Figure - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Talent identification research in soccer comprises the prediction of elite soccer performance. While many studies in this field have aimed to empirically relate performance characteristics to subsequent soccer success, a critical evaluation of the methodology of these studies has mostly been absent in the literature. In this position paper, we disc...
Contexts in source publication
Context 1
... the criterion used in most studies is the selection decision itself, which is usually a categorical variable indicating performance or skill level. Examples of performance-level indicators that have been used in studies are elite, sub-elite, and non-elite level [40][41][42]; professional, semi-professional, or non-professional level [43][44][45]; first team or reserves [46]; elite, club level, or dropouts [47,48]; national or regional level [49][50][51]; selected and non-selected players [52][53][54][55]; and nationally drafted or non-drafted players [56] (see Table 1). ...Context 2
... the empirical value of cross-sectional studies is limited compared with those with longitudinal designs, the methodological topics that are addressed in this paper also apply to those studies [43,52,56,74]; self-regulation, motivation, task and ego orientation, and cognitive functions (psychological) [3,21,50,52,[75][76][77][78][79][80]; and hours of practice and perceived social support (sociological) [44,76]. Other predictors that are derived from this classification scheme are technical skills, such as dribbling and passing technique, and self-assessed tactical skills [3,45,48,[81][82][83][84] (see Table 1). Given the multifaceted nature of soccer performance, it makes sense to investigate the extent to which these variables combined predict success and individual performance. ...Context 3
... restriction of range is common in talent identification research, but is rarely considered explicitly when the generalizability of predictive validities is discussed (see Table 1). ...Context 4
... date, the talent identification literature has not generally taken this into account. We were able to identify one study within the talent identification literature that considered utility [43], whereas the effect of the base rate on the usefulness of the examined predictors was not discussed in the other studies in Table 1. ...Citations
... This multidisciplinary approach introduces a range of factors that are potential indicators of future success or antecedents of high-level performance , however, these indicators are suggested to yield typically low prognostic validity (Güllich & Larkin, 2023). Whilst classifying these predictors within a specific discipline is the traditional approach within research and practice (e.g., Williams et al., 2020), an alternative approach is to develop a deeper understanding of how these predictors are assessed and emerge in practice (Barraclough et al., 2022;Bergkamp et al., 2019;Den Hartigh et al., 2018). ...
... Considering these potential issues, it has been suggested that a more ecologically valid approach is to also take samples of actual soccer performance, as these are expected to provide measures that are more representative of the prediction of future soccer performance (Bergkamp et al., 2019;Den Hartigh et al., 2018). For example, assessing a player's technical ability for the skill of passing would take a sample of their actual passing ability from match situations (e.g., a percentage of completed/accurate passes). ...
... For example, assessing a player's technical ability for the skill of passing would take a sample of their actual passing ability from match situations (e.g., a percentage of completed/accurate passes). A sample approach is thus theorised to provide greater predictive value, particularly in homogeneous groups (Bergkamp et al., 2019;Den Hartigh et al., 2018). Therefore, using samples seems a suitable tool for player profiling in groups such as male youth soccer players, but should also be combined with further multidisciplinary information (e.g., signs and subjective information; Barraclough et al., 2022;Bergkamp et al., 2019). ...
... Dribbling speeds, as measured along the curved paths, are only correlated with an individual's attacking and defensive ability in 1v1 competitions [20,24,25] and we do not know the causal relationship between our metric of dribbling and the actual activity utilised in gamerealistic situations. Our measure of dribbling performance relies on the execution of closed skills, which have often been criticised for de-contextualising skills from the performance setting [38][39][40]. However, we do expect that improvements in dribbling speed along the testing paths will be associated with improvements in a player's attacking and defensive capabilities during match-realistic conditions, but this still remains to be tested. ...
Background/Objectives: Dribbling is a fundamental skill in soccer, but assessing the performance of youth players in this skill is complicated by the confounded effects of age and physical development. In this study, our aim was to quantify the interactive effects of age, height, and mass on the dribbling performance of 180 players between 10 and 21 years old from an elite Brazilian junior academy. Methods: For each player, we quantified their dribbling and sprinting speed along four different paths with varying curvature, and their ability to perform specific, directed dribbling drills using one or both feet. To characterise patterns of variation among player's age, height, and mass-and to control for their confounding effects-we used a principal component analysis (PCA) to create a multivariate index of age and size (ASI). Results: Dribbling, sprinting, and directed dribbling were all positively associated with ASI; however, age alone was a better predictor of performance than ASI. Using multi-model inference, we found that a player's overall dribbling was best predicted by models that included sprint speed and overall directed dribbling ability (p < 0.0001). When performing subsequent analyses that separate each of the directed dribbling drills into using dominant, non-dominant, or both feet, we found the best predictors of overall dribbling performance were sprinting and directed dribbling activities that use both feet. Conclusions: These results provide the first set of normative data for a detailed metric of dribbling performance and soccer-specific foot coordination that can allow players and coaches to compare and assess their performances relative to a single population of high-quality junior players
... attainment of senior elite soccer or not) serves as a commonly used and understandable performance indicator, attainment ultimately relies on the judgement of a coach. This decision-making may not accurately reflect an individual's performance and thus limits a more nuanced understanding of what is indicative of future performance (Bergkamp et al., 2019). However, as objective soccer performance indicators are still lacking, such selection decisions are still necessary. ...
