A technical evaluation was achieved by Partridge et al., (1993). A specialised computer analysis system was developed, using 38 key events entered in real time by a trained analyst. The system was used to provide a comprehensive technical evaluation of performance by comparing the results of 2 distinct levels of performance, the 1990 FIFA World Cup and the 1990 World Collegiate Soccer Championships. From the results it can be inferred that collegiate coaches must be selective when presenting World Cup teams as an appropriate model of performance as many differences do occur, which makes any comparison invalid.
At International football level, where games are decided by small margins, the team that is superior in physiological and motor abilities will have the advantage (Reilly and Holmes, 1983). This places a high emphasis on a team at the elite level possessing high levels of technical ability.
Within the literature, there are very few examples of technical analysis, in particular skill analysis involving association football. In this respect, the following study is innovative in two ways:
i) By analysing at the exact technical requirements of each position
- By using qualitative data within a quantitative system.
1.1 Aim of Study
The aim of this study is to analyse every individual’s technical ability that competes in the European Football Championships of 2004. This measure will be based on a subjectively drawn continuum that analyses a player’s technical movement throughout the game.
It will be investigated if technical differences occur between player positions and between successful and unsuccessful teams.
Data will be gathered from matches within the European Championships of 2004, which were held in Portugal. This tournament has been chosen as it provides an ideal environment for a comparison to be made between elite level players competing in an elite sporting environment. Data will be collected using a hand notation system, in the form of a table. The table will consist of six columns: player number; technique performed; technique rating, pitch position, time of action and any outcome (if applicable).
1.2 Hypothesis
The hypothesis states that there is a significant difference in the distribution of technique between the different playing positions and between successful and unsuccessful teams at the European Championships 2004.
1.3 Assumptions
Certain assumptions are considered during this study. It is assumed that all players selected are of international standard and that all players are competing in the tournament to win.
1.4 Limitations
Limitations of the study that depict the nature and length of the investigation include the word limit imposed and the submission date. Variations from the tournament itself may also not give the data a true reflection of reality. Incidents such as injury to key players, conditions and the environment may impact on the manner in which a team or certain individuals perform. As the matches were all recorded from terrestrial television, some data cannot be collected due to its absence from the showing of action replays and other inappropriate footage (Winkler, 1996). The choice of some camera positions and angles also makes some forms of player identification difficult.
1.5 De-Limitations
A de-limitation from our research design is that it is impossible to make broad generalisations from our data. The obtained data is so specific to the tournament and elite level players that any such generalisation would be inaccurate and invalid. The analysis is also produced in a very subjective manner using operational definitions and a likert scale, which may also not be universally agreed by people other than the researchers.
1.6 Definition of Terms
Technique: The method of performance of an individual against a normative scale.
Player Position: An individual’s role and function within a team’s structure.
- Literature Review
2.1 Notational Analysis in Football
Some of the earliest analysis of association football was carried out by Reep and Benjamin (1968). They analysed over 3000 competitive games from 1953 to 1967 within top division English football and World Cup matches using a simple hand notation system. The position of player’s actions, the amount of passes before goals and the way in which possession was gained were all notated. From this extensive analysis it was reported that:
- 80% of goals resulted from 3 passes or less
- 50% of goals result from regaining possession in the final attacking ¼
- It takes 10 shots to score 1 goal
Many coaches and managers throughout the English game saw these results as an immediate formula for success. The transfer of these findings into a game situation saw the arrival and development of the traditional long ball game within British soccer.
The work of Reep and Benjamin was later taken on further by Bate (1988) in an attempt to disprove a modern day notion that maintaining possession of the ball was the key to accomplishment. From analysing games in the English 3rd division, right up to World Cup level competition, Bate concluded that in order to be a success, teams must portray the following characteristics in their play:
- Play the ball forward as often as possible
- Reduce the number of square and backwards passes
- Increase the number of forward passes and forward runs
- Play the ball into the space behind the defenders as early as possible
From Bate’s (1988) analysis, it can be seen that the determinants of success across a wide range of football were relatively unchanged for a long period of time and that the notion of possession football was not the key to success.
Further characteristics of successful play were noted by Winkler (1996). He stated that good teams are capable of defending their own goal well as well as creating more goal scoring opportunities than the opposition. His findings were however contradictory to the research of both Reep and Benjamin (1968) and Bate (1988) by suggesting that successful teams should keep possession of the ball for longer, as opposed to getting the ball as forward as quickly as possible.
More recently Luhtanen et al., (2001) adapted a methodology from Luhtanen (1993) to study selected offensive and defensive variables of individuals within the European Championships of 1996 and 2000. The study aimed to examine if any correlation occurred between performance and final position reached in the tournaments. A computerised notation system was used to notate over 2000 actions per game. The actions of the individuals could then be calculated to give means for each of the teams per selected variable. France, who were the winners of Euro 2000 were found to be the highest ranked nation in passing, receiving the ball, running with the ball and tackling. Luhtanen et al. (2001) concluded that this consistent high ranking suggested that France were worthy winners of the tournament. The winners of Euro 1996 however, Germany, were not found to be the highest ranked in any selected variable. Luhtanen et al. (2001) explained this concept by stating that the Germans performed very ordinarily throughout the tournament, and a great deal of luck brought them success. The selection of performance indicators for use in the study may however have only been conducive to certain patterns and styles of play. Although these may have highlighted the attacking nature of France’s performance, they do not take into account factors such as levels of team cohesion and team organisational structure. It may have been high levels of factors such as these which brought the Germans success, not merely luck, as was stated in the review.
Luhtanen et al., (2001) summarised that comparisons between team performances across tournaments or time spans couldn’t be made absolute due to the numerous changes in team selections, opposition, tactics used and managerial changes.
2.2 Factors determining success within association football
Vast amounts of literature are apparent concerning the various aspects of a team’s performance, especially those, which bring success.
Hughes et al., (1988) developed a methodology to differentiate between successful and unsuccessful teams from the 1986 World Cup by examining the main characteristics of play when in possession of the ball. Successful teams were defined as those teams who progressed to the semi final stages, whereas unsuccessful teams were those who were eliminated after the 1st group stage. Analysis was conducted using 24 performance variables. It was found that successful teams have more touches in possession, had more shots at goal from within the penalty area and approached the final sixth of the pitch by playing predominantly in the central areas. Unsuccessful teams dribbled more and played the ball to the wide areas in their own defensive zones more frequently.
The findings of Hughes et al., (1988) study were contradicted by Ali (1988) who conducted a statistical analysis of patterns of play in 18 Scottish Premier Division matches. It was concluded that attacks which proceeded along either wing were more successful, rather than those which were played more centrally.
Teams using low pass combination moves combined with centrally based attacks were the factors deciding success in Bishovets et al., (1993) study of 52 World Cup matches. This manner of performance was related to winning teams have a more consistent and reliable understanding between players.
Pearce and Hughes (2001) conducted analysis of the perceived successful impact of substitutions in the 2000 European Championships. Data was gathered in the 15 minutes preceding a substitution and in the 15 minutes afterwards. Each performance variable was evaluated and ranked depending on its value of influence. This allowed a show of ratings along a continuum to be created, both of the individual substitute and of the team’s performance following the substitute’s introduction. Although it was found that 15 minutes of analysis did not produce enough data, midfield was seen to be the most hazardous player position in which to introduce a substitute, due to the high work intensities required immediately by the incoming player.
The idea of coding actions according to their difficulty and their success can also been seen in Rico and Bangsbo (1997) study of Denmark’s performance in the 1992 European Championships. Actions were noted according to the amount of pressure applied on the performer and conclusions were drawn from the results. The decline of Denmark’s passing in their final 3 games was put down to the increased pressure and closeness of the score line related to the latter stages of the tournament.
