical
performance.
Within the present study, higher frequency values observed in 3v3 SSGs regarding the
principles of Penetration and Delay may be due to the fact that the reduced number of
outfield players might have led opponents to perform more actions of direct opposition
to the player in possession, who, in turn, was able to easier place himself in situations
that favoured goal attempts or dribbling. Such inference was confirmed by frequency
values of the variables Shoot at Goal and Take a Shot at Own Goal. These results are
similar to those found in literature that indicates higher number of goal attempts in
smaller fields (Jones and Drust, 2007; Katis and Kellis, 2009).
On the other hand, higher frequency values of the principle of Offensive Unity in 5v5
SSGs may be explained by the increased field dimensions and by the number of players
involved, what may have enabled a higher amount of actions in which participants play
in spaces further back in the field, acting as support players instead of taking part of the
action closer to the centre of play or ahead of the ball line. Similar findings were
reported in previous research (Silva et al., 2014). Also, the higher occurrence of the
principle of Balance may be eXplained by the wider distribution of players over the
playing field, including corridors surrounding the epicentre of play (Little and Williams,
2007).
Both in 3v3 and 5v5 SSGs we observed a low frequency value for the principle of
Depth Mobility, what reveals a low tendency of performing actions between the last
defender and the opponents' goal.
In the categories Place of Action in the Game Field and PARP, we observed a
distribution of actions over the field that was not significantly different in any of the
SSGs. A similar distribution of offensive actions within the variables Offensive
810

Midfield and Defensive Midfield was also observed. On the other hand, defensive
actions displayed higher values within the variable Defensive Midfield in both SSGs.
These results indicate that in both SSGs teams converged to perform offensive
sequences based on direct play (Barreira et al., 2010; Garganta, 1997), thus emphasizing
ball possession through few passes, that begin in defensive midfield and passing
through offensive midfield while players organize themselves within the team. This is
highlighted when we take into consideration the high frequency of the variable Ball
Possession of the Opponent and the low frequency of the variable Regain Ball
Possession observed in both SSGs. On the other hand, defensive actions tend to be
performed when pressure is exerted down the field, that is, when players are drawn back
in their defensive halves. Results similar to those described in this study were found in
earlier researches (Teoldo et a/., 2010a).
In 3v3 SSGs constant positional changes between players were identified, while in SvS,
fewer movements were observed. Regarding the Tactical Performance Index, the
highest values displayed by players three, four, five and six in 3v3 and SvS SSGs
occurred due the highest Offensive TPI achieved by them, while Defensive TPI
displayed values that were closer to the means of the rest of the players involved.
Findings within the present study might allow coaches to improve players' tactical
skills, by considering the advantages of each SSG to the modelling of tactical
behaviour. Thus, the suitability of the amount of the players involved in the drill might
take into account the tactical principles that underlie the behaviours that the coach wants
to modify/develop.
Further researches regarding SSGs should examine players' tactical skills in different
contexts such as the use of floaters or neutral players, in order to investigate how
different types of constraints influence behaviour and performance.

5. Conclusion
This study aimed to compare tactical behaviour and performance of soccer players in
3v3 and SvS small-sided games (SSGs)
It was observed that in 3v3 and SvS SSGs, players' behaviours displayed statistical
significant differences with respect to the tactical actions performed during the test.
Players proved to perform considerably more frequently the principles of Penetration
and Delay in 3v3 and Offensive Unity and Balance in SvS SSGs.
No significant differences were found in the category Tactical Performance Index (TPI),
suggesting that tactical performance did not vary significantly between 3v3 and Sv5
SSGs.

