Identification of key performance parameters and development of a youth badminton performance index in male youth players

Abstract

Youth badminton performance involves an interplay of structural, physiological, training, and experiential factors more than a stand-alone measure like competition ranking. Yet, in youth programmes, athlete assessment is based on such siloed indicators such as isolated fitness tests or competition results and therefore, fails to appreciate the complexity of performance progression. Consequently, the objective of this study was to ascertain the fundamental multi-focal parameters relevant to youth badminton performance and construct a descriptive performance profiling index in the form of a Youth Badminton Performance Index (YBPI). The study employed a cross-sectional design and the subjects were 170 youth male badminton players aged 9–16 years, drawn from state-level teams and national development programmes. The components of the study were anthropometry, physical fitness, training exposure, competition and injury-related functional activity. To determine the dimensions of performance that lie underneath each of the components, Principal Component Analysis (PCA) with Varimax rotation was performed. Ten components accounting for 76.43% of the total variance were retained in the analysis. The domains of variance included anthropometric make-up, aerobic-agility, training exposure, competition experience, physical fitness, and injury-related functional readiness. The components were standardised and the YBPI was constructed from the sum of the standardised components, allowing athletes to be positioned within a relative performance grouping based on their multidimensional profile within the sample. These groupings are intended to support descriptive profiling and athlete monitoring rather than talent identification or prediction of future performance. Although the YBPI offers a multi-faceted approach to athlete profiling and monitoring and helps coaches to discern the athlete’s characteristics and traits across multiple performance domains, the index is designed for performance profiling at a single point in time and not for forecasting or talent scouting. Practically, the YBPI may assist coaches in identifying whether an athlete’s current profile reflects lower, moderate, or higher relative multidimensional performance status within a comparable youth cohort.

Introduction

Badminton is classified as a racket sport, and is characterized by multiple short, high-intensity actions, rapid directional changes, as well as explosive jumps and other actions that make it extremely demanding physiologically and neuromuscularly on players^1,2. During a game, players need to make precise and quick tactical decisions as well as perform accelerations, lunges, and overhead strokes (skills that require a high level of technical precision). When it comes to young players, the matter is even more complicated since a combination of a multitude of variables has to be taken into account and integrated to evaluate the effects of growth, biological maturation, and exposure to training during the critical period of adolescence^3,4. The youth athletes case is even more complicated and presents multiple challenges, particularly with respect to young people in sports. In practice, the young athletes’ case is even more complicated and presents multiple challenges due to the young age of the participants, and the evaluation of young athletes is largely based on arbitrary and fragmented criteria, such as separate physical fitness tests, results from competitions, or training volume.

Although the aforementioned indicators are significant, unfortunately, the current literature tends to treat them in isolation, leading to an overly fragmented understanding of youth athletes’ performance profiles^5–7. A continuous body of evidence in youth sport, provides insights to emphasise that the understanding of an athlete’s capability, especially of an adolescent athlete, cannot be understood in isolation, particularly when biological maturation in that age cohort can have a pronounced effect on the physical performance outcome^5,6,8. Therefore, the range of performance outcomes in youth athletes is often a function of developmental advantages rather than sport-specific competencies. The modern athlete development frameworks reiterate the need to assess youth athletes in multiple, integrated developmental domains. In their most common format, these domains are concerned with the athlete’s physical capacity, structure, training, and functional readiness^4,5. In performance evaluation, this means that the evaluation of youth athletes should go beyond their tournament placement or the outcome of their competitive performance. It should capture a comprehensive athlete profile that reflects physical, structural, and experiential characteristics and their interplay to facilitate the athlete’s progress in the sport^4,5,9.

In regard to previous studies in badminton, it is possible to list a number of specific anthropometric and physical factors related to the performance of badminton players. As described in the article by¹. Attributes such as muscular strength, agility, speed, and explosive power are considered essential elements for good performance in court coverage, movement in different directions, and good execution of different strokes in badminton. Other anthropometric factors, such as body height, limb length, body composition and limb reach, and other biomechanical and efficient movement components, also influence performance in badminton. Furthermore, the occurrence of injuries and musculoskeletal problems may alter the training calendar and also alter the physiological preparedness of the badminton player, restricting the player’s development for a considerable time. However, despite the importance of such factors, many studies choose to review some of the factors related to performance in badminton, without building a performance profile of a badminton player^1,10,11. Hence, previous studies published in badminton have largely examined performance-related constructs through isolated domains such as anthropometric and physical capacities or psychological predictors, with relatively little emphasis on integrating these dimensions within a uniform multidimensional descriptive profile framework^12,13.

In the field of sports science, and particularly in the field of sports performance, multivariate analysis is often used to simplify very large sets of data and identify the most important factors for different groups of athletes¹⁴. Using principal component analysis, the most related set of variables can be grouped and simplified, converting the variables to a smaller number of factors. This helps to identify the most dominant factors in a very complex performance analysis of athletes. In an index-based model, the components can be condensed into a composite score that aids in the practical profiling of athletes and their performance grouping^15–17. Such methods might allow coaches and practitioners a means to manage athlete improvement and help in evidence-based methods of talent recognition and training. Therefore, the primary objectives of this study were (i) to determine the primary multidimensional components that are pertinent to youth badminton performance and (ii) to construct a Youth Badminton Performance Index (YBPI) aimed at providing a multidimensional descriptive performance profiling framework and performance grouping within the study sample.