Dooan et al., (1996) conducted a further study to determine the factors, which promote success within soccer. The importance of playing at a high tempo was recognised, due to the constant opponent and time pressures that are placed upon performers. He aimed to compare the efficiency of pass in elite versus non-elite Turkish performers. All passes were recorded as either positive or negative, according to their perceived degree of difficulty and success. It was concluded, as expected that speed, fluency and intensity of pass are all more apparent for elite performers.
A further technical study on elite Turkish footballers was reported by Eniseler et al., (1996). The researchers found that Galatasary’s failure in European football competition was due to technical and tactical inadequacies as well as a lack of physical conditioning. This method of performance profiling also included recording variables as either positive or negative, according to their execution.
A tactical and technical inadequacy leading to failure was also reported by Acar (1996) who analysed the performance of teams playing in derby matches, using both computer and hand notation.
The stated research has all described the relative success of play in relation to entire team performances. A team is defined as a system in that a group of players interact in a dynamic fashion concerning a single purpose (Pinto, 1998). Despite a team participating as a homogenous unit, it is also vital to remember that the different individuals comprising a single team make it heterogeneous at the same time (Pinto, 1998).
2.3 Evaluation of Individuals Performance
Success in football is very much judged on a team’s ability to win matches (Luhtanen et al., 2001). The study of individual’s performance can be regarded as invaluable as it is in essence these individuals which comprise any team. Success at any level collectively cannot be achieved without the performance of individuals within this team unit.
A great deal of individual analysis has been conducted on the physical demands of football competition and the necessary physiological state for optimum performance (Reilly and Thomas, 1976; Bangsbo, 1997; O’Donoghue et al., 2001). Wells and Reilly (2002) attempted to research into the demands of playing position within women’s soccer also taking into account performance variables. However the only performance variable measured was kicking distance. This variable seems very irrelevant to determining successful performance within a game situation, particularly within elite level International football.
The lack of literature relating to the importance of individuals levels of skill was identified by Reilly and Holmes (1983). Reilly applied 2 methods in order to investigate the notion of skill distribution within soccer. Match analysis of 6 non professional games was carried out looking at skill performances as either successful or unsuccessful. In addition to this analysis, a group of 40 adolescent males, from a variety of outfield positions performed a cross section of skills tests. From the 2 tests, Reilly concluded that significantly:
- Success rate of each skill depends upon pitch location
- Defensive area provides highest skill success rate
- Midfield players show more superior test scores to defenders
Reilly explained that as the space available from the opponent’s goal increases, the time available on the ball also increases. This explains why the most successful skills are performed in the defensive area, where less pressure is applied. It was also suggested that a common trend is to place the most poorly equipped players in defensive roles, hence the defenders performing worst on the skills test.
James at al., (2002) recognised the importance of studying individuals within a team. He stated that this level of deeper analysis allows a much finer grained overall team examination. In a study of the same team over 21 matches, across various competition types, the analysis aimed to identify the different roles individuals may take across differing circumstances. It was noted that in European competition players played more defensively and played passes involving much less risk.
Dufour (1993) completed a technical analysis of outfield players, summarising the percentage of time spent in each action category. For an outfield player it was reported that on the ball playing time was divided into:
50.6% Intercepting, 22.4% Passing, 18.7% Controlling ball, 4.5% Tackling, 2.4% Shooting and 1.4% other activities. Although this analysis enabled a template of the aspects contributing to an outfield player’s role, no specification was made relating to playing position, taking into account the obvious differences which occur between the outfield positions.
The lack of individual analysis relating to the highly specialised position of a goalkeeper was identified by Wooster and Hughes (2001). Goalkeepers were seen as vitally important, as they provide the last line of defence and the first line of attack for any team. From studying 1126 goalkeeper distributions at Euro 2000 with a hand notation system, it was concluded that successful goalkeepers used an equal distribution of kicks and throws, with variety which enabled possession to be maintained in the attacking third of the field. 24% of the 67 goals scored within the tournament all originated from successful goalkeeper distribution.
Researchers have also used other methods, other than standardised hand and computer notation systems in order to investigate the performance of individuals.
A qualitative analysis of individual movement patterns was performed by Grehaigne et al., (2001). A player’s effective play space was considered by drawing polygonal lines to create an individuals players action zone. This action zone represented the areas in which 80% of the players activities were performed.
Graphic modelling and statistical calculations were used by Chernenjakov and Dimitrov (1988) to help detail individual players performances. By entering specific data into a computer, it allowed players to be arranged in order of their playing effectiveness.
Erdmann (1993) described how due to the imprecise, subjective and ambiguous nature of qualitative observations, all quantitative analysis should be performed based upon mechanical properties. Erdmann introduced how by looking at the Kinematics of movement, such as the displacement, velocity and acceleration of movements, individual performance profiles could be built.
2.4 Use of Match Analysis by Coaches
As stated by Coghlan (1990) there is enormous pressure on football mangers throughout the world to succeed. In an attempt to bring about such success, notational analysis is being increasingly used within the modern game (Partridge and Franks, 1997). It can help to provide coaches with detailed analysis of observations that would otherwise be missed (Coghlan, 1990). Many coaches now consider information derived from such technological advances to be invaluable (Liebermann et al., 2002). With the pace of the modern day game, team strategies and tactics should be based on something more substantial than opinion (Bate, 1988).
Match analysis enables accurate, objective and relevant feedback to be applied to a coach or a performer about past performances (Franks and McGarry, 1996). Due to coaches being active, biased observers of their team’s performances, their observations are often not accurate. In a study by Franks and Miller (1986), International level coaches were only able to recollect 30% of the key elements which determined performance with the use of their memory alone (cited in Franks and McGarry, 1996).
The presentation of accurate results takes away any subjective views and opinions and allows a coach to build up a portfolio of all future opponents in order to prepare to play against them (Pollard et al., 1988).
Olsen and Larsen (1997) showed how the outcome of analysis can be a tool for evaluation and for the development of team tactics. The system responsible for such analysis must however be valid, accurate and easy to use, without large tables of complicated figures and graphs (Gerisch and Reichelt, 1993). Such simple but detailed analysis systems have enabled Norway to maximise its limited resources and to compete on the International football arena (p.220).
Franks and Goodman (1986) stated that an objective quantification of critical events during a game is critical for a complete post match analysis. The generated analysis must be used by coaches to instigate an observable change in behaviour, and so an improvement in performance. Liebermann et al., (2002) and Franks and McGarry (1996) reinforced this notion by stating that appropriately timed feedback can significantly improve motor skill acquisition and performance.
Reporting to the results of both qualitative and quantitative feedback to a performer can increase performance (Partridge and Franks, 1997). Following matches, a subject was shown analysis results, and video clips from relevant expert performances which related to his previous performance. Over the course of 6 games, the subject’s performance across 15 measured variables did improve. This performance increase was directly related to the analysis procedures the player was subjected to between performances.
2.5 Individual Roles Within a Team Framework
Although a team unit is comprised of 11 individuals, all 11 players must assume certain roles and functions in order to make such a team unit a success.
Subconsciously, players and coaches alike have a universal knowledge of which technical components are required in order to play in each position within association football. There is however very little research to either reinforce or question these concepts. Having an exact technical analysis of the precise playing requirements of each position would allow accurate training schedules and more accurate player profiles to be established.
Many coaching publications state the necessary credentials to play in certain positions within soccer (Smith, 1973; Cook, 1982). These publications are however based upon opinion, as opposed to exact epidemiological research. Much ambiguity does exist between these opinions and the reported differences. Wiemeyer (2003) in interviewing 14 coaches, across varying participation levels in order to establish positional technical demands emphases this. In only one case did all coaches agree of the exact functions of a position. Many common features were however apparent of the requirement of players:
Table 1. Technical requirements of positions (Wiemeyer, 2003)
The goalkeeper’s roles and responsibilities were also highlighted by Hughes (1981). The 5 main areas of responsibility were seen as:
- Dealing with shots
- Dealing with crosses
- Supporting the defence
- Distribution
- Organising defence of set plays.
