Skip to Content
MDPIMDPI

Advanced

BiomechanicsBiomechanics
PDF
  • Article
  • Open Access

3 October 2026

15 Pages

Biomechanical Characteristics of the Jump Shot in Elite Handball: Transitioning from Kinematic Parameters to Expert Performance Models

,
,
and
1
Faculty of Health Studies, University of Split, 21000 Split, Croatia
2
Faculty of Kinesiology, University of Split, 21000 Split, Croatia
3
Department of Social Sciences and Humanities, University of Slavonski Brod, 35000 Slavonski Brod, Croatia
*
Author to whom correspondence should be addressed.
Biomechanics2026, 6(4), 91;https://doi.org/10.3390/biomechanics6040091
This article belongs to the Section Sports Biomechanics
Version Notes
Review Reports
Order Reprints

Abstract

Background: The jump shot is a critical technical element in modern handball, requiring the complex integration of speed, power, and precision. As the game evolves toward higher intensities, traditional coaching models require validation against objective biomechanical data. This study aimed to investigate the kinematic characteristics of the jump shot in elite players and establish a methodological framework for transitioning raw kinematic parameters into expert-based performance models. Methods: Kinematic analysis was conducted on five elite male handball players (n = 5, 15 total trials), focusing on the approach, flight, and release phases. Additionally, ten highly qualified handball experts (N = 10) independently established a theoretical performance model. To enable direct comparison between objective measurements and expert assessments, kinematic frequencies were linearly normalized onto a five-point ordinal continuum. The relationship between the kinematic model and the hypothetical expert model was determined using Spearman’s rank correlation coefficient. “Kinematic analysis was conducted on elite handball players, focusing on the approach, flight, and release phases. Results: Elite players predominantly utilize a two-step approach, with the first step slightly longer than the second to optimize the kinematic chain. Kinematic data and expert evaluations showed high congruency regarding body posture, highlighting the necessity of diagonally aligned shoulders and a forward-leaning trunk during the flight phase. A statistically significant correlation was observed between objective kinematic parameters and expert ratings, confirming the model’s validity. Conclusions: This study demonstrates a novel approach for evaluating technical performance by bridging the gap between biomechanical data and expert intuition. These findings provide a valuable exploratory foundation for automated technical evaluation systems and offer practical insights into contemporary technical trends for elite-level coaching in the rapid tempo of modern handball.

1. Introduction

Handball is a complex sport characterized by high technical, tactical, physical, and psychological demands [1,2]. Abilities such as speed, agility, endurance, and explosive power are essential, as they directly enable players to meet the sport’s rigorous physical requirements [3,4]. While handball shots can be executed in nearly ten different ways, two techniques predominate: the standing shot and the jump shot. Although standing shots generally produce higher ball velocity [4], the jump shot is the most frequent and effective scoring technique in modern handball [5,6,7,8,9]. By allowing players to release the ball at a higher point, it bypasses defenders and goalkeepers more effectively. Indeed, Wagner et al. (2010) reported that jump shots account for approximately 73–75% of all attempts in competitive matches, highlighting the need for comprehensive kinematic investigation [10].
Following established classifications [6,10,11,12,13], this study divides the jump shot into five functional phases: run-up, take-off, flight, throw, and landing. The run-up serves as the preparatory phase, where effectiveness depends on optimal step length and a controlled lowering of the center of mass [6,11]. Research indicates that a three-step approach significantly increases shooting velocity compared to a one-step approach [14]. “The take-off phase involves a progressive transfer of force with strong activation of the ankle, knee, and hip extensors [11,15]. During this phase, the free leg swings forward and laterally at approximately 90°, contributing to take-off efficiency [11,16]. Expert assessment in handball involves the systematic evaluation of performance by specialists such as coaches and kinesiologists [11,16,17].” In the flight phase, players adjust their body position in response to situational demands [10,11]. The throw phase is then initiated by trunk rotation and forward shoulder movement, ending at ball release [18]. High performance in this phase relies on coordinated proximal-to-distal segmental sequencing [19]. Finally, the landing phase is critical for impact absorption and injury prevention [8,11].
Expert assessment in handball involves the systematic evaluation of performance by specialists such as coaches and kinesiologists [5,11,16,17]. This qualitative approach captures nuances that objective measurements alone might miss. While recent research has integrated artificial intelligence to enhance data analysis and reliability [20], this study takes an innovative path by directly integrating 3D kinematic analysis with expert modeling.
Consequently, the aim of this study is to quantify the kinematic parameters of the jump shot in elite handball players and compare these findings with a hypothetical model of optimal performance defined by international experts.

