Whole,half and peak characteristics of women's international flag football match-play during the 2024 world championships

Abstract

This study quantifies and describes the full, half and peak running characteristics during women's international flag football match-play. A total of 54 player match observations were taken from eight players, consisting of four offence and four defence. A repeated measures observational cohort design was used. Commonly used microtechnology variables including total-, high-speed- (3.46–5.29 m·s⁻¹), very high-speed- (5.29–6.26 m·s⁻¹) and sprint-distance (>6.26 m·s⁻¹) alongside high acceleration and deceleration (>2 m·s⁻² and <-2 m·s⁻², respectively) distance and count is documented. Whole, half and peak game characteristics were calculated for offence, defence and players who played both ways. On average, the whole game total distance for offence and defence was 1627 ± 386 m and 1514 ± 343 m respectively. For offence and defence ∼16% and 9% of total distance was high speed running (3.46–5.29 m·s⁻¹) and ∼3% was very high-speed running (5.29–6.26 m·s⁻¹) and <1% was covered by sprinting (>6.26 m·s⁻¹). The median number of accelerations and decelerations per match was 21 and 16 respectively for offence, and 8 and 10 for defence. The whole-match average speed was 29.8 m·min⁻¹ for offence and 27.0 m·min⁻¹ for defence. Average speeds increased as temporal durations decreased to 5 s (average speed = 294 m·min⁻¹). Findings highlight the intermittent nature of flag football, with players required to cover distance at various speeds and frequently change speed throughout. This is the first study to highlight the running characteristics of flag football tournaments. Such data can be used by coaches and performance staff to design appropriate and informed training stimuli to help optimise athlete performance for future tournaments.

Keywords

Acceleration high-speed running performance analysis physical conditioning playing position

Introduction

Flag Football (FF) is a non-contact, pass-orientated variation of American Football, eliminating blocking, tackling, and kicking. Instead, players “tackle” opponents by pulling off fabric flags attached to their hips.^1,2 Matches are played for a total time of 40 min on a 50 × 25-yard (45.72 × 22.86 m) field, with five players per team on the field at one time, with players either on offence (O) or defence (D).¹ There are different versions of FF played across the globe, however annual international tournaments such as the 2024 World Championships are governed by International Federation of American Football (IFAF) regulations. During IFAF tournaments, players are expected to play multiple games per day with the effects of cumulative load on performance currently unknown. Ahead of the 2028 Olympics, there is a global drive to push the sport competitively.³ At present, despite FF's growing prominence, there is currently no peer-reviewed academic literature quantifying and evaluating the match characteristics of FF. This is in direct contrast to well-established research and practices in other team sports such as American football, soccer, and rugby.⁴

The quantification of match characteristics is a foundational component in athlete monitoring, informing training prescription, load management, and injury risk mitigation strategies.^5,6 Such data can be used by coaches and support staff to design appropriate and informed training stimuli to optimise athletic performance.^4,7,8. Furthermore, understanding both the average (whole-match, and half-match) and peak running demands enables practitioners to design training drills that prepare athletes for the most physically demanding periods of competition. It is likely that the whole and half match characteristics of flag football are lower than those of other team sports e.g., soccer⁹ and rugby¹⁰ due to shorter game durations, however relative (per minute) and peak characteristics may be comparable those observed in these sports, reflecting the intermittent, high-intensity nature of match-play. Fluctuations in match characteristics across games and tournaments are common in intermittent team sports,^9,11 this variation is often impacted by contextual (e.g., match opposition) and internal factors (e.g., player physical capabilities).¹² It is important to understand match characteristic variability to support individualised preparation. Currently, there is an absence of this information in FF cohorts.

Given the sport's aim to increase participation and level of competition, establishing an understanding of match demands is critical to help develop physical athletic performance and provide insights for coaching, medical and sports science staff.¹³ Therefore, the primary aim of the study was to quantify and describe the full, half and peak running characteristics during match-play in women's international FF. The secondary aim of the study was to describe the variation in running characteristics between matches and players.

