Artificial turf systems require sustainable infill materials capable of maintaining adequate functional performance. The aim of this study was to develop recycled polyethylene-based infill materials for artificial turf systems and evaluate the relationship between their thermal behaviour and system-level performance. Recycled polyethylene streams from greenhouse and mulching films were processed through melt compounding using compatibilizers, elastomeric copolymers, mineral fillers, and foaming agents to obtain sustainable infill materials. The resulting blends were characterized by Fourier Transform Infrared Spectroscopy (FTIR), Differential Scanning Calorimetry (DSC), Thermogravimetric Analysis (TGA), and Melt Flow Rate (MFR) analyses. Laboratory-scale artificial turf systems were evaluated according to FIFA reference procedures, including shock absorption, vertical deformation, ball rebound, and rotational resistance. The results indicated that differences in crystalline structure and chain mobility were accompanied by differences in system-level mechanical behaviour. AF200 blends exhibited lower crystallinity and greater flexibility, achieving values within FIFA reference ranges, whereas AF2000 blends showed higher rigidity and lower overall performance. These findings support the potential of recycled polyethylene as a sustainable alternative to conventional End-of-Life Tires (ELT)-derived rubber infill materials.
This field-based case series developed and applied a low-cost, high-speed two-dimensional video workflow to analyse the barrier-step action in elite futsal goalkeepers. Three male professional goalkeepers with international experience (age 28.3 ± 3.2 years; height 182.4 ± 4.1 cm; body mass 78.6 ± 5.3 kg) completed three valid dominant-side trials each in a standardised indoor task (nine trials). Video was recorded at 240 frames·s −1 and analysed frame by frame in Kinovea. The action was operationally segmented into initiation-propulsion, lateral transfer, and terminal support-stabilisation. Projected ankle, knee, hip, and shoulder angles were extracted at the end of initiation-propulsion and at first terminal-support foot contact. Temporal organisation was goalkeeper-specific, particularly during initiation-propulsion. Between-goalkeeper coefficients of variation, calculated from the three goalkeeper mean values, decreased from 36.1% to 4.1% for ankle angle, 40.1% to 3.2% for knee angle, and 15.2% to 4.6% for hip angle, whereas shoulder dispersion remained high. Within this case series, the workflow distinguished individual preparatory organisation from reduced lower-limb angular dispersion at the blocking instant. The findings support the feasibility of accessible video-based sports engineering workflows when operational definitions and interpretive limits are explicit.
This study aimed to develop and validate a supervised machine learning algorithm based on a Random Forest (RF) model to automatically detect collision events in elite rugby union players using microelectromechanical sensor (MEMS) data. Inertial data from triaxial accelerometers and gyroscopes were collected from 36 professional players across four competitive matches. A total of 2,100 collision events were labelled through synchronised video analysis and categorised as tackles, rucks, mauls, or scrums. Data were segmented using 2-s sliding window, and time- and frequency-domain features were extracted. Three matches were used for training and one for testing. Model performance was evaluated using precision, recall, and F1-score. The RF classifier achieved a precision of 88.4%, a recall of 87.2%, and an F1-score of 87.8%, demonstrating robust discrimination between collision and non-collision events. Performance remained consistent across player roles and contact types, indicating strong external validity. These results validate the integration of MEMS-derived data with supervised RF learning as a reliable framework for automated impact detection in professional rugby. This approach offers a scalable tool for workload quantification, and tactical performance assessment, with potential integration for broader performance analysis systems.
