A Scientific Basis for Talent Identification and Elite Player Development October 2006

1
A Scientific Basis for Talent Identification and
Elite Player DevelopmentOctober 2006
2
Questions we asked..

• Is there any evidence in the scientific
  literature on the ability (or inability) to
  identify sporting giftedness/talent?
• Are there any known indicators of potential (but
  which are invariant to age and training)?
• Are there specific age-related developmental
  stages in developing players?
• What time and/or effort is required to master a
  task (in this instance, football)?
• Are we making appropriate use of the entire
  talent pool available to us?

3
Gagnés Differentiated Model of Giftedness and
Talent (DMGT)
Under the partial influence of genetic endowment
4
Gagnés Differentiated Model of Giftedness and
Talent (DMGT)

• Giftedness is defined as the possession and use
  of untrained and spontaneously expressed natural
  abilities (called aptitudes or gifts), in at
  least one ability domain, to a degree that places
  an individual at least amongst the top 10 of age
  peers
• Talent is defined as the superior mastery of
  systematically developed abilities (or skills)
  and knowledge in at least one field of human
  activity to a degree that places an individual at
  least amongst the top 10 of age peers who are or
  have been active in that field or fields
• Where giftedness is the beginning of the learning
  process, talent is the result of it.

5
Gagnés Differentiated Model of Giftedness and
Talent (DMGT)

• Three components influence the learning/practice
  process that changes natural athletic ability
  into systematically developed sporting skills
  (talent)
• Intrapersonal catalysts (physical and
  psychological factors)
• Environmental catalysts
• Chance factors

• There is an element of both nature and nurture in
  this model
• Without an innate ability no amount of training
  will create a top elite player, and
• Without the appropriate quantity and quality of
  training a player will not develop into a top
  elite player

6
Questions we asked..

• Is there any evidence in the scientific
  literature on the ability (or inability) to
  identify sporting giftedness/talent?
• Are there any known indicators of potential (but
  which are invariant to age and training)?
• Are there specific age-related developmental
  stages in developing players?
• What time and/or effort is required to master a
  task (in this instance, football)?
• Are we making appropriate use of the entire
  talent pool available to us?

7
Ability or inability to identify or predict
giftedness/talent

• Is it possible to identify a priori which persons
  might become talented footballers, and which not?

8
Talent Prediction

• A number of studies have been undertaken to
  discriminate between elite and non-elite soccer
  players on the basis of anthropomorphic,
  physiological, psychological and soccer-specific
  skills tests
• Univariate differences
• Endomorphy (fatness), skinfolds (7 areas) and
  body fat
• Sprint speed, speed endurance, aerobic power,
  agility, standing vertical jump
• Task-orientation, anticipation, somatic anxiety
• Dribbling skill
• Multivariate (discriminant) analysis
• Agility, 30m sprint time, ego orientation and 1
  vs 1 anticipation were the set of variables that
  provided the best predictors of talent.
• Results indicate that
• Advanced biological maturity status is associated
  with slightly better technical performance on the
  tests specifically, older, more mature players
  perform better in dribbling and in ball control

9
Talent Prediction

• Leaves us wondering if
• these variables/model are only useful in talent
  discrimination rather than talent
  identification/prediction
• these measures are able to discriminate among
  players already selected and exposed to
  systemised training, because the sensitivity of
  the tests tends to decrease once players reach
  the elite level

10
Talent Prediction

• No consensus about the relative importance of
  physical, psychological and physiological
  qualities in predicting football talent (Williams
  and Reilly, 2000)
• Talent ID programs around the globe are not
  firmly based on scientific rationale (Williams
  and Franks, 1998) and rely heavily on intuition
  or eye of expert coaches and talent scouts to
  identify talented performers

11
Talent Prediction

• It is probable that talent detection,
  identification and development are not amenable
  to a reductionist process that can permit
  ultimate potential to be defined with much degree
  of certainty. The problem is especially complex
  in sports such as soccer where performance itself
  is multifactorial. At present anthropomorphic and
  physiological profiling is best viewed as an
  objective means of monitoring young players,
  while emphasis should be placed on technical
  skills and engagement in teamwork.
• (Reilly, T., Bangsbo, J. and Franks, A. (2000).
  Anthropometric and physiological predispositions
  for elite soccer. Journal of Sports Sciences, 18,
  669-683.)
• A coaches judgement is the best solution for
  identifying talent, not some scientific detection
  model.
• (Tranckle, P. (2004). Understanding giftedness
  and talent in sport. The Coach, 21, March/April
  2004, 61-73.)