Materials and methods

Study design and participants

The purpose of this research was to devise a Youth Badminton Performance Index (YBPI), as well as recognise the principal, multidimensional elements that are pertinent to youth badminton performance, using a cross-sectional research study design. Youth badminton players aged 9 to 16 years old, who are actively participating in structured training and competitive programmes, comprise the study sample of 170 male players.

All participants are part of the national development squad programmes and the state badminton teams. Participants were classified as state-level players if they were registered participants in state-level badminton competitions, and as national development players if they were participants in the organised youth development programmes of the national badminton talent pathway. The participants were classified into the youth badminton competition categories of Under-12 (U12), Under-14 (U14), and Under-16 (U16), which represent the selected age range. This age range represents the interstage of the adolescent developmental period associated with a succession of rapid growth, biological maturation, and neuromuscular development, which impacts performance in the sport during youth sport development^4,6. However, biological maturation status was not directly assessed in this study, and therefore, maturation-related differences were not controlled for in the analysis.

Although indicators such as maturity offset, peak height velocity (PHV), or predicted adult stature are commonly used to provide developmental context in youth sport research, their implementation was beyond the practical scope of the present field-based testing design, which prioritised standardized, non-invasive, and operationally feasible assessments across multiple performance domains. In addition, methods such as maturity offset, predicted peak height velocity, and percentage of predicted adult stature may require assumptions or supplementary developmental information, which can introduce estimation error when applied to heterogeneous adolescent athletic populations. Furthermore, commonly used non-invasive maturity estimation approaches require cautious interpretation when applied across heterogeneous youth athletic populations⁶. Accordingly, the present study was designed to reflect real-world applied performance monitoring conditions commonly used in youth badminton programmes, where direct assessment of biological maturation is rarely implemented.

All methods were conducted in accordance with ethical guidelines for research involving human participants. Ethical approval for this study was obtained from the Research Ethics Committee of the National Defence University of Malaysia (Jawatankuasa Etika Penyelidikan, JKEP), Universiti Pertahanan Nasional Malaysia (Approval No.: JKEP 17/2025). Written informed consent was obtained from all participants and/or their parents or legal guardians prior to data collection. Participation was voluntary, and participants were free to withdraw at any time. All data were collected and analysed anonymously in accordance with the Declaration of Helsinki¹⁸.

Variables and data collection

Performance profiling was carried out on three domains that reflect structural, physical, and functional characteristics related to badminton performance. These domains comprised anthropometric characteristics, physical fitness components, and injury-related functional^1,19.

Anthropometric measurements

Anthropometric measurements aimed to define the structural body characteristics of youth badminton players that may affect biomechanical leverage, reach, and movement efficiency in the performance of badminton. All the measurements were carried out in accordance with the International Society for the Advancement of Kinanthropometry (ISAK) standardized protocols²⁰ by assessors who have received ISAK accreditation.

To verify the reliability of the anthropometric assessment, test–retest reliability was evaluated using the Intraclass Correlation Coefficient (ICC). The results demonstrated excellent to near-perfect reliability across all anthropometric variables, with ICC values ranging from 0.999 to 1.000 (p < 0.001). These very high ICC values indicate a high degree of measurement consistency and reproducibility, which may be attributed to the use of standardized ISAK procedures, calibrated instruments, and measurements conducted by trained assessors, thereby minimizing measurement error across repeated trials.

The anthropometric variables included assessment of body mass (kg), stature (cm), body mass index (BMI; kg·m⁻²), shoulder width (cm), upper limb length (cm), arm span (cm), leg length (cm), femur length (cm), upper arm circumference (cm), forearm circumference (cm), calf circumference (cm), and thigh circumference (cm). Body mass and stature were recorded by means of a digital weighing scale and a portable stadiometer, and subsequently, BMI was computed from the measurements of body mass and stature. Linear body length and limb circumference were measured with a flexible anthropometric tape according to standard procedures to guarantee measurement reliability²¹.

Physical fitness assessments

A field-based testing battery was employed to evaluate physical fitness in the context of badminton performance, capturing the physiological and neuromuscular components. Participants’ muscular endurance was evaluated through the Sit-Up Test and the Push-Up Test, both performed over a 60-second duration, with results recorded as total repetitions²². Lower-body explosive power was assessed using the vertical jump test. Relative power (W/kg) was calculated using the Sayers Eq. 2³. Power (W) = (60.7 × jump height (cm)) + (45.3 × body mass (kg)) − 2055. Relative power was subsequently obtained by normalizing peak power to body mass. The Sayers equation was selected as a practical field-based estimator of lower-body peak power because it incorporates vertical jump height and body mass, both of which are readily obtainable in applied sport settings²³. Although prediction equations for muscular power may vary in accuracy across developmental stages, the present study used the resulting value as a relative indicator of lower-body explosive performance for multidimensional profiling purposes rather than as a laboratory-equivalent measure of peak mechanical power²⁴. Flexibility was assessed by the V-Sit and Reach Test, maximum distance was recorded in centimeters (cm); while static balance was measured by the Stork Stand Test, the result was recorded in seconds (s)²⁵. Upper body strength was determined by the Handgrip Strength Test, measured in kilograms (kg) using a hand dynamometer²⁶ and hand-eye coordination was measured by the Hand Wall Toss Test, the result was recorded by the number of catches made²⁷.