Many lay equestrian texts describe the use of curb bits in horses, but understanding of their mechanism of action is limited with respect to the complex interaction between the curb bit, the horse and the rider. Forces within the rein-bridle system may be generated by the rider and or by motion or movement of the horse. Understanding the interaction between the curb mouthpiece, the reins, the cheekpiece and the curb chain is an important step towards understanding the interaction with the horse. This study aimed to quantify the forces on the reins, the bridle cheekpieces to the horse’s occiput and the curb chain. A high-level dressage horse was ridden in walk, trot and canter, and in transitions between gaits. A double bridle was equipped with load cells in the left curb rein, left cheekpiece, and centre of the curb chain and synchronised with video (100 Hz). In walk, trot and canter, curb rein and curb chain tensions were of similar magnitude, and of lower magnitude than curb cheekpiece tensions. Peaks in curb cheekpiece tension (>10 N) frequently occurred in the absence of increases in rein tension, indicating that the horse can increase cheekpiece tension in the absence of rider-created rein tension. Using the principles of classical static mechanical equilibrium, intra-oral forces were estimated using the tension data for the curb chain, cheekpiece and rein when in canter. These findings show that it is possible for increased cheekpiece tensions to be created by horse interaction with the bit in the absence of rider rein forces.
This study introduces DataGoal, an open-source MATLAB-based toolbox designed to process, analyze, and visualize positional data in soccer obtained from Global Positioning System (GPS/GNSS), Local Positioning System (LPS), and video-tracking technologies. The increasing availability of tracking data has created new opportunities for investigating the spatial-temporal dynamics of team sports; however, the complexity of these datasets often limits their effective use. DataGoal addresses this challenge by providing an integrated analytical framework that combines data importation, spatial calibration, preprocessing, metric computation, and graphical visualization within a single workflow. The toolbox implements a comprehensive set of linear and non-linear metrics to quantify individual and collective behaviors, including physical demands, team compactness, spatial organization, and dynamic coordination patterns. Through an interactive graphical user interface, DataGoal enables users to configure analyses and process positional datasets without extensive programming requirements, while its modular MATLAB-based architecture allows users to extend or customize analytical routines. By integrating multiple analytical procedures within a unified environment, DataGoal facilitates reproducible positional-data analyses and supports investigations of collective dynamics in soccer. The framework is suitable for applications in scientific research, performance analysis, and sports science education. As an open-source project, DataGoal is intended to evolve through community contributions and future extensions, contributing to the advancement of data-driven approaches in soccer performance analysis.
Novel technologies that mitigate head impact severity can contribute to reducing TBI risk in helmeted activities. This study evaluated the efficacy of prototype equipment that couples a helmet and shoulder pads during head impacts representative of American football. A custom test fixture accelerated the head, neck, torso, and pelvis of an anthropomorphic test device (ATD) into a second, stationary ATD so that head and neck loading could be assessed for both players involved in an impact. Three equipment configurations were tested, each at 6.7 m/s and 4.4 m/s: (i) both ATDs wore a standard helmet and shoulder pads, (ii) the accelerated ATD wore the prototype while the stationary ATD wore a standard helmet and shoulder pads, and (iii) both ATDs wore the prototype. The prototype yielded significant reductions in peak linear accelerations of the head for both ATDs at 6.7 m/s, but only one ATD at 4.4 m/s. Reductions in angular head kinematics were inconsistent across the different equipment configurations and test conditions. Significant reductions in upper neck loading (both force and moment) were reduced in both ATDs as a result of wearing the coupled equipment in all but one scenario.
Outdoor archery, especially in long-range disciplines such as Bhutanese archery, is influenced by environmental factors, including air density, which affects arrow trajectory through aerodynamic drag. Although the effects of air density on projectile motion have been studied in ballistics, they have not been empirically examined in archery. In this study, we conducted online surveys with Bhutanese players and developed iterative arrow-trajectory models for four equipment setups to investigate how diurnal changes in air density affect arrow range. To validate these models, we conducted an experiment using a shooting machine alongside real-time weather data to quantify how air-density variations affected arrow range at a 145 m Bhutanese archery field in Thimphu, Bhutan. Archers' subjective estimates of air-density effects on arrow range averaged 0.69 m, while the iterative models predicted range changes of 0.11-0.25 m for a 0.05 kg/m3 air-density shift across the examined equipment setups. The range change observed during the shooting-machine trial (0.19 m) closely matched the predicted range change from the corresponding equipment setup in the iterative model (0.17 m). Overall, the findings suggest that archers can enhance accuracy by adjusting their aim according to expected air-density fluctuations throughout the day, with temperature serving as a practical on-field proxy for air-density changes. The study offers a framework for integrating air-density effects into archery performance, reducing reliance on mid-play trial-and-error.
