No.47 May 2006 |
|||
print view |
Abstract
The majority of professional football clubs and National teams now use
match analysis in order to provide objective information to guide the
decision-making process. In this article, we highlight how technological
enhancements have greatly improved the efficiency and value of the match
analysis process. In particular, we consider how National coaches may
use this technology during, and as part of their preparations for, the
2006 FIFA World Cup. Finally, we consider the future of match analysis
and provided some examples of how technology is likely to play an increasing
role in the game in future years.
Key Words: performance analysis, soccer, strategy, tactics, motion
analysis Introduction
The continual drive for success has led coaches to search for different
means of evaluating and improving performance at all levels of the game.
There is a permanent need for objective, accurate and pertinent feedback
on match performance (Carling et al., 2005). Coaches are unable to recall
every sequence of events from a match accurately and may fail to appreciate
where successful plays originated or mistakes began. Match analysis
provides a factual record of events provided the data collection methods
are reliable, accurate and objective. Moreover, the process of match
analysis provides a means of evaluation through quantifying and qualifying
various characteristics of individual and team performance.
Match analysis systems can be designed to collect data on several aspects
of performance, embracing physical, technical, tactical and behavioural
factors (Carling, 2003). These data can help the coach to determine
the most effective strategy and tactics that should be employed to maximise
the chance of success by providing objective data upon which to base
the decision-making process. Feedback on performance can be presented
to players during team-talks in various formats using, for example,
edited video sequences, graphs, tables, animated game reconstructions
or simple descriptive reports. These team-talks provide a collective
opportunity to appraise performance critically and suggest areas for
performance enhancement. Finally, relevant coaching sessions can subsequently
be planned and implemented to prepare optimally for future competition.
The typical coaching cycle is presented in Figure 1, with the importance
of observation and analysis highlighted. A match analyst will observe
and analyse competition to collect information on various aspects of
performance, before collating and presenting the results. The coach
can then examine and interpret this information before planning and
preparing future match strategy/tactics. Probably the most important
stages in the analysis process are the interpretation of data and how
findings are integrated and applied on the training field. The process
invariably depends on the type and quantity of data collected and the
way these are presented as well as the time available to the coach for
optimum interpretation. The analysis cycle can be repeated following
each match or at any time during the season. ![]() Figure 1. The coaching cycle highlighting the importance
of observation and analysis (adapted from Carling et al., 2005).
It is only in the last decade or so that formal match analysis has
gained widespread acceptance among soccer coaches, with most professional
clubs now enjoying access to match analysis in some form or other. Professional
clubs frequently employ computerised analysis systems and specialist
consultants or scientific support staff with expertise in this area.
The systems used by clubs often exploit the very latest in state-of-the-art
video and computer technology to record and evaluate their own team’s
performance to identify strengths that can be built upon and weaknesses
to be remedied during practice. Similarly, systems can be employed for
surveillance of upcoming opponents to provide information that can be
used to suggest ways of countering the opposition’s strengths
and to exploit weaknesses. A detailed knowledge of the opposition’s
strengths and weaknesses can help a team prepare effectively, leaving
nothing to chance; at the highest levels this factor may be the fine
line between success and failure. In this article we highlight how match
analysis systems have evolved over recent years and by way of illustration
provide examples to suggest ways in which match analysis techniques
will be employed by coaches in preparation for, and during, the 2006
FIFA World Cup Finals. The Evolution of Match Analysis Technology
Match analysis systems are becoming increasingly more sophisticated
and complex, mirroring quite closely the rapid developments in technology
generally. Over the last two decades, analogue video and, more recently,
digital technology combined with specialist statistical packages have
changed the face of the match analysis process and have been progressively
adopted by many top professional soccer teams across the globe. It is
now possible to have feedback about athletic performance available throughout
the cyclical process of competition, analysis and practice (Franks,
2004). The methods of collecting, storing, analysing and presenting
data have developed from using pen, paper and a calculator to semi-automated,
computer-based systems (Olsen and Larsen, 1997). Previously, simple
manual, hand-based notation systems were used by coaches at all levels.
These can be developed cheaply (pen and paper) and relatively easily,
and provide answers to questions posed by the majority of coaches, particularly
at lower levels of the game. However, the symbols used to code action
need to be learnt, a somewhat lengthy and tedious process, and the quantity
of information that may be collected in this manner is limited and time-consuming.
