Relative Age Effect in Football
Disclaimer: This work has been submitted by a student. This is not an example of the work written by our professional academic writers. You can view samples of our professional work here.
Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
Published: Tue, 20 Feb 2018
The purpose of the study was to investigate the performance level at which birth date effects selection for performance pathways in English football, as well as examining whether coaches are currently implementing arrangements to limit the relative age effect. The study comprised of 2450 players from performance levels including community, grass root and academy. The birth dates of each player within every performance level were analysed through the use of statistical tools within Microsoft Excel, with interviews analysed through transcription and the highlighting of recurrent themes. The sub-groups were viewed by age group, month of birth and the total percentage of players born within each quartile of the selection year to analyse the birth bias within specific performance levels. The statistical data of each sub-group were then collated to view differences in progressing through each performance level.
The main results found an over-representation of players born in the first quartile throughout each performance level. The bias within the community and grass root subgroup was 4.1%, with a 39.9% bias towards the eldest players at academy standard. The evidence highlighted that birth date only has significant impact on selection once the academy standard of play is reached, with minimal difference in impact when progressing through inferior performance levels. Coaches in the study showed high awareness of the effect, with implementation of two strategies to reduce the relative age effect being implemented. The two strategies were found to be ineffective in the reduction of the relative age effect, through implementation occurring after selection.
To conclude the academy pathway highlighted the most significant bias, with birth date having minimal impact at inferior performance levels. Strategies to reduce the relative age effect are currently ineffective requiring further research into reducing the bias prior to selection.
- To investigate the performance level at which birth date may affect selection for performance pathways in English football.
- To examine if football coaches are making arrangements to limit the Relative age effect in football
In English football is there a starting point to the relative age effect and if this is the case then how are football coaches currently taking this bias into consideration when working with children at all levels of football.
Relative age effect is the difference in ages between children in the same age group. An example being, a child born in the start of the selection period in football i.e. 1st September will be 11 months older than a player who falls in the same age group born on the 1st of August. (Barnsley et al, 1992) Throughout the study research has referred to the relative age effect as birth date and birth bias, all meaning the same.
The football world is competitive and making sure that your team are developing young athletes to progress into the first team and national team is very important. This has made the selection and development of children an important aspect in youth football.
Studies have progressively shown that in football there are children not given the opportunity, due to a simple aspect such as their age. (Brewer et al 1995; Cobley et al 2008; Delorme et al 2010) The research that has previously been carried out has rarely been specified around the English game with only a minority being carried out in this area. (Simmons & Paull, 2001; Musch & Grondin, 2001) The research has predominantly been aimed towards a number of different nations across the world. (Glamser & Vincent, 2004; Jimenez, 2008; Delorme et al 2010; Campo et al 2010) The limited amount of research on the English game highlighted an area in which further study could be carried out in order to fully understand the impact the relative age effect has within English football. Correspondingly the research into the affect performance level has on the impact of the relative age effect within English football has been under-represented by preceding research. The research specific to this area often views different nations or sports. (Mujika et al, 2007; Cobley et al, 2009; Till et al, 2010) The knowledge of how performance level could affect selection will allow understanding of where birth bias is present and predominant, furthermore highlighting the level at which change is needed to reduce the relative age effect. The results will be of great value to coaches within the performance level that the relative age effect is most predominant by raising awareness and creating knowledge for change.
Preceding research has also viewed how different organisations and football associations have tried to reduce the impact of the relative age effect in different countries. (Helson et al 2000; Vaeyens et al 2003) Although this research has shown how interventions have been made there has not been a study on how the coaches within the football clubs in these associations are practically trying to reduce the bias or in fact if they are. This is an area of research that is being analysed. This will help compare the current tools being put in place and to see if there is a working intervention to help reduce the birth bias. The context of the study will include raising awareness of the relative age effect to the coaches in which the relative age effect is most predominant within the standard at which they coach. Research has previously highlighted this to be an advantage in reducing the effect, increasing the value of the study. (Baker et al, 2010; Cobley et al, 2009)
The study begins viewing previous research on the relative age effect reviewing specifically topics around the aims and secondary topics in which can affect the predominance of the effect. The second section views the method in which the study carried out the research with reasoning and description, progressing onto the results in which are presented with the discussion following. A conclusion in relation to the aims of the study is carried out in the penultimate section, ending with self reflection discussing the learning throughout the study.