Prior exposure to spring surfaces may shape the biomechanical interactions between the lower limbs and joints during movement. This study investigates how surface-specific training influences joint stiffness and mechanical energy expenditure (MEE) during hopping tasks. Thirty male athletes with relevant training backgrounds participated in controlled trials on both elastic and rigid surfaces, conducted before and after a structured familiarization period. Kinematic and kinetic analyses were employed to evaluate alterations in leg and joint stiffness, alongside changes in MEE. The results indicate that while leg stiffness remained consistent overall, surface-specific training prompted distinct joint-level adaptations. Notably, knee stiffness increased, and hip stiffness decreased across both surface conditions post-training, suggesting a redistribution of mechanical load. Additionally, ankle MEE rose following spring-surface training, signifying heightened energy demands, whereas hip MEE declined, pointing to improved energy efficiency. These outcomes emphasize the importance of familiarity in refining neuromechanical strategies for surface adaptation and highlight the dominant roles of the ankle and knee in regulating stiffness. The findings also underscore the critical need to consider prior exposure when evaluating the biomechanical effects of surface properties on lower limb function.
Rowing blades (oars) transfer power from athletes to the water to propel a boat forward. The complex nature of fluid forces around a blade has only recently started to be understood with previous studies only investigating blades that are commercially available. The full motion of a blade spoon through water is complex, with motion both longitudinal and perpendicular to the boat motion. This multidisciplinary research uses a combination of computational models with experimental analysis to evaluate how spoon shapes affect the fluid forces on a blade during rowing. The computational model used 3D computational fluid dynamics (CFD) to evaluate the drag and lift forces around four different spoon shapes as they rotate through moving water. Each design had the same surface area, with different positions of maximum depth along their length. Experimental analyses used telemetry to measure blade forces during a series of rowing runs, for two different Concept2 blades, over different stroke rates and blade gearing. The results showed there was a correlation of R2 = 0.78 between the location of the spoon's maximum depth and the lift force generated at the catch. The study also found that, a resultant fluid force change of 47.3 N, at the spoon, can promote a -10.7 degrees +/- 1.7 degrees shift in the maximum force angle in the real world. The shift in maximum force angle highlighted how spoon design, especially regarding position of the maximum depth, affected the profile of the rowing stroke.
Understanding team performance in professional football increasingly benefits from network science, which models players as nodes and their interactions as functional links. While Social Network Analysis (SNA) provides valuable structural insights, it often fails to capture the probabilistic nature of possession. This exploratory study introduces an integrated methodological framework that combines static SNA with a Markov-spectral model of ball circulation to analyse the Portuguese National Team across two high-stakes 2025 Nations League finals matches. This approach moves beyond describing who is connected to quantifying how possession flows through the team network. Match event data were obtained via the Wyscout (R) platform, with all passing actions by Portugal extracted. The framework combined SNA (using uPATO (R)) with Markov-spectral analysis (implemented in MATLAB (R)) to compute indices of passing uncertainty, diffusion speed, navigability, and network robustness. The results suggested that (i) SNA identified key players as crucial hubs for ball circulation at the micro level, while indicating a cohesive team structure with adaptable macro-level properties; and (ii) Markov-spectral quantification showed descriptive between-match variations, with the second match displaying higher Entropy (2.84 vs 2.77 bits/pass), suggestive of greater unpredictability, and a larger Spectral Gap (0.53 vs 0.45), indicative of faster potential diffusion of possession. Overall, this integrated approach demonstrates the feasibility of profiling team coordination through both structural configuration and stochastic flow properties. The Markov-spectral framework complements traditional SNA by providing quantifiable indices related to passing variability, network navigability, and structural cohesion, offering a multi-layered, proof-of-concept toolkit for analysing collective performance.
