INTRODUCTION

High-intensity locomotor actions (HILAs) such as high-speed running (HSR), sprinting, accelerations, and decelerations, typically defined using absolute or relative velocity or acceleration thresholds, are central to performance in adult male professional soccer [1, 2]. These actions have become increasingly frequent and intense over recent decades [3–5], with clear links to match-defining moments [6, 7] and team outcomes [8]. Forecasts suggest that HSR distances (> 25.1 km · h⁻¹) in the English Premier League may increase by up to 40% between 2020 and 2030 [9], alongside rises in sprinting and high-intensity accelerative and decelerative efforts [8]. These trends reflect the increasing speed and tactical complexity of the modern game, which vary substantially by position. For example, wide players typically accumulate greater HSR, whereas forwards perform more repeated high-intensity accelerations and decelerations [5, 7, 10, 11]. Such positional heterogeneity challenges uniform threshold application and cross-study comparability, as fixed cut-offs may misrepresent the true intensity demands of specific roles.

There is a strong link between HILAs, injury risk, and performance load management [12–15], although evidence regarding sprintingrelated hamstring injury risk remains mixed. Some studies identify maximal sprinting as a key precipitating factor, whereas others emphasise the multifactorial nature of risk, highlighting eccentric strength, previous injury, and exposure history [16–19]. High-intensity decelerations generate ground reaction forces up to six times body weight, approximately twice those during maximal sprinting and elicit greater creatine kinase and soreness responses than steadystate running [18, 20, 21]. These eccentric demands have been associated with anterior cruciate ligament (ACL) injuries [22], and both over- and under-exposure may elevate injury risk [23]. Accelerations similarly impose substantial strain during transitional and pressing actions [24, 25] and are metabolically demanding [26]. Accordingly, accurate monitoring and profiling of HILAs across positions and contexts are essential for optimising training design, managing player load, and mitigating injury risk [27, 28].

A range of technologies are used to monitor HILAs. Global Positioning Systems (GPS) remain most common, offering portable and relatively low-cost solutions for quantifying external load and identifying peak exertion periods [29]. A sampling frequency of ≥ 10 Hz is recommended for reliable distance and speed measurement; however, accurately capturing accelerations and decelerations remains problematic. Rapid velocity changes may occur within time intervals shorter than the device’s sampling window, producing signal smoothing and underestimation of peaks. Moreover, acceleration and deceleration metrics are often derived from tri-axial accelerometers integrated within GPS units rather than solely from positional differentiation, which may introduce additional noise and limit sensitivity to brief, high-magnitude movements [30–34]. Local Positioning Systems (LPS) provide higher spatial and temporal resolution in environments where GPS accuracy may be compromised, such as stadiums or indoor facilities [35], and demonstrate superior validity for quantifying rapid speed changes during intense accelerations and decelerations [36, 37]. Video-based optical tracking systems (OTS), including Second Spectrum, TRACAB, and STATS SportVU, are widely used for tactical and physical analysis [38, 39]. While generally reliable for distance-based metrics, OTS often overestimate sprint demands compared with GPS [40, 41]; HSR and sprint distances can be 12–18% higher despite strong correlations (r² > 0.99) [42]. However, no standardised agreement currently exists between GPS, LPS, and OTS technologies, each using distinct algorithms, filters, and smoothing parameters. Consequently, data cannot be used interchangeably across devices or competitions [43–45].

Threshold selection can be conceptualised along a continuum from fixed to fully individualised approaches, each with distinct implications for interpreting HILAs. Early studies primarily employed absolute thresholds (e.g., > 19.8 km · h⁻¹ for HSR, > 25.2 km · h⁻¹ for sprinting, > 3 m · s⁻² for accelerations, < −3 m · s⁻² for decelerations), which facilitated between-study standardisation but failed to account for inter-individual differences in physical capacity [21, 45–48]. Consequently, absolute thresholds may underestimate demands on faster players and overestimate those on slower players, obscuring meaningful workload differences across individuals and positions. Relative thresholds, defined as percentages of a player’s maximum sprint speed (MSS) or test-derived velocities (e.g., VIFT), improve individualisation and contextual accuracy but require regular testing and lack universal reference values, limiting comparability [24, 47, 48]. More recently, adaptive thresholding models that dynamically adjust based on rolling match or training data have emerged. These models aim to balance individual sensitivity with longitudinal consistency, offering a pragmatic bridge between absolute and relative approaches, though adoption remains limited [4, 14, 47, 49, 50].