811

6. References
Barreira, D., Garganta, J., and Anguera, T. (2010). In search of nexus between attacking
game-patterns, match status and type of ball recovery in European Soccer
Championship 2008. In M. Hughes, H. Dancs, K. agyvaradi, T. Polgar,
James, G. Sporis & G. Vuckovic (Eds.), Research Methods and Performance
Analysis. Szombathely: University of West Hungary.
Casamichana, D., Castellano, J., and Castagna, C. (2012). Comparing the physical
demands of friendly matches and small-sided games in semiprofessional soccer
players. Journal of Strength and Conditioning Research, 26(3),837-843.
Dellal, A., Owen, A., Wong, D. P., Krustrup, P., van Exsel, M., and Mallo, J. (2012).
Technical and physical demands of small vs. large sided games in relation to
playing position in elite soccer. Human Movement Science, 31(4), 957-969.
doi: 10.1016/j .humov.2011.08.013
Garganta, J. (1997). Modela~ao tactica do jogo de futebol - estudo da organiza~ao
da fase ofens iva em equipas de alto rendimento. [Tactical modelling in soccer
- study about the attacking phase in top level teams]. (Unpublished Ph.D.
Thesis), University of Porto, Porto, Portugal.
Grehaigne, J. F. (1992). L'organisation du jeu en football. Joinville-le-Pont: Actio.
Grehaigne, J. F. (2007). Configurations de jeu: Debat d'idees & apprentissage du
football et des sports collectifs. Paris: Presses Universitaires de FrancheComte.
Grehaigne, J. F., Mahut, B., and Fernandez, A. (2001). Qualitative observation tools to
analyse soccer. International Journal of Perfomance Analysis in Sport, 1(1),
52-61.
Jones, S., and Drust, B. (2007). Physiological and Technical Demands of 4v4 and 8v8
Games in Elite Youth Soccer Players. Kinesiology, 39(2), 150-156.
Katis, A., and Kellis, E. (2009). Effects of small-sided games on physical conditioning
and performance in young soccer players. Journal of Sports Science and
Medicine, 8(3), 374-380.
Lames, M., and Hansen, G. (2001). Designing observational systems to support toplevel teams in game sports. International Journal of Perfomance Analysis in
Sport, 1(1), 83-90.
Landis, R., and Koch, G. G. (1977). The Measurement of Observer Agreement for
Categorical Data. Biometrics, 33(1),159-174.
Little, T., and Williams, A. G. (2007). Measures of exercise intensity during soccer
training drills with professional soccer players. Journal of Strength and
Conditioning Research, 21(2), 367-371.
McPherson, S. L. (1994). The Development of Sport Expertise: Mapping the Tactical
Domain. Quest, 46(2),223-240.
Oslin, J., Mitchell, S. A., and Griffin, L. L. (1998). The Game Performance Assessment
Instrument (GPAI): development and preliminary validation. Journal
of
Teaching in Physical Education, 17(2),231-243.
Pers, J., Bon, M., Kovacic, S., Sibila, M., and Dezman, B. (2002). Observation and
analysis of large-scale human motion. Human Movement Science, 21(2), 295311.
812

Platt, D., Maxwell, A., Hom, R., Williams, A. M., and Reilly, T. (2001). Physiological
and Technical Analysis of 3v3 and 5v5 Youth Football Matches. Coaching,
4(4),42-45.
Reilly, T. (2005). An ergonomics model of the soccer training process. Journal of
Sports Sciences, 23(6), 561-572.
Robinson, G., and O'Donoghue, P. (2007). A weighted kappa statistic for reliability
testing in performance analysis of sport. International Journal of Perfomance
Analysis in Sport, 7(1), 12-19.
Silva, B., Garganta, l, Santos, R., and Teoldo, 1. (2014). Comparing Tactical Behaviour
of Soccer Players in 3 vs. 3 and 6 vs. 6 Small-Sided Games. Journal of Human
Kinetics, 41(1), 191-202. doi: 10.2478/hukin-2014-0047
Silva, C. D., Impellizzeri, F. M., atali, A. 1., Lima, 1. R., Bara-Filho, M. G., Garcia, E.
S., and Marins, 1. C. (2011). Exercise intensity and technical demands of smallsided games in young Brazilian soccer players: effect of number of players,
maturation, and reliability. Journal of Strength and Conditioning Research,
25(10), 2746-2751.
Tabachnick, B., and Fidell, L. (2012). Using Multivariate Statistics: International
Edition (6 ed.). London: Pearson Education.
Teodorescu, L. (1977). Theorie et methodologie des jeux sportifs. Paris: Les Editeurs
Franc;ais Reunis.
Teoldo, 1, Garganta, J., Greco, P., and Mesquita, 1. (2009a). Avaliac;ao do Desempenho
Tatico no Futebol: Concepc;ao e Desenvolvimento da Grelha de Observac;ao do
Teste "GR3-3GR". [Evaluation of Tactical Performance in Soccer: Conception
and development of Framework of Tactical Behavior Analysis of "GK3-3GK"
Test]. Revista Mineira de Educaf;3o Fisica, 17(2), 36-64.
Teoldo, 1, Garganta, J., Greco, P. J., and Mesquita, I. (2009b). Tactical Principles of
Soccer Game: concepts and application. [in Portuguese]. Motriz, 15(3), 657668.
Teoldo, 1., Garganta, 1., Greco, P. 1., Mesquita, 1., and Afonso, 1. (2010a). Assessment
of tactical principles in youth soccer players of different age groups. Revista
Portuguesa de Ciencias do Desporto, 10(1), 147-157.
Teoldo, I., Garganta, 1., Greco, P. 1., Mesquita, I., and Maia, 1. (2011). System of
tactical assessment in Soccer (FUT-SAT): Development and preliminary
validation. Motricidade, 7(1),69-83. doi: 1O.6063/motricidade.7(1).121
Teoldo, 1., Garganta, 1., Greco, P. J., and Muller, E. (2010b). Influencia do tempo de
jogo nos comportamentos tilticos de jogadores de futebol, no Teste "GR3-3GR".
[Influence of game time on tactical behaviours of soccer players in the "GK33GK" test]. Revista Mineira de Educal;3o Fisica, 18(1), 7-25.
Worthington, E. (1974). Learning & Teaching Soccer Skills.
orth Hollywood:
Wilshire Book Company.

813