The force perpendicular to a limb is often measured for muscle strength assessment. However, muscles act in three-dimensional (3D) space, exerting a 3D force and moment. The validity of hip adductor and extensor strength measured from the perpendicular force versus 3D forces were compared. Ten participants performed hip adductor and extensor strength tests on two visits, with 5-7 days between visits. Limb orientations were determined from 3D motion capture and reaction forces exerted against the thigh using a multi-axis force platform. 3D hip net joint moments (NJM) were calculated using an inverse statics rigid body model. Hip adductor and extensor strength were also calculated using a planar assumption from the force vector perpendicular to the thigh's long axis, with and without gravity correction. Planar hip adductor and extensor strength without gravity correction were greater than with gravity correction and from the 3D NJM (p < 0.001; Cohen's d = 0.84-1.62). Gravity corrected planar hip adductor strength was not different than the hip adductor NJM (p = 0.032; d = 0.04), but gravity corrected planar hip extensor strength was less than the hip extensor NJM (p = 0.002; d = 0.10). This effect size difference was small; however, it indicates that measurement error may be present when ignoring the non-perpendicular forces during strength testing. The presence of 3D forces, and their influence on muscle strength measurement should be considered for different muscles, joints, and limb orientations.
This study presents a data-driven exploratory analysis that combines experimental measurements with statistical modelling to investigate how the mechanical properties of a football, specifically stiffness and energy loss, influence a football's impact behaviour. Experimental testing of twelve FIFA-certified footballs, impacted at low (5.93 +/- 0.06 m s-1) and high (19.2 +/- 0.4 m s-1) velocities, revealed three distinct characterisations of impact behaviour. These groupings were primarily defined by contact time and deformation, variables that were strongly correlated with the stiffness properties of the balls, measured in both static and dynamic environments. Multivariable models were developed using inputs of impact conditions and ball properties. For the selected outcome variables (contact time and deformation) and under the conditions tested, these inputs could explain up to 96% of the differences observed in impact behaviour in the best-performing model. The models also demonstrated agreement with observed behaviour in an independent dataset, in this instance errors of 2.8% (corresponding to 0.3 ms) were observed for contact time and errors of 3.4%-9.7% (corresponding to 1-5.5 mm) were observed for deformation. Properties of the footballs measured at the low velocity also demonstrated reasonable consistency in their relationship with dynamic impact behaviours at the higher velocity with errors below 7%. This study demonstrates the use of statistical modelling to classify and explain aspects of football impact behaviour using a small set of measurable properties, within the conditions investigated.
This study compares minimalist and comprehensive inertial measurement unit (IMU) configurations for monitoring rider posture in equestrian sport. The minimalist setup employed three IMUs positioned at the pelvis, sternum, and saddle, while the comprehensive configuration used 18 sensors distributed across the rider and saddle. Data collected during real riding conditions were processed in MATLAB to evaluate posture angles, symmetry measures, and center of mass (CoM) trajectories. The minimalist system achieved sub-degree accuracy for pelvis and trunk orientations, exhibited no substantial differences in key posture metrics, and required approximately 40% less setup time than the comprehensive system. However, CoM estimation showed greater variability, with an average RMS deviation of 0.185 m compared to the full-body configuration. In addition, the reduced sensor configuration demonstrated improved practicality for field deployment due to lower system complexity and reduced preparation requirements. These results indicate that minimalist IMU systems can provide accurate and field-deployable posture monitoring, while comprehensive configurations remain preferable when high-precision whole-body biomechanical measures, such as CoM estimation, are required.