The Sideway Agility Test and Four-Corner Agility Test are used to measure agility and change-of-direction ability. Both tests are used to measure specific movement and rapid change of direction during badminton play^1,28. For both tests, performance is measured in time, which is recorded in seconds (s). Estimation of aerobic capacity was determined using the multistage shuttle run test, which is used to determine maximal oxygen uptake (VO₂max was expressed in ml·kg⁻¹·min⁻¹)²⁹. All tests were conducted following standardized procedures to ensure consistency and reliability across participants.

Injury related functional status

Injury-related functional status was used to include contextual elements that potentially affect training continuity and physical preparedness of youth badminton players. Injury details were gathered using a self-reported Google Form questionnaire capturing injury history, injury recovery status, injury recurrence, injuries and training, and the perceived impact of injuries on training. These elements are considered critical in the context of athlete monitoring and performance evaluation³⁰. The questionnaire achieved an acceptable level of internal consistency, with a Cronbach’s alpha coefficient of 0.859.

The combination of injury-related elements with the anthropometric and physical fitness-related attributes was used to define the multifaceted performance profile of youth badminton players. These attributes were utilized in Principal Component Analysis (PCA) to determine performance-related components and develop the Youth Badminton Performance Index (YBPI).

All variables were numerically coded prior to analysis. Training-related variables (technical/tactical training, physical conditioning, and match play) were categorised into ordinal scales based on weekly hours (1 = 1–2 h, 2 = 3–4 h, 3 = 5–6 h). Training volume was categorised into ordinal bands to reduce reporting inconsistency and improve response standardisation across youth participants and coaches, particularly given the variability in how weekly training exposure is recalled and reported in field-based settings. This approach was intended to improve response consistency rather than to represent precise load quantification. Competitive exposure variables included the number of competitions (count), multi-sport participation (1 = no, 2 = yes), total medals/titles (count), and competition level (1 = recreational to 6 = international).

Injury-related variables were coded as follows: injury frequency (count), injury recurrence (1 = no, 2 = yes), injury context (1 = non-badminton, 2 = badminton-related), medical treatment status (1 = no, 2 = yes), and training restriction (1 = no, 2 = yes). Injury activity type and injury mechanism were classified into categorical groups (activity type: 1–5; mechanism: 1–10). Perceived recovery status was measured on a 10-point scale (1 = poor recovery, 10 = full recovery), while performance impact was coded as a binary variable (1 = no impact, 2 = impact present).

Statistical analysis

All statistical analyses were performed on XLSTAT version 2023.5 (Addinsoft, Paris, France). Prior to analysis, the dataset was screened for implausible values and data entry inconsistencies to ensure data accuracy. Any suspected extreme values were cross-checked against the original field records before inclusion in the final dataset. Descriptive statistics were performed for all variables and described as mean ± standard deviation. The adequacy of the dataset for Principal Component Analysis (PCA) was evaluated using the Kaiser–Meyer–Olkin (KMO) test and Bartlett’s Test of Sphericity before performing any multivariate analyses. Subsequently, PCA with Varimax orthogonal rotation was performed to determine the core multidimensional structure of the dataset.

According to³¹, components with eigenvalues ≥ 1.0 were kept, while variables with factor loadings ≥ 0.70 were highlighted as key significant variables pertaining to the components extracted. Prior to PCA, to standardize and to make variables measured on different scales more comparable, all variables were organized to also use the z-score method. Following this process, these standardized component scores were summed to form the Youth Badminton Performance Index (YBPI), which illustrates the multidimensional performance profile of all athletes. Based on the outcome of the distribution of YBPI scores, athletes were divided into three categorical performance levels: Novice, Amateur, and Elite. Age was not included as an independent variable in the PCA to avoid redundancy with growth-related anthropometric variables, which inherently capture age-related variation across the entire developmental spectrum (U12–U16), rather than within individual age categories.

Result

Table 1 presents descriptive statistics for all anthropometric dimensions, physical fitness components, training exposure, and injury-related variables. The dataset consisted of 170 youth badminton players, which is sufficient for conducting multivariate analysis. Descriptive analysis results illustrate the dispersion of empirical, functional, structural, and training characteristics, demonstrating a diverse range of performance levels in the sample of young badminton players. Furthermore, descriptive statistics show a wide range of values for the major physical fitness components, VO₂max values, for example, ranged from 29.02 to 73.4 ml·kg⁻¹·min⁻¹ and the number of push-up repetitions performed by the players ranged between 3 and 66. These results show huge discrepancies in aerobic and upper-body endurance among youth badminton players.

Table 1.

summarizes the descriptive statistics of variables used for youth badminton performance profiling (n = 170).