The tasks and functions of individuals within a team can differ according to whether a team has possession of the ball or not (Van Lingen, 1997).
Table 2. Individual tasks when in possession of the ball (Van Lingen, 1997)
Table 3. Individual tasks when not in possession of the ball (Van Lingen, 1997)
- Relation of literature to study
Despite the literature present in this field of notational analysis, there is an obvious opening for further research to be conducted. Reilly and Holmes (1983) identified that a further study could incorporate an analysis of skill performance within a competition context. Luhtanen (1988) also recognised how by using a similar notation system to his own an evaluation of individual and team skills in match conditions could occur.
The need for further research in this area is also highlighted by Dufour (1993) stating that “a techno-tactical profile may be generated in relation to an ideal profile corresponding to players place and function in a team” (p.165).
As stated by Franks and McGarry (1996), the ability to provide information about individual’s technical performance and the profiles of such players can significantly modify playing behaviour and promote successful performance. Information about technical performance is also much preferable to cursory comments made by coaches following competition (Franks and Goodman 1986). The recording of events in some coded form can help such coach observations to be formed, especially by defining each skill performed as successful or unsuccessful (Franks and Goodman, 1986).
If an accurate analysis of the technical attributes of each player position was able to be established then the results could significantly influence team selections and coaching sessions (Bate, 1996).
- Aims of study
This research aims to deliver such a technical analysis of the exact requirements of playing positions in elite football from the European Championships 2004. A comparison will also be made between successful and unsuccessful team’s profiles from the tournament.
It will involve a detailed quantitative analysis based on a qualitative notation system. It will aim to specify the technical differences between playing positions and the relative success in the performance of these techniques.
- Methodology
3.1 Introduction
The study consisted of a post-event hand notation match analysis system which was utilised by four researchers in order to collect data from every match from the European Championships 2004. The system was devised to give an accurate evaluation of the technical requirements of each playing position at this elite level of competition.
- Equipment
The matches which were analysed, were taped from terrestrial television (both BBC and ITV), onto an E-180 VHS videocassette using a Matsui video recorder VX1100. A Panasonic Quintrix television and a Panasonic NV-HS930 video cassette recorder were used to watch the tapes back in order for the analysis to be carried out. The pause and rewind functions of the cassette recorder allowed optimum accuracy whilst notating. The analysis was recorded onto paper, using black and blue ink to differentiate between the two teams involved.
- Pilot Study
A pilot study was conducted in order to familiarise the researcher with the system and the procedure of the analysis. As multiple researchers conducted the analysis, a series of operator training sessions were conducted to ensure all analysists were consistent and accurate in their analysis of events. After this period of training, 2 analysts analysed the first ten minutes of the Championship final alone after having talked through the system beforehand. The results were then compared and discrepancies were discussed and reasoned through. This process was repeated three times until the results across the researchers became universal. (See section 3.7 for inter and intra observer reliability).
- Data
Data were collected from all 31 matches which were played at the 2004 European Championships. The number of participants varied across the study due to changes in team selections and substitutions, which were made during the matches. Each of the four analysts were required to analyse a minimum of seven matches in order to guarantee all matches were notated. For this study the analyst was required to notate the following Group B’s games, along with one quarter final:
Group B:
Switzerland v Croatia France v England
England v Switzerland Croatia v France
Switzerland v France Croatia v England
Quarter Final:
France v Greece
Due to the imposed time restraint and the enormity of data gathered, only the above mentioned games were utilised for further analysis within the study.
- Procedure
The required recording of a game was inserted into the video recorder. When the team selections for the game were shown from the television coverage, the video was paused allowing the researcher to record the participating players and their proposed positions of play. Following this, the first ten minutes of the game were then viewed with no analysis taking place to ensure that the stated player’s were playing in their proposed positions.
Once the participating players and positions were established, analysis of the match began. The match was played and after every action performed by a player, the video was paused. This action was then notated in the following order onto a pre-composed data collection sheet:
- Player involved
- Technique performed
- Technique’s rating
- Pitch position of where technique occurred
- Any possible outcome resulting from the technique e.g. Goal is scored
- Time the technique occurred
Different colour pens were used to differentiate between the two teams on the data collection sheets; this allows instant recognition of periods of possession for each team. This method of collection also allowed the sequentially of the data to be maintained.
- Operational Definitions
It was important to establish operational definitions into which all players actions could be grouped in order to record all data. The following definitions were universally agreed across all researchers to ensure data was collected consistently.
In order to notate the different techniques performed by the players during data gathering, a key was used to denote which particular technique was used:
Table 4. Outfield player’s operational definitions used for study
Table 5. Goalkeeper’s operational definitions used for study
The continuum of technique ratings, which were given to each technique performed, were also agreed across the researchers:
Table 6. Continuum of technique ratings
The playing surface was divided by the researchers into distinct areas to allow a more detailed analysis of the techniques performed within a game. Having a perception of an actions pitch location gives a more accurate assessment of the conditions in which the action was completed.
Figure 1. Pitch division example as used in analysis (adapted from James et al, 2002).
Figure 1. Pitch division example as used in analysis (adapted from James et al., 2002).
This pitch division was chosen as it breaks down the playing surface into strategic, fairly broad, easily identifiable areas. The subdivisions of the penalty area were added from James et al., (2002) pitch division to allow for a greater depth of analysis in this critical area of the field.
- Data Processing
Data were tabulated and then entered into Microsoft Excel computer package for analysis.
The Chi squared test of independence was used to determine whether differences between the results were statistically significant. The Chi-Squared test of independence is a statistical test which provides the significance value of the difference between the observed and expected results. It provides the probability of a Type 1 Error (Thomas and Nelson, 2001). For an identified difference to be significant, the reported P value must be 0.05 or lower (p<0.05) to be accepted at the 95% level of significance (Vincent, 1999).
- Validity and Reliability
It is vital that the reliability of a data gathering system is clearly demonstrated (Hughes et al., 2002). As stated in section 3.2.1. all researchers collectively went through a series of operator training sessions in order to increase familiarity and reliability of the data. As a test of inter-observer reliability, 2 of the 4 researchers completed an inter-observer analysis of the opening 10 minutes of the championship final, between Portugal and Greece, allowing a test-retest-retest method (Hopkins, 2000).
The results of both the inter and intra observer reliability can be seen within the results section, in tables 7 to 12 respectively. This test of both intra and inter-observer reliability was achieved by using the Chi-Squared test of independence and the percentage error statistic, within the Microsoft Excel package.
For trials to be reliable, the calculated P value from a Chi squared test are required to be of a value of 0.95 or higher, to indicate that there is agreement between the trials and therefore also reliability at the p>0.95 level of significance (Bland and Altman, 1986).
The percentage error statistic is used to calculate the amount of error for each variable involved in the observation system (Hughes et al., 2002). The calculated values are required to be of 5% or lower in order to fall below the accepted 0.05% level.
- Results
4.1 Reliability
The Chi-Squared test of independence and the percentage error statistic were used to calculate the both the intra and inter-observer reliability.
Table 7. Intra-observer action observation reliability between T1 and T2
The Chi square P value of 0.95 indicates that reliability does occur between T1 and T2 for intra-observer action observation at the 95% level of significance.
The percentage error statistic however shows that reliability cannot be assumed as 10.6% is above the accepted 5% error threshold.
Table 8. Intra-observer action observation reliability between T2 and T3
Intra-observer reliability has increased for action observation between T2 and T3 to a P value of 0.99. The percentage error between the 2 data sets has also improved to an accepted value of 1.56%.
Table 9. Inter-observer action observation reliability between T3
A P value of 0.99 indicates strong inter-observer reliability between action observations at the 95% level of significance. The percentage error statistic shows however that reliability cannot be assumed at the 95% level of significance as the overall percentage error value is 5.32%.
Table 10. Intra observer technique rating reliability between T1 and T2
Reliability cannot be assumed at the 95% level between T1 and T2 as the Chi squared P value is below the required 0.95. The percentage error statistic confirms this as the figure of 13.02% is above the accepted 5% level.