Beyond threshold selection, temporal and contextual framing plays a crucial role in accurately capturing HILAs [49, 51]. Traditional analyses using whole-match averages underestimate physical intensity by including stoppages, substitutions, and other low-activity periods [52, 53]. In contrast, ball-in-play (BiP) analyses isolate active phases, providing a more accurate representation of true physiological and mechanical load and allowing clearer identification of peak and fluctuating demands. By excluding inactive time, BiP-derived values for HSR and sprinting are typically 10–20% higher than wholematch estimates [54–58], reflecting methodological superiority for quantifying genuine intensity. BiP analysis also facilitates integration with tactical phases (e.g., in-possession, out-of-possession, transitions), enhancing ecological validity and supporting more applied load prescriptions. Similarly, the choice of moving average window substantially influences how peak demands are quantified and interpreted. Shorter epochs (e.g., 30 s–1 min) are more sensitive to transient high-intensity bursts and provide greater ecological validity for capturing the most demanding passages of play [59, 60]. Longer epochs (e.g., 3–5 min) smooth fluctuations and may underestimate instantaneous load, though they better represent sustained work rates. Recent approaches combining rolling averages with cumulative time above key thresholds offer a more complete view of the volume–intensity relationship during match play [61, 62]. Despite their potential, few studies have contextualised HILAs through tactical phases or integrated GPS with video analysis to align physical outputs with match events [63], limiting ecological validity and applied utility.

Given the evolution of match demands, variability in measurement technologies, inconsistent thresholding practices, and limited contextualisation, there is a clear need to consolidate and critically evaluate current approaches to quantify HILAs in adult male professional soccer. A scoping review design was selected to comprehensively map this emerging evidence base, capturing methodological diversity and conceptual developments since the rapid expansion of wearable and optical technologies post-2021. Unlike systematic reviews, which assess intervention efficacy or effect size, the scoping approach enables identification of methodological inconsistencies, evidence gaps, and emerging trends across heterogeneous study designs. This review therefore aims to clarify how HILAs are operationalised, highlight underexplored areas such as adaptive thresholding and contextual integration, and provide a conceptual framework to guide future research and applied monitoring practices. Specifically, this scoping review seeks to (1) evaluate contemporary methodologies used to quantify HILAs in adult male professional soccer; (2) identify methodological inconsistencies, evidence gaps, and emerging trends; and (3) provide recommendations to support more standardised, individualised, and context-sensitive monitoring frameworks.

MATERIALS AND METHODS

RESULTS

Search Results

A total of 894 records were identified across the five databases, with 349 duplicates removed prior to screening. Following title, abstract, and methods screening, 23 full-text articles were assessed for eligibility. Six studies were excluded because participants were semiprofessional (n = 5) or under 18 years old (n = 1). An additional three studies were identified through hand-searching of reference lists. Consequently, 20 studies met all inclusion criteria and were retained for data extraction and synthesis (Figure 1).

FIG. 1

PRISMA flow diagram illustrating the identification, screening, eligibility, and inclusion process for study selection.

DISCUSSION

The aim of this study was to identify and synthesise prevailing methods for quantifying HILAs in adult male professional soccer. Findings revealed widespread reliance on GPS technologies, limited use of relative or adaptive thresholds, and inconsistent data-processing procedures. While GPS remains the predominant technology, variability between device manufacturers, firmware versions, and proprietary filtering algorithms further complicates cross-study comparisons. Even when sampling frequencies are identical, data output can differ substantially depending on the manufacturer’s smoothing, interpolation, and satellite correction procedures, leading to systematic bias, particularly for short-duration, high-acceleration movements where latency or signal noise is more pronounced. Standardising validation protocols and promoting manufacturer transparency regarding data-processing algorithms would improve reliability and facilitate more meaningful comparisons across studies and contexts.