Weight (kg)	Min	Max	Mean	Std. deviation
	22.9	64.9	44.41	8.35
Height (cm)	123.7	177.7	153.82	11.03
BMI	12.96	23	18.51	1.73
Shoulder Width (cm)	21	48	38.01	3.52
Upper Limb Length (cm)	34	79	48.91	6.06
Arm Span (cm)	120	181	153.15	12
Leg Length (cm)	62	105	83.96	8.17
Femur Length (cm)	30	59	44.42	4.88
Upper Arm Circumference (cm)	5	35	22.46	3.37
Forearm Circumference (cm)	4	29	21.02	2.74
Calf Circumference (cm)	23	43	32.11	2.89
Thigh Circumference (cm)	30	59	43.05	4.57
Sit Up (repetition)	27	64	42.93	6.61
Push Up (repetition)	3	66	33.76	9.1
Relative Power (w/kg)	13.86	117.18	45.55	11.67
V Sit And Reach (cm)	34	97	60.24	9.98
Stork Stand (Second)	31	65	59.29	3.1
Handgrip (kg)	8.3	48	26.56	6.99
Hand Wall Toss (number of catches)	3	40	26.59	4.71
Sideway Agility (Second)	8.2	16.51	11.31	1.17
Four Corner Agility (Second)	14.28	43.81	24.85	3.69
VO2Max (mL/(kg·min))	29.02	73.4	55.62	8.71
Technical/Tactical Training Volume (hours/week)	1	3	2.33	0.61
Physical Conditioning Volume (hours/week)	1	3	1.91	0.61
Match Play Volume (hours/week)	1	3	1.92	0.59
Number of Competitions (past 12 months)	1	49	13.59	10.64
Multi-Sport Participation (past 5 years)	1	2	1.13	0.37
Total Medals/Titles (count)	0	26	3.72	5.14
Competition Level (ordinal scale)	1	6	3.22	1.07
Injury Frequency (count)	0	3	1.14	0.52
Injury Recurrence (categorical)	1	2	1.06	0.24
Injury Context (badminton-related vs. non-badminton)	1	2	1.48	0.50
Injury Activity Type (categorical)	1	5	2.60	1.20
Injury Mechanism (categorical)	1	10	2.82	1.61
Medical Treatment Status (categorical)	1	2	1.47	0.50
Training Restriction (ordinal scale)	1	2	1.45	0.50
Perceived Recovery Status (1–10)	3	10	9.19	0.94
Performance Impact (ordinal scale)	1	2	1.06	0.26

As shown in Table 2, the Kaiser–Meyer–Olkin measure for insampling adequacy was good (KMO = 0.799), and Bartlett’s test of sphericity was statistically significant (p < 0.001), which indicates that the data set was suitable for conducting principal components analysis^31,32. Principal component analysis (PCA) serves to decipher the complexity of data by cutting down its dimensions and determining important metrics related to performance³³. Figure 1 shows the scree plot. At component 10, there is a visible inflection point. This shows that the first ten components constitute the bulk of variance from the data set. According to the Kaiser criterion (eigenvalues > 1.0), 10 components that cumulatively explained 76.43% of the variance were selected³¹.

Table 2.

Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s Test of Sphericity in principal component analysis.

Measure	Value
Kaiser–Meyer–Olkin (Overall)	0.799
Bartlett’s Test of Sphericity	χ²(df = 703), p < 0.001

Fig. 1.

Displays the scree plot, indicating an elbow point at the tenth component.

As shown in Table 3, Principal Component Analysis with Varimax rotation yielded the retention of ten components (D1–D10) with eigenvalues greater than 1.0, per the Kaiser criterion^31,32. Collectively, the components captured 76.43% of the variance, affirming a substantial and appropriate capture of the complex structure of the data. The eigenvalues were evenly distributed, showing a consistent drop across the components, and from this, we conclude that the retention of all components above the threshold of 1.0 was justified.

Table 3.

Ten components with eigenvalues bigger than 1.0 are preserved, accounting for 76.43% of the total variance.