Materials and methods

Study design and participants

A repeated measures observational cohort research design was used to quantify and describe the time-motion characteristics of FF match-play during the 2024 World Championships. Data was collected from one team, across seven matches, over a 4-day period during the tournament. Time-motion characteristics of match-play were calculated using microtechnology (Catapult S5, Catapult Innovations, Melbourne, Australia). Multiple metrics were used to calculate the match demands including total distance (m), average speed (m·min⁻¹), high-speed running (HSR) distance (m) (3.46–5.29 m·s⁻¹), very high-speed running (vHSR) (5.29–6.26 m·s⁻¹) distance and sprint distance (m) (>6.26 m·s⁻¹).¹⁴ Acceleration distance and deceleration distance (>2 m·s⁻²) alongside acceleration and deceleration counts were also selected for analysis.¹⁵ A total of 54 player match observations from eight players across one national team ranked within the top 5 following the 2024 World Championship (age = 27.9 ± 4.1, height = 167.4 ± 5.5, body mass = 66.5 ± 6.0), consisting of four offence (O) and four defence (D) (mean = 6.75 ± 0.46, range = 6 to 7 observations per player). Ethical approval was granted by the Leeds Beckett ethics committee (reference number: 135619). Participation in the study was voluntary, with all participants providing written informed consent.

Data collection

During the tournament all participants wore a micro-electrical mechanical system (MEMS) device with enabled Global Navigation Satellite System (GNSS) technology (Catapult S5, Catapult Innovations, Melbourne, Australia). These devices provide geospatial positioning at a 10 Hz sampling frequency encompassing both Global Positioning System (GPS) and Global Navigation Satellite System (GLONASS) satellites. The device also contains an embedded 100 Hz tri-axial accelerometer, gyroscope, and magnetometer. The validity and reliability of these devices have previously been undertaken, demonstrating adequate reliability and validity to measure instantaneous speeds across multiple starting velocities (CV% = 2.0% to 5.3%).¹⁶ The device was worn between the participants scapulae in a tightly fitted company made vest to reduce device movement. Each participant wore the same device and an appropriately fitted vest throughout data collection to reduce inter-device variation.¹⁷ The devices were turned on and placed outside 15 min prior to the match warm up to ensure adequate satellite connection. Familiarisation to the devices took place prior to the tournament during a training camp. The units possessed adequate average signal quality for each player match, with an average of 15.58 ± 2.75 satellites connected during each match and an average horizontal dilution of precision (HDOP) of 0.70 ± 0.16.¹⁸

Data analysis

The start and end time of each half of the match were recorded. Following each match, data were extracted and analysed using propriety software (OpenField, Catapult Innovations, Melbourne, Victoria). Descriptive statistics for the whole and half game time-motion characteristics (mean ± standard deviation for distance data, median [IQR] for acceleration and deceleration counts) were calculated for offence, defence and players who played both ways i.e., both offence and defence a single game. Match duration was calculated using whole match time as opposed to on-field duration during matchplay. Variability between players and matches was calculated using the back-transformed standard deviation of the random effects from a random-effects-only model, expressed as a percentage, in R studio (Version 2024.12.1, R Foundation for Statistical Computing, Vienna, Austria) using the lme4 package (version 1.1–36).¹⁹ To calculate the peak running characteristics of match-play, instantaneous speed data (10 Hz) were also extracted from the manufacturer's software. Only data between the specified start and end time of each half were included, therefore warmups and half time were excluded. A custom-built R script using the zoo package (version 1.8–13)²⁰ computed moving averages of speed over 5 s, 10 s, 30 s, 60 s, 180 s, 300 s, 600 s and calculate the peak distance covered within each time interval. The maximum value for each player, for each duration, per match, was determined. The relative distance covered per minute (average speed) for each time interval was then calculated (m·min⁻¹). The Power-Law relationship was calculated by calculating the intercept and slope using the 10 to 180-s peak durations.²¹ Due to the small sample size of the cohort; no statistical comparisons including effect sizes and confidence intervals for whole, half and peak demands were performed on the data.