2. Materials and Methods

2.1. Participants

2.1.1. Elite Players (Kinematic Sample)

The sample consisted of five elite male players from the Croatian national handball team, including individuals decorated as multiple-time world’s best players. To ensure a highly homogenous and world-class elite sample, strict selection criteria were established. The inclusion criteria required participants to: (1) be active, long-term members of the senior national handball team at the time of testing; (2) possess a minimum of 10 years of professional playing experience; (3) have actively participated in major international tournaments within the current competitive season; and (4) be verified multi-medalists across all three major global competitions: the Olympic Games, the World Championships, and the European Championships. Conversely, the exclusion criteria strictly omitted players if they: (1) had not been completely cleared by medical staff for full, unrestricted competitive training and match-play for at least six consecutive months prior to the testing session following any musculoskeletal injury of the upper or lower limbs; (2) presented with any subjective chronic pain or functional deficit that restricted maximum-effort jumping or throwing; or (3) failed to complete the full standardized warm-up protocol. Participant characteristics were: body mass 100.00 ± 8.00 kg; body height 196.00 ± 4.64 cm; age 30.6 ± 3.91 years; and professional experience 12.4 ± 4.10 years (four back players and one wing player). To ensure data stability and capture technical consistency, each participant performed three successful maximum-effort jump shots, yielding a total pool of 15 kinematic trials for subsequent analysis. All participants provided written informed consent, and the study was conducted in accordance with the Declaration of Helsinki.

2.1.2. Handball Experts (Evaluative Sample)

Ten highly qualified experts were recruited based on extensive professional and scientific experience. Inclusion criteria required: (1) a Master’s or PhD degree in handball; (2) experience as lecturers in academic or international coaching programs; and (3) at least 10 years of active coaching with documented competitive success. To ensure a strict blinding protocol, these experts operated entirely independently of the kinematic data collection phase. They did not view the video recordings or trials of the five elite players. Instead, they were tasked with establishing a general, theoretical performance model based purely on elite technical standards and coaching taxonomy.

2.2. Experimental Design and Procedures

2.2.1. Testing Protocol and Standardization

To ensure consistency, all players were tested under standardized conditions at 10:00 a.m. in a sports hall with a wooden parquet floor and an ambient temperature of 22 °C. Testing was conducted at the start of the post-season. Testing was conducted at the start of the post-season. To mitigate any residual physical or neuromuscular competitive fatigue, all testing sessions were scheduled after a mandatory 48-h complete recovery window following the final match of the competitive season, ensuring that the participants were sufficiently rested and capable of producing maximum-effort technical trials. According to foundational motor learning paradigms, long-term athletic practice deepens the consolidation of autonomous motor programs, resulting in high movement stereotypy even in the presence of physical fatigue [21,22]. Biomechanical and longitudinal tracking in elite handball confirms that technical execution patterns and kinematic stability reach their highest consolidation toward the end of the competitive cycle [23]. To mitigate physical fatigue, all testing sessions were scheduled after a mandatory 48-h complete recovery window following the final match, ensuring that the participants were sufficiently rested and capable of producing maximum-effort technical trials.
Before the jump shot execution, participants completed a standardized 14-min warm-up protocol consisting of: 3 min of light running; 2 min of core stability exercises; 2 min of dynamic stretching; 2 min of neural activation; 5 min of handball-specific drills (passing and shooting).
Following the warm-up, players performed the jump shot aimed at an empty goal. They were instructed to execute the movement at match-speed while focusing on their natural, technically correct in-game technique. Each participant performed three trials with a 30-s rest interval between attempts to prevent fatigue and ensure technical consistency.