Weight (kg)	D1	D2	D3	D4	D5	D6	D7	D8	D9	D10
	0.69	0.05	0.15	0.03	0.06	0.57	0.06	−0.08	0.02	0.10
Height (cm)	0.82	0.00	0.24	0.01	−0.01	0.25	0.01	0.08	−0.01	0.06
BMI	0.13	0.16	−0.03	0.05	0.06	0.83	−0.09	−0.14	0.08	0.03
Shoulder Width (cm)	0.59	0.02	−0.01	−0.08	0.04	0.48	0.26	−0.04	0.01	0.04
Upper Limb Length (cm)	0.78	0.11	−0.15	0.06	−0.03	0.30	0.06	0.00	0.07	−0.06
Arm Span (cm)	0.85	−0.02	0.25	0.01	0.09	0.25	0.00	−0.06	0.03	−0.02
Leg Length (cm)	0.86	0.00	0.17	−0.04	0.15	0.07	0.16	0.08	−0.06	0.16
Femur Length (cm)	0.76	−0.01	0.30	0.01	0.02	0.06	−0.05	0.09	0.09	−0.02
Upper Arm Circumference (cm)	0.17	0.10	−0.18	0.03	0.00	0.79	−0.02	0.17	−0.04	0.05
Forearm Circumference (cm)	0.31	−0.01	0.04	0.04	0.02	0.74	0.02	0.32	−0.09	0.03
Calf Circumference (cm)	0.38	−0.07	0.10	−0.11	−0.07	0.67	−0.07	0.08	−0.16	−0.07
Thigh Circumference (cm)	0.43	−0.09	0.11	0.11	0.04	0.75	−0.03	0.00	−0.01	0.06
Sit Up (repetition)	0.27	−0.01	0.21	0.05	−0.13	−0.03	0.66	0.21	−0.06	0.21
Push Up (repetition)	0.06	−0.01	0.18	0.02	0.20	−0.15	0.73	−0.14	0.12	0.02
Relative Power (w/kg)	0.53	0.10	0.49	−0.22	0.18	0.13	0.01	−0.02	0.03	0.06
V Sit And Reach (cm)	0.14	−0.02	0.36	0.14	0.20	0.30	0.34	−0.27	0.12	−0.06
Stork Stand (Second)	0.10	0.03	0.12	0.05	0.05	0.18	0.01	0.89	0.05	−0.01
Handgrip (kg)	0.70	−0.10	0.13	0.02	0.01	0.43	0.08	0.00	0.04	0.04
Hand Wall Toss (number of catches)	0.53	0.03	0.20	0.13	0.10	0.10	0.35	0.25	0.05	−0.05
Sideway Agility (Second)	−0.28	−0.01	−0.73	−0.13	−0.10	−0.01	−0.24	−0.23	−0.08	−0.15
Four Corner Agility (Second)	−0.30	−0.03	−0.70	−0.07	−0.07	−0.02	−0.32	−0.27	−0.07	0.01
VO2Max (mL/(kg·min))	0.30	−0.06	0.75	0.02	0.08	−0.07	0.10	−0.08	0.08	0.07
Technical/Tactical Training Volume (hours/week)	0.02	−0.04	−0.26	0.36	−0.13	−0.03	−0.16	0.02	0.44	0.52
Physical Conditioning Volume (hours/week)	0.06	0.06	0.00	0.80	−0.19	−0.03	0.10	−0.01	0.01	0.32
Match Play Volume (hours/week)	−0.03	0.04	0.11	0.88	0.02	0.13	0.01	0.06	−0.04	−0.18
Number of Competitions (past 12 months)	0.04	0.94	0.09	0.05	−0.08	0.04	0.04	0.02	−0.01	0.01
Multi-Sport Participation (past 5 years)	0.06	−0.09	0.17	−0.02	0.10	−0.15	0.16	0.05	0.80	−0.03
Total Medals/Titles (count)	0.01	0.99	0.00	0.01	−0.01	0.04	−0.01	0.00	−0.06	0.00
Competition Level (ordinal scale)	0.42	0.18	0.45	0.06	0.19	−0.22	−0.39	−0.08	−0.08	0.18
Injury Frequency (count)	0.10	−0.05	0.40	−0.19	−0.17	0.28	−0.20	−0.08	0.42	−0.19
Injury Recurrence (categorical)	0.05	−0.05	0.14	−0.31	0.71	0.10	0.11	0.02	0.27	0.09
Injury Context	0.01	0.99	0.00	0.01	0.01	0.01	−0.02	0.00	−0.02	0.00
Injury Activity Type (categorical)	0.00	0.99	−0.03	0.01	0.01	0.01	−0.03	0.01	−0.02	0.00
Injury Mechanism (categorical)	0.12	0.02	0.25	−0.02	0.09	0.15	0.19	−0.01	−0.12	0.80
Medical Treatment Status (categorical)	0.00	0.99	−0.02	0.01	0.00	0.02	−0.01	0.00	−0.02	0.01
Training Restriction (ordinal scale)	0.00	0.99	−0.02	0.01	0.00	0.01	−0.02	0.00	−0.03	0.00
Perceived Recovery Status (1–10)	0.02	−0.06	−0.17	−0.07	−0.75	−0.07	0.02	0.02	0.13	−0.05
Performance Impact (ordinal scale)	0.22	−0.10	−0.05	−0.05	0.74	−0.07	0.06	0.06	0.03	−0.07
Eigenvalue	9.54	6.00	3.54	2.24	1.67	1.37	1.32	1.23	1.09	1.04
Variability (%)	25.11	15.79	9.31	5.89	4.39	3.62	3.46	3.24	2.88	2.75
Cumulative %	25.11	40.90	50.21	56.10	60.48	64.10	67.56	70.80	73.68	76.43

Distinct principal components contributed differently to explained variance. The first four components (D1 - D4) showed comparatively higher eigenvalues (from 9.54 to 2.24) and together represented 56.10% of the overall variance, suggesting that these components were able to capture the prevailing patterns in the dataset. These components were mainly characterized by the variables of anthropometric structure, competitive exposure, aerobic–agility capacity, and training volume, respectively.

The remaining components (D5-D10) accounted for 2.75–4.39% of the variance individually, yet all of them passed the eigenvalue keeping criterion, and each one of them exhibited unique factor structures with obvious factor loadings (≥ 0.70). These components represent distinct domains including injury recurrence and recovery status, body composition and muscularity, muscular endurance, balance ability, cross-sport involvement, and injury mechanisms, although several moderate cross-loadings were observed, the dominant loadings remained clearly associated with their primary components, which confirms these dimensions as unique independent sources of variance and are not statistically redundant³³.

Accordingly, the ten-component solution reveals a particular order structure. Several variables demonstrated moderate cross-loadings across components. For example, body weight loaded at 0.69 on D1 and 0.57 on D6. In such cases, variables were retained within the component where the loading was conceptually more consistent with the majority of variables defining that component. In the case of body weight, retention under D1 was considered more appropriate due to its stronger conceptual association with overall body size and structural morphology, which aligned more closely with the dominant loading pattern of that component despite a secondary loading on D6. This approach follows recommended practices in exploratory multivariate analysis when secondary loadings remain lower than the primary loading³¹. There are a small number of components that explain the majority of the variance, while the rest contain a contextual/function specific detail pertaining to youth badminton performance. These components, and the separate, secondary dimensions they contain, maintain the construct integrity by embedding these performance-related dimensions within the more prominent components. This coheres to the best practice for multivariate analysis of complex performance data sets³¹.