Studies using relative thresholds demonstrated greater sensitivity to individual differences by scaling metrics to player capacity (e.g., maximal sprint speed, VIFT, or peak match-play outputs) [69, 70, 81, 87]. These approaches enhanced contextual relevance in load monitoring [48, 63], yet only four studies implemented them, and none consistently across all HILAs. Most research still applied absolute cut-offs, overlooking inter-individual variability and underestimating player-specific demands. Although widely advocated, individualisation remains infrequently applied, revealing a clear gap between methodological recommendations and practice.

Contextual and tactical factors such as match status, opposition strength, and tactical phase substantially influence the occurrence of HILAs [63, 81, 87]. However, few investigations incorporated these variables when quantifying match demands [73, 76, 78, 80]. Of the 20 studies included, only one [63] explicitly integrated GPS data with video-coded tactical information, while three others [73, 76, 78] reported partial contextual descriptors (e.g., opposition level or match outcome) without formal analytical integration. The remaining studies did not include contextual variables, confirming that contextual integration remains the exception rather than the norm. Environmental and situational conditions also modulate external load, with greater sprint volumes observed on artificial turf and higher high-speed running against weaker opponents [73, 80]. Positional and tactical demands further shape activity profiles; wingers and full-backs typically accumulate greater distances during inpossession phases, whereas defenders perform more recovery sprints when out of possession [55, 88].

The limited use of integrated GPS–video systems likely reflects several practical and logistical barriers. Synchronising positional and tactical datasets requires advanced software infrastructure and technical expertise, while access to high-fidelity optical data is often restricted by commercial rights holders. Data integration also increases analytical complexity and time requirements, particularly when aligning event and tracking timelines. Overcoming these barriers will require collaborative partnerships between researchers, technology providers, and professional clubs, alongside open-source frameworks to streamline multi-source integration. To facilitate systematic inclusion of positional and tactical-phase variations, future research should employ multi-level monitoring frameworks that align physical metrics with contextual game states. For example, combining GPS and video-derived data enables event-based segmentation (e.g., in-possession, out-of-possession, transitions) and position-specific profiling. Standardising positional groupings (e.g., central defenders, wide players, forwards) and linking these to phase-specific demands would enhance comparability and advance applied load modelling. Recent benchmark analyses by Bradley [89] further contextualise HILA profiles across major international tournaments, providing reference points for interpreting current findings.

A consistent discrepancy between training and match demands was evident, particularly for sprinting and high-intensity accelerations. Four studies [74, 75, 79, 82] reported that training underreplicated match intensities, with non-starters experiencing reduced exposure to high-intensity workloads [63]. Many drills also failed to reproduce peak match demands [79, 85], although some evidence suggested that training-to-match ratios for specific HILA metrics (e.g., sprint distance) may optimise adaptation [83]. Insufficient exposure to match-level intensities can impair readiness and elevate soft-tissue injury risk, highlighting the importance of targeted training and compensatory conditioning. Incorporating contextual elements such as tactical phase and positional role could improve training specificity and ecological validity.

Positional and temporal variations in HILA output were also evident. Strikers and full-backs generally recorded higher high-speed running and sprint distances than central defenders and midfielders [76], while greater frequencies of HILAs occurred during the early phases of matches [72]. These patterns indicate that positional role and match period influence physical intensity, with implications for load management, substitution timing, and recovery planning. Training interventions could replicate the most demanding positional and temporal scenarios observed in match play – for example, exposing wide players to repeated sprint sequences reflecting in-possession transitions, and central defenders to short, high-force decelerations during out-of-possession phases. Embedding such contextual conditioning within weekly micro-cycles and small-sided games would align tactical intent with physical stimulus and optimise performance transfer.