Computer and video technology has many advantages and contributes to
the match analysis process by facilitating data input and output. The
main advantages of computer- over hand-based systems are listed below.
It should also be noted that the price of audio-visual and computer
equipment has dramatically decreased over the years, although coaches
at the lower end of the game and in poorer countries may still not be
able to purchase the best systems or the most recent video equipment.
Furthermore, systems must comply with strict quality control specifications.
These include reliability, accuracy and objectivity (Carling, 2001a).
Only tailored, good quality and well-exploited match analysis systems
and equipment can provide a sound basis to analyse, evaluate and better
understand performance.
In performance analysis, analogue video cassette recordings were initially
used as a means of visualising and analysing performance before being
gradually phased out by advancements in digital technology. The use
of digital technology offers several advantages, notably:
Most modern systems combine a digital video feed into a computer with
statistical input using a specialised software interface. As the match
progresses, the analyst will enter or “code” information
on specific match actions often pre-defined by the coach. For each action,
the player’s name and pitch position, and the time and type of
and action, are manually inputted. This process can be carried out in
‘real-time’ or post-match depending on the quantity and
type of information required and the needs of the coach. Small handheld
or palm top computers have also been developed to code match actions
using specially adapted analysis software. Their portability and user-friendliness
mean they can be used to analyse performance from any position in and
around the pitch.
These video-based statistical systems do have certain limitations.
For example, pitch positions of player actions are often determined
by clicking on a schematic pitch representing the playing area. This
procedure will to a certain extent lead to the production of inaccurate
positional data. Furthermore, as video is often restricted to a single
camera viewpoint of the action, analysis of performance is generally
limited to only the player in possession and those around the ball.
There are further problems if broadcast television footage is used due
to the prevalence of action replays. As the replay is being shown, the
viewer cannot analyse the on-going action and may miss several key incidents.
Finally, these systems do not allow analysts to monitor physical performance
through player motion analysis. In other words, it is not possible to
recreate and analyse the movements and actions of every single player
during a match. This potential lack of measurement precision and sensitivity
has led to the development of high-tech video tracking analysis systems
(e.g., ProZone by ProZone Group Ltd and AMISCO by Sport-Universal Process
Ltd) which automatically and accurately track, calculate and recreate
players’ positions and movements.
The AMISCO system initially developed in 1996 requires
the installation and optimal positioning of several cameras to cover the
whole pitch so that every player is always captured on video, whatever
their position on the pitch and the moment in time (Brulé et al.,
1998). This process is achieved through the calibration of stadium and
pitch information (height, length, width) which is then transformed into
a two-dimensional model to allow the calculation of players’ positions
from the raised camera viewpoints. Using complex trigonometry, mathematical
algorithms and digital video/image processing techniques, each player’s
positions and movements can be calculated and tracked at every single
moment of the game as illustrated in Figure 2. AMISCO was the pioneer
player-tracking system and is now used by many top European clubs. It
allows analysis, post-match (although if a coach chooses to track a limited
amount of players e.g. for comparing two midfielders, this can be done
in ‘real-time’) the movements of all players, referee and
the ball, ten to twenty-five times a second during the whole 90 minutes.
This process leads to a database for each match containing around 4.5
million positions as well as 2500 ball touches. Some human input is still
required to check and complete the data. For example, the system may encounter
difficulties in identifying and tracking players during corner kicks (due
to the sheer number of players in a confined area) and various match actions
evidently cannot be automatically calculated by the system (e.g., red
cards, tackles, headers). The major advantage of this type of player-tracking
system is that players do not have to be equipped with an electronic tag
which up to now, is forbidden by all soccer authorities. Although other
‘real-time’ tracking systems are currently being developed
and tested around the world using electronic tracking equipment such as
microprocessors placed on players), the AMISCO and Prozone systems are
currently accepted as the standard in contemporary match analysis. ![]() Figure 2. An example of player tracking using the AMISCO system (Carling et al., 2005). One major problem encountered by coaches over the years is the storage,
retrieval and comparison of data from previous matches. A major evolution
in the development of computerised systems is the incorporation of databases
to store automatically the information obtained from match analysis
and allow trends in performance over defined periods of time to be analysed.