Based on the extracted PCA factors, a composite Youth Badminton Performance Index (YBPI) was created. Table 4 shows the YBPI score distribution and the associated performance grouping. Based on performance, participants were descriptively grouped into three relative performance categories (Novice, Amateur, and Elite) based on the distribution of YBPI scores within the sample. These categories represent relative positioning within the study cohort rather than fixed performance standards or indicators of future potential. Out of the entire sample (n = 170), there were 25 Novice participants (YBPI score 0.00–39.060), 97 Amateur participants (39.061–63.804), and 48 Elite participants (63.805–100), making a total of 170. Out of the sample, 122 participants (71.76%) were classified as Novice or Amateur, with 48 (28.23%) being classified as Elite.

Table 4.

Performance grouping of youth badminton performance index (YBPI) status based on composite score distribution.

YBPI performance status	Freq	Cum. Freq.	%	Group range	Performance category
Training phase
0.00–39.060.00.060	25	25	14.71	−90.29 ≤ Novice < −26.49	Novice
39.061–63.804	97	122	57.16	−26.5 ≤  Amateur < 13.94	Amateur
63.805–100	48	170	28.23	Elite ≥ 73.06	Elite

Table 4 presents the grouping thresholds and the corresponding participant counts for each performance grouping. Since YBPI is a composite score, the YBPI score for each participant was calculated by summing his or her component scores for each of the principal components. The retained components were combined using equal weighting as part of an exploratory multidimensional profiling approach. This decision does not imply that each domain contributes identically to badminton performance in a physiological or practical sense. Rather, equal weighting was applied to preserve the multidimensional structure identified through Principal Component Analysis (PCA) while avoiding arbitrary or subjective prioritisation of selected performance domains in the absence of a validated theoretical weighting framework^31,33. In this context, the YBPI was designed to provide a balanced descriptive summary of athlete characteristics across multiple independent domains rather than to establish a hierarchically weighted model of competitive success.

Performance grouping into Novice, Amateur, and Elite categories was based on the empirical distribution of composite YBPI scores within the sample. Cut-off values were determined using a distribution-based segmentation approach, whereby thresholds were derived based on percentile-based distribution of YBPI scores within the sample, allowing for relative grouping across performance levels. The classification thresholds were derived from the empirical distribution of composite YBPI scores across the sample, allowing relative differentiation of performance levels within the study cohort^15,31. Accordingly, these performance grouping represent descriptive performance status at a single time point and are intended to support performance profiling and monitoring purposes only. The negative values observed in the group range are a statistical consequence of z-score standardization used during composite score construction. In this context, negative scores should not be interpreted as “negative performance,” but rather as values falling below the sample mean relative to the multidimensional distribution of the study cohort. Therefore, the group ranges reflect relative positioning within the sample rather than absolute performance deficits.

Discussion

This study aimed to determine and analyse key parameters related to performance and to create an index of multidimensional performance for youth players of badminton. The results confirmed that the performance of youth badminton players occurs in a number of interconnected areas. This confirms that the development of performance in youth athletes can only be captured through a multi-domain approach as opposed to a single domain^5,6. In the present study, performance refers to a multidimensional athlete profile derived from anthropometric characteristics, physical fitness attributes, training exposure, competitive experience, and injury-related functional indicators rather than competition ranking alone. Accordingly, the YBPI was developed to describe current multidimensional performance characteristics within the sample rather than to predict long-term competitive success.

Importantly, the extracted components illustrate varying degrees of performance characteristics with different actionable implications, which should be understood as not having equal importance vis-a-vis training or intervention implications. For example, components with structural and anthropometric variables of stature, body mass, limb length, segment length proportions, and arm span, reflect either non-modifiable characteristics, or those minimally modifiable. These characteristics are largely the result of biological growth and maturation processes^5,6. While these characteristics are biomechanically contextualized, and help understand movement efficiency, they should not be construed to be targets for training adaptations especially in populations where there are significant maturation-related differences and variabilities⁷.

Nevertheless, these characteristics remain relevant in performance profiling because they provide structural context that may influence reach advantage, movement efficiency, and stroke mechanics in badminton. Therefore, their inclusion in the index serves as contextual information rather than a direct target for training adaptation. Although biological maturation was not directly assessed, the YBPI was not derived solely from anthropometric variables, but from a multidimensional combination of physical fitness, training exposure, competition-related, and injury-related functional indicators. Therefore, while maturation-related structural advantages may have influenced selected components, the overall profiling framework should not be interpreted as a maturation-only model, but rather as a broader descriptive representation of multidimensional youth badminton performance. In contrast, components with physical fitness, training exposure, neuromuscular abilities, injuries and the function related to them, that are also present, reflect attributes that are modifiable, and can be addressed through systematic training, load management, and rehabilitation strategies^15,34. It is this distinction that aids the contextualized and appropriate interpretation of performance profiling outcomes in the framework of youth multidimensional development^4,35.

Although the PCA-derived components were equally weighted in the composite index, this should not be interpreted to mean that each domain holds equal physiological or practical importance in badminton performance. Instead, equal weighting was adopted as a pragmatic and conceptually neutral approach to preserve the multidimensional structure identified by PCA while avoiding arbitrary prioritisation of selected performance domains in the absence of a validated weighting framework^15,31. In this context, the YBPI was intended to function as a descriptive profiling tool that captures the relative contribution of multiple independent dimensions, rather than as a hierarchically weighted model of performance prediction or athlete ranking. Accordingly, the practical interpretation of YBPI scores should focus on overall multidimensional athlete profiling rather than if higher scores in any single domain necessarily represent a more important determinant of future performance.