2.2.2. 3D Kinematic Data Collection

Kinematic analysis was performed using the GAIT–LaBACS software system (v1.0, Split, Croatia), developed at the Faculty of Electrical Engineering, Mechanical Engineering, and Naval Architecture in Split [24]. Twelve symmetrical anatomical reference points were defined for facial tracking and digitization: the lateral malleolus, the patella (utilized as a superficial landmark for estimating the knee joint center), the iliac crest, the acromion, the olecranon, and the distal ends of the ulna and radius (serving as the wrist joint landmark for tracking terminal segment velocity). To ensure biomechanical accuracy, the functional 3D knee joint center of rotation was not equated directly with the patella; instead, it was mathematically projected within the GAIT–LaBACS (v1.0) software system. Following the system’s validated framework [24], the software calculated the instantaneous 3D geometric center of the knee joint rather than a polycentric functional center of rotation. This was achieved by applying a predictive anthropometric offset model based on standard multi-segment biomechanical conventions. Specifically, a posteriorly directed translation vector was applied perpendicular to the frontal plane from the digitized superficial patella landmark. The depth of this intra-articular projection was calculated as half of the participant’s physical knee width (Wknee/2) obtained during baseline screening, effectively embedding the coordinate origin mid-way between the femoral condyles to establish a stable local coordinate system. This mathematical displacement vector follows the universally validated 3D anthropometric regression frameworks established for multi-segment lower-limb tracking [25,26] ensuring that the geometric center aligns precisely with the intercondylar midpoint within the local joint coordinate system.
Data recording took place in a sports hall following the calibration of the measurement system. Two camera systems were utilized to ensure continuous visibility of all reference points: BASLER-402-FC cameras (Basler AG, Ahrensburg, Germany) operating at 100 frames per second and a PANASONIC VW-D5100 (Panasonic Corporation, Kadoma, Japan), camera operating at 50 frames per second. The 100 fps cameras served as the primary tracking systems for high-velocity segments, while the 50 fps camera functioned as an auxiliary system to prevent marker occlusion. Synchronization was achieved a posteriori by utilizing an external electronic LED strobe flash trigger positioned within the overlapping fields of view. The synchronization frame was defined as the exact video frame corresponding to the instantaneous onset of the LED illumination. To eliminate temporal mismatch errors prior to 3D DLT reconstruction, the 2D coordinates from the 50 fps camera were upsampled to 100 Hz directly within the GAIT-LaBACS software environment utilizing its native cubic spline interpolation function. This mathematical alignment effectively compensated for the temporal offset (maximum hardware variance of 20 ms), reducing the residual temporal alignment error to under 10 ms. Given that the auxiliary camera primarily supported tracking during stable flight phases, the interpolation process did not significantly affect the peak coordinate accuracy or degrade the resolution of ball-release kinetics.
The spatial coordinates were reconstructed using the Direct Linear Transformation (DLT) method. Three-dimensional calibration was achieved using eight static markers: four positioned vertically on a metal rod of known height and four placed on the floor in a proportional square arrangement. The calibrated control volume was precisely set to encompass the entire phase sequence of the jump shot, measuring approximately 4.5 m (length times) 2.5 m (width times) 3.8 m (height). Video recordings were digitized frame by frame, and the DLT method was used to compute the spatial coordinates for every time frame of the analyzed movement. The accuracy of the 3D reconstruction framework was verified by calculating the Root Mean Square (RMS) errors of the calibration markers. The mean reconstruction errors were 4.2 mm for the X-axis (mediolateral), 4.6 mm for the Y-axis (anteroposterior), and 6.8 mm for the Z-axis (vertical). These dimensions and sub-centimeter error margins confirm high spatial accuracy and reproducibility for tracking elite athletic performance. Each jump shot was recorded three times, and recording quality was evaluated immediately after testing to allow for any necessary corrections or repetitions. All three successful trials for each participant were fully digitized and analyzed, subsequently undergoing individual-level aggregation within the statistical phase to control for intra-subject dependency.

2.3. Variables

To establish a rigorous analytical structure, the variables investigated in this study were divided into two hierarchical categories based on their role in answering the primary research question.
(1)
Main Outcome Variables (Core Performance Model): These comprise the observed frequencies (f(i)) of the 30 qualitative technical execution variants across the five functional phases (approach patterns, take-off foot mechanics, trunk-shoulder alignment, and landing strategies). These parameters represent the core outcome variables, as they were directly normalized via Equation 1 and subjected to Spearman’s rank correlation analysis against the expert evaluation model to validate the performance framework.
(2)
Supplementary Descriptive Variables: These consist of the 25 continuous objective kinematic parameters (including step lengths, segment velocities, joint angles, and center of gravity displacements). These parameters serve an auxiliary role, providing comprehensive descriptive context regarding the physical profile and performance boundaries of the elite sample, without directly entering the rank correlation matrix.
A detailed integrated taxonomy of both variable sets across the functional phases of the jump shot is provided in Table 1.
Table 1. Methodological framework of objective kinematic parameters and qualitative expert evaluation variants across functional jump shot phases.
To ensure spatial and mechanical reproducibility, specific kinematic variables were operationally defined relative to fixed baselines. Lower leg incline (LLI) was computed as the angular orientation of the longitudinal tibia axis relative to the horizontal global coordinate axis (the floor line), where 90° denotes a vertical position. Lowest center of gravity displacement (LCG) was defined as the maximum vertical downward displacement of the center of gravity, measured as the difference between the subject’s baseline static standing COG and the minimum COG height reached during the take-off amortization phase. Horizontal kinetic energy of the center of mass KEhoriz at the onset of take-off was derived using the classical equation: KE horiz = 1 2 ×   m   ×  v2 where m is the participant’s whole-body mass (kg) and v is the instantaneous horizontal run-up speed (m/s). This parameter isolates the horizontal translational energy of the center of mass, omitting vertical or segmental rotational energies. The large variance reported in KEhoriz results directly from the quadratic dependence on velocity within the function, amplified by the inclusion of both high-velocity backcourt profiles (velocities > 7 m/s) and a position-specific low-velocity wing profile (velocity < 3 m/s) within the elite sample. The throwing phase was temporally bounded between the instant of maximum shoulder external rotation and the exact frame of ball release. Angular variables were calculated using specific segment vectors: Trunk Angle (TA) utilized the iliac crest-to-acromion vector relative to the vertical global axis; Shoulder Angle (SA) evaluated the humerus vector relative to the trunk; and Lower Leg Incline (LLI) utilized the malleolus-to-patella vector relative to the floor line. Reported continuous parameters represent discrete metrics captured at critical mechanical events: take-off parameters at maximum amortization, flight configurations at peak flight height, and throwing dynamics at the exact instant of ball release. All upper limb angular and velocity metrics were evaluated strictly on the dominant (throwing) side of the participants. The wrist landmark was operationally defined as the mathematical midpoint calculated between the digitized distal ends of the ulna and radius. To account for the multi-planar nature of the movement, all angular variables were computed as true 3D spatial angles between segment vectors within the global three-dimensional coordinate system, rather than being projected onto a single 2D anatomical plane. Hand velocity was quantified via the distal wrist marker, operationally defined as Wrist Speed (WS).