Table 11. Intra observer technique rating reliability between T2 and T3
Reliability has however improved between T2 and T3. A P value of 0.99 indicates strong reliability at the 95% level of significance. This is reinforced with an accepted percentage error score of 4.8%.
Table 12. Inter-observer technique rating reliability between T3
The P value of 0.99 indicates strong reliability between researchers at the 95% level of significance.
The acceptance of reliability from the Chi squared statistic is however contradicted by the percentage error score. A percentage error of 9.37% between researchers suggests that reliability cannot be assumed at the 95% level.
- Processed Data
4.2.1 Distribution of techniques according to positional role
Figures 2 to 5 respectively show the frequency distribution of player’s actions across the 7 games according to their positional role.
Figure 2. Distribution of goalkeeper’s on the ball actions
Figure 3. Distribution of defenders on the ball actions
Figure 4. Distribution of midfielders on the ball actions
Figure 5. Distribution of strikers on the ball actions
Figure 6. Comparison of the techniques used between defenders, midfielders and strikers
Figure 6 shows a frequency calculation for the total number of actions performed across each position. It shows how midfielders consistently have the highest frequency count, except for heading, tackling and throw-ins, all of which are displayed by the defenders.
The Chi squared test of independence indicates that significant differences occur between the frequency distributions of defenders and midfielders (p<0.05), defenders and strikers (p<0.05) and between midfielders and strikers (p<0.05).
This trend could perhaps be explained due to the varying number of subjects analysed and the varying patterns of play used. In order to get a more accurate representation, a mean distribution of the quality of techniques performed per position will be more accurate.
Figure 7. Average quality rating of techniques used across playing positions
Displaying a mean technique rating per position shows a distinct deviation from the overall frequency distribution. From not being the highest ranked in any variable in Figure 6, Strikers are now found to be the highest ranked in receiving the ball (0.95), running with the ball (1.29), dribbling (1.3) and in the execution of free kicks (1.0). Defenders however still rank highest for heading the ball (1.10), tackling (1.28) and now also shooting (0.71). Midfielders have however retained their high rank for both passing (0.95) and crossing (0.90).
The reported P values from the Chi squared test of independence report that no differences between the average quality of ratings across playing positions are significant (p>0.05).
- Distribution of technique ratings across player position for selected performance variables
4.2.2.1 Passing
Figure 8. Distribution of defenders passing technique ratings
Figure 9. Distribution of midfielders passing technique ratings
Figure 10. Distribution of strikers passing technique ratings
All positions show a trend of +1and 0 rated passes being the most frequent. Apart from +2 rated passes, all other ratings show a low frequency of occurrence.
A significant difference was found between the passing distribution of defenders and midfielders (p<0.05) and defenders and strikers (p<0.05). No significant difference was found between midfielders and strikers (p>0.05).
- Shooting
Figure 11. Distribution of defenders shooting technique ratings
Figure 12. Distribution of midfielders shooting technique ratings
Figure 13. Distribution of strikers shooting technique ratings
The frequency distributions for shooting are far more diverse than for passing. More data can be seen at the 2 extreme ends of the continuum, especially for midfielders and strikers. A Chi squared test revealed that significant differences do occur between defenders and midfielders (p<0.05) and defenders and strikers (p<0.05).
- Heading
Figure 14. Distribution of defenders heading technique ratings
Figure 15. Distribution of midfielders heading technique ratings
Figure 16. Distribution of strikers heading technique ratings
Although defenders have the highest frequency of headers, they also are the most consistent with very few headers possessing a minus rating. Striker’s headers appear to be far more inconsistent, with a much greater data spread.
Heading the ball also revealed significant differences between defenders and midfielders (p<0.05) and defenders and strikers (p<0.05).
4.2.2.4 Crossing
Figure 17. Distribution of defenders crossing technique ratings
Figure 18. Distribution of midfielders crossing technique ratings
Figure 19. Distribution of strikers crossing technique ratings
Figures 17 to 19 show the inconsistent nature of crossing the ball across all playing positions at the European Championships 2004. All positions show occurrences of both excellent and unacceptable crossing technique on a number of occasions. Defenders and midfielders (p<0.05) and defenders and strikers (p<0.05) again showed significant differences between the sets of data.
4.2.2.5 Tackling
Figure 20. Distribution of defenders tackling technique ratings
Figure 21. Distribution of midfielders tackling technique ratings
Figure 22. Distribution of strikers tackling technique ratings
The majority of defenders and midfielders tackles occur at either +1 or +2. Striker’s tackles appear to be much more inconsistent and are positioned right across the ratings scale. A Chi squared statistical test emphasized this by revealing significant differences between strikers tackling against both defenders and midfielders (p<0.05).
4.2.2.6 Receiving the ball
Figure 23. Distribution of defenders receiving the ball technique ratings
Figure 24. Distribution of midfielders receiving the ball technique ratings
Figure 25. Distribution of strikers receiving the ball technique ratings
Receiving the ball shows a much greater distribution of the data across technique ratings between all player positions, compared to other performance variables. No significant difference was found between midfielders and strikers running with the ball ratings (p>0.05). Significant differences were however found between defenders and midfielders (p<0.05) and defenders and strikers (p<0.05).
- Distribution of techniques in successful versus unsuccessful teams
The tournament winners, Greece were denoted as a successful team. Switzerland were used an example of an unsuccessful team as they failed to progress beyond the group stages, finishing bottom of Group B (). Data were collected from a single Championship game of each team.
Figure 26. Frequency of techniques used between successful and unsuccessful teams
Figure 27. Average quality rating of techniques used between successful and unsuccessful teams
Figure 26 displays how Switzerland (unsuccessful team) ranks higher in the frequency of all performance variables but one against Greece (successful team).
Despite their low frequency rating, Figure 27 represents how Greece have a higher average technical rating across all but 1 technique compared to Switzerland. Neither technical distribution between successful or unsuccessful teams was found to be significant under the Chi squared statistic (p>0.05).
Figure 28. Frequency of goalkeeper techniques used between successful and unsuccessful teams
Figure 29. Average quality rating of goalkeeper techniques used between successful and unsuccessful teams
Figure 28 shows how there are very little deviation in the frequency of actions between successful and unsuccessful team’s goalkeepers, apart from goalkeeper’s kicks. This difference between goalkeepers kicks was found to statistically significant (p<0.05).
The average technical ratings of each of the respective actions (Figure 29) do also not show much deviation between successful and unsuccessful, apart from the aspect of goalkeeper’s catching the ball. These differences were not found to be statistically significant (p<0.05).
4.2.4 Distribution of technique ratings across selected performance variables between successful and und unsuccessful teams
Figure 30. Distribution of passing technique ratings between successful and unsuccessful teams
Figure 31. Distribution of receiving the ball technique ratings between successful and unsuccessful teams
Figure 32. Distribution of shooting technique ratings between successful and unsuccessful teams
Figure 33. Distribution of running with the ball technique ratings between successful and unsuccessful teams
Figure 34. Distribution of dribbling technique ratings between successful and unsuccessful teams
Figure 35. Distribution of heading technique ratings between successful and unsuccessful teams
Figure 36. Distribution of crossing technique ratings between successful and unsuccessful teams
Figure 37. Distribution of tackling technique ratings between successful and unsuccessful teams
Figure 38. Distribution of goalkeeper kicks technique ratings between successful and unsuccessful teams
Figure 39. Distributions of goalkeeper pass technique ratings between successful and unsuccessful teams
None of the distributions of technique ratings across selected performance variables between successful and unsuccessful teams were found to be significant (p>0.05).
- Discussion
5.1 Reliability
Atkinson and Nevill (1998) recommend that any research conducted within sport should incorporate the use of a number of statistical methods for analysing reliability.