Databases can store and combine any type of information on physical,
tactical and technical performance. For example, the database can be
queried to know how often over a 10-game period, the player’s
high-intensity work-rate drops at the end of a game (e.g., number, duration,
length and speed of sprints). The database can then provide technical
information on the passing success rate as well as positional data on
the same period of the match. This information can then be graphically
displayed to show, for example, that a midfielder’s performance
tended to decrease in the final ten minutes of match (increased recovery
time between sprints, lower passing success rate and stayed in more
defensive areas hence also contributed little to attacking play). A
specific fitness training programme may then be implemented and an analysis
of the next 10 games can be made to evaluate whether the problem has
been remedied. Match Analysis and Preparation for the FIFA World Cup Finals Tournament
The vast majority of coaches will use match analysis in some form or
other when preparing for a major tournament such as the World Cup Finals.
The coaches’ analyses are likely to be based around three major
areas:
It is important to mention that coaches of National teams have a major
advantage compared to club coaches in that they have more time spare
to observe and analyse performance, whether it be that of their own
team or of future opponents. The coaches of teams playing in domestic
and international competition will have a limited amount to spend on
analysing team performance and even less time to focus on the opposition
due to the higher frequency of matches. On the other hand, the club
coach has the advantage of seeing the players on a daily basis in both
training and competition and will probably have a better idea of how
they are currently performing. However, most National team coaches will
try to keep close contact with club coaches so as to receive regular
updates on player performance. We now provide a brief overview of the
type of information that may be collated in each of the three areas
highlighted above. Analysis of the performances of one’s own team during the World
Cup qualification campaign and international warm-up matches
Most coaches will employ match video recordings to observe, analyse
and evaluate previous team and player performances. Players can relate
better to video than pure statistics due to its familiarity and the
notion that ‘a picture paints a thousand words’ (Robertson,
2002). Video recordings supply valuable information that may have been
missed or forgotten by coaches and players during the match. More importantly,
they provide an opportunity to play back match actions repeatedly to
demonstrate errors or identify areas that may require improvement. Also,
this medium may be used to highlight aspects of successful performance
such as defensive structure or delivery at set plays (Kormelink and
Seeverens, 1999). These examples of good performance can be memorised
and re-used or adapted by coaches to create ‘motivational’
videos. The coach can present an edited video sequence of important
match actions and repeat the action sequence as often as necessary to
ensure that the players have absorbed and understood the required information.
The action can be played back at different speeds (e.g. slow-motion
replay of all the opposition’s goals from open play) or paused
to highlight a particular issue (e.g. the positioning of defenders and
goalkeeper during opposition set-plays). Slow-motion replay is extremely
useful when breaking down individual game sequences. For example, the
team may have been caught out several times by opposition counter-attacking
situations. Using a slow-motion facility, the coach can ‘walk’
the players through the critical movements and freeze-frame to highlight
player positions and actions. This information can then be presented
against an example of good performance (where the counter-attack was
successfully broken down).
The player-tracking systems mentioned earlier are sometimes utilised
by national teams. Analysis of the physical performance of a player
obtained by means of motion analysis is more limited in value in one-off
matches at international level compared to the high frequency of matches
played at club level. At club level, a coach can build up a profile
or benchmark of a player’s physical capacities and compare performance
against this yardstick on a game to game basis. For example, comparison
of the total distance run which is representative of the overall severity
of exercise (Reilly, 1994) by a midfielder over several games may provide
a useful picture of his current efforts. However, unless data from club
performances are available, it may be difficult to determine whether
a midfield player is performing at optimal level during an international
match. A coach could compare data from the same match against those
obtained on fellow midfield players (own or opposition), albeit bearing
in mind that various tactical choices may have limited the player’s
contribution (e.g. a player who ran less may simply have had a more
defensive or ‘holding’ role). The coach may prefer to evaluate
high-intensity actions, to see if the player sprinted less at certain
moments in the game, although again it is difficult to draw conclusions
from a one-off performance. The coach may even have data from several
previous international matches or even better (but more unlikely) access
to data supplied by clubs which can be used for more accurate and objective
comparisons.