2.4. Statistical Analysis

Descriptive statistics and the Kolmogorov–Smirnov test for normality were applied. Inter-rater reliability was assessed via Cronbach’s alpha. To allow direct comparison between kinematic data and expert models, frequencies were normalized to an ordinal scale using the formula
NO = 1 + f ( i ) × 4 max f
where NO is the normalized value, f(i) the observed frequency, and max f the maximum frequency. This specific scaling function is based on the standard Min-Max normalization framework, adjusted to linearly map continuous frequency distributions onto a bounded 1–5 ordinal range [27]. The choice of this method is justified as it preserves the exact proportional variances of the raw kinematic data while transforming them into a format directly comparable to the 5-point expert evaluation scale, which is a prerequisite for the Spearman’s rank correlation analysis. To account for potential intra-subject correlation and prevent the nesting of data from inflating the statistical power, trials were not treated as completely independent data points. Instead, individual-level technical kinematic variants were condensed using a categorical mode-aggregation algorithm across the three trials performed by each player before entering the final group-level distribution.
f i = ∑ j = 1 5 δ V j , i
where δ(Vj, i) = 1 if Vj = i, and 0 otherwise. This reduction workflow strictly enforces that the final statistical degrees of freedom are grounded on the (N = 5) unique participants rather than the raw pool of 15 observations. Descriptive statistics and the Kolmogorov–Smirnov test for normality were subsequently applied to these aggregated participant profiles, ensuring that intra-subject dependency did not bias the distributional assumptions or the subsequent Spearman’s rank correlation analysis.

2.5. Declaration of Generative AI in the Writing Process

In the preparation of this paper, the authors used ChatGPT (OpenAI, GPT-4o model, accessed on 25 June 2026) to assist in data visualization and technical formatting. Specifically, GenAI was utilized to generate Python code within the Google Colab (Google LLC, Mountain View, CA, USA, accessed on 25 June 2026) environment to create graphical representations and to structure and refine data tables. Additionally, the tool was used to improve the linguistic clarity and phrasing of the manuscript. It is important to emphasize that the AI tool was used solely for technical and linguistic execution; it was not involved in the collection of primary biomechanical data, the formulation of research hypotheses, or the interpretation of the final scientific results. The authors have critically reviewed all outputs and take full responsibility for the content of the final manuscript.

3. Results

3.1. Kinematic Parameters and Normality Testing

Descriptive statistics and the results of the Kolmogorov–Smirnov (K-S) test for all RUN-UP, TAKE-OFF, FLIGHT, THROW, and LANDING kinematic variables are presented in Table 2. The majority of continuous variables followed a normal distribution (p > 0.20). The third step length (3SL) was excluded from the K-S normality testing procedure because it exhibited an absolute variance of zero. This is attributed to the fact that elite players in this sample utilized either a two-step approach (TSOPP, performed consistently by the backcourt players) or a single-step approach (SSOPP, performed by the wing player), resulting in an average step number of 1.80 ± 0.41 and leaving the third step completely unexecuted (0.00 ± 0.00 cm) across all analyzed trials.
Table 2. Descriptive statistics and normality testing (Kolmogorov–Smirnov) for kinematic parameters of the jump shot (n = 5 players, 15 trials).
The analysis reveals that the first RUN-UP step (131.54 ± 33.42 cm) is slightly longer than the second. Elite players achieved an average run-up speed of 6.48 pm 2.06 m/s, which contributed to a mean horizontal kinetic energy KEhoriz of 2315.21 pm 1189.29 J prior to take-off. Elite players achieved an average flight height (FH) of 46.26 ± 7.32 cm and a ball release height (TH) of 354.99 ± 7.36 cm. The throwing time (TT) was characterized by a brief duration of 0.08 ± 0.05 s, demonstrating high-velocity movement execution during the THROW phase.

3.2. Expert Model Reliability

The reliability of the ten handball experts is detailed in Table 3 The Cronbach’s alpha coefficient of 0.92 and an inter-item correlation of 0.60 demonstrate excellent internal consistency and a high degree of consensus among evaluators, confirming the validity of the expert-defined model.
Table 3. Descriptive statistics of expert assessments and inter-rater reliability (n = 10 experts).