Recognizing the predominance of structural and morphological dimensions underlines the importance of anthropometric and strength variables that differentiate youth badminton players. The growth and maturation processes result in significant alterations in the adolescent’s height, limb proportion and body weight. Such changes, which could offer possible mechanical benefits of reaching and moving more efficiently during play, are due to the growth and maturation processes. This finding corresponds with the literature in badminton and racket sports which explored the body structure-performance relationship in the growing phases of the players’ sporting life cycle^1,6.

While addressing the importance of this aspect/structural dimension, it should be regarded as more of a contextual factor associated with performance rather than a predictor of success due to the changes attributed to growth and maturation^7,36,37. However, this interpretation should be considered with caution, as the present study did not include direct indicators of biological maturation such as peak height velocity or maturity offset. Given that the participants were youth badminton athletes aged 9–16 years, this developmental period is characterised by substantial inter-individual variation in growth and maturation status. Consequently, anthropometric characteristics such as stature, limb length, and body mass may partly reflect maturational advancement rather than stable sport-specific performance advantages^6,38. This consideration is particularly relevant in youth badminton, where structural and physical attributes may influence movement efficiency, reach-related mechanics, and selected performance outcomes during adolescence³⁹.

Although the multidimensional structure of the present framework integrates anthropometric, physical fitness, training exposure, competition participation, and injury-related functional indicators, the absence of maturation data means that the potential influence of biological development cannot be entirely ruled out. Nevertheless, the PCA-derived structure incorporated multiple independent domains, including physical fitness, training exposure, competitive experience, and injury-related functional status, indicating that the performance profile was not based solely on structural characteristics. Therefore, the Youth Badminton Performance Index (YBPI) should be interpreted as a relative multidimensional performance profile within the sampled cohort rather than a maturity-adjusted or predictive model of athlete potential. This interpretation was further supported by an additional Pearson correlation analysis, which demonstrated that age was only weakly associated with YBPI (r = 0.195, p = 0.011). Although developmental progression may contribute to performance profiling, the observed relationship suggests that the composite score was not strongly driven by age alone, but instead reflected a broader integration of structural, physiological, experiential, and injury-related factors.

Furthermore, although performance grouping may be influenced by biological maturation, the PCA results demonstrated that performance-related variance was distributed across several domains rather than being dominated by anthropometric variables alone. This suggests that the multidimensional structure captured a broader range of performance-related attributes beyond structural characteristics. Although chronological age is closely associated with physical development and performance capacity in youth athletes, participants in this study were grouped within standard competitive age categories (U12, U14, U16), which partially controls for age-related variation. Nevertheless, variation in biological maturation may still exist within these categories and should be acknowledged as a limitation of the present study.

Competitive exposure was recognized as a distinct aspect of performance, indicating that the frequency of competitive involvement and the degree of competitive achievement are associated with the different elements of a performance profile. Regularly competing may help young people learn certain abilities, improve their tactical choices, and make the mental changes that are needed to meet the demands of the match, which are all important for improving their performance. This supports the ideas behind long-term athlete development models, which say that structured exposure to competitive situations is an important part of an athlete’s structured growth paths^4,40.

Another important dimension of performance was determined to be physical fitness and movement capabilities, specifically, attributes of aerobic capacity and agility. In the present dataset, aerobic capacity and agility were grouped within the same performance dimension, which may reflect the integrated physiological and movement demands of badminton. Badminton is characterised by repeated high-intensity rallies, rapid directional changes, explosive movements, and intermittent recovery periods, where aerobic support and agile movement capabilities operate concurrently to sustain performance throughout match play^1,41. Therefore, the clustering of these attributes within the same component likely reflects the functional interaction of these capacities during badminton play rather than suggesting that they represent identical physiological constructs. These findings are consistent with the well-documented physiological demands of badminton, which require repeated accelerations, rapid directional changes, and sustained rally performance. The role of aerobic and agility-related attributes is aligned with the literature that demonstrates these capacities contribute to the demands of competitive play and assist in sustaining performance throughout the duration of the competitive event^1,28,42.

The impact of training exposure was seen as a critical performance dimension, incorporating elements of applied technical-tactical training, physical conditioning, and match-play participation. The accumulated training stimuli during the youth developmental stage is critical in the development of physical, technical, and tactical capability. It reinforces the need for planned training and competitive participation at all levels, in order to achieve the comprehensive growth of youth badminton players. Developmentally, there are numerous colliding elements that stimulate improvement at the adolescent stage, such as the volume of training, growth and maturation, and stage-related performance specific exposure^6,35.

Even beyond the performance domains, the injury-related functional parameters were significant for performance differentiation. In particular, functional readiness and health status were examined in connection with injury recurrence, recovery status, and performance injury perception, as well as sustained performance. The injury burden and under-recovery is a well-known barrier to the training of young athletes⁴³. Principal Component Analysis (PCA) is one of the most common techniques for dimensionality reduction of complex data sets in sport science, and has been successfully used to reveal hidden performance factors among athletes. Previous studies have applied PCA in understanding relationships between certain anthropometric, physiological, and neuromuscular variables, thus assisting in athlete profiling and performance assessment^1,6. Other studies have highlighted the usefulness of PCA in identifying key determinants of the physical and performance-related attributes of individuals in various sports¹⁴. For this study, PCA was useful in consolidating the structural, physical fitness, training exposure, competition participation, and injury variables of youth badminton players into a multidimensional performance framework.