1. Relative Age Effect
Children are split into age groups throughout school and whilst they are in education. In England the children are split into age groups running from nursery, primary school with years 1-6, then into secondary schools with years of 7-11. In England the school year starts in early September and runs to August (Direct.gov, 2009). This means that two children within the same year participating in educational studies and sports could have a difference of more than eleven months between them.
In sport, the relative age effect was first noticed in Canadian ice-hockey and volleyball. Grondin et al, (1984) found unequal birth-date distributions for males and females at recreational, competitive and senior professional levels for both sports during the 1981/1982 season.
2. What age does the relative age effect occur?
Simmons & Paull (2001) are a set of researches who have previously viewed the relative age effect in England. They found that there was a bias within centre of excellences in England. In 1997 in the age groups of U-15 – U16 there was seen to be a large difference within the birth dates of players participating, players oldest within the year consisted of 58.7% with just 12.7% of younger children being within these centres. Glamser, Vincent, (2004); Musch & Grondin (2001) found specifically that ‘players in the United Kingdom, Sweden, and Belgium approximately 70% of elite youth players had birthdays in the first half of the soccer year.’ This shows that there is a bias within English football and shows that the age effect occurs highly in these ages.
The relative age effect does not just occur within England. Del Campo et al (2010) viewed the relative age effect within Spain. The research that was carried out found that the relative age effect occurred through age groups including under 11’s to 18’s. The players within the teams included within the research consisted of a minimum of 45% of players born within the first trimester, with only a maximum of 15% of the players coming from the fourth trimester. This shows that throughout each age group the difference between the players born early in the year and the ones later in the year was 30% in favour to the players born in the early stages of the selection period.
Williams, (2009) study on the U-17 World Cup also suggests that the relative age effect is continued into not just the older age groups but also into senior international teams. Williams (2009) looked at all the players participating within the tournament and found a large difference in the months of when the players were born. The study found that,
‘Nearly 40% of the players are born in the first three months of the year while only 16% are born in the last quarter’.
Dudink (1994) research supports the research carried out by Williams (2009) into evidence that the relative age effect progresses through all age groups. Dudink (1994) claimed that both Dutch and English players born early in the competition year are more likely to participate in national soccer leagues. This research not only suggests that the relative age effect occurs through childhood and adolescence but also occurs through to adulthood.
3. Why does the relative age effect occur?
Research has shown many reasons behind why there are biases towards players who are older than their peers. (Delorme & Raspaud, 2009: Musch and Grondin, 2001). Musch and Grondin, (2001) suggests that ‘as children are separated into age groups there are regularly cognitive, physical and emotional differences between the youngest and oldest’. Research backing up this is seen from Malina et al,(2004) who states there are ‘advantages in body size, fat free mass and several components of physical fitness including aerobic power, muscular strength, power, endurance, and speed’. This means that there is a difference within all aspects of a character within players in the same year. It has been suggested that the size of the relative age effect may be affected by additional maturational variation at ages associated with the onset of puberty, generally applicable at the ages of 13-15 in boys and 12-14 in girls (Musch & Grondin, 2001). This meaning that puberty is a large section of a player’s selection/development process in which could affect the number of players who are chosen. Helsen et al (2000) found that relative age effect was present within children aged as young as 8.
As players develop differently at different stages this would suggest the younger players would have a larger disadvantage at the stage of puberty. Research from Gil et al, (2007b) found when looking at the selection of young soccer players in terms of anthropometric and physiological factors found that during puberty the players selected were taller, heavier, leaner and faster than the non-selected players and that a high percentage of those chosen were found to be born within the first 6 months of the year.
Helsen et al, (2000) looked at the possible difference between two players within the same selection year:
‘A 10-year-old child in the 5th percentile is likely to be 1.26 m tall with a body mass of 22 kg, whereas a child in the 95th percentile who is almost 11 years of age is likely to be 1.54 m tall and 49 kg in mass.’
This shows that one player could be as much as 0.3m taller and 27kg heavier than a player placed in the same selection year showing a clear advantage physically towards the older player.
Along with maturation levels studies have shown that the playing position of a player also has an effect. Ashworth and Heyndels (2007) noted the relative age effect varied according to playing position in elite German soccer. The strongest effect sizes were found for goalkeepers and defenders, with relative age effects not evident for forwards. Research carried out by Gil et al, (2007a) found that goalkeepers and defenders are on average are the tallest players being five centre meters taller than both the midfielders and attackers. This with the research found from Malina et al., (2004) shows that it could be very difficult for the younger players to achieve selection within these positions. Research by Gil et al, (2007a) also progresses on to further back up research from Malina et al, (2004) as when viewing players who were in the selection process, players who had better endurance, were faster and in some instances taller were primarily selected. Although there were statistical information found within research from Gil et al, (2007a) looking further into the study there were also instances in which the research challenged the statement by Malina et al,(2004) in which they stated players who were faster and taller for example had an advantage. Gil et al (2007a) found that when viewing players who were selected and those not, it turned out that the non selected players were taller faster and had superior endurance.
