https://orcid.org/0000-0001-6614-9914
ABSTRACT
We analysed peak 1-, 2- and 5-min periods and the associated 5-min recovery period in matches from three consecutive seasons in the Danish Superliga. A semi-automatic multicamera system was used to collect high-speed running distance (≥5.5 m/s; HSRD), sprint distance (≥7.0 m/s; SpD) and distance covered during intense acceleration (≥3 m/s2; AccD). Analysis included 479 players and 6042 to 9671 match observations using rolling average. Distances covered per minute during the peak periods were significantly higher than match averages: HSRD (207–772%), SpD (447–1793%), and AccD (383–1096%). Distances covered per min were lower during 1-min recovery periods than match average for HSRD following peak 1-, 2- and 5-min period (29%, 6%, 3%, 2%, 2%; 35%, 11%, 0%, 2%, 3%; and 45%, 29%, 13%, 8%, 4%; p < 0.05, respectively), and for SpD (20%, 3%, 7%, 3% (4% higher in the 5th min); 24%, 12%, 3%, 0%, 7%; and 39%, 29%, 18%, 17%, 12%; p < 0.05, respectively). Opposite, AccD increased in the following 1-min recovery periods following peak 1-, 2- and 5-min periods (68%, 89%, 94%, 88%, 90%; 47%, 86%, 93%, 90%, 88%; 23%, 56%, 76%, 85%, 87%; p < 0.05) compared to match averages. Intensity was higher during shorter periods, whereas performance decrements were largest after longer peak periods for HSRD and SpD, whereas no decrement was observed in AccD.
Introduction
Association football (soccer) matches consist of high-intensity periods interspersed with activities of lower intensity (Bangsbo et al., Citation1991). The overall demands as well as the length and frequency of these high-intensity periods vary throughout the match (Mohr et al., Citation2005) and are affected by several contextual factors such as e.g., game status, quality of opponent, playing position, individual capacity, tactical role (Baptista et al., Citation2019; Delaney et al., Citation2018; Oliva-Lozano et al., Citation2020; Oliva-Lozano, Fortes, et al., Citation2021). Studies investigating the demands of the game have shown increases in game demands during the last 10–20 years (Allen et al., Citation2024; Barnes et al., Citation2014; Bradley et al., Citation2016; Bush et al., Citation2015; Lago- Peñpeñas et al., Citation2023). Due to the extensive interest in understanding the most intense periods, game demands of 1- to 10-minute periods have been described (Fereday et al., Citation2020; Fransson et al., Citation2017; Oliva-Lozano et al., Citation2020; Schimpchen et al., Citation2021).
Studies have indicated that the most intense 5-min periods in football matches elicit periods of temporary fatigue, a concept originally introduced by Mohr et al. (Citation2003), as the players’ physical performance has been observed to diminish during the 5-min period following the most intense period (DiMascio & Bradley, Citation2013; Fransson et al., Citation2017; Mohr et al., Citation2003). The first studies who investigated these intense periods used fixed time periods (0–5 min, 5–10 min, etc.), which, compared to rolling averages, has been shown to underestimate the physical demands during the most intense periods (Fereday et al., Citation2020; Oliva-Lozano, Martin-Fuentes, et al., Citation2021). Therefore, recent studies have used rolling averages to investigate the most intense periods and the demands of the game (Fransson et al., Citation2017; Novak et al., Citation2021; Oliva-Lozano, Martin-Fuentes, et al., Citation2021; Schimpchen et al., Citation2021), which is of importance for understanding the real peak loading in games. These insights have enabled coaches to design exercises with a specific focus on meeting these peak demands, but training for the peak demands has been questioned due to the fact that the various locomotor variables (high intensity running, sprinting distance, accelerations etc.) do not necessarily peak during the same epochs (Novak et al., Citation2021). Nevertheless, these methods have increased the understanding of the demands of these most intense periods.
The following period (post peak intervals) and the associated fatigue response has primarily been investigated as performance in 5-min intervals (DiMascio & Bradley, Citation2013; Fransson et al., Citation2017; Mohr et al., Citation2005). A recent study by Schimpchen et al. (Citation2021) reported peak intensities periods as 1-min, 5-min and 10-min windows and the subsequent first, third and fifth 1-min period. Measurements included in the analysis were total distance covered, high-speed running distance >5.0 m/s and number of accelerations–decelerations (acc–dec). The authors observed that physical performance was limited in the first minute after all peak periods and for all metrics but total distance covered and acc–dec were recovered to match average in the subsequent third minute, whereas high-speed running distance was limited in the third and fifth minute after the most intense period (Schimpchen et al., Citation2021). The study by Schimpchen and colleagues is limited by solely including one elite team and it is, therefore, unknown whether these findings can be generalized, as one team may be specific in terms of physical performance, training status and tactics, which all may impact the outcomes. Moreover, only the first, third, and fifth recovery minute were evaluated, limiting the understanding of the recovery kinetics. To be able to better understand the recovery period after a peak period, analysis of a larger cohort of football players is warranted in order to avoid the specificity of one single team.