12
Talent Prediction

• Psychological factors are what distinguishes
  successful elite players from their non-elite
  counterparts (Morgan 1979). Little work done on
  psychological factors in talent ID, or even which
  psychological factors, or how to measure them
• Cognitive factors and game intelligence
  perceptual skill in soccer is a promising area
  for talent ID
• Skilled players are better than lesser skilled
  players at recognising structured patterns of
  play because their knowledge allows them to
  chunk perceptual information into larger and
  more meaningful units (cf. Ross (2006) regarding
  expert chess players). This enables them to
  recognise emergent features of a pattern of play
  early in its initiation, thereby facilitating
  anticipation
• Skilled players also use advance cue
  utilization (i.e. ability to make accurate
  predictions based on information arising from an
  opponents posture and bodily orientation). This
  has been tested extensively in the penalty
  situation.

13
Talent Prediction

• It is conjectured that skilled players use their
  expert knowledge to dismiss many potential
  situational events as being highly improbable
  and attach a hierarchy of probabilities to
  remaining events, i.e. likelihood of occurrence.
• Can perceptual skill be developed through
  training?
• Teach young players about important
  features/structures of play. Set pieces are
  obvious starting points.
• Focus on postural cues, and the relationship
  between these and subsequent performance (via
  video simulation).
• Decision making skills can be taught from as
  young as 7 years, i.e. earlier than expected.

14
Talent identification as it currently happens

• Vrljic and Mallet interviewed five youth coaches
  responsible for selecting and coaching Under
  15/16 Queensland boys teams at the national
  championships.
• Results show that four categories are considered
• Physical skills
• Running speed (over short distances, say 30-40
  meters) most important
• Physical strength
• Skill-based (i.e. being able to win the ball)
• Physical (i.e. physical appearance of
  athleticism)
• Aerobic fitness
• Technical skills
• Ball control at speed
• Ball control under unpredictable circumstances
• Ball control with various body parts

15
Talent identification as it currently happens

• Cognitive-perceptual skills
• Reading the game / anticipating play
• Decision making / choosing the best option /
  vision (usually have to be taught this in the
  development process, so should not be part of the
  identification process)
• Personal qualities
• Hunger to succeed
• Mental toughness
• Unselfish, i.e. team player rather than
  individual
• Coaches preferred to identify talent in authentic
  playing situations (i.e. a match) than in
  skill/drill tests.

16
Body shape

• Is there a particular body shape that is best
  suited to elite football?

17
Body shape (somatotype)

• 3 components
• Endomorphy (fatness)
• Measured by skinfolds, adjusted for height
• Ectomorphy (linearity/tallness)
• Measured by height/weight ratio
• Mesomorphy (skeletal/muscular)
• Measured by limb circumferences corrected for
  fatness, skeletal breadths/widths, height
• An individuals physique includes all 3
  components e.g. A typical adult elite player
  might have Endo / Meso / Ecto 2.8 / 5.3 / 3.3
• Each component is expressed numerically from 1
  (low) to 7 (high)
• NB. Components change with growth and maturation
  during adolescence

18
Body shape (somatotype)

• Youth soccer players tend to have physiques
  similar to those of elite adult players, with
  adult elite players more mesomorphic, probably
  reflecting growth of muscle and skeletal mass
• Potential effects of training on muscularity at
  more elite levels
• Youth soccer players tend to have more fat and be
  relatively speaking taller than elite adult
  players
• None of this research is predictive, i.e. the
  shape of body that might indicate a potential
  elite player

19
Questions we asked..

• Is there any evidence in the scientific
  literature on the ability (or inability) to
  identify sporting giftedness/talent?
• Are there any known indicators of potential (but
  which are invariant to age and training)?
• Are there specific age-related developmental
  stages in developing players?
• What time and/or effort is required to master a
  task (in this instance, football)?
• Are we making appropriate use of the entire
  talent pool available to us?

20
Invariant indicators of potential
Is it possible to identify the subset of the
population that are gifted in the domains that
are relevant to sporting ability? i.e. have the
right nature in the nature/nurture debate? The
measure would have to be invariant to age and to
training (If so, we might concentrate our talent
identification/development process on that
subset, assuming that very few others will ever
become talented) Proviso Any procedure we
use will have to be practical and easy enough to
implement
21
Measuring innate ability in sport

• Ability in sports is highly related to male
  physical competitiveness
• It is well researched that the ratio of the
  length of the 2nd to the 4th fingers (2D4D) is a
  negative correlate of prenatal and adult
  testosterone
• Men with lower 2D4D ratios reported higher
  attainment in a range of sports and had higher
  mental rotation scores (a measure of
  visual-spatial ability) than those with high
  2D4D ratios.
• Implication Testosterone promotes the
  development and maintenance of traits which are
  useful in sports, and the 2D4D ratio is a way to
  measure this ability, independent from previous
  exposure to or training in the sport.
• Recent research (Spector, T. (2006). British
  Journal of Sports Medicine) shows that this holds
  in the case of females as well. It is conjectured
  that genetic factors play a major role.