Examples being that the goalkeepers non-selected were four centre meters taller, 0.3 seconds faster than the selected players. The research found that in midfield where players are seen to run the most which in turn means they need to have greater endurance levels the non selected players were found to have lower heart rates after an endurance test. (Gil et al, 2007a) This research highlights evidence contesting the advantages older children are perceived to possess.
4. The effect on participation levels
Researchers have also viewed that the relative age effect can make players drop out of sport. Delorme et al (2010), suggests that the players born later within the year ‘experience inferiority and failure within their practice and may be reduced to less playing time.’ Vaeyens et al (2005) also had similar thoughts and stated that the reason why the relative age effect relates to the players dropping out is due to the older players receiving more playing time than the younger players. This leading to the younger players feeling less competent and increasing the possibility of them dropping out of the sport. Cobley et al (2009) noted that the size of the relative age effect increased with age until late adolescence, but then decreased in adult sporting contexts meaning that if the players who are born in the younger part of the year, who carry on in sport have a good chance of being selected to play at a high standard, contesting research carried out by Williams (2009) and Dudink (1994).
While Musch and Grondin (2001), stated that;
‘The relative age effect is not only thought to generate discrimination in the selection process, but also to lead to dropout among less advantaged players’ (i.e. those born at the end of the year)
Delorme et al, (2010) found that the rates of drop outs in French football were highest within players in the last two quarters of the year. This means that although Cobley et al, (2009) found that the relative age effect decreases into adulthood the number of the late born players progressing through to that stage is low. Research by Delorme et al, (2010) supports research by Cobley et al (2009) as they found the number of players dropping out in French adult football was higher within the players born early within the year with a number of 1,612 players dropping out more than the late born players.
Although this can be seen from the research, Delorme et al (2010) also show that the number of players born late in the years that are dropping out is higher throughout the ages of 9 – 15. This means that a high number of players born within the last part of the year have already dropped out implying that as the years progress the number of players that can drop out have reduced significantly.
5. Does the level of play affect the relative age effect?
Research from Mujika et al, (2007) views similar areas to the one carried out in this study with the difference of them viewing this within Spanish football. They viewed the difference between the relative age effect at different levels of football within Spain. The levels they viewed were players from La Liga (Spain’s highest division) club AC Bilbao, Elite youth from AC Bilbao, Regional Youth and School Youth. The research found that players born in the first quarter of the selection period decreased as did the level of football, after the La Liga players group who had 43.9%. Elite youth players consisted of 46.6% of players in the first quarter, the regional youth group consisted of 28.6% whilst the school youth group had the lowest percentage at 27.1%. These statistics show that throughout youth football the relative age effect increases, slightly decreasing when reaching the highest level of football although only by 2.1%. They also found that players in the last quarter were found mostly in the School Youth subgroup consisting 22.9%, decreasing to 21.2% in the regional youth group, then significantly decreasing to just 10% of players within the Elite youth subgroup, finishing with a very slight increase in the number of players within the La Liga group of 2.2% to a total of 12.2% of players being within the last quarter. This again shows that there is a bias throughout the progression in performance level within Spanish football.
Cobley et al (2009) viewed performance level in relation to the size of the relative age effect among similar levels to the current study. Cobley et al (2009) found that the largest bias towards the oldest players was found within players that participate within the representative stage. This stage related to the level below the elite stage which was viewed to be the highest level in the study. Cobley et al (2009) progressed to suggest that the level that players partake within has an effect on the size of the relative age effect. The relative age effect was found to increase within each progression in performance level until the optimal performance level is reached comparable to the findings by Mujika et al (2007)
Till et al (2010) viewed the relative age effect within rugby league players, similar to Mujika et al (2007) they found that as the performance level increases as does the impact of the relative age effect. In the study throughout each increase in performance level there was an increase in the size of the relative age effect. The highest bias towards the eldest players found was 61.34%, this statistic was found within the under 13’s age group. The research highlights that the birth bias is affected by skill/performance level not just within football but also other sporting environments.