The collection of precise positional data via tracking systems may
be more important than information on physical performance at international
level. These data can be used to create animated, two-dimensional reconstructions
of key action within a match as illustrated in Figure 3. These reconstructions
can be used to visualise and dissect the movements, positions and actions
of each player at every moment of the game. Similarly, team tactics
and strategies can be easily visualised and broken down by a coach and
then presented to players. In contrast to conventional film footage,
these systems provide valuable ‘off the ball’ information.
This latter factor may be particularly important given that players
are often only in possession of the ball for around 2-3% of the match
duration. For example, the coach may want to know how an opposition’s
goal was scored. Unfortunately, the normal video footage was concentrated
on the player in possession and those around him and the scoring midfield
player’s run from deep was not picked up. This focus on the ball
to the exclusion of information relating to ‘off the ball’
activity is particularly frustrating for coaches. ![]() Figure 3. The software presented above developed by the French company Sport-Universal
Process is used by many top European Clubs to reconstruct and dissect their own teams and opponents play.
The graphical line helps the coach to visualise team shape (adapted from Carling, 2001b).
The overhead, animated viewpoint offers a fascinating insight and unlimited
possibilities to visualise and analyse match performance.
Analysis of the current performances and form of one’s own players
for their home clubs during domestic league and cup games and international
competitions
The methods of analysis used for evaluating the current performances
and form of national team members when playing for their domestic clubs
tend to be more simplistic than that used for analysing the opposition.
The coach or one of his support staff will often see a player perform
‘live’ and simply take notes or draft up a descriptive scouting
report. Video from broadcast television coverage of a player’s
performance may also be employed if further analysis is required or
if the coach has not been able to see the match ‘live’.
In these cases, statistics are rarely used and the coaches tend to create
descriptive reports, subjectively using their own experience, knowledge
and judgement of player performance. These reports often contain analyses
on how well a player is performing tactically and may also consider
players’ work-rates and behaviours (e.g., body language, decision
making). For example, is a centre-forward making the right tactical
runs behind the defence, is (s)he tracking back to help the defenders
when possession is lost, or does the body behaviour suggest a lack of
confidence when a scoring opportunity is missed?
It is possible the national coaches may access data
from support staff at a player’s professional club, but this would
depend on the nature of the relationship between the club and national
coach and the sensitivity of the data. As the World Cup usually takes
place at the end of the domestic league season, some players may have
participated in up to sixty matches for their club and country and fatigue
may take its toll. Motion analysis data on the physical performance of
players could be useful in determining the current workload level at which
certain players are operating. It may be necessary to access detailed
information on work-rates over the whole season to check if there is any
reduction in physical performance over recent club matches. For example,
has the overall distance run dropped or does the time spent in high-intensity
exercise tend to decrease in the second half compared to the first half
or towards the end of matches. The output from this type of analysis is
presented in Figure 4. The importance of high-intensity efforts in soccer
cannot be over-emphasised and a close evaluation of a player’s ability
to perform maximal or near-maximal sprints repeatedly (speed-endurance)
may be necessary. Soccer players may experience fatigue during a match
regardless of fitness level but, with appropriate training, the periods
of fatigue during match-play can be shortened or delayed (Bangsbo, 1994).
![]() Figure 4. Real-time analysis every five minutes of the maximal sprint
speed performance of an international midfielder over a whole match
(from Carling, 2002). It can be seen that the player is suffering from
a decline in high-intensity exercise performance during the last few
minutes of this match as maximal sprint speed has decreased; an opponent
may be able to exploit this weakness. A coach may want to know if this
apparent weakness is a common occurrence by comparing data obtained
from other matches. As is often the case before a World Cup, some players
suffer from injury and may just be returning to competitive action. Work-rates
can be used to analyse performance by checking the total distance run
or number of sprints in matches made before and after injury. Such analyses
can help a coach objectively identify if a player is performing at pre-injury
levels. When players meet up for National duty, physical performance data
previously obtained from match-play can be combined with quantitative
information from fitness tests and/or heart rate measurements taken in
training sessions to provide vital indicators on the current fitness level
of players and help determine the intensity of future training sessions. Analysis of future opponents including group stage adversaries and
possible opposition in view of qualification for the next round
Traditional descriptive scouting reports are often the first step used
by coaches to provide an introductory picture of general opposition
play. A national coach may send an observer to see a team play ‘live’,
or to watch match video footage, draft up a report which can be complimented
by further analysis using video and/or statistical information. The
coach can immediately identify strong and weak points from the report
before starting to implement a ‘game plan’. These reports
are traditionally used to outline opposition strengths and weaknesses
in strategies, playing systems, tactics, techniques and even behavioural
characteristics from both an individual, unit and team viewpoint. Skills
such as space creation, penetration, width, improvisation, and switch
from defence to attack may be considered. Similarly, defending skills
such as support play, reducing space, depth, delay, balance, and the
ability to regain possession can be covered. The use of edited video
sequences showing individual and team strengths and weakness is commonly
used during team-talks to illustrate key points. Some of the typical
footage presented in these team sessions is highlighted in Table 1.