22
Measuring innate ability in sport

• This is promising research, but a little
  impractical at this stage
• Could there be other such measures?

23
Questions we asked..

• Is there any evidence in the scientific
  literature on the ability (or inability) to
  identify sporting giftedness/talent?
• Are there any known indicators of potential (but
  which are invariant to age and training)?
• Are there specific age-related developmental
  stages in developing players?
• What time and/or effort is required to master a
  task (in this instance, football)?
• Are we making appropriate use of the entire
  talent pool available to us?

24
Age-related development stages

• When is the best age to concentrate training?

25
Age-related development

• Coaches worldwide currently design long and
  short-term athlete training models as well as
  competition and recovery programs based on their
  athletes chronological age.
• Research has shown that chronological age is not
  a good indicator on which to base athlete
  development models for athletes between the ages
  of 10 to 16. There is a wide variation in the
  physical, cognitive and emotional development of
  athletes within this age group.
• Ideally, coaches should determine the biological
  age of their athletes and use this information as
  the foundation for athlete development models.
  Unfortunately, there is no reliable procedure to
  identify biological age non-invasively.

26
Age-related development

• Use the onset of Peak Height Velocity (PHV) as a
  reference point for the design of optimal
  individual programs with relation to critical
  or sensitive windows of trainability during the
  maturation process.
• Prior to the onset of PHV, boys and girls can
  train together and chronological age can be used
  to determine training, competition and recovery
  programs.
• The average age for the onset of PHV is 12 and 14
  years for females and males respectively.
• The onset of PHV is influenced by both genetic
  and environmental factors, including climate,
  cultural influences, and social environment.

27
Age-related development

• Using simple measurements, PHV can be monitored
  and training can be related to and optimised to
  exploit the critical periods of trainability.
• All energy systems are always trainable, but
  during the so-called critical periods
  accelerated adaptation will take place if the
  proper volume, intensity and frequency of
  exercise are implemented.

28
Questions we asked..

• Is there any evidence in the scientific
  literature on the ability (or inability) to
  identify sporting giftedness/talent?
• Are there any known indicators of potential (but
  which are invariant to age and training)?
• Are there specific age-related developmental
  stages in developing players?
• What time and/or effort is required to master a
  task (in this instance, football)?
• Are we making appropriate use of the entire
  talent pool available to us?

29
Time and effort to develop mastery in football

• How much / long does it take to master a sport
  (i.e. to attain elite level)?

30
The 10 year / 10,000 hours rule

• The Theory of Deliberate Practice is predicated
  on the notion that it is not simply training of
  any type, but rather a minimum of 10 years
  engagement in deliberate practice that is the
  necessary condition for the attainment of
  expertise. Deliberate practice refers to practice
  activities done with the specific instrumental
  goal of improving performance, and which
• Are performed in a daily, work-like manner
• Require effort and attention
• Do not lead to immediate social or financial
  rewards
• Are frequently not enjoyable to perform
• (i.e. the hard yards)
• First suggested by Simon and Chase (1973) in
  relation to skill in chess
• Subsequent work has shown this to apply to
  expertise in other domains, such as music,
  mathematics, swimming, distance running, tennis,
  soccer, hockey etc

31
The 10 year / 10,000 hours rule

• Subsequent work tested subjects (experts and
  non-experts) in hockey, netball and basketball in
  Australia.
• This research concludes that international
  standard players in almost all sports took
  between 10 and 13 years to develop fully to
  international level, with around 6,000 hours of
  sport-specific training. The balance of hours
  could be attributed to
• Maintaining the level of expertise
• Other sports training (both before attaining
  expert status in the chosen sport, and
  non-specific sports training, e.g. running to
  maintain fitness).
• This implies a strong leaning towards the
  nurture side of the nature/nurture debate, i.e.
  its all in the training

32
The 10 year / 10,000 hours rule

• Other studies have shown a negative correlation
  to exist between hours of sport-specific training
  required to reach international expertise and
  number of prior sporting activities experienced,
  i.e. participation in other sporting activities
  may aid development of expert decision-making
  skills, and transfer of learning may take place
  from one sport to another.
• Significantly, the results point to an
  implication that early specialisation may not be
  a necessary requirement for expert level
  performance in decision-making sports.