6. Is being young an advantage?
While previous research has shown a bias towards the players born early within a selection period, there is research suggesting that if the later born players successfully progress through and become professional players they can be at an advantage. Ashworth and Heyndels (2007) found that players who were seen as being born in the later period of selection had higher wages than the players born in the early period. When looking at German football players during the 97-98 and 98-99 seasons, players born in the cut off month of August 1st earned 2 million deutschemarks where players born later in the selection period earn up to 2.8 million deutschemarks. Ashworth and Heyndels (2007) stated that this occurred when the later born players played in a high standard soccer education programme. The later born players benefited playing with the early born players or perceived better players, enhancing their development as young players benefit from playing alongside or against superior players. They further progress to imply that for the later born players to succeed throughout the selection process, when younger they must have above average talent.
7. Can the relative age effect be decreased?
Research has viewed the possibilities of whether a change in selection dates will correct the bias that occurs within football.
The Royal Belgian Football Association changed their cut off date in 1997 to reduce the impact of the relative age effect, but the shift from August the 1st to the 1st January just meant a shift in the Bias. (Vaeyens et al, 2005) Similarly research has viewed Japan’s competition year which begins on 1st April and the bias is observed in May and July, Germany and Brazil produce similar distributions with a start date of 1 August. In each case, the season-of-birth bias aligns with whichever quarter is earliest in the competition year.
(Simmons and Paull, 2001) Fifa and Uefa have also been seen to prolong the selection period for players in order to make it fairer. Research found that the number of players within teams that were looked at had more players in a wider range of months but there was still a bias to the younger players in the selection process. (Helson et al, 2005)
Although footballs attempt to shift the selection dates has seen little or slight improvements in producing equality into the selection process, varying the cut off dates for selection in sports has before been seen as a way of being successful in reducing if not preventing relative age effect. For example in swimming they have no cut off date. Ryan (1989) stated this would be successful if key competitions were avoided within certain months. Although this may work for individual sports such as swimming has been seen to not be applicable in team sports. (Musch & Grondin, 2001) This shows that there is a possibility of reducing the relative age effect.
The results and interventions found and used in other nations will help when aiming to reduce the impact which may be found within the English game and to see if there are any differences between England and the other Nations. Although these have been used to try and reduce the relative age effect in the sport few research have gone into what the coaches can do. Cobley et al (2009) suggested that just raising awareness of those responsible for the infrastructure and coordination of youth sport may be effective. Baker et al (2010) also suggested that increasing awareness and understanding of the relative age effect, as part of coach training and education programs, may help centre coaches attention to the potential selection bias.
The study comprised of a total of 2540 football players, ranging from the under 9 to under 15 age category within England. The players were allocated to one of three sub groups relating to their standard of play within the sport. The Academy group perceived as the highest level within the study consisted of a total number of 416players who played for an academy football team at the time of the study. The Grass root group consisted of 354players who played for a FA Chartered team. The third subgroup contained players who played recreational football within a community scheme that related to the lowest standard of play within the study. The total number of players in the recreational sub group was 1770.
The study required the birth dates of football players within age groups from U-9 to U-15. The players were chosen from community football, grass root football to academy football. The players were then sub-divided into secondary groups of grass root team players, community players and academy players. The details of the players were collected through contacting teams from the respective leagues through the use of letters. These letters detailed the information required from the clubs and how the results of the information would be used.
There were a number of coaches selected to undertake an interview. The coaches were chosen from the category in which the largest relative age effect was found from area one which was within the academy standard. The number of coaches chosen was four and this was due to limited time. The coaches selected were based on accessibility. Prior to interview the coaches were provided information regarding to the research in which is to be carried out. (Appendix 1.1)
There are two types of research, these are qualitative and quantitative. Qualitative research involves ‘researches describing kinds of characteristics of people and events without the use of measurements or amounts.’ (Thomas, 2003). Quantitative research involves ‘measurements and amounts of the characteristics displayed by people and events.’ (Thomas, 2003).