The coach will want to present and discuss the footage in detail using
the playback techniques mentioned earlier and present ways of countering
these strengths or exploiting any apparent Achilles’ heel. Individual
players may be provided with a report and edited film sequences, perhaps
on a DVD, of his/her direct opponent(s). For example, an opponent may
be singled out as lacking pace or discipline. Coaches often avoid using
match statistics such as percentages, totals or averages unless really
pertinent. If a team scores many goals or creates chances from set-plays,
a percentage for success rates may be given that is easy to understand. Match Analysis during the World Cup
Most National coaches will use some of the various match analysis technologies
mentioned above, and in particular video due to its efficiency in examining
performance, during the tournament itself. Some coaches may be in an
unfamiliar position as a result of having less time to spend analysing
performance due to the higher frequency of matches. Therefore, time
is at a premium, the match analysis process needs to be streamlined
as coaches must be able to evaluate and assimilate information on performance
quickly and easily. A coach will tend to concentrate on simply identifying
areas requiring improvement, highlighting both errors and instances
of good performance and then presenting these to players in short, efficient
edited video sequences. Coaches will also want to remind players of
the strengths and weaknesses of their next opponents and demonstrate
the optimal countering strategy and tactics. Care will be taken during
team-talks not to alienate players who have made errors, and not to
over-emphasise the strengths of the opponents and make players anxious,
and not to underestimate the opposition. Many National Federations will
also want to carry out a post-World Cup review on their team’s
performance, or on the tournament as a whole, using one or more of the
methods of analysis previously discussed. The Future of Match Analysis
Soccer coaches especially at elite levels are constantly looking for
new ways to improve performance and consequently, they now need to be
at the forefront of developments in technology. The technology currently
used to analyse players’ performances will no doubt continue to
play an important part in the coaching process and move forward in the
way it already has done over the last few years. However, coaches must
always be aware of how state-of-the-art systems and technology can be
efficiently implemented into the busy working practices of a club or
national team and what they can really offer in terms of enhancing performance.
The technology used should enhance the process of match analysis and
support the coaching process with the eventual output always being that
of improving performance (Robertson, 2002). Future systems will certainly
be much more complex and powerful in terms of reducing the amount of
time required to analyse matches and in the quantity and precision of
results obtained. The ultimate problem facing the coach and the analyst
is how best to transform oceans of data into meaningful interpretations
(Hughes and Franks, 2004). Contemporary use of such systems will go
beyond the classic performance analysis cycle by enabling the prediction
and modelling of forthcoming contests which is a fundamental function
of match analysis (Franks and McGarry, 1996). In the near future, styles
of play and patterns of movement may be simulated either as physical
models or as computer-driven virtual reality scenarios based on actual
match data.
More recently, systems have been developed using Global Positioning
techniques and ultrawideband radio frequency procedures and these have
much promise for the systematic tracking and recording of players’
locomotion during training and matches. This approach requires tagging
individual players electronically and tracking their movements by means
of satellites or radio transmitters and receivers positioned around
a pitch. Speeds and distances can be measured in ‘real-time’
from the co-ordinates obtained, for both players and the ball, providing
a level of precision and quantity of data not yet seen. These tracking
systems will provide ‘real-time’ feedback on performance
enabling coaches to make split-second decisions at any moment in the
match. Although the technologies are currently available for such tracking
systems to be used in soccer, existing rules restrict their development
and hence they are only used in training sessions.