3.3. Integration of Kinematic Data and Expert Evaluation

Table 4 displays the frequencies of technical variants observed via kinematic analysis, their corresponding normalized scores, and the experts’ qualitative ratings.
Table 4. Comparison of kinematic frequencies, normalized scores, and expert evaluation ratings for jump shot technical variants.
In the RUN-UP phase, the most frequent execution was the two-step approach ending on the leg opposite to the throwing arm (TSOPP, 12 frequencies), which achieved the maximum normalized score (5.00). This alignment suggests that elite players prioritize rapid execution over the traditional three-step approach. In the TAKE-OFF phase, the heel-to-toe roll on the opposite leg (HTROPP) was dominant (14 frequencies), showing strong agreement between elite performance (5.00) and expert preference (4.60).
During the FLIGHT phase, all recorded attempts (15 frequencies) featured a forward-leaning trunk with diagonal shoulders (TLFWDDIAG). Although this achieved a maximum normalized score (5.00), experts indicated a slightly higher preference for an upright trunk alignment (TRUPDIAG = 4.50). This represents a minor “Gap*”, where experts value the biomechanical advantage of an upright posture for vision and power, while players in practice maintain a forward lean to preserve horizontal momentum gained during the explosive two-step approach.
In the THROW phase, the elbow lead technique with the elbow above shoulder level (ELBABVLEAD) was the most common (12 frequencies), aligning with the high expert rating (4.10). The measured upper arm and trunk angle (UATA = 13.72 ± 7.43°) confirms this positioning. Furthermore, the extremely short throwing time (TT = 0.08 ± 0.05 s) and wrist speed (FS = 1.87 ± 0.74 m/s) reflect high-velocity movement execution, indicating significant explosive power during the THROW phase. This suggests that elite players focus on “whip-like” shoulder-to-fist acceleration, which is essential for bypassing modern, aggressive goalkeepers. Finally, the LANDING phase revealed a significant discrepancy. While players frequently utilized a parallel two-foot stance (LDBOTHPAR, 7 frequencies), experts rated this variant poorly (2.40). This Gap* suggests that while experts prefer a single-leg landing (LDTO) to facilitate a faster transition back to defensive positioning, elite players often default to a parallel landing for increased stability and injury prevention (ACL protection) after high-velocity shots.
Both kinematic analysis and expert evaluation emphasize that a diagonal shoulder orientation and a forward-leaning trunk provide the most favorable biomechanical conditions for modern shot execution. The high correlation between objective kinematic data and subjective expert ratings validates the proposed model as a benchmark for elite handball performance.

3.4. Correlation Between Models

A Spearman’s rank correlation was conducted to assess the relationship between the normalized kinematic frequencies and expert evaluation scores. A statistically significant correlation was found (r = 0.67; p < 0.001), confirming that the kinematic performance of elite players is highly consistent with the expert-defined model of technical excellence.

4. Discussion

This study aimed to investigate the kinematic parameters of the jump shot in elite handball players and compare them with a hypothetical model of optimal performance defined by experts. The results yielded several key findings: elite players predominantly use a two-step approach (TSOPP); the first step of the run-up is slightly longer than the second; and both kinematic analysis and expert evaluation emphasize a diagonal shoulder position with a forward-leaning trunk (TLFWDDIAG) during the flight phase. Furthermore, a statistically significant correlation was found between the objective kinematic data and subjective expert ratings, validating the proposed model.
The jump shot remains one of the most researched technical skills in handball, traditionally focused on the three-step approach [19,26]. In this study, the first step was found to be slightly longer than the second, which aligns with findings by Šibila [8]. Conversely, Burger (2016) observed a longer second step in first-division players [8,11]. These discrepancies likely stem from the qualitative level of the sample. Crucially, the final take-off step must be rhythmic and synchronized with the lowering of the body’s center of mass to initiate an effective kinematic chain. Previous research suggests that optimizing horizontal velocity in the final step is vital for maximizing ball release velocity [10,13].
Interestingly, all players in this study utilized a two-step approach (TSOPP, total length: 260.74 ± 56.61 cm). While older studies favored the three-step approach as most effective [6,8,28], those findings often reflect a slower era of handball. Modern elite handball is characterized by increased speed and aggressive, deep defensive movements, leaving players with minimal time and space. This trend is supported by Toquet et al. (2021), who noted that over a quarter of goals are now scored from one-step shots [29]. Our results reflect this evolution, where elite players must initiate their approach rapidly with fewer steps to bypass modern defenses.
Regarding the FLIGHT phase, the recorded flight height (FH: 46.26 ± 7.32 cm) was slightly lower than in previous studies [9,11], likely due to the laboratory setting. The upper arm and trunk angle (UATA) averaged 13.72°, indicating the elbow was positioned slightly above the shoulder at release (ELBABVLEAD), which is consistent with established literature [12,13]. This positioning, combined with the explosive strength of the shoulder girdle, is fundamental for shot velocity.
To bridge the gap between raw kinematic data and expert assessments, we transformed frequencies into a normalized scale (1–5). For instance, the most frequent technique—rolling from the heel to the toes of the take-off foot (HTROPP) resulted in a normalized value of 5.00, closely matching the expert value of 4.60. This confirms that elite players instinctively adopt biomechanically optimal patterns [21]. However, a discrepancy (Gap*) was observed in the LANDING phase; while players frequently utilized a parallel two-foot stance (LDBOTHPAR, 7 frequencies), experts rated this variant poorly (2.40), likely favoring a single-leg landing for faster defensive transition.
The analysis of body posture further highlighted the synchronization between kinematics and expertise. Our results suggest that elite players are systematically trained to adopt a diagonal shoulder alignment (TLFWDDIAG). This configuration allows for an uninterrupted kinematic chain while protecting the ball from defenders. Both the expert evaluations and the normalized kinematic parameters confirm that a diagonal shoulder orientation and an upright (TRUPDIAG) or slightly forward-leaning trunk (TLFWDDIAG) provide the most favorable biomechanical conditions for effective shot execution [6,10,11].