The purpose of the present study was, therefore, to investigate the most intense periods and the time course of the subsequent performance changes in the minutes following the most intense 1-min, 2-min, and 5-min periods of high-speed running distance, sprint distance, and acceleration distance using rolling averages in a large cohort of football players. The hypotheses were that peak periods have a large negative impact on the physical game performance in the subsequent recovery period for all physical performance metrics, but the physical performance return to normal after five minutes.
Methods
Spatio-temporal data from all matches in the Danish Superliga throughout three seasons (2015/16; 2016/17; 2017/18) were collected using a semi-automatic tracking system (Prozone TM; Stats LLC, Chicago, IL, USA). Prozone have been independently validated to verify accurate measurements and have shown high inter-observer reliability with kappa index of 0.938 ± 0.035 (Di Salvo et al., 2009, Bradley et al., Citation2009). Data from 479 (range 478–480) Danish Superliga players were included in the analysis. Only players who completed a full match were included in the analysis to avoid interference with different playing time.
For further analysis, distance covered with speed >5.5 m/s (high-speed running distance, HSRD) and >7 m/s (sprint distance; SpD) as well as distance covered with acceleration >3 m/s2 (AccD) were chosen. Peak periods were identified using rolling averages for 1-, 2- and 5-min periods and data from the following 5-min periods were extracted in 1-min periods, as previously described by Fransson et al. (Citation2017). To be able to analyse the 5-min recovery period, peak periods were identified ending no later than in the 40th and 85th minute for first and second half, respectively. In addition, 1-min mean was calculated by extracting the peak and recovery period from the 90-min game data and dividing with 84, 83 or 80 for 1-, 2- and 5-min periods, respectively.
For the analysis, a total of 478–480 players and 9585–9671, 9054–9249, and 6042–6111 match observations for HSRD, SpD and AccD were included, respectively. The differences in number of match observations between the HSRD and SpD are due to differences in the time of the match that peak periods for each variables occur. Furthermore, the fewer peak periods in AccD compared to HSRD and SpD were due to a data error in some of the match observations for AccD.
Peak periods are abbreviated as P1, P2 and P5 for peak 1-min, peak 2-min and peak 5-min period, respectively, whereas the recovery periods are abbreviated as R1, R2, R3, R4 and R5 for the 1st, 2nd, 3rd, 4th and 5th minute after the peak period, respectively. Recovery periods are followed by a subscripted 1, 2 or 5 indicating which length the previous peak period had (e.g., R11). Data were sorted and analysed with multifunctions in Excel.
Statistical analysis
Data are presented as mean±SD. Kolmogorov Smirnov test and visual inspection of histogram and qq plots were used to check normality. Data with normal distribution were analysed using one-way ANOVA and Student paired t-test, whereas non-parametric data were analysed using sign test. All statistical analysis were performed using SPSS (IBM version 27). A significance level of 0.05 was chosen. Effect size was calculated as Z value divided by square root of N and interpreted as suggested by Fritz et al. (Citation2012). Effect sizes were set as followed 0.1 small effect, 0.3 medium effect and 0.5 large effect (Fritz et al., Citation2012).
Results
High-speed running distance
HSRD was 62 ± 18, 76 ± 23 and 110 ± 35 m during P1, P2 and P5, respectively, accounting for 9, 11 and 16% of HSRD of the entire match (704 ± 258 m) equalling a HSRD of 62 ± 18, 38 ± 11 and 22 ± 7 m/min during P1, P2 and P5, respectively. HSRD during P1, P2, and P5 was 772, 431, and 207% higher than the mean match value (7.2 ± 2.8, 7.2 ± 2.8, 7.1 ± 2.8 m/min) (see ) and 852%, 490%, and 283% higher, respectively, than during the 5-min recovery periods (6.6 ± 5.1, 6.4 ± 5.0, 5.7 ± 5.6 m/min).
Figure 1. Peak 1-, 2-, and 5-min high-speed running distance and following 5-min recovery period, divided into 1-min interval. Dotted line expresses average high-speed running distance for the whole match. * Significantly lower high-speed running distance than average for the whole match (p < 0.01).