The Youth Badminton Performance Index (YBPI) is not intended to function as a talent identification or prognostic tool. Instead, it serves as a cross-sectional multidimensional profiling framework designed to describe current athlete characteristics across multiple performance domains. Its practical value lies in assisting coaches and sport scientists in identifying multidimensional strengths and weaknesses within an athlete’s present profile, thereby supporting more informed training, monitoring, and developmental decision-making.

For example, two athletes may achieve similar overall YBPI scores through different multidimensional performance profiles. One athlete may demonstrate stronger anthropometric and strength-related characteristics but lower competition exposure, whereas another may present better agility, aerobic capacity, and competitive experience despite less structural advantage. This illustrates that the YBPI is not intended to reward a single “ideal” athlete profile, but rather to summarise different combinations of performance-related characteristics across youth badminton players.

Accordingly, the performance groupings derived from the YBPI should be interpreted as relative descriptive categories within the sampled cohort rather than fixed indicators of athlete ranking, selection status, or long-term potential. This interpretation is consistent with youth athlete development literature, which emphasises that performance during adolescence is shaped by a combination of growth, biological maturation, training exposure, competition experience, and injury history, and therefore should not be reduced to deterministic or predictive judgments based on isolated indicators^5,15,35. Due to the cross-sectional composition of this study, there is no cause-and-effect reasoning to be drawn from the results, nor from the performance profiles, as they were captured at one point in time.

Additionally, injury data were self-reported and therefore susceptible to recall bias. Although the injury-related questionnaire demonstrated acceptable internal consistency (Cronbach’s α = 0.859), indicating reliable measurement across items, younger athletes may not accurately recall the timing, severity, or mechanism of previous injuries, which may have influenced the representation of injury-related variables within the PCA components. This limitation may have been more pronounced among younger participants (e.g., U12), who may possess lower recall precision and less complete injury awareness than older adolescent players in the U16 category. However, the structured format of the questionnaire and the inclusion of multiple injury-related indicators (e.g., recurrence, recovery status, and perceived impact) help to minimise random response variability and enhance the robustness of the injury-related components within the PCA structure. Furthermore, the multidimensional structure of the PCA may help reduce overreliance on any single variable or reporting source when interpreting the overall performance profile.

Technical–tactical performance was not directly evaluated in this study, as the research was specifically designed to examine underlying physical, anthropometric, and contextual dimensions of athlete profiling. While training exposure variables included elements of technical and tactical practice, no direct sport-specific skill assessments were conducted. In this context, monitoring should be understood as repeated descriptive profiling conducted across multiple time points rather than prediction of future performance outcomes. Future studies should investigate the YBPI as a descriptive profiling tool longitudinally at various different developmental stages, without predictive or prognostic outcomes, in order to determine its stability and sensitivity over varying time periods. Such studies would provide insight on the performance profile shifts associated with growth, maturation, training, injuries, and the cycle of all three.

Conclusion

This research investigates youth badminton performance using an integrated approach across multiple performance domains rather than using the traditional, segmented performance assessments. The findings yield a detailed, systematic, and integrated landscape description of youth badminton players’ functional performance related to anthrometrics, physical fitness, training, competition, and injury. This approach aligns with the current athlete development models, that emphasize holistic assessments during the adolescent years^6,30.

A novel aspect of this study is the integration of functional variables pertaining to injuries, perceived recovery, and performance impact within a single performance profile. These variables are typically separate in children’s sport research, yet collective impact on the training continuum and performance expression is often overlooked. Incorporating injury and physical training related components into performance profiles offers a more comprehensive insight into the functional health of the performance⁴³.

The Youth Badminton Performance Index (YBPI), developed from this framework, is intended to be used as a cross-sectional performance profiling instrument and should not be used for predictive analysis or talent identification. Rather than predicting future performance, the index focuses on contextualising and analysing multidimensional performance characteristics at a specific developmental stage. In this context, monitoring refers to repeated descriptive profiling conducted across different time points rather than prediction of future performance outcomes. While future research may examine the stability and sensitivity of this approach over time, the findings of the present study are limited to descriptive profiling and do not establish causal relationships or developmental pathways. Additionally, future studies should incorporate objective indicators of biological maturation, such as peak height velocity or maturity offset, to better distinguish structural developmental effects from performance-related attributes in youth badminton athletes^6,12.

Author contributions

Mohammad Firdaus Mohd Israj¹ contributed to the conceptualization of the study, data collection, statistical analysis, and preparation of the original manuscript draft. Nur Syazwani Ibrahim² assisted in data curation, visualization, and manuscript editing. Nor Ikhmar Madarsa³ contributed to the methodology design, validation, and critical review of the results. Siti Nor Alia Kartika Mohamed Shaihan⁴ contributed to the literature review support, resources, and coordination of data acquisition. Rozita Abdul Latif⁵ contributed to the review of the theoretical background, editing, and refinement of the manuscript structure. Mohd Syaiful Nizam Abu Hassan⁶ contributed to data interpretation, manuscript review, and academic refinement of the discussion and conclusion sections. Ahmad Bisyri Husin Musawi Maliki⁷ provided supervision, project administration, conceptual guidance, and final approval of the manuscript as the corresponding author. All authors have read and approved the final version of the manuscript.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.