The Chi-squared test of independence and the percentage error statistic were used to determine both intra and inter-observer reliability within the study. Chi squared was chosen as it provides a significance value of the difference between the expected and observed result (Thomas and Nelson, 2001) between 2 or more sets of nominal data that have been arranged by categories into frequency counts (Vincent, 1999). Using percentage error allowed the amount of error for each variable between trials to be established (Hughes et al., 2002).
5.1.1 Intra-observer reliability
Table 7 indicates that intra-observer action observation reliability between T1 and T2 can be assumed according to the Chi squared statistic (p≥0.95) but not according to the percentage error statistic (p>0.05). The major differences occurred in the researcher’s identification of players receiving the ball and running with the ball. Often players would receive the ball whilst travelling in a forwards direction, this action caused confusion as to whether the player was simply receiving the ball or making a conscious attempt to run with the ball. These discrepancies of working within the operational definitions were resolved between T2 and T3 as reliability was assumed by both statistical tests.
Neither statistical test could assume reliability of technique ratings between T1 and T2. The researcher graded performance more negatively throughout the first trial, hence contributing to a number of recognised differences when in comparison with the second trial. Reliability however again increased between T2 and T3 to an acceptable level from both statistical tests. The researcher was far more consistent with the technique ratings as familiarity with the definitions became apparent.
5.1.2 Inter-observer reliability
Table 9 indicates inter-observer action observation reliability between researchers T3 trials. Reliability is determined as strong by the Chi squared statistic (p>0.95) but not apparent by the percentage error statistic (p>0.05).
The largest deviation was again witnessed in the 2 researcher’s identification of players running with the ball. Some instances of players running with the ball were recorded as players dribbling. This definition of a player mastering the ball had to be made absolute before further analysis could take place.
A Chi squared test also revealed a high level of agreement between researchers technique rating reliability (p>0.95). An overall percentage error difference (p>0.05) however contradicted this. Technique ratings across researchers were however fairly consistent with only minor idiosyncrasies between ratings. This opposed Luhtanen (1988) who identified that a major problem in using multiple researchers is in the evaluation of the quality of classified variables. A mean standard across players had to be first be established in order to differentiate between standard technique and good technique of each performance variable.
5.1.3 Evaluation of reliability
The ability to collect data with minimum measurement error is vital during any research (Atkinson and Nevill, 1998). Having data which possess reliable characteristics implies that individual measurements can be taken and used precisely with confidence for future reference (Hopkins, 2000).
Tests of the researcher’s intra-observer reliability of both action observation and technique ratings proved to be reliable from both statistical tests. Inter-observer reliability was however found to be present by a Chi squared test but not by the percentage error statistic.
Hughes et al., (2002) suggest that the percentage error statistic and Chi squared test often indicate little sensitivity to large differences in sets of data. The tests only show differences in the overall shape of the distribution. As differences between variables at the highest level are so slight (Dufour, 1993), these statistical tests may not have been appropriate to highlight such small changes. The percentage error scores could have perhaps been used for each performance variable instead of for the accumulated total with a larger set of data in order to increase reliability (Hughes et al., 2002).
- Distribution of techniques according to positional role
Figures 2 to 5 show the exact distribution of on the ball techniques according to a player’s positional role as a percentage value. This type of distribution is useful as it allows for the relative distribution of techniques across positions. Simply using a frequency count would be not be representative as it does not allow for the different numbers of subjects participating in different positions. This type of representation has been used across other sports (Hughes and Bell, 1998), but is innovative within association football. Dufour (1993) previously summarised the actions of outfield players as a percentage value but without giving any differentiation as to player’s positions of play.
The Chi squared test of independence indicated that differences in the frequency distributions between all playing positions are statistically significant (p<0.05). There is no possible comparison which could be made between the frequency distribution of goalkeepers as the performance indicators are infinitely different to those of an outfield player.
5.2.1 Goalkeepers
It was important not to exclude goalkeepers from the study as other researchers have (Dufour, 1993; Rico and Bangsbo, 1997) as goalkeepers can play a vital role in a both a teams defensive and attacking features (Luxbacher and Klein, 1993; Wooster and Hughes, 2001). Some of the many qualities required of an elite level goalkeeper such as concentration and confidence (Wilson, 1980) are however immeasurable by this type of analysis.
The most frequent on the ball event for a goalkeeper is using a clearing kick, either from the floor or from the hands (42%). They also perform saves (21%), throws (16%), passes (14%), catches (5%) and punches (2%).
Of the analysed variables, 72% of goalkeeper’s actions involve distributing the ball. This high value echoes Hughes (1981) and Welsh (2004) who saw the ability of a goalkeeper to distribute the ball as a main area of responsibility. As nearly three quarters of a goalkeeper’s on the ball actions are spent distributing the ball, it is vital that this aspect it sufficiently focussed on during coaching sessions. Failure to address this issue within training could be unproductive for the entire team’s patterns of offensive play (Welsh, 2004).
5.2.2 Defenders
More than half of a defenders on the ball time is spent passing the ball (53%). 13% of the time is spent competing for the ball in aerial duals, with 10% spent competing for the ball in a tackle. Throwing the ball into play accounts for 7%, with receiving the ball and running with the ball both accounting for 5%. 3% is spent crossing the ball, 2% taking free kicks and 1% spent both dribbling with the ball and attempting to shoot at goal.
Contradictory to the research, 53% of a mean defenders time is spent passing the ball. Wiemeyer (2003) identified the main tasks of a defender to be contesting in defensive play against opposition and competing for the ball aerially. These 2 aspects of play (heading and talking) do rank highly at 13% and 10% respectively but fall way below the most frequent occurrence of passing (53%).
Only designated ‘sweepers’ within a defensive formation are said to be responsible for ball possession and circulation of the ball (Van Lingen, 1997; Wiemeyer, 2003). The analysis however did not distinguish between defensive positions. Full backs, centre backs and sweepers were all grouped within the broad classification of defenders. The reported results could therefore be misleading as a variety of positional characteristics may have been hidden under the group heading of defenders.
Individual full backs or wing backs may have received instruction to provide crosses and portray the characteristics of wingers whilst in the attacking phase. Idiosyncrasies such as these may result in the mean defenders frequencies of certain variables such as running with the ball (5%) and crossing (3%) becoming inconsistent and invalid.
5.3.3 Midfielders
54% of midfielder’s on the ball actions were spent passing the ball with receiving the ball in order to gain control accounting for 9%. Running with the ball (8%), crossing (6%), tackling (6%) and dribbling (6%) also ranked highly. Heading the ball accounted for 5%, with free kicks (3%), shots (2%) and throw-ins (1%) also featuring.
Differing from defenders, the percentage of time midfielders pass a ball (54%) corresponds with the literature. Smith (1973), Van Lingen (1997) and Wiemeyer (2003) and all state that midfielders, those both attacking and defensive in their nature must be efficient passers of the ball. Smith (1973) also iterates that especially attacking midfielders must be proficient at receiving the ball. From the analysis, this aspect can be witnessed as receiving the ball accounts for 9% of the overall actions.
As was explained with defenders, the features of many discrete midfield based positions have all been accumulated into a broad title of midfielders. Attacking midfield players, defensive midfield players and wide player’s performance variables have all been accumulated together as one positional group. A combination of these discrete player positions data explains the occurrence of many performance variables having an equal distribution within the overall frequency distribution.
5.2.3 Strikers
The majority of a striker’s on the ball time is also spent in passing (42%). They however spent 21% of the time attempting to receive the ball. 11% is spent competing for the ball aerially with 3% competing for the ball on the ground. Running with the ball and dribbling account for 7% each, with 5% spent attempting to shoot on goal and 4% crossing the ball to provide others with the opportunity to score.
The passing attributes of a striker are often overlooked. Smith (1973) rates the ability of a striker to pass as only the 7th most important component of a striker’s overall priorities. Although the analysis provides no estimation of the importance of passing, a high value of 42% deems it to be the most frequently used attribute. A striker who doesn’t possess the ability to pass the ball will naturally be very individualistic and not transgress very successfully into an overall team’s pattern of play.