In basketball, slip-related injuries pose a significant risk, often caused by reduced friction between the court surfaces and shoes due to outsole wear. While numerous studies have investigated the grip characteristics of basketball shoes on various surfaces and under different conditions, the impact of outsole wear on friction has not been thoroughly examined. This study evaluates the frictional performance of 14 basketball shoe outsoles under progressively worn conditions using a belt sander across three common playing surfaces: polyurethane, synthetic, and wood. The available coefficient of friction (ACOF) was measured under both dry and wet conditions using a portable slip-testing device. Results indicate a progressive decline in ACOF with increasing outsole wear, regardless of the surface or environmental condition. A higher ACOF was observed in outsoles with herringbone treads, especially in wet conditions. Shoes with deeper treads and wider grooves exhibited consistent friction across all testing scenarios. Additionally, the outsoles demonstrated uniform friction on wooden surfaces in dry conditions. These results are expected to provide valuable insights into basketball shoe design and selection, and to help establish suitable replacements that optimize performance and safety.
The purpose of this study is to develop an optimized prediction model for Champions League qualification using machine learning and SHAP methods. The research data comprises 14 match statistics from 300 teams over 15 seasons, from the 2009-2010 season to the 2023-2024 season of the English Premier League. The prediction models were evaluated by combining four algorithms with six data balancing techniques, resulting in the evaluation of 24 models. Additionally, the hyperparameters of the top-performing models were tuned to evaluate the optimized models' performance. Finally, SHAP methodology was applied to assess the importance of variables in the optimized models. The results of this study are as follows. First, the performance of the SMOTE XGBoost model was the highest, followed by the SMOTE-Tomek Gradient Boosting Trees model. Second, the most important variable in the SMOTE XGBoost model was the Shots OT pg, followed by Six Yard Box. These results can contribute not only to predicting team performance in advance but also to improving operational efficiency and can serve as foundational data for teams aiming for long-term success.
The purpose of this study was to analyze scoring outcomes according to manufacturer differences in official Taekwondo Protector and Scoring Systems (PSS). To this end, match data were collected from 77 women's bouts at the 2024 Paris Olympic Games using the D-company PSS and 361 women's bouts at the 2025 World Taekwondo Championships using the K-company PSS. Due to differences in weight class structures between the two competitions, all data were integrated and analyzed based on the Olympic weight class criteria. The results showed that, across all weight classes, the K-company PSS exhibited a higher proportion of trunk scoring, whereas the D-company PSS demonstrated a relatively higher proportion of head scoring. For both manufacturers, roundhouse kicks were identified as the primary scoring technique; however, statistically significant differences in scoring outcomes were observed between manufacturers even when technically comparable actions and match situations were applied. These differences were particularly pronounced in the lightweight divisions, where a wider range of techniques and competitive situations contributed to scoring, while in the middleweight and heavyweight divisions, differences were more limited and primarily associated with high-scoring techniques.
This paper proposes a real-time framework to estimate the energy expenditure (EE) in gym exercises by combining the information provided by IoT wearables and machine learning (ML). The classical techniques, like indirect calorimetry are both expensive and immobile whereas only the heart rate model mostly fails to work in dynamic situations. To overcome these drawbacks, we are going to rely on a multimodal dataset featuring a combination of physiological, personal, and environmental measurements and captured with a Zephyr BioHarness 3 device. We learn and train supervised models of learning, such as recurrent (LSTM, CNN-LSTM) and two new hybrids: an LSTM-LightGBM ensemble and a CNN-LSTM-LightGBM cascade. The proposed system identifies gym activities, and at the same time, the estimated values of EE are accurate and real time, which is a viable alternative to the lab-based methods. We have shown that our experimental assessment provides a high level of predictive power with the Random Forest model, which gives us an R2 = 0.94 and RMSE = 0.39 MET, and with the LSTMLightGBM ensemble, which gives us R2 = 0.80 and RMSE = 0.71 MET.Key contributions include a subject-independent training methodology for robust generalization and a feedback mechanism for monitoring athletic performance. Experimental results validate the precision and responsiveness of the framework, demonstrating its potential in both sports science and clinical healthcare. By integrating IoT sensing with ML, this work advances adaptive data-driven approaches to optimize training and recovery.