Results

Descriptive statistics of the whole and half time-motion characteristics of match play are shown for O, D and players who played both O and D during a single match (Both) are shown in Table 1. Distance data is shown as mean ± SD, whilst count data is shown as median (IQR).

Table 1.

Descriptive statistics of the whole- and half time-motion characteristics of match play are shown for offence (O), defence (D), and players who played both O and D during a single match (both). Distance data is shown as mean ± SD, whilst count data is shown as median (IQR).

	Whole			1st half			2nd half
Variable	Offence	Defence	Both	Offence	Defence	Both	Offence	Defence	Both
Match Duration (minutes)	55 ± 5			27 ± 1			28 ± 5
Average Speed (m·min⁻¹)	29.8 ± 6.0	27.0 ± 5.6	38.3 ± 6.2	31.1 ± 9.5	27.4 ± 4.4	39.9 ± 6.7	28.7 ± 7.9	26.7 ± 9.9	36.4 ± 7.7
Total Distance (m)	1627 ± 386	1514 ± 343	2191 ± 405	830 ± 246	743 ± 123	1086 ± 167	796 ± 228	771 ± 306	1105 ± 329
HSR (3.46–5.29 m·s⁻¹) distance (m)	260 ± 85	143 ± 53	344 ± 69	139 ± 67	69 ± 20	175 ± 29	120 ± 50	74 ± 43	169 ± 53
vHSR (5.29–6.26 m·s⁻¹) distance (m)	51 ± 35	46 ± 44	124 ± 59	27 ± 26	25 ± 22	57 ± 22	23 ± 17	22 ± 24	67 ± 39
Sprint (> 6.26 m·s⁻¹) distance (m)	9 ± 15	4 ± 7	24 ± 22	2 ± 6	2 ± 4	12 ± 9	7 ± 11	2 ± 4	12 ± 18
Acceleration (> 2 m·s⁻²) distance (m)	83 ± 30	48 ± 64	133 ± 41	45 ± 22	24 ± 34	70 ± 22	38 ± 15	24 ± 31	64 ± 22
Acceleration (> 2 m·s⁻²) count	21 (11 to 26)	8 (7 to 14)	37 (29 to 38)	10 (5 to 15)	3 (2 to 6)	18 (13 to 20)	9 (4 to 13)	6 (2 to 9)	17 (16 to 28)
Deceleration (> 2 m·s⁻²) distance (m)	9 ± 4	7 ± 6	11 ± 3	4 ± 2	4 ± 3	5 ± 1	4 ± 2	4 ± 3	6 ± 2
Deceleration (> 2 m·s⁻²) count	16 (12 to 19)	10 (12 to 19)	18 (16 to 21)	8 (5 to 10)	5 (3 to 10)	11 (7 to 11)	6 (5 to 8)	5 (3 to 7)	10 (7 to 11)

n.b. match duration reflects the whole match duration, regardless of playing time for each positional group.

Table 2 shows the variance of the model explained by each random effect for whole match microtechnology variable. (Player-to-Player = 13% to 170% and Match to Match = 8% to 22%).

Table 2.

The variance of the model explained by each random effect for all whole match microtechnology variable. Expressed as a percentage derived from the back-transformed standard deviation of the random effects.

Variable	Between player variability (%)	Between match variability CV (%)	Residual variability (%)
Average Speed (m·min⁻¹)	13	10	18
Total Distance (m)	13	17	19
HSR (3.46–5.29 m·s⁻¹) distance (m)	43	13	31
vHSR (5.29–6.26 m·s⁻¹) distance (m)	80	22	63
Sprint (> 6.26 m·s⁻¹) distance (m)	138	9	120
Acceleration (> 2 m·s⁻²) distance (m)	53	13	39
Acceleration (> 2 m·s⁻²) count	170	N/A	60
Deceleration (> 2 m·s⁻²) distance (m)	49	8	26
Deceleration (> 2 m·s⁻²) count	91	14	39

Nb match to match variability was not able to be calculated for Acceleration count but was maintained within the model for consistency with other models.