HSRD for the most intense 1-, 2-, and 5-min periods and the following five 1-min recovery periods are shown in and . HSRD in R21 was 6% higher than in R22 (p < 0.05, ES = 0.06). Furthermore, HSRD in R11, R21, and R31 was 29%, 32% and 11% higher compared to R15, R25, and R35, respectively (all significant, p < 0.05, ES = 0.07, 0.05, and 0.04). Moreover, HSRD in R12, R22, R32, R42, and R52 was 20%, 25%, 16%, 6% and 1% higher compared to the comparable R15, R25, R35, R45, and R55. HSRD in the 5-min recovery period after 1-, 2- and 5-min peak periods was 9, 10 and 19% lower, respectively, (p < 0.001, ES = 0.17, 0.19, and 0.36) than the average 5-min period of the game.
Table 1. Illustrates peak periods, recovery period divided into 1 min intervals, total 5 minutes post peak period and mean match value. Values are expressed in both median with 25% and 75% interquartile range (IQR) and average with SD. All values are m/min.
Sprint distance
SpD was 53 ± 18, 60 ± 22 and 75 ± 32 m during P1, P2 and P5, corresponding to 17, 19 and 25% of the total match SpD (308 ± 161 m) and equalling a SpD of 53 ± 18, 30 ± 11, 15 ± 6 m/min during P1, P2, P5. SpD during P1, P2 and P5 was 1793, 964, and 447% higher (p < 0.05) than the mean match value (2.9 ± 1.7, 2.8 ± 1.7, 2.8 ± 1.7 m/min) (see ) and 1725, 1045 and 519% higher (p < 0.05) than during the entire subsequent 5-min recovery period (2.9 ± 8.4, 2.6 ± 8.4, 2.5 ± 8.7 m/min).
Figure 2. Peak 1-, 2-, and 5-min sprint distance and following 5-min recovery period, divided into 1-min interval. Dotted line expresses average sprint distance for the whole match. * Significantly lower sprint distance than average for the whole match (p < 0.01).
SpD for the most intense 1-, 2-, and 5-min periods and the following five 1-min recovery periods are shown in and . In the 5-min recovery periods following 1-, 2- and 5-min peak periods, SpD was 7, 9 and 21% lower (p < 0.001, ES = 0.07, 0.10, 0.23) than in the mean 5-min period, respectively.
Intense accelerations
AccD was 14 ± 6, 18 ± 8 and 28 ± 16 m during P1, P2, and P5, respectively, accounting for 11, 15 and 22% of AccD during the entire match (126 ± 44 m). These values equal 14 ± 6, 9 ± 4 and 6 ± 3 m/min during P1, P2 and P5, which were 1096, 678, and 383% higher (p < 0.05) than the mean match value (1.2 ± 0.5, 1.2 ± 0.5, 1.1 ± 0.5 m/min) (see ). Moreover, AccD during P1, P2, and P5 was 543, 331 and 192% higher (p < 0.05) than during the 5-min recovery period (2.2 ± 3.2, 2.2 ± 3.2, 2.2 ± 3.1 m/min).
Figure 3. Peak 1-, 2-, and 5-min acceleration distance and following 5-min recovery period, divided into 1-min interval. Dotted line expresses average acceleration distance for the whole match. * Significantly higher acceleration distance than average for the whole match (p < 0.01).
AccD for the most intense 1-, 2-, and 5-min periods and the following five 1-min recovery periods are shown in and . Distance covered with intense acceleration in the 5-min recovery period after P1, P2, and P5 was 78, 80, and 73% higher (p < 0.001, ES = 0.78, 0.77 and 0.67) compared to mean 5-min periods for the game, respectively.
Discussion
The main findings of our study were that we demonstrate an decrement in physical performance in all the recovery periods following all the peak periods for HSRD and SpD, but with trivial to small effect sizes. The decrement in HSRD and SpD lasted longer after longer peak periods. On the contrary, distance covered with intense acceleration was higher than match average during all five subsequent 1-min periods following peak 1-min, peak 2-min and peak 5-min periods.
The intensity during peak periods decreased with increasing epochs for all three variables in accordance with previous studies (Delaney et al., Citation2018; Fereday et al., Citation2020; Fransson et al., Citation2017; Schimpchen et al., Citation2021). The most intense 1-, 2- and 5-min period accounted for 9, 11 and 16% of the total HSRD even though only accounting for ~ 1, ~2 and ~ 6% of the playing time. Similar high percentages of total match performance were observed in the most intense 1-, 2- and 5-min period for SpD (17, 19 and 25%, respectively) and intense AccD (11, 15 and 23%, respectively) highlighting the importance of preparing for these intense periods. It should, however, be noticed that the peaks not necessarily overlap, and data can therefore not just be accumulated across variables.