33
Effortful study is a prerequisite for success

• A recent study (in Scientific American) has
  looked at how chess grand-masters are able to
  assimilate so much information so quickly and
  accurately.
• Concludes (based on academic research) that
  effortful study is the key to achieving success
  where effortful study. entails tackling
  challenges that lie just beyond ones
  competence, i.e. to get good you must extend
  yourself.
• This explains
• why you can only be the best at competitive teams
  sports if you are playing with the best and
• why a very strong RAE among youth may be somewhat
  counterbalanced at an older age (if the youngest
  in an age group have to engage in more effortful
  study than others of their age in order to excel
  and if those differences become habits then
  persist into the professional years while the
  early innate benefits of being slightly older
  disappear)
• the characteristics of a developmental pathway
  (i.e. it must provide substantial opportunity for
  effortful study at all ages)

34
The 10 year / 10,000 hours rule

• Generally accepted categorisation of sports
  participation
• Sampling years deliberate play (6 to 12 years)
• Broad range of sporting activities
• Emphasis on fun
• Specializing years (13 to 15 years)
• Decrease in other sporting activity, starting to
  specialize in one sport
• Growing emphasis on skill development
• Investment years (16 and over)
• Devotion to one primary activity
• Strategic and competitive skills training
• Teachers become more demanding as one moves
  through the stages

35
Questions we asked..

• Is there any evidence in the scientific
  literature on the ability (or inability) to
  identify sporting giftedness/talent?
• Are there any known indicators of potential (but
  which are invariant to age and training)?
• Are there specific age-related developmental
  stages in developing players?
• What time and/or effort is required to master a
  task (in this instance, football)?
• Are we making appropriate use of the entire
  talent pool available to us?

36
Squandering the talent pool the Relative Age
Effect

• Systematic exclusion of groups of players, some
  of which might be highly talented

37
Relative Age Effect (RAE)

• The relative age effect (RAE) in sport was first
  noted among elite level ice hockey players in the
  USA. These findings demonstrated that for major
  junior leagues and the NHL player birth dates
  decreased in frequency from January through
  December.
• The RAE is strikingly evident in activities that
  are competitive and where performance is highly
  correlated with age and maturity, (e.g. football
  in Australia)
• It was theorized that the RAE arose from the
  consequences of grouping young boys for entry
  into organized sport, thereby producing a
  one-year age range for the participants. As size,
  speed, and coordination are highly correlated
  with age, older players within the age-group
  will, on average, show superior performance.

38
Relative Age Effect (RAE)

• Thus maturity has been mistaken for ability by
  coaches, peers and the individuals themselves.
  The age-advantaged children are imbued with
  greater self-confidence and regard by others. The
  opposite will likely hold for those younger than
  their group-mates, who are likely to suffer from
  lowered self confidence and self-esteem.
• A consequence is an increased drop-out rate for
  those disadvantaged by age.
• Predictably, the RAE has also been found in a
  host of other competitive sports such as
  baseball, basketball, soccer, rugby, American
  football, cricket etc etc.

39
Relative Age Effect (RAE)

• RAE has been observed in activities other than
  sport as well the obvious one is the cut-off
  for school entering in all schools, but also
  musical ability, youth suicide etc
• There are some sports and activities in which no
  RAE is evident (e.g. dance, table tennis,
  gymnastics to name a few). These sports depend
  heavily on the technical ability (or motor skill)
  of the participant.

40
Relative Age Effect (RAE) Australian data

• Boys Under 14 and 15 NYC 2006 (n351)

Very highly statistically significant trend
41
Relative Age Effect (RAE) Australian data

• Boys Under 14 and 15 NYC 2006 (n351)

Very highly statistically significant trend
42
Relative Age Effect (RAE) Australian data

• Under 17 and 20 national teams 2006 (n47)

Statistically significant trend
43
Relative Age Effect (RAE) Australian data

• Under 17 and 20 national teams 2006 (n47)

Statistically significant trend
44
Relative Age Effect (RAE) Australian data

• HAL and overseas professionals (includes
  Socceroos and Futsalroos) (n278)

Statistically significant trend
45
Relative Age Effect (RAE) Australian data

• HAL and overseas professionals (includes
  Socceroos and Futsalroos) (n278)