... Also, to detect elite soccer players using one-class support vector machine (one-class SVM) on a dataset (N = 951) collected from 14-year-old junior players. According to the results, this approach is useful to make decision [2,19]. To evaluate the performance of a team and identify talents, various machine learning algorithms are used in beach soccer and sepak takraw [3]. ...
... 39 Finally, it is important to note that using the current playing level of players as a criterion only provides limited information on interindividual differences and neglects prognostic validity. 40 The latter has been evaluated for the 360° video decision-making test in a sample of U17 and U19 football players, 28 but should also be assessed in younger age groups and across a wider range of cognitive skills in future research. ...
... Při výběru talentů hrají právě antropometrické vlastnosti (výška, tělesná hmotnost) velkou roli. Antropometrické vlastnosti, jako je výška, tělesná hmotnost a index tělesné hmotnosti jsou často zdůrazňovány při výběru talentů (Bergkamp et al., 2019). ...
My master thesis focuses on the analysis of the level of maximum speed and agility of elite football players in the U17 category. The testing was conducted with a team play-ing the highest competition in the given age category. The testing involved 21 players including 3 goalkeepers, 7 defenders, 6 midfielders and 5 strikers. The average age was 16.41 ± 0.22 years, the average height was 176.4 ± 6.3 cm and the average weight was 63.4 ± 6.1 kg. Players were tested for top speed and agility. Maximal speed was meas-ured over a 50-meter stretch, with each proband completing 2 trials and a full rest be-tween trials. Agility was measured using the 505 agility test. The test was completed twice on both legs, so probands completed a total of 4 trials. Only the best result from each test was taken. For both tests, probands were measured using a Catapult GPS vest that recorded top speed and a photocell that measured time for the agility test. The theo-retical part of the thesis builds on my undergraduate thesis where I describe the anthro-pometric and physical characteristics of soccer, somatic characteristics and playing positions, I also describe the U17 category and specialize in measuring and testing top speed and agility.
... Furthermore, in this same context, using machine learning, factors have been added and confirmed that allow the identification of sporting talent through the evaluation of performance in stages prior to the all-competitive category, determining the optimal time of specific dedication to the discipline in formative stages [9][10][11]. Although outstanding performance in lower categories of a sport does not ensure success in allcompetitive events [6,12], physical and morphological assessment with respect to sex, age, and maturity could be an initial parameter to be applied in the selective process through a multidimensional work [3,[6][7][8]13,14], since, during biological development, differences are observed in biomechanical, physiological, and psychosocial aspects that make the process even more difficult [2,6,13,15]. ...
... Furthermore, in this same context, using machine learning, factors have been added and confirmed that allow the identification of sporting talent through the evaluation of performance in stages prior to the all-competitive category, determining the optimal time of specific dedication to the discipline in formative stages [9][10][11]. Although outstanding performance in lower categories of a sport does not ensure success in allcompetitive events [6,12], physical and morphological assessment with respect to sex, age, and maturity could be an initial parameter to be applied in the selective process through a multidimensional work [3,[6][7][8]13,14], since, during biological development, differences are observed in biomechanical, physiological, and psychosocial aspects that make the process even more difficult [2,6,13,15]. ...
The purpose of this study was to assess the methodology for selecting sports talent in the Chilean Promising Athletes Program (CPAP). Methods: The standing broad jump (SBJ), medicine ball chest throw (MBCT), thirty-meter running sprint (RS-30), Agility-T (T-test), and height were analyzed in 3060 schoolchildren from Chile, grouped by sex, age, geographical area (north, central and south) and maturity status at peak height velocity (PHV) in pre-, circa- and post-PHV. Results: Superior differences were found in boys for all age ranges in SBJ, RS-30, T-test, and MBCT (p < 0.05). Girls were taller than boys until the age of 12 years (p < 0.01) with a reversal of this trend after 13 years old (p < 0.001). In geographical grouping, higher scores in SBJ, RS-30, and T-test were found in the central zone for boys and girls (p < 0.05). In both sexes, higher scores for the northern zone in MBCT (p < 0.01) and height (p < 0.05) are observed. Among selected talents, when comparing post-PHV vs. pre-PHV the differences are superior in all tests for post-PHV in girls and for MBCT and height (p < 0.05) in boys. In circa-PHV vs. post-PHV in MBCT, T-test, and RS-30 only girls (p < 0.05) were superior in post-PHV. For pre-PHV vs. circa-PHV, RS-30, and height (p < 0.05), both sexes were superior in circa-PHV. There are differences between talents selected by physical tests and technical aspects (p < 0.05). Conclusions: The results show variations in physical and anthropometric tests in schoolchildren by age, sex, geography, and maturity status that should be considered for talent selection.