The understanding of association football head collision mechanics is of interest in the context of increased neurodegenerative disease risk in ex-professional players. This study sought to measure and characterise micromechanical pressure wave propagation from football head collisions; a distinct energy transfer mechanism from, and occurring before the start of, macromechanical gross head kinematics. A surrogate head setup, instrumented with a hydrophone pressure sensor, was exposed to footballs travelling at match realistic velocities. Pressure waves emanating from the region of first contact were measured within the cranial cavity. Their mean peak-to-peak magnitudes were up to 31.0 kPa. Across a range of 20 footballs of different materials, constructions and assembly methods, all of which satisfied the Laws of the Game, a 9.1-fold difference in peak-to-peak pressure and 54.7-fold difference in energy transfer, was observed. A 5-fold increase in peak-to-peak pressure was observed for a 77% increase in ball inbound velocity. Hydrophilic leather ball types demonstrated up to 4.2-fold greater peak-to-peak pressures when tested in wet, compared to dry conditions. These data highlight that pressure wave energy transfer may be influenced by the materials and construction of the ball, with scope to affect this without requiring alteration to the game or its Laws. The influence of these pressure waves on acute and long-term brain health requires further exploration.
Accurate and reliable performance monitoring is crucial in elite team sports, where wearable technologies, such as GPS systems, are widely utilized. However, the impact of Bluetooth transmission distance between wearable devices and receiver units on measurement quality has received limited attention. This study aimed to evaluate the reliability and validity of running distance data collected using the Polar Team Pro GPS at three different Bluetooth connection distances-20 m, 50 m, and 100 m-during a standardized 20-meter shuttle run test in elite youth football players (n = 20). Participants were randomly assigned to perform the test at each of the specified distances. Reliability was assessed using intraclass correlation coefficients (ICC), coefficient of variation (CV%), standard error of measurement (SEM), smallest worthwhile change (SWC), and minimal detectable change (MDC). Validity was evaluated through Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Square Error (RMSE), and Bland-Altman plots. Results showed that data reliability and validity declined substantially as the Bluetooth distance increased, with the highest ICC (0.73) and lowest measurement error occurring at a distance of 20 m. These findings underscore the importance of maintaining close receiver proximity to ensure the accuracy of GPS-derived data in both research and applied athlete monitoring settings.
As one of the most widely played and watched contact sports in the United States, football presents a significant risk of blunt chest trauma. Thus, effective chest protection is essential to prevent rare but fatal injuries like commotio cordis. Despite the risk of fatal outcomes, current test methods for certifying chest protective equipment fail to account for injury metrics relevant to commotio cordis and neglect body-to-body collisions, common in contact sports such as football. This study presents the development of a method that simulates shoulder-to-chest impacts via a pneumatic linear impactor, and the evaluation of football chest protectors, with the goal of informing cardiac safety. This statistically verified method considers two impact locations vulnerable to commotio cordis at minimum and moderate football tackle testing speeds to evaluate chest forces, rib deflections, and viscous criterion values of seven different football chest protector combinations on a Hybrid III Upper Torso Assembly. Additionally, this study investigates the quasi-static behavior of the chest to evaluate the predictive capability of quasi-static tests for the dynamic performance of these protector combinations. Researchers show that current football chest protectors fail to reduce viscous criterion values below the established 25% risk threshold for commotio cordis risk at moderate football tackle testing speeds. Additionally, researchers observed a moderate trend between quasi-static stiffness and dynamic viscous criterion ( R 2 = 0.580, p = 0.0468, slope = 0.059, 95% CI [0.02, 0.963]). Ultimately, a statistically verified test method was developed that differentiates between protector combinations, locations, and speeds, to inform certification standards for football chest protective equipment.