Nearly statistically significant trend
46
Relative Age Effect (RAE) Australian data

• Socceroos 2006 FIFA World Cup squad (n23)

Not statistically significant (sample size)
47
Relative Age Effect (RAE) Australian data

• Socceroos 2006 FIFA World Cup squad (n23)

Not statistically significant (sample size)
48
Relative Age Effect (RAE) Australian data

• Girls NTC U15 and U17 2006 (n268)

Not statistically significant
49
Relative Age Effect (RAE) Australian data

• Girls NTC U15 and U17 2006 (n268)

Not statistically significant
50
Relative Age Effect (RAE) Australian data

• Matildas and Young Matildas 2006 (n52)

Nearly statistically significant trend
51
Relative Age Effect (RAE) Australian data

• Matildas and Young Matildas 2006 (n52)

Statistically significant trend
52
Relative Age Effect (RAE) Australian data

• Observations on the Australian data
• RAE is greatest at youngest age levels for boys
  (U14 and 15), and decreases as ages (and
  abilities) increase
• Estimated loss of the talent pool is around 42
  at the U14/15 age group!
• The RAE effect reduces until at the professional
  level it is fairly small
• Socceroos negative RAE! we conjecture that
  this is caused by very talented youngsters having
  to push themselves and play above themselves to
  keep up with the older kids, and this work ethic
  has remained with them throughout their career,
  leading to them staying ahead of their peers

53
Relative Age Effect (RAE) Australian data

• Observations on the Australian data (contd)
• Socceroos seems to indicate an August cut-off
  (43 in Aug to Oct, same proportion as the
  U14U15 (42) and U17U20 (44)) slipped into
  the European cycle?
• Amongst the girls the RAE is not significant,
  although the fourth quarter is still
  under-represented we conjecture that this is
  because competition for places is less fierce
  than in the boys competition
• In the elite senior women the RAE is again
  strongly evident competition is stronger at
  this level

54
Relative Age Effect (RAE) Australian data

• We are not alone!
• Football 1st January 31st December
• AFL 1st January 31st December
• Rugby Union 1st January 31st December
• Rugby League 1st January 31st December
• Netball 1st January 31st December
• Cricket 1st September 31st August
• This means that a significant proportion of kids
  born later in the year are, on average, dropping
  out of all organised winter team sports because
  of maturity issues (RAE) rather than ability.
  There is thus a huge pool of talent out there
  which could be tapped.
• Question Does this mean that competitive and
  specialised team sports (i.e. football) should
  start later, once the maturity/ability factors
  have aligned (say only at age 13?)

55
Relative Age Effect (RAE)

• Potential remedy offered
• Classification system based on biological age
  (e.g. anthropomorphic or physiological
  measurements)
• Classification systems based on chronological age
• Variation of cut-off date within the competition
  year from year to year.
• Rotation of RA advantage by using a period
  different from a 12-month period - this is
  probably unmanageable and impractical, and would
  reduce the number of players per age group
• Change soccers cut-off date away from January to
  take advantage of drop-outs from other
  footballing codes.
• Create different squads of developing players who
  are technically well-developed, but are currently
  lacking in terms of physical development.
• Start competitive competition at a later age,
  e.g. 14 or 15, after physical development has
  equalised.

56
Relative Age Effect (RAE) Australian data

• A variation on RAE occurs in the later years due
  to FIFAs policy of having the Under 17 and Under
  20 World Youth Cup in odd years only (e.g. 2005,
  2007 etc)
• This means that every second year of players
  misses out on this pinnacle event, but each group
  should play in one of the two events (but
  unlikely to play in both)
• It is conjectured that the Under 20 competition
  favours those born in odd years, and the Under 17
  competition favours those born in even years

57
Relative Age Effect (RAE) Australian data

• A variation on RAE Under 17 and 20 national
  teams Year of birth

Disadvantaged too young for 2005, but too old
for 2007
Disadvantaged too young for 2005, but too old
for 2007
Under 17
Under 20
58
Summary Some key assertions about elite coaching

• Sustained success comes from training and
  performing well over the long-term rather than
  winning in the short-term.
• There is no short-cut to success in athletic
  preparation.
• Overemphasizing competition in the early phases
  of training will always cause shortcomings in
  athletic abilities later in an athletes career.
• (Balyi, I. and Hamilton, A. (2004). Long term
  athlete development trainability in childhood
  and adolescence. Olympic Coach, 13, No.3 (Spring
  issue). http//coaching.usolympicteam.com/coaching
  /kpub.nsf/v/2ltad04.