Figure 1a. Shows the mean (SD) of peak average speed (m·min⁻¹) of temporal durations from 5 s, 10 s, 30 s, 60 s, 180 s, 300 s, 600 s. Along the peak values for each player match, coloured by position played. Intercepts and slopes and power-law relationship for average speed are presented Figure 1b, the intercept and slope for the power-law relationship are 445.9 and −0.36, respectively.

Figure 1.

(A) the mean (SD) of peak average speed (m·min⁻¹) of temporal durations from 5 s, 10 s, 30 s, 60 s, 180 s, 300 s, 600 s. Along the peak values for each player match, coloured by position played within the match. (B) Power Law relationship for maximal duration-specific average speed.

Figure 2. illustrates the tournament structure during the 2024 FF World championships. The tournament consisted of three group games played across days 1 and 2 along with four knockout games/play-off games played across days 3 and 4. During the tournament, the number of matches played per day ranged from 1–3.

Figure 2.

Illustration of the tournament structure and fixture scheduling for one team during the 2024 FF world championships.

Discussion

The aim of the study was to quantify and describe the full, half and peak running demands of women's international FF. To the authors’ knowledge this is the first study to measure the match characteristics of this sport. The findings from the study provide initial insights, which may help practitioners consider how to manage future tournaments and optimise training practices. The findings confirm the intermittent nature of FF, characterised by a combination of high and moderate intensity forms of locomotion including sprinting, jogging, walking along with frequent changes in speed.

The whole game total distance covered during women's international FF was 1627 ± 386 m and 1514 ± 343 m for O and D respectively. For those players who were utilised in both O and D positions, an increased total distance of 2191 ± 405 was recorded. This is substantially lower than values seen during other women's sports such as soccer (8940 ± 1094 m)⁹ and rugby union (4177 ± 2066 m).¹⁰ As common with other team sports, this total distance is broken down into high-speed running, very high-speed running and sprint distance. On average ∼16% and ∼9% of total distance was covered between 3.46–5.29 m·s⁻¹, ∼3% was covered 5.29–6.26 m·s⁻¹, and ∼0.6% and ∼0.2% of distance was covered whilst sprinting (>6.26 m·s⁻¹) for offence and defence respectively. This is comparable to women's soccer, with previous research showing an average ∼3% total distance (regardless of position) was covered at speeds >5.29 m⋅s⁻¹.⁹ However, during FF when players play both ways, ∼21%, 8% and 1.5% of total distance were covered at speeds of 3.46–5.29 m·s⁻¹, 5.29–6.26 m·s⁻¹, and >6.26 m·s⁻¹ respectively. It is likely that the lower distances seen within FF are attributable to shorter match durations and less time spent on field, rather than lower intensity of actions.

FF players are also exposed to frequent changes of speed throughout match-play, with median numbers of accelerations and decelerations of 21 and 16 respectively for offence, and 8 and 10 for defence, increasing to medians of 37 accelerations and 18 decelerations when for players who played both ways. The number of whole game accelerations and decelerations for offensive players are similar to seen within women's soccer (accelerations mean = 13.6 to 24.7; decelerations mean = 11.4 to 24.7), whilst defensive players are slightly below these values in a shorter match duration.²² The number of accelerations and decelerations reflect the tactical demands of FF, with players repeatedly tasked with offensive routes and defensive manoeuvres that include performing high intensity efforts from a stationary stance and often include changes of direction. Given the frequency and tactical necessity of the accelerations and decelerations during match-play, it is necessary for practitioners working with FF athletes to ensure players have adequate capabilities in order to perform such actions repeatedly^23,24 and avoid neuromuscular fatigue which could contribute to an increased risk of injury.²⁵