The ability of a striker to receive balls played forwards by team-mates is a vital part of their game (Smith, 1973; Van Lingen, 1997). By the very nature of their position on the field, strikers must be able to receive the ball efficiently whilst under pressure from opponents. The data shows how nearly a quarter of all on the ball time is spent attempting to receive the ball. This high percentage of incidence reinforces the need to spend sufficient time in training improving this aspect of play.
Strikers often are required to challenge for the ball aerially, either in helping to maintain possession of the ball or in an attempt to score (Smith, 1973). 11% of time is spent competing aerially and surprisingly 3% of time is spent attempting to dispossess an opponent, an aspect of forward play which is frequently overlooked.
Only a small fraction of a striker’s on the ball time (5%) is spent attempting to shoot for goal, something which is vital for both the individual and the team (Smith, 1973; Van Lingen, 1997; Wiemeyer; 2003). With such a small fraction spent performing this variable, the importance of a shot being a successful one is obvious. This aspect can justify the large percentage of time spent by strikers working on their shooting during practice.
- Quality of techniques performed across different playing positions
The ability to illustrate the mean quality of techniques performed eliminates the biased that simply showing a frequency distribution provides. The idea of coding performance indicators according to their successful or unsuccessful implementation has been previously used before (Rico and Bangsbo, 1997; Pearce and Hughes, 2001). No exact quantitative analysis has however been completed showing the exact successful execution of techniques across positions. This ability to conduct an exact objective quantification of critical events is vital for a complete analysis (Franks and Goodman, 1986). Showing an exact technical distribution also shows a cross section of performance which would otherwise be hidden within a calculated average.
Figure 7 demonstrates the average technical rating of techniques used across outfield playing positions. The highest mean average rating for passing was held by the midfielders (0.95). This ability of midfielders to be the most efficient passers within a team framework is reinforced by the literature (Smith, 1973; Van Lingen, 1997 and Wiemeyer, 2003). This passing rating was found to be statistically significant than the mean passing rating score for defenders (p<0.05). This echoes the finding of Reilly and Holmes (1983) where midfielders were significantly found to perform better than defenders in specific outfield performance tests. The passing distribution graphs (figures 8-10) show how across all positions a trend of +1 and 0 rated passes are apparent. It is however the extreme ends of the continuum which may directly affect the result of games. The relative occurrence of either a +3 or -3 pass may be the critical pass which can either create or concede a goal. As an accumulative total, more -3 passes occurred (n = 33) than +3 passes (n = 27). This may suggest that the nature of many of the analysed goals occurred by opposition’s mistakes other than individual ability from the team in possession.
Strikers were the highest ranked position in receiving the ball (0.95) compared to defenders (0.91) and midfielders (0.89). The difference between strikers and defenders was found to be statistically significant (p<0.05). Both Smith (1973) and Van Lingen (1997) identify that this ability for strikers to able to receive the ball is a technical necessity. The distributions of receiving the ball ratings were however inconsistent throughout all positions, especially strikers (figures 23-25). Having negative ratings for receiving the ball in the attacking third of the pitch is not uncommon due to the close attention of opposition defenders (Reilly and Holmes, 1983). Midfielders possessing such a low receiving the ball rating could be explained by the lack of time in possession present in the midfield areas in order to effectively control the ball (Reilly and Holmes, 1983).
Strikers were also the highest ranked for running with the ball (1.29) and dribbling (1.3). Both scores confirm the findings of Wiemeyer (2003) who identified that an effective striker is one who possesses good one to one play against an opponent. The ability to run with the ball both unopposed and under pressure can create problems for the oppositions defensive and increase the probability of creating an attacking option. There is also limited risk for attackers moving with the ball up field as any resulting dispossession which may occur happens in the opponents own defensive third, leaving time and space for the ball to be regained.
Defenders surprisingly had the highest average technical rating for shots on goal (0.71) with a result which was statistically significant (p<0.05) to the shooting scores of both midfielders (0.15) and strikers (0.55). The ability for defenders to be able to shoot towards goal is not included across any of the literature. Even though midfielders and strikers had a higher frequency of +3 shots, the erratic nature of their overall shooting led to a lower average score than defenders. As previously stated, it is often these +3 scores which lead to goals being scored within games. It may be universally accepted by coaches that strikers could have 2 erratic shots for every goal scored that could win the game, as opposed creating a high frequency of mediocre shots which have no general influence.
Defenders have the highest average technique rating for heading (1.10), and for tackling (1.28). These results are as expected, as part of a defenders role is stopping opposition’s attacks both aerially and on the ground (Smith, 1973; Van Lingen, 1997; Wiemeyer, 2003). This result may also be related to the fact that less pressure is also often applied within the defensive area leading to a high success rate of a skilful performance (Reilly and Holmes, 1983). The nature of these tackles and headers is also very consistent, with very few negative ratings (figures 14 and 20). Negative ratings for tackling or heading in the defensive regions could be very costly as any vital error could result in the opposition having an attempt on goal.
Strikers tackling distribution was extremely erratic (figure 22) with a significant difference being found against the proficient defenders and midfielders (p<0.05). An identical trend is also apparent between strikers heading capabilities (figure 16). A large data spread across all heading ratings resulted in a significant difference between defenders and strikers (p<0.05).
The low rating of strikers tackling technique is not surprising. Although the frequency of tackling was unusually high and gave an indication of the possible high work ethic of forward players, the attempted tackles were however poor in their execution. The low rating which strikers received for heading could be a definite area for improvement as the ability to be a good header of the ball is a priority (Smith, 1973).
Defenders mean crossing rating (0.25) was significantly lower (p<0.05) than both midfielders (0.90) and strikers (0.75). Although defenders had a high frequency of crosses (n = 91), a large spread of the data resulted in a low mean rating. All 3 outfield positions showed a low consistency of crossing, with data dispersed across all of the technique ratings (figures 17-19). This reveals that although many positive opportunities to cross were created, very few were delivered effectively. From the results it can be inferred that crossing is a technique which requires sufficient improvement, particularly from defenders.
- Distribution of techniques between successful and unsuccessful teams
Figure 26 shows how Switzerland (unsuccessful team), although not significant (p>0.05) have a higher frequency of techniques when compared to Greece (successful team) in all but one of the performance indicators.
These findings are contradictory to Bishovets et al., (1993) who reported that winning teams displayed a higher number of collective actions than unsuccessful teams. The low frequency rating of Greece in both receiving and passing the ball also contradicts a theory that successful teams are in possession of the ball for longer durations, with more touches than unsuccessful teams (Winkler, 1996; Hughes et al., 1988; Eniseler et al., 1996).
Greece failed to win a game by more than 1 goal and much of their success was based upon defensive organisation (). This relatively low frequency of Greece maintaining possession of the ball could however be related to the increased pressure placed upon them during the latter tournament stages and the closeness involved of defending these narrow score lines (Rico and Bangsbo, 1997).
Eniseler et al., (2001) also identified that unsuccessful teams crossed, shot at goal, tackled and dibbled less that successful teams. The reported data vastly differentiates from this as it shows the unsuccessful team possessing a higher frequency count for all of the described performance indicators. The low amount of goals scored in the tournament by Greece (n = 7) could be an explanation for the low frequency of crosses and shots. It however seems uncharacteristic that such a defensive minded team (Greece) produced such a low frequency of tackles.
The Greeks based many of their attacking and defensive strengths upon the heading ability of their team, an area in which they were superior (). Having a high frequency of headers may reveal one of the Greeks most reliable tactics during both the attacking and defensive phases of play.
- Average technique ratings between successful and unsuccessful teams
The frequency trend appears to have been reversed in figure 27, when the average technical ratings of the two teams are compared. Greece are now rated ahead of Switzerland in all but 2 of the performance variables. It appears a common trend for successful teams to display a more efficient technical performance (Acar, 2001).