Marked methodological heterogeneity persisted across studies. Threshold definitions, filtering procedures, and reporting standards varied considerably, with key methodological details often omitted. This variability echoes longstanding calls within sports science for greater methodological consensus. Comparable standardisation initiatives, such as the GPS consensus statement and injury surveillance frameworks, have shown how shared definitions and reporting checklists improve reproducibility. A Delphi-based consensus process involving researchers, practitioners, and technology providers could harmonise HILA thresholds, filtering practices, and contextual reporting to strengthen both transparency and applied translation. At present, optical tracking systems remain unvalidated by FIFA for measuring accelerations and decelerations, although ongoing improvements in sampling frequency and algorithm precision may help close this gap. It is also important to note that acceleration and deceleration metrics are derived from tri-axial accelerometers integrated within GPS devices rather than from the GPS signal itself, ensuring more accurate capture of high-frequency movement changes [45].

Only four investigations [69, 70, 81, 87] applied relative thresholds, and just two extended these to accelerations and decelerations [81, 87], limiting comparability and interpretability. Given their strong association with mechanical load and neuromuscular stress, accelerations and decelerations represent critical indicators of external load that warrant greater emphasis in future research. A major limitation concerns the limited integration of internal load measures with external data. Despite growing recognition that “intensity” cannot be fully captured through mechanical metrics alone [90], only one study included HR data ≥ 90% HR_max [69]. This overreliance on external metrics provides an incomplete representation of physiological strain. Furthermore, contextual, perceptual, and tactical dimensions were inconsistently reported, and longitudinal designs were scarce, restricting understanding of internal–external load relationships and adaptive responses to repeated high-intensity exposure. Finally, the restriction to English-language publications may have introduced language bias, potentially excluding relevant non-English studies.

CONCLUSIONS

This scoping review highlights the central role of HILAs in adult male professional soccer while exposing substantial methodological fragmentation. Although GPS remains the dominant monitoring tool, absolute speed thresholds are the most commonly applied across studies, yet inconsistent threshold definitions, filtering procedures, and reporting standards continue to limit cross-study comparability and applied impact. A critical gap concerns the limited inclusion of internal load metrics. Reliance on external outputs alone neglects the physiological strain associated with HILAs and weakens understanding of player adaptation and fatigue. Integrating heart rate, perceptual, and biochemical measures alongside external data is essential to capture the full spectrum of player demands.

From an applied perspective, this integration could be achieved by synchronising internal and external metrics within unified monitoring platforms. For instance, combining GPS-derived external loads with heart rate or session-RPE data in the same time domain would enable practitioners to assess physiological efficiency (internal-to-external load ratios) and detect early signs of maladaptation. Incorporating biochemical or hormonal markers, where feasible, could further contextualise recovery and fatigue status. Developing standardised protocols for multi-modal data capture and aligning collection windows across technologies would improve consistency across clubs and research settings. However, feasibility and cost remain key considerations, particularly regarding GPS–video integration, which, despite its analytical value, may be limited by resource availability and technical expertise.

Future research should prioritise multi-modal, context-aware monitoring frameworks that combine internal, external, and tactical dimensions. To address practical constraints in elite soccer, hybrid research–practice collaborations embedded within existing club workflows are encouraged. Longitudinal observational designs across competitive seasons would enable repeated internal–external load integration without disrupting team operations. Cross-club data-sharing initiatives or anonymised central databases could enhance sample diversity and ecological validity, while lightweight, automated data pipelines linking GPS, RPE, and tactical video data may reduce practitioner workload and improve feasibility.

Recent technological and organisational advances are enhancing the practicality of such approaches. Commercial platforms such as Catapult OpenField [31], STATSports Sonra [32], and Kinexon PERFORM [33] now enable simultaneous synchronisation of GPS, inertial, and physiological data. Moreover, initiatives such as UEFA’s Football Research Programme [91] and FIFA’s EPTS Working Group [92] are developing standardised data protocols and validation frameworks that will facilitate harmonised, context-rich monitoring systems in professionalsoccer.

A progressive methodological roadmap should begin with the standardisation of threshold definitions, followed by systematic contextual integration and large-scale validation of hybrid tracking systems. Such integration, supported by transparent data reporting, is vital for advancing interpretation, training precision, and performance optimisation. Among these priorities, establishing consistent threshold definitions represents the most immediate and critical step, without a shared quantification basis, subsequent contextual or multimodal advances risk inconsistency and limited comparability. Once methodological alignment is achieved, contextual and tactical integration can meaningfully enhance ecological validity and applied translation.