The average speed (m·min⁻¹) for during a whole match of FF was 29.8 ± 6.0 m·min⁻¹, 27.0 ± 5.6 m·min⁻¹ for O and D respectively, and 38.3 ± 6.2 m·min⁻¹ for players who played both ways. This reflects the intermittent nature of FF whereby players may only be on the field for very short periods at a time whilst O or D plays occur. These values are also much lower than the values seen during more continuous sports such as soccer (92 m·min⁻¹)⁹ and rugby union (55.0–64.4 m·min⁻¹).¹⁰ Along with the whole and half demands, peak average speeds across temporal durations of 5 s, 10 s, 30 s, 60 s, 180 s, 300 s, 600 s were calculated. An inverse relationship between temporal duration and peak average speed exists, aligning with previous research assessing peak demands of intermittent sports²⁶ (3), with the greatest peak average speed seen at 5 s durations (294.0 ± 49.4 m·min⁻¹) and the lowest seen at 600 s durations (45.7 ± 6.97 m·min⁻¹). The peak covered during 60, 180, and 500 s in FF are lower than those seen within previous research into other sports such as soccer,⁹ this is again attributable to the intermittent nature of FF whereby players are likely to spend more time stationary or moving at low velocity between plays and drives. This study also presents shorter time intervals (5–60 s) in an attempt to capture the intermittent nature of the sport, seen within both plays and drives throughout the match.

The present study also presents potential sources of variability in match demands. The match demands experienced by players differ between individual players, with running demands varying 13% to 138% between players. Sprint distance has the greatest variation and total distance the least, which is consistent with other team sports including soccer,⁹ rugby league²⁷ and rugby union.²⁸ The acceleration demands varied by 50% to 81% dependent on metric, with accelerations showing greater variability than decelerations. The within player differences in match characteristics highlights the importance of individualised monitoring, and adjustment of training and match loads. A possible explanation for this between player variation is differences in physical capabilities, this is particularly true for threshold-based metrics such as sprint, acceleration and deceleration distance given the use of absolute thresholds within this study. However, the physical capabilities have not yet been quantified in international women's FF players. Another reason for these differences is tactical roles of individual players (which is dictated by a team's playbook) and positions during match-play, which is also highlighted by variance between positions highlighted within the present study. Whole-match running and acceleration demands also varied between-matches, with numerous potential contextual factors affecting the between-match demands of FF. Both quality of opposition and match-result are likely contributing factors to the variation in demands, which have previously been found to influence the running demands of match-play in other team sports such as soccer²⁹ and rugby league.²⁷

Despite FF having lower whole match demands than those seen in other sports, the FF world championships are condensed, meaning a total of 7 games are played across 4-days, with the number of games ranging between 1 and 3 games per day (Figure 2). During team sports such as soccer, fixture congestion has been observed to have equivocal effects of physical and technical performance.³⁰ Such congestion is generally observed with less than or equal to 72 h between games. However, this is different to flag football, as IFAF tournaments require players to play multiple games, of less duration per day. As such the potential implications of accumulated load and limited recovery on physical and technical performance between games on the same day is currently unknown. It is therefore important that practitioners monitor match exposure and potential fatigue responses closely to ensure players are adequately prepared for the demands of repeated games throughout the tournament, although at present more research is required to determine dose-response relationships in elite women's FF.

Practical applications

The whole and half time-motion characteristics within this study contribute to coaches and performance staff understanding of the demands and characteristics of FF tournament match-play. The variability data presented within the study can be used by practitioners as reference values when assessing if meaningful change exists between players and matches. The power law relationship reported within this study can also act as a reference value to evaluate training drill intensity for female international FF. For example, when prescribing a 120 s second (2 min) technical-tactical drill with the intention of replicating peak average speed observed during international FF match-play, the intercept and slope values of the power law relationship can be used to calculate peak intensity where (t) is drill time in seconds and (I) is peak intensity.,^9,21

I = - 445.9 t^{- 0.36}

I = - 445.9 \times 120^{- 0.36}

I = 83 m \cdot \min^{- 1}

This example provides an estimated target average speed of 83 m·min^−1, allowing coaches and performance staff to monitor exposure to match intensities where appropriate.