Although the difference between the passing ratings is not significant (p>0.05), the successful teams passing showed an elevated value (0.87) when compared to the unsuccessful team (0.60). An identical trend can be witnessed between the receiving the ball ratings at 1.13 and 0.73 respectively. The frequency distributions (figures 30 and 31) reinforce this point by revealing how the unsuccessful team have a far greater frequency of negatively ranked data, compared to the successful team who have the majority of their data positively ranked.
Although the Greeks tackling frequency was below the frequency of Switzerland, the average rating of 1.4 was far higher than the Swiss at 0.8. This implies that the Greek mentality of defending was not based upon the frequency of tackles performed but upon the quality of such tackles. During the analysis the Greeks did not perform one negatively ranked tackle, thus emphasising the efficiency of tackles performed. The findings of the analysis correlate with those of Luhtanen et al., (2001) who in analysing performance variables of the 2000 European Championships found the eventual winners France to be highest ranked in passing, receiving and tackling. Although such an in depth analysis has not been performed, a similar trend seems to be apparent.
Despite the data differing from that of Eniseler et al., (1996) in that the successful team shot and dribbled less than the unsuccessful team, the mean quality of a shot and dribble was considerably higher in both cases. Greece’s dribbling (1.8) and shooting (0.75) ratings were rated more positively than Switzerland’s at 1.12 and 0.37 respectively. The decision making capabilities of when are where to dibble may have been enhanced (Hughes et al., 1988) for the successful team with a far better ability to shoot accurately should the chance arise. This condition was also apparent for the successful team running with the ball.
The aerial ability of the Greeks was emphasised by their positive mean technique rating (1.33) compared to the Swiss (0.83). The heading distribution (figure 35) shows how the Greeks never rated below 0 for a header, compared to the inconsistent Swiss who showed examples of headers as far down the scale as –2.
A significant difference was found between the goalkeeping kicking distributions of successful and unsuccessful team’s goalkeepers (p<0.05). Having an accurate distribution of a goalkeeper is vital as it can shape the nature of any attack (Wooster and Hughes, 2001; Welsh, 2004). The accurate and consistent nature of the Greeks goalkeepers’ distribution (figure 38) may have been a factor leading to success further up the field. The lack of the Greek goalkeepers’ throws (n=1) and passes (n=0) restricts the risk of dispossession occurring in a dangerous area of the pitch which may lead to an attempt on goal from the opposition (Luxbaxcher and Klein, 1993).
This decision-making quality may suggest a mentality, which brought the Greeks such success.
- Conclusion
6.1 Findings of study
The study consisted of a hand notation system with an accepted level of both intra and inter-observer reliability under the chi squared statistic for both action observation and technique rating observation (p>0.95).
Significant differences (p<0.05) were found between the frequency distributions of all 3 outfield positions, namely defenders, midfielders and strikers, thus accepting the original hypothesis. No significant differences (p>0.05) were found between the accumulated mean of technique ratings across all of the performance indicators.
Individual variable analysis however showed significant differences (p<0.05) between defenders and midfielders passing, crossing, shooting, heading and receiving the ball. Significant differences (p<0.05) were also reported between defenders and strikers across all performance variables and also between midfielders and strikers tackling technique distribution.
No significant differences (p>0.05) were discovered between either the frequency distribution or the technical rating of outfield players between successful and unsuccessful teams, this rejects the hypothesis that a significant difference would occur. A significant difference (p<0.05) was however apparent between the frequency distribution of goalkeepers actions between successful and unsuccessful teams.
- Recommendations for players and coaches
From the reported results it can be inferred that players and coaches alike must recognise that significant differences do occur between the outfield playing positions of association football.
The defined results can be of impact to a coach in 2 ways:
- By helping to align certain players into specific playing positions according to their individual technical attributes
- By revealing specific aspects of performance across positions which could be either stabilised or enhanced within a training context (Bate, 1996).
- Future recommendations
There are many direct extensions of the study available to enhance this work further. A Kruskal-Wallis and Mann Witney post hoc test (Vincent, 1999) could be utilised to calculate the significance of differences between variables. This type of statistical test may establish further significance between either the average technical ratings or between successful and unsuccessful teams. This test could be assisted by analysing more games, consequently producing more data which may influence any significance values (Hughes et al., 2002).
A detailed within position analysis could also be devised to break down the positional subdivisions even further. For example defenders could be further grouped into exact roles as full backs, centre backs and sweepers.
An alteration and further detailing of the operational definitions themselves may also generate further findings. Heading the ball could have been detailed further as either attacking or defensive, hence giving a new dimension to any generated data.
Studying the changes in techniques used in different areas of the pitch or in different time periods would enable extra data to be generated, also enabling a comparison between successful and unsuccessful teams.
By using the presented system a multitude of differing comparisons are instantly available:
- Comparing each team within the Championship to investigate which team is the most technically astute.
- Comparing different levels of competition.
- Investigating the differences between the men’s and women’s game.
References
Acar, M. F. (1996). A computerised, post event analysis of the technical and tactical events during the derby matches in the second half of the Turkish soccer first division. In Notational Analysis of Sport III (edited by M.Hughes.), pp. 117-120. CPA: UWIC: Cardiff.
Ali, A.H. (1988). A statistical analysis of tactical movement patterns in soccer. In Science and Football (edited by T, Reilly., A. Lees., K. Davids. and W.J. Murphy.), pp. 302-308. London: E&FN Spon.
Atkinson, G. and Nevill, A.M. (1998). Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Medicine, 26 (4), 217-237.
Bangsbo, J. (1997). The physiology of intermittent activity in football. In Science and Football III (edited by T. Reilly, J. Bangsbo & M. Hughes), pp. 43-53. London: E&FN Spon.
Bate, D. (1996). Soccer skills practice. In Science and Soccer (edited by T. Reilly.), pp. 227-241. London: E&FN Spon.
Bate, R. (1988). Football chance: Tactics and Strategy. In Science and Football (edited by T, Reilly., A. Lees., K. Davids. and W.J. Murphy.), pp. 293-301. London: E&FN Spon.
Bishovets, A., Gadjiev, G. and Godik, M. (1993). Computer analysis of the effectiveness of collective technical and tactical moves of footballers in the matches of 1988 Olympics and 1990 World Cup. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.232-236. London: E&FN Spon.
Bland, J.M. and Altman, D.G. (1986). Statistical methods for assessing agreement between two methods of clinical agreement. The Lancet, Feb: 8, 307-310.
Chervenjakov, M. and Dimitrov, G. (1988). Assessment of the playing effectiveness of soccer players. In Science and Football (edited by T, Reilly., A. Lees., K. Davids. and W.J. Murphy.), pp. 288-292. London: E&FN Spon.
Coghlan, A. (1990). How to score goals and influence people. New Scientist, 2, 54-59.
Cook, M. (1982). Soccer Coaching and Team Management. EP Publishing Ltd:Leeds.
Dooan, B., Eniseler, N., Aydin, S., Morali, S. and Vakkas Ustun, S. (1996). The effects of the number of touching ball in the passing efficiency in professional and amateur soccer game and comparing them according to league levels. In Notational Analysis of Sport III (edited by M.Hughes.), pp. 163-166. CPA: UWIC: Cardiff.
Dufour, W. (1993). Computer - assisted scouting in soccer. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.160-166. London: E&FN Spon.
Eniseler, N., Camliyer, H., Vakkas Ustun, S. and Aydin, S. (1996). The analysis of some technique elements in a soccer game on determined areas of the football field and on certain time periods of the game. In Notational Analysis of Sport III (edited by M.Hughes.), pp. 121-131. CPA: UWIC: Cardiff.
Erdmann, W.S. (1993). Quantification of games – Preliminary kinematic investigations in soccer. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.174-179. London: E&FN Spon.
European Championships 2004 Website (2004). (accessed 05 January 2005).