Differences in physical performance have been shown between standard of play (Bradley et al., Citation2016; Mohr et al., Citation2003). In the present study, Danish Superliga players covered HSRD of 62 and 110 m during the most intense 1- and 5-min period which is similar to previously reported in English Premier League players (DiMascio & Bradley, Citation2013; Fransson et al., Citation2017; Ju et al., Citation2021), English Championship players (Fereday et al., Citation2020) and Spanish La Liga players (Oliva-Lozano, Fortes, et al., Citation2021), but higher than reported in Norwegian elite players (Baptista et al., Citation2019; Dalen et al., Citation2021) and Spanish Segunda B division players (Martin-Garcia et al., Citation2018). Schimpchen et al. (Citation2021) reported slightly higher HSRD for 1- and 5-min peak periods, but they applied slightly lower cut-off level for HSRD (5.5 vs. 5.0 m/s) than the present and abovementioned studies, which may account for some of this difference. Similar differences are observed for SpD, whereas data for intense AccD are scarce.
In the period following the most intense period, a decrement in performance was observed for HSRD and SpD as the distance covered for both variables were lower than the 1-min match average. Independent of the length of the peak period, the drop was highest in the first minute after the peak period and the distance covered increased towards match average in the following 1-min periods. This indicates that the physiological mechanisms causing fatigue during peak period have a relatively fast recovery, and the recovery kinetics from the peak period may resemble the recovery phases of muscle PCr (see for example Bowtell et al., Citation2016). Thus, it may be speculated that muscle metabolism may play a role in peak period recovery. Moreover, in a study by Mohr et al. (Citation2016) in trained male football players, one of the strongest predictors of performance during a 5-min peak period was skeletal muscle Na+-K+-ATPase protein expression. Moreover, correlations have been observed between other muscle ion regulatory proteins and peak period performance (Mohr et al.. Citation2016, Citation2022). Since the perturbations in muscle ion homeostasis are rapidly restored after high-intensity exercise (Juel et al. Citation2000; Mckenna et al. Citation1985), it may be also speculated that the temporary fatigue that occurs in short peak intervals in football could be associated with muscle ion handling capacity.
How fast the performance returned towards match average depended on the length of the peak period and thus the accumulated HSRD and SpD with slower recovery observed after the longest peak period epochs. Indeed, for HSRD in all peak period epochs and SpD in 1- and 2-min peak periods performance returned to match average within the 5-min recovery period, whereas SpD was at least 10% below match average for all 5 recovery minutes after peak 5-min. In accordance with the present study, Schimpchen et al. (Citation2021) found faster recovery after 1-min peak periods than 5-min periods for HSRD. The longer distances covered with high-intensity exercise causes greater muscle metabolic disturbances (Mohr et al., Citation2007), which demand longer recovery.
Several studies have found that the performance in the 5-min period following the most intense period is lower than match average (DiMascio & Bradley, Citation2013; Fransson et al., Citation2017; Mohr et al., Citation2005). We found 9–19% lower HSRD and 7–21% SpD during the 5-min period following the peak period with trivial to moderate effect sizes (ES = 0.07–0.36). Since the inter-individual exertion during the peak period intervals is large (HSRD range 5.9–142.4; 2.9–89.7; 1.2–51.6 m/min, SpD range 8.4–139.5; 4.7–89.8; 1.9–44.5 m/min), the trivial-to-small ES are expected because of the distribution of the data, where around 65% of the data included in the following 5 min period has the value 0. Since the many 0 values have a large impact on the Z value, it is expected that the effect sizes are trivial-to-small, since the Z value have a large influence on the calculation of effect sizes which is suggested by Fritz et al. (Citation2012). Inter-individual variability will always be present in sports such as football, so the use of effect sizes in these sports needs to be interpretated with care.
Surprisingly, intense AccD was higher in all five 1-min periods following the most intense period independent of the length of this intense period. During the 5-min recovery period players covered 73–80% more distance with intense acceleration than the average 5-min period. Thus, in contrast to findings by Akenhead et al. (Citation2013), our findings indicate that players do not experience temporary decrement in the ability to perform accelerations >3 m/s2. At least not to an extent that influence the ability to perform in accordance with the demands of the game, which also contrasts with the demands for HSRD and SpD. A study by Varley and Aughey (Citation2013) showed that less than half of the maximal accelerations (2.78 m/s2) led to runs with speed above 4.2 m/s, thus being able to cover distance with maximal acceleration may not necessarily be associated with covering distance with high running speed. Moreover, the fact that more distance was covered with intense acceleration in the recovery period than the mean match value indicates that intense accelerations are performed in clusters in the same periods of the match. Furthermore, distances covered per minute with intense acceleration were rather low both during peak periods (6–14 m/min), recovery periods (~2.2 m/min) and match average (~1–2 m/min) supporting these assumptions.
The study has strengths and limitations. Firstly, it is a strength that we apply a large data set spanning over three seasons. Secondly, we