Limitations

The present study is not without limitations. Firstly, all data was collected from a small sample size (one team) during a single tournament, therefore, statistical analysis was not possible to investigate differences between groups (e.g., O vs D), thus the findings may lack generalisability to the wider sport. This is because the results are specific to the players within the team and due to variation in the tactical game plan. Due to sample size and the need to avoid type 2 error, it was beyond the scope of the current study to investigate position-specific differences, as such future studies with a larger sample should aim to address positional-specific demands to identify any unique role requirements.

Conclusion

The present study describes the whole, half and peak demands of international women's FF match play. The findings highlight the intermittent nature of FF, with players required to cover distance at various speeds and frequently change direction throughout the match, suggesting that coaches should prioritise repeated short-duration sprints, change of direction drills and adequate recovery strategies during congested tournament schedules. Variation in match demands is also present, with between-player, and -match variation being highlighted. These initial findings form a foundation for future research exploring positional and fatigue-related performance trends in elite female FF. This study provides an initial assessment of match characteristics within flag football. Future research within this space should look to investigate the relationships between physical characteristics, cumulative load, fatigue and match-characteristics across a tournament.

Footnotes

Acknowledgements

We would like to thank the FF team and BAFA for supporting this research study. We also wish to express our gratitude to Jo Clubb for offering insightful feedback on the draft.

ORCID iDs

Eleanor Travis

Chris Brogden

Ashley Jones

James Parmley

Ethical considerations

This study was granted ethical approval by Leeds Beckett University ethics committee (reference number: 135619). Participants gave informed consent to participate in the study before taking part.

Consent to participate

All participants were asked to read the participant information sheet and sign the institutionally approved consent form.

Consent for publication

All participants consented to participate to their data being used anonymously in a scientific journal.

Authors contributions

ET: conceptualisation, funding acquisition, investigation, methodology, project administration, resources, supervision, visualisation, writing – original draft preparation. CB & AJ: conceptualisation, methodology, writing – review and editing. JP: investigation, methodology, data curation, formal analysis, visualisation, writing – original draft preparation.

Funding

The authors received a small internal grant from the lead authors institution to carry out this research project. This funding helped finance travel expenses and the researcher's time.

Declaration of conflicting interest

During the data collection period ET and JP were employed on a contractual basis with the FF team. Two of the authors worked with the team on a contractual basis at the time of data collection.

Data availability

All data is presented within the results section of the manuscript.

References

International Federation of American Football. International flag football rules 2023. Online: https://americanfootball.sport/wp-content/uploads/2023/05/FlagRules2023.pdf (2023, accessed 9 June 2025).

British American Football Association. What is flag football?. Online: https://www.britishamericanfootball.org/what-is-flag-football/ (2025, accessed 9 June 2025).

International Federation of American Football. IFAF history. Online: https://www.americanfootball.sport/history/ (2025, accessed 9 June 2025).

Torres-Ronda

Beanland

Whitehead

, et al. Tracking systems in team sports: a narrative review of applications of the data and sport specific analysis. Sports Med Open 2022; 8: 15.

Eckard

Padua

Hearn

, et al. The relationship between training load and injury in athletes: a systematic review. Sports Med 2018; 48: 1929–1961.

Nielsen

Bertelsen

Møller

, et al. Training load and structure-specific load: applications for sport injury causality and data analyses. Br J Sports Med 2018; 52: 1016–1017.

Campbell

Peake

Minett

. The specificity of rugby union training sessions in preparation for match demands. Int J Sports Physiol Perform 2018; 13: 496–503.

Edwards

Spiteri

Piggott

, et al. Monitoring and managing fatigue in basketball. Sports 2018; 6: 19.

Myhill

Weaving

Barron

, et al. Variability of whole and peak match physical performance in highly trained female academy soccer players: a multi-club analysis. PLoS One 2025; 20: e0318642.