... However, it would be incorrect to assume that these traits are unimportant for success in soccer in general (Ackerman, 2014). It is likely that the sample of academy players were exposed to the same training routine, had similar practise histories, and were (directly or indirectly) preselected on at least some of the variables in this study (Bergkamp et al., 2019). There is less volatility in the predictors and the criterion as a result of this pre-selection of athletes within a physically homogenous group of athletes. ...
Background: Physiological characteristics offer a potent approach to understanding the underpinnings of human behaviour and provide information about the internal state of an individual. It involves the function of the circulatory, respiratory, and related systems during the activity. Some physiological characteristics, such as pulse rate, force vital capacity, slow vital capacity, maximum voluntary ventilation, aerobic capacity, and anaerobic capacity, have long been allied with accomplishment in certain sports. The main objective was to study how the physiological characteristics are assessed among soccer players in talent identification in Manipur. The minor objectives were to 1) assess the global physiological characteristics of soccer players; 2) investigate the influence of age on the physiological characteristics of soccer players; and 3) to develop a norm and grading scale for talent identification in physiological characteristics among soccer players. Method: The subjects were a group of soccer players in Manipur (N = 45) who regularly attend morning practise programmes organised by various clubs and academies affiliated to the All Manipur Football Association (AMFA). A simple random sampling technique was used to select the subjects. The age of the subjects ranged between 12 and 16 years. Descriptive statistics such as mean and standard deviation and inferential statistics such as one-way ANOVA were used as statistical techniques for data analysis. Data
... The test power is a critical issue in talent research in general (Bergkamp et al. 2019) and specifically in GK-specific talent research . Therefore, sensitivity analyses were performed to determine the size of a possibly detectable effect utilizing G*Power (version 3.1.9.7; Faul et al. 2009) with the predetermined parameters for ANCOVAs (α = 0.05, 1 -β = 0.80, N = 120 across three groups, and two covariates). ...
Talent identification and development (TID) in football can be enhanced through objective assessments of talent predictors. Yet, available instruments rarely consider the unique demands of goalkeepers (GKs). During early phases of talent development, considering a GK’s giftedness relating to, for example, different
abilities (e.g. motor coordination), can complement views on highly specialized GK-specific attributes (e.g. technical skills). Therefore, this study aimed to evaluate a GK-specific motor coordination assessment to support TID in football at early developmental phases. Six tests were designed to assess ball control relevant to GKs. Their content and face validity were confirmed by independent experts (N = 8). The assessment was
evaluated with GKs selected for the German Football Association’s TID program (U12-U15; N = 120). This study examined the assessment’s test–retest reliability and agreement, the structural validity, and the concurrent validity. The overall test score demonstrated good test–retest reliability, although some individual tests showed lower coefficients. Additionally, limited agreement between repeated measurements due to considerable measurement error as well as issues with the structural validity of the test battery were identified. Nevertheless, the findings support the assessment’s concurrent validity as higher-rated (i.e. more talented) GKs outperformed lower-rated individuals. Yet, the test battery’s diagnostic accuracy is not high enough to justify individual decisions for talent selection. Overall, these findings support the consideration of motor coordination as a talent predictor in youth GKs and emphasize the assessment’s potential to enhance coaches’ evaluations. The identified psychometric weaknesses in some individual tests provide impetus to further optimize the test battery.
... 3 There is, however, a lack of longitudinal assessment to consider the validity of these practices across an academy cohort. 4,5 Previous research has identified physical differences between those in a professional academy and recreational academy players, highlighting the influence of physical attributes on successful progression. 6 Additionally, physical differences have been shown between selected and non-selected international U14-U17 players. ...
Background
Increases in high-intensity locomotor activity of match play have been recorded in elite soccer. This places an onus on academy practitioners to develop players for the future demands of the game. At an academy level, locomotor data are not available for analysis over a longitudinal period, and thus changes can only be assessed with physical attribute assessment. The aim of the present study is to establish if changes in physical capacity were observed in a professional Scottish soccer academy over a ten-year period.
Methods
A retrospective analysis was completed where linear mixed effect (LME) models were individually fitted to explain variation across each measure of physical capacity. Model selection was undertaken with likelihood ratio tests where the initial complex models were compared to simpler nested models to arrive at the final model by maximum likelihood.
Results
The main findings were that most recent players' sprint test data revealed a significant improvement in 5m, 10m and 20m sprint performance, greater increases in CMJ performance in older age groups, and greater increases in YYIR1 performance at U13 and U14. Most physical results showed increased performance with greater relative height and weight.
Conclusion
Players recruited more recently to academies are fitter than they were previously. Reference values within clubs that establish first team requirements will contribute to appropriate planning and implementation of training.