Study Limitations

Despite the significant findings and the novel methodological approach, several limitations must be acknowledged. First, the research was conducted in a controlled laboratory environment rather than under real match conditions. Although this allowed for high precision kinematic measurements, the absence of defensive pressure and the psychological stakes of a competitive game may have influenced performance intensity. This is particularly reflected in the slightly lower flight height observed, suggesting that elite athletes may not reach their maximum physical output without situational motivation.
Second, the sample size, while consisting of high-level elite players, is relatively specific (N = 5) participants, 15 trials). While the results provide a clear insight into the technique of top-tier players, they may not be fully generalizable to younger age categories or lower competitive levels where technical and physical requirements differ. Although analyzing 3 trials per player increased the total dataset to 15 jump shots—thereby stabilizing the observed kinematic frequencies—the limited number of individual participants inherently reduces the statistical power to detect broader demographic variances. Future studies should attempt to expand this cohort, though recruiting Olympic-level athletes remains a structural challenge in elite sports science. Finally, a structural limitation involves the technological framework utilized for data collection. The kinematic recording was conducted using video-based, frame-by-frame digitization and Direct Linear Transformation (DLT), which represents an older technological standard compared to modern, automated markerless tracking or multi-camera infrared systems. This temporal gap is due to the long-term retrospective nature of tracking and validating data from this highly specific cohort of Olympic and World Championship medalists. Access to such elite athletes is exceptionally rare, making the data structurally unique. While the hardware and manual digitization reflect the standards of the data collection period, the mathematical precision was tightly controlled via high-order cubic spline interpolation and rigorous calibration, ensuring that the biomechanical integrity of the performance model remains highly accurate and valuable for modern comparative analysis. Furthermore, although the integration of expert evaluations with kinematic data represents a methodological strength, such assessments inherently contain a degree of subjective judgment. While we mitigated this through the normalization process and the Spearman’s rank correlation coefficient, the “hypothetical model” remains based on a specific expert consensus. This is evident in certain observed “gaps,” such as in the landing phase, where expert preferences for defensive transition may conflict with players’ prioritized stability and injury prevention.

5. Conclusions

In conclusion, this study provides a comprehensive kinematic analysis of the jump shot in elite handball, introducing a novel methodological framework that bridges the gap between objective biomechanical data and qualitative expert assessments. By utilizing a normalized ordinal scale and the Spearman’s rank correlation, this research offers a unique approach to evaluating technical execution that has seldom been applied in team sports.
The findings highlight the current technical execution patterns of the jump shot in world-class elite handball players, demonstrating how live execution styles adapt to the high-intensity demands of the modern game. The shift toward a two-step approach (TSOPP), characterized by a longer first step and a high-velocity run-up, reflects the necessity for rapid execution under defensive pressure. Furthermore, the results confirm that diagonal shoulder alignment and a forward leaning trunk during the flight phase are biomechanical imperatives that optimize both ball protection and the efficiency of the kinematic chain.
Ultimately, the integration of these kinematic criteria with expert defined models provides a robust empirical foundation for future developments in sports technology. The methodological procedures established here serve as a precursor to computer-based, automated systems capable of delivering objective, real-time evaluations of technical performance. Such tools will be invaluable for coaches in optimizing player development and refining the execution of complex technical elements in high-performance handball.

Author Contributions

Conceptualization, A.B. and H.K.; methodology, A.B.; software, I.V.; validation, A.B., H.K. and K.Š.; formal analysis, I.V.; investigation, A.B.; resources, H.K.; data curation, H.K.; writing—original draft preparation, A.B.; writing—review and editing, H.K.; visualization, K.Š.; supervision, H.K.; project administration, I.V.; funding acquisition, A.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki. All participants were informed of the minimal potential risks and participated voluntarily. According to the regulations of the Faculty of Kinesiology and Faculty of Health studies at the University of Split, formal ethical approval was CLASS:029-03/26-08/01, URBROJ: 2181-228-103/1-6, approved 3 February 2026. The study involved healthy adult participants.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors on request.

Acknowledgments

The authors thank all subjects for their participation in this study. Special thanks are extended to the Croatian Handball Federation for facilitating the participation of elite handball players during their preparations for the Olympic Games. During the preparation of this work, the authors used ChatGPT (OpenAI, GPT-4o model, accessed on 25 June 2026) to assist in data visualization, technical formatting, and language editing. Specifically, the tool was utilized to generate Python code within the Google Colab environment for creating graphical representations and structuring data tables, as well as to improve the linguistic clarity of the manuscript. The authors reviewed and edited the output as needed and take full responsibility for the content of the publication.