10.

Nolan

Curran

Brady

, et al. Physical match demands of international women’s rugby union: a three-year longitudinal analysis of a team competing in the women’s six nations championship. J Funct Morphol Kinesiol 2023; 8: 32.

11.

Dalton-Barron

Whitehead

Roe

, et al. Time to embrace the complexity when analysing GPS data? A systematic review of contextual factors on match running in rugby league. J Sports Sci 2020; 38: 1161–1180.

12.

Gregson

Drust

Atkinson

, et al. Match-to-match variability of high-speed activities in premier league soccer. Int J Sports Med 2010; 31: 237–242.

13.

Hill

. Flag football’s Olympic ascent: perspectives on growth, safety, and analytics. Int J Sports Sci Coach 2025; 17479541251385653: 1–9.

14.

Park

Scott

Lovell

. Velocity zone classification in elite women’s football: where do we draw the lines? Sci Med Football 2019; 3: 21–28.

15.

Mara

Thompson

Pumpa

, et al. The acceleration and deceleration profiles of elite female soccer players during competitive matches. J Sci Med Sport 2017; 20: 867–872.

16.

Varley

Fairweather

Aughey

. Validity and reliability of GPS for measuring instantaneous velocity during acceleration, deceleration, and constant motion. J Sports Sci 2012; 30: 121–127.

17.

Buchheit

Simpson

. Player-tracking technology: half-full or half-empty glass? Int J Sports Physiol Perform 2017; 12: S2–35–S2–41.

18.

Tahsin

Sultana

Reza

, et al. Analysis of DOP and its Preciseness in GNSS Position Estimation. In: International conference on electrical engineering and information communication technology (ICEEICT), 2015, pp.1–6.

19.

Bates

Mächler

Bolker

, et al. Fitting linear mixed-effects models using lme4. J Stat Softw 2015; 67: 1–48.

20.

Zeileis

Grothendieck

Ryan

, et al. zoo: S3 infrastructure for regular and irregular time series (Z’s ordered observations). Online: https://cran.r-project.org/web/packages/zoo/index.html (2025, accessed 18 July 2025).

21.

Delaney

Thornton

Rowell

, et al. Modelling the decrement in running intensity within professional soccer players. Sci Med Football 2018; 2: 86–92.

22.

Ramos

Nakamura

Pereira

, et al. Movement patterns of a U-20 national women’s soccer team during competitive matches: influence of playing position and performance in the first half. Int J Sports Med 2017; 38: 747–754.

23.

Castaño-Zambudio

Repullo

Jiménez-Reyes

. Enhancing acceleration capabilities in professional women’s football players: a comparative analysis of game-based versus resisted sprint trainings. Appl Sci 2024; 14: 10327.

24.

Harper

Kiely

. Damaging nature of decelerations: do we adequately prepare players? BMJ Open Sport Exerc Med 2018; 4: e000379.

25.

Carling

Le Gall

Reilly

. Effects of physical efforts on injury in elite soccer. Int J Sports Med 2010; 31: 180–185.

26.

Whitehead

Till

Weaving

, et al. Whole, half and peak running demands during club and international youth rugby league match-play. Sci Med Football 2019; 3: 63–69.

27.

Dalton-Barron

Whitehead

Roe

, et al. Time to embrace the complexity when analysing GPS data? A systematic review of contextual factors on match running in rugby league. J Sports Sci 2020; 38: 1161–1180.

28.

McLaren

Weston

Smith

, et al. Variability of physical performance and player match loads in professional rugby union. J Sci Med Sport 2016; 19: 493–497.

29.

Castellano

Blanco-Villaseñor

Álvarez

. Contextual variables and time-motion analysis in soccer. Int J Sports Med 2011; 32: 415–421.

30.

Julian

Page

Harper

. The effect of fixture congestion on performance during professional male soccer match-play: a systematic critical review with meta-analysis. Sports Med 2021; 51: 255–273.