Franks, I.M. (1997) Use of feedback by coaches and players. In Science and Football III (edited by T. Reilly, J. Bangsbo & M. Hughes), pp. 267-278. London: E&FN Spon.
Franks, I.M. and Goodman, D. (1986). A systematic approach to analysing sports performance. Journal of Sports Sciences, 4, 49-59.
Franks, I.M. and McGarry T. (1996). The science of match analysis. In Science and Soccer (edited by T. Reilly.), pp. 363-375. London: E&FN Spon.
Garganta, J. (1998). Tactical modelling in soccer: a critical view. In Notational Analysis of Sport IV (edited by M.Hughes. and F. Tavares), pp. 58-64. Centre for team sports studies: Porto.
Gerisch, G. and Reichelt, M. (1993). Computer and video – aided analysis of football games. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.167-173. London: E&FN Spon.
Grehaigne, J.F., Mahut, B. and Fernandez, A. (2001). Qualitative observation tools to analyse soccer. International Journal of Performance Analysis in Sport - E, Vol. 1: No 1, pp.52-61.
Hopkins, W.G. (2000). Measures of reliability in sports medicine and science. Sports Medicine, 30 (1), 1-15.
Hughes, C. (1981). Soccer Tactics and Teamwork. EP Publishing Ltd: England.
Hughes, M. (1993). Notation Analysis in Football. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.151-159. London: E&FN Spon.
Hughes, M. (1996). Notational analysis. In Science and Soccer (edited by T. Reilly.), pp. 343-361. London: E&FN Spon.
Hughes, M. (1998). The application of notational analysis to racket sports. In Science and Racket Sports II (edited by A. Lees, I. Maynard, M. Hughes and T. Reilly), pp. 211-220. E and FN Spon: London.
Hughes, M. and Bell, K. (1998). Performance profiling in cricket. In Notational Analysis of Sport IV (edited by M.Hughes. and F. Tavares), pp. 176-183. Centre for team sports studies: Porto.
Hughes, M. and Franks, I, M. (2004). Notational Analysis of Sport (2nd Edition). Routledge: London.
Hughes, M., Robertson, K. and Nicholson, A. (1988). Comparison of patterns of play of successful and unsuccessful teams in the 1986 World Cup for soccer. In Science and Football (edited by T, Reilly., A. Lees., K. Davids. and W.J. Murphy.), pp. 363-367. London: E&FN Spon.
Hughes, M., Cooper, S.M. and Nevill, A. (2002). Analysis procedures for non parametric data from performance analysis. International Journal of Performance Analysis in Sport - E, Vol. 2: No 1, pp.6-20.
James, N., Mellalieu, S.D. and Hollely, C. (2002). Analysis of strategies in soccer as a function of European and domestic competition. International Journal of Performance Analysis in Sport - E, Vol. 2: No 1, pp.85-103.
Leibermann, D.G., Katz, L., Hughes, M.D., Bartlett, R.M., McClements, J. and Franks, I.M. (2002). Advances in the application of information technology to sport performance. Journal of Sports Sciences, 20, 755-769.
Luhtanen, P.H. (1988). Reliability of video observation of individual techniques used in soccer. In Science and Football (edited by T, Reilly., A. Lees., K. Davids. and W.J. Murphy.), pp. 356-360. London: E&FN Spon.
Luhtanen, P.H. (1993).A statistical evaluation of offensive actions in soccer at World Cup level in Italy 1990. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.215-220. London: E&FN Spon.
Luhtanen, P., Belinskij, A., Hayrinen, M. and Vanttinen, T. (2001). A comparative tournament analysis between the Euro 1996 and 2000 in soccer. International Journal of Performance Analysis in Sport - E, Vol. 1: No 1, pp.74-82.
Luxbacher, J. A. and Klein, G. (1993). The Soccer Goalkeeper (2nd edition). Human Kinetics: USA.
O’Donoghue, P.G., Boyd, M., Lawlor, J. and Bleakley, E.W. (2001). Time –motion analysis of elite, semi-professional and amateur soccer competition. Journal of Human Movement Studies, 41, 1-12.
Olsen, E. and Larsen, O. (1997). Use of match analysis by coaches. In Science and Football III (edited by T. Reilly, J. Bangsbo & M. Hughes), pp. 209-220. London: E&FN Spon.
Partridge, D. and Franks, I.M. (1997). The use of computer video interactive analysis in the sport of soccer: changing individual performance by providing quantitative and qualitative feedback. In Notational Analysis of Sport I & II (edited by M.Hughes.), pp. 99-111. CPA: UWIC: Cardiff.
Partridge, D., Mosher, R.E. and Franks, I.M. (1993). A computer assisted analysis of technical performance – a comparison of the 1990 world cup and intercollegiate soccer. In Science and Football II (edited by T. Reilly., J. Clarys. and A. Stibble), pp.221-231. London: E&FN Spon.
Pearce, M. and Hughes, M. (2001). Substitutions in Euro 2000. In pass.com (edited by M.Hughes. and I.M. Franks.), pp. 303-315. CPA: UWIC: Cardiff.
Pinto, J. (1998). Performance Factors in Soccer. In Notational Analysis of Sport IV (edited by M.Hughes. and F. Tavares), pp. 93-97. Centre for team sports studies: Porto.
Pollard, R., Reep, C. and Hartley, S. (1988). The quantitative comparison of playing styles in soccer. In Science and Football (edited by T, Reilly., A. Lees., K. Davids. and W.J. Murphy.), pp. 309-315. London: E&FN Spon.
Reep, C. and Benjamin, B. (1968). Skill and Chance in Association Football. Journal of the Royal Statistical Society, 131, 581-585.
Reilly, T. and Holmes, M. (1983). A Preliminary analysis of selected soccer skills. Physical Education Review, 6(1), 64-71.
Reilly, T. and Thomas, V. (1976). A motion analysis of work-rate in different positional roles in professional football match play. Journal of Human Movement Studies, 2, 87-97.
Rico, J. and Bangsbo, J. (1997). Coding system to evaluate actions with the ball during a soccer match. In Notational Analysis of Sport I & II (edited by M.Hughes.), pp. 85-90. CPA: UWIC: Cardiff.
Smith, M. (1973). Success in Football. Butler and Tanner: London.
Thomas, J.R. and Nelson, J.K. (2001). Research Methods in Physical Activity (4th Edition. Human Kinetics:USA.
Van Lingen, B. (1997). Coaching Soccer. Reedswain, Spring City: USA.
Vincent. W.J. (1999). Statistics in Kinesiology (2nd Edition). Human Kinetics: USA.
Wells, C. and Reilly, T. (2002). Influence of playing position on fitness and performance measures in female soccer players. In Science and Football IV (edited by W. Spinks., T. Reilly. and A. Murphy. ), pp. 369-373. London: Routledge.
Welsh, A. (2004). The Soccer Goalkeeping Handbook (2nd edition). A&C Black: London.
Wiemeyer, J. (2003). Who should play in which position in soccer? Empirical evidence and unconventional modelling. International Journal of Performance Analysis in Sport - E, Vol. 3: No 1, pp.1-18.
Wilson, B. (1980). The Art of Goalkeeping (2nd edition). Pelham Books Ltd: London.
Winkler, W. (1996). Qualitative and quantitative match analysis in soccer. In Notational Analysis of Sport III (edited by M.Hughes.), pp. 43-56. CPA: UWIC: Cardiff.
Wooster, B. and Hughes, M. (2001). Playing patterns ensuing from the distribution of goalkeepers in elite association football. In pass.com (edited by M.Hughes. and I.M. Franks.), pp. 317-324. CPA: UWIC: Cardiff.
Yamanaka, K., Liang, D,Y.and Hughes, M. (1997). An analysis of the playing patterns of the Japan national team in the 1994 World Cup qualifying match for Asia. In Science and Football III (edited by T. Reilly, J. Bangsbo & M. Hughes), pp. 221-228. London: E&FN Spon.