Expert Systems (an application that uses a knowledge base of human
expertise to aid in solving problems) using artificial intelligence
(which is the branch of computer science concerned with developing computerised
systems to mimic human behaviour) will probably play a major role in
the modelling and prediction of performance. Expert Systems using a
database containing the very best in human coaching knowledge and expertise
will be designed to analyse, understand and interpret the masses of
data obtained from match analysis. The Expert System will propose the
most optimal strategies/tactics and corresponding training sessions
as well as attempting the prediction of the future outcome of matches
(Carling, 2003). It is also worth mentioning that the Internet is another
modern facet already playing an important part in the match analysis
process due to its flexibility in the way information and data about
match performance are readily transferred and made available for viewing
by coaches. With the advent of broadband connection, video based performance
analysis and viewing will soon be possible from anywhere in the world.
Finally, whatever the system and technology used in the future by coaches,
it is always important to weigh up the expenditure compared to the results
obtained. Do improvements in performance outweigh the initial cost of
the system and the time and effort required to measure match performance?
Professional teams in English football that have used one of the current
high-tech player-tracking systems have reportedly improved on average
by 5 points per season leading to a higher final league position and
much greater prize money which has seemingly more than justified the
investment in match analysis (Prozone Group Ltd, 2005). It will be interesting
to see whether the future developments in technology mentioned above
have similar or even greater success in helping improve elite performance
in soccer.
References
Bangsbo, J. (1994). The physiology of soccer-with
special reference to intense intermittent exercise. Acta Physiologica
Scandinavica, 15 (suppl 619), 1–156.
Brulé, P., Carling, C., David, A. (1998).
The development of a computerised match analysis to automatically track
the movements of soccer players. In Book of Abstracts. Notational
Analysis of Spor,. IV World Congress, Porto, p.36.
Carling, C. (2003). Outils d'évaluation des
stratégies de jeu en sports collectifs. Cahiers de l’INSEP,
34, 147-151.
Carling, C. (2001a). Choosing a Computerised Match
Analysis System. Insight, The Football Association Coaches Association
Journal, 3 (4), 30-31.
Carling, C. (2001b). Getting the most out of Video
and Match Analysis. Insight, The Football Association Coaches Association
Journal, 1(5), 16-17.
Carling, C. (2002). Football: A game of chance or
does match analysis have the answers? Insight, The Football Association
Coaches Association Journal, 2 (5), 41-43.
Carling, C. Williams, A.M. and Reilly, T. (2005).
The Handbook of Soccer Match Analysis: A Systematic Approach to
Improving Performance. Routledge: London.
Franks, I. (2004). The need for feedback. In Notational
Analysis of Sport: Systems for Better Coaching and Performance in Sport.
2nd edition. (edited by M. Hughes and I. Franks), pp. 8-15.
Routledge: London.
Franks, I. and McGarry, T. (1996). The science of
match analysis. In Science and Soccer, (edited by T. Reilly),
pp. 363-375. E. & FN Spon: London
Hughes, M. and Franks, I. (2004). From analysis
to coaching. In Notational Analysis of Sport: Systems for Better
Coaching and Performance (edited by M. Hughes and I. Franks), pp.
257-271. Routledge: London.
Kormelink, H. and Seeveren, T. (1999). Match
Analysis and Game Preparation Book. Reedswain: Pennsylvania.
Olsen, E. and Larsen, Ø. (1997). Use of match
analysis by coaches. In Science and Football III (edited by
T. Reilly, J. Bangsbo, M. Hughes), pp. 209–20. London: E &
FN Spon.
Prozone Group Ltd (2005) Why Prozone ? http://www.pzfootball.co.uk
(accessed 25/3/06).
Reilly, T. (1994). Motion characteristics. In
Football (Soccer), (edited by B. Ekblom), pp. 31-43. London: Blackwell
Scientific.
Robertson K. (2002). Observation, Analysis and Video. The National
Coaching Foundation: Leeds, UK. Table 1. Some match factors on opposition teams commonly looked at during team-talks using video analysis (adapted from Carling et al., 2005). ![]() Contact: A. Mark Williams Research Institute for Sport and Exercise Sciences, Liverpool John Moores University, The Henry Cotton Building, 15-21 Webster Street, Liverpool, L3 2ET, UK, Tel: +44 151 231 4334, Fax: +44 151 231 4353, E-mail: m.williams@livjm.ac.uk ![]() http://www.icsspe.org/portal/index.php?w=1&z=5 |