Conflicts of Interest

The authors declare no conflict of interest.

Abbreviations

The following abbreviations are used in this manuscript:
STNRStep number
1SL1st step length
2SL2nd step length
3SL3rd step length
SLTTotal length
SLAAverage length
RUSRun-up speed
KEKinetic energy
RUTRun-up time
LLILower leg incline
LCGLowest center of gravity
TOTTake-off time
FHFlight height
BPLBody plane length
SMBHSpeed to max height
TATrunk angle
SAShoulder angle
FTFlight time
THThrow height
UATAUpper arm & trunk angle
WSWrist speed
UAASUpper arm angle
TTThrow time
DBMBody movement distance
LTLanding time
SSOPPSingle-step/Opposite leg
SSSAMESingle-step/Same leg
TSOPPTwo-step/Opposite
TSSAMETwo-step/Same
THSOPPThree-step/Opposite
THSSAMEThree-step/Same
TOOPPToes/Opposite leg
TOSAMEToes/Same leg
HTROPPHeel-to-toe/Opposite
HTRSAMEHeel-to-toe/Same
BTTOESTwo-foot/Toes
BTROLLTwo-foot/Roll
TLFWPARTrunk: Forward/Parallel
TLFWDIAGTrunk: Forward/Diagonal
TRUPPARTrunk: Upright/Parallel
TRUPDIAGTrunk: Upright/Diagonal
TLBWPARTrunk: Backward/Parallel
TLBWDIAGTrunk: Backward/Diagonal
ELBABVLEADElbow: Above shoulder/Lead
ELBABVNOElbow: Above/No lead
ELBSHLLEADElbow: At shoulder/Lead
ELBSHLNOElbow: At shoulder/No lead
ELBBLWLEADElbow: Below/Lead
ELBBLWNOElbow: Below/No lead
LDTOLanding: Take-off leg
LDSWLanding: Swing leg
LDBOTHPARLanding: Both legs/Parallel
LDDIAGTOFLanding: Diagonal/Take-off forward
LDDIAGSWFLanding: Diagonal/Swing forward
LDBOTHWPARLanding: Both legs/Wide

References

  1. Rogulj, N. Differences in situation-related indicators of the handball game in relation to the achieved competitive results of teams at 1999 World Championship in Egypt. Kinesiology 2000, 32, 63–74. [Google Scholar]
  2. Wagner, H.; Finkenzeller, T.; Würth, S.; Von Duvillard, S.P. Individual and team performance in team-handball: A review. J. Sports Sci. Med. 2014, 13, 808. [Google Scholar] [PubMed]
  3. Chaouachi, A.; Brughelli, M.; Levin, G.; Boudhina, N.B.B.; Cronin, J.; Chamari, K. Anthropometric, physiological and performance characteristics of elite team-handball players. J. Sports Sci. 2009, 27, 151–157. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  4. Karcher, C.; Buchheit, M. On-court demands of elite handball, with special reference to playing positions. Sports Med. 2014, 44, 797–814. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  5. Almeida, G.P.L.; Silveira, P.F.; Rosseto, N.P.; Barbosa, G.; Ejnisman, B.; Cohen, M. Glenohumeral range of motion in handball players with and without throwing-related shoulder pain. J. Shoulder Elb. Surg. 2013, 22, 602–607. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  6. Pori, P.; Bon, M.; Šibila, M. Jump Shot Performance in Team Handball-A Kinematic Model Evaluated on the Basis of Expert Modelling. Kinesiology 2005, 37, 40. [Google Scholar]
  7. Póvoas, S.C.; Seabra, A.F.; Ascensão, A.A.; Magalhães, J.; Soares, J.M.; Rebelo, A.N. Physical and physiological demands of elite team handball. J. Strength Cond. Res. 2012, 26, 3365–3375. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  8. Šibila, M.; Pori, P.; Bon, M. Basic kinematic differences between two types of jump shot techniques in handball. Univ. Palacki. Olomuc. Gymnica 2003, 33, 19–26. [Google Scholar]
  9. Yamada, E.; Aida, H.; Nakagawa, A. Notational analysis of shooting play in the middle area by world-class players and Japanese elite players in women’s handball. Int. J. Sport Health Sci. 2011, 9, 15–25. [Google Scholar] [CrossRef] [Scilit][Green Version]
  10. Wagner, H.; Buchecker, M.; von Duvillard, S.P.; Müller, E. Kinematic description of elite vs. low level players in team-handball jump throw. J. Sports Sci. Med. 2010, 9, 15. [Google Scholar] [PubMed]
  11. Burger, A. Kinematičko i Ekspertno Modeliranje Odabranih Elemenata Tehnike u Rukometu; Faculty of Kinesiology, University of Split: Split, Croatia, 2016. [Google Scholar]
  12. Van den Tillaar, R.; Ettema, G. A three-dimensional analysis of overarm throwing in experienced handball players. J. Appl. Biomech. 2007, 23, 12–19. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  13. Wagner, H.; Pfusterschmied, J.; Von Duvillard, S.P.; Müller, E. Skill-dependent proximal-to-distal sequence in team-handball throwing. J. Sports Sci. 2012, 30, 21–29. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  14. Belcic, I.; Ocic, M.; Dukaric, V.; Knjaz, D.; Zoretic, D. Effects of One-Step and Three-Step Run-Up on Kinematic Parameters and the Efficiency of Jump Shot in Handball. Appl. Sci. 2023, 13, 3811. [Google Scholar] [CrossRef] [Scilit]
  15. Zvonarek, N.; Vuleta, D.; Hraski, Z. Kinematic Analysis of Two Different Jump Shot Techniques in Handball. In Proceedings of the 1st International Scientific Conference “Kineziologija-Sadasnjost I Buducnost”, Dubrovnik, Croatia, 25–28 September 1997; pp. 180–182. [Google Scholar]
  16. Rogulj, N.; Foretić, N.; Srhoj, V.; Čavala, M.; Papić, V. Utjecaj nekih motoričkih sposobnosti na brzinu lopte kod udarca u rukometu. Acta Kinesiol. 2007, 1, 71–75. [Google Scholar]
  17. Burger, A. Ekspertna Procjena za Varku Jednostruke Promjene Smjera Kretanja u Rukometu. In Proceedings of the 32 Ljetna Škola Kineziologa Republike Hrvatske, Zadar, Croatia, 26–29 June 2024; pp. 564–567. [Google Scholar]
  18. Fradet, L.; Botcazou, M.; Durocher, C.; Cretual, A.; Multon, F.; Prioux, J.; Delamarche, P. Do handball throws always exhibit a proximal-to-distal segmental sequence? J. Sports Sci. 2004, 22, 439–447. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  19. Belcic, I.; Rodić, S.; Dukarić, V.; Rupčić, T.; Knjaz, D. Do blood lactate levels affect the kinematic patterns of jump shots in handball? Int. J. Environ. Res. Public Health 2021, 18, 10809. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  20. Marquina, M.; Lozano, D.; García-Sánchez, C.; Sánchez-López, S.; de la Rubia, A. Development and validation of an observational game analysis tool with artificial intelligence for handball: Handball. ai. Sensors 2023, 23, 6714. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  21. Fitts, P.M.; Posner, M.I. Human Performance; Brooks/Cole Publishing Company: Pacific Grove, CA, USA, 1967. [Google Scholar]
  22. Schmidt, R.A. A schema theory of discrete motor skill learning. Psychol. Rev. 1975, 82, 225. [Google Scholar] [CrossRef] [Scilit]
  23. Marques, M.C.; van den Tillaar, R.; Vescovi, J.D.; González-Badillo, J.J. Relationship between throwing velocity, muscle power, and bar velocity during bench press in elite handball players. Int. J. Sports Physiol. Perform. 2007, 2, 414–422. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  24. Blažević, Z.; Cecić, M.; Grujić-Šupuk, T.; Papić, V.; Pavelin, A.; Zanchi, I.; Zanchi, V. Dislocated Measurements of Human Gait and Central Data Processing. In Proceedings of the SoftCOM 2001, Dubrovnik, Croatia, 9–12 October 2001. [Google Scholar]
  25. Vaughan, C.L.; Davis, B.L.; O’Connor, J.C. Dynamics of Human Gait; Human Kinetics Publishers: Champaign, IL, USA, 1992. [Google Scholar]
  26. Aloui, G.; Hermassi, S.; Hammami, M.; Gaamouri, N.; Bouhafs, E.G.; Comfort, P.; Shephard, R.J.; Schwesig, R.; Chelly, M.S. Effects of an 8-week in-season upper limb elastic band training programme on the peak power, strength, and throwing velocity of junior handball players. Sportverletz. Sportschaden 2019, 33, 133–141. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  27. Dizdar, D. Kineziometrija [Kinesiometrics]; Faculty of Kinesiology, University of Zagreb: Zagreb, Croatia, 2006. [Google Scholar]
  28. Davis, R.B.; Õunpuu, S.; Tyburski, D.; Gage, J.R. A gait analysis data collection and reduction technique. Hum. Mov. Sci. 1991, 10, 575–587. [Google Scholar] [CrossRef] [Scilit]
  29. Tuquet, J.; Cartón, A.; Marco-Contreras, L.A.; Mainer-Pardos, E.; Lozano, D. Analysis of the steps cycle in the action of throwing in competition in men’s elite handball. Sustainability 2022, 14, 5291. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
Identifying Asymmetrical Gait Using Neural Network with Marker Coordinates Around the Pelvis and Implication of AI Application in Motion Analysis
Previous
Spinal Kinematics During Wheelchair Fencing Lunges Under Natural and Coach-Guided Conditions: An Exploratory IMU-Based Study
Next