The data that will be collected will be both qualitative and quantitative. The quantitative data will be collected through primary research. Individual clubs from each area and age group were contacted through a letter which included the details of what the study will involve, the information needed and ethical considerations. The letter was sent to the coaches of the grass root teams, the academy managers and the chief executive of the community scheme. This data collection method was chosen to save time which is limited and through previous research having successfully acquired similar data. (Diaz Del Campo, 2010) Grass root team players are players from teams who were found to be FA chartered and within division A of their respected leagues. The recreational players came from a local community scheme located in South Yorkshire. The academy players were selected from a number of professional academy teams. The players and teams that were selected were based on accessibility and convenience. The teams that were chosen were local teams based within South Yorkshire, as money and time limitations would not enable collecting data from teams located in different regions. The teams were presented with a sample research response sheet in which they entered the necessary information required for the study. (Appendix 1.2, 1.3, 1.4)
To collect the data from the coaches structured interviews were carried out involving a number of open and probing questions. (Appendix 1.5) Open ended questions were used to allow the interviewee to provide more detail, rather than a one word answer from a closed question. The open ended questions allowed the interviewee to communicate using their own language and this takes you into their own world to view the area from their perspective. (Johnson & Christensen, 2011). This would add value to the study information being reliable. (Johnson & Christensen, 2011). The interviews were recorded through the use of a Dictaphone to reduce the risk of missing information and this enabled re analysis to ensure all important data was processed. Although a Dictaphone can help in recording the interviews, they can also have changed the behaviour of the interviewee and the answers they gave. (Silk et al, 2005) The interviewer made sure the coach was comfortable before progressing with the interview to enable reliable results could be collected and the coach’s responses weren’t systematic and fictitious.
Structured interviews will be used as the reliability of the interviews will be increased. (Hersen et al 2007) When looking at the purpose of the study which is to see if the coaches are aware of the relative age effect and what they are doing about it, the coaches could diverse into different areas if a structured path is not in place similar to a semi structured interview. (Hersen et al 2007) Although when designing this interview considerations such as making sure all areas needed are covered were considered to ensure the responses did not divulge into unnecessary areas. (Hersen et al 2007). The questions started with short and easy questions in which they could comfortably answer in order not to scare them and make them feel comfortable. (Johnson & Christensen, 2011) The interview then progressed onto more sensitive questions in which were placed once the interviewee felt comfortable and had given alot of their time to the interview reducing the possibility of fictitious responses being given. (Johnson & Christensen, 2011). Questionnaires were not used as a tool to collect this data as questionnaires could be returned incomplete and also could be found as being ambiguous. This would lead to incomplete data and unreliable sources. (Gratton & Jones, 2005)
Analysing data from the birth dates of players, each teams data that was collected were placed into the categories assigned for them (Community, grass root and academy). The different age groups were separately analysed to highlight which age group had the largest relative age effect. The birth dates were organised into sub categories, these were the birth months of the players. These sub categories are; September to November, December to February, March to May and June to August. These categories have been used in previous studies, (Simmons & Paull, 2001) and using these will give an area of comparison. The statistics will then show in which area the birth bias is evident and the different impact of the bias within different levels of the sport. Percentages of which players are born within each month will be produced giving a statistic which can be easily be compared. The statistical analysis tools within Microsoft Excel were used to create the data throughout the study.
The data collected through the structured interviews were analysed by transcription of the interviews, in which throughout this key themes could be formed to help seek whether there are current similarities in how coaches are selecting players and strategies minimizing the impact of the relative age effect. (Appendix 1.6) Any interventions being inputted by coaches were highlighted and used to compare what coaches are doing to overcome the relative age effect in their teams currently. Direct quotes seen as aiding research and results of the research are highlighted within the main body to provide evidence.
Throughout the research of the study ethical issues will be considered throughout. When collecting data for the birth dates of players, the managers of the grass roots teams, chief executive of the community scheme and academy managers were told specifically what they will be partaking within and what information is needed from them. (Appendix 1.7)To keep the players details confidential all that was required were the birth dates of the child, as this will keep personal details which are not required safe and ensured the research could not be related back to any specific person. Consent forms were included to evidence their cooperation within the study. (Appendix 1.2, 1.3, 1.4) When giving information the use of a data template in which the teams filled out to make sure only the necessary information is given was used.
The collection of data for the interviews will include specifically explaining verbally and documenting what the coaches will be partaking within, and where the results of the research will be used with consent forms highlighting their cooperation within the study. (Appendix 1.9)The necessary resources needed to carry out the interviews were accessed prior to the interviews, such as Dictaphones and interview rooms. (Appendix 2.0). The information collected from the coaches was specific to the research needs and the only information needed personally from the coach was of what club they are involved with. The information gathered through the interview was only viewed by the researcher and the MIS Supervisor. Questions were designed prior to the interviews to enable ethical approval on them. (Appendix 1.5)
It was made aware to all parties involved within the research project that there was ethical approval approved by an appropriate representative of the Faculty Research Ethics Committee at Leeds Metropolitan University and that if any issues arise they will be informed to ensure confidence in the divulgence of research. This was done through the completion of necessary forms, such as risk assessment (Appendix 2.1) local level approval .
Cite This Work
To export a reference to this article please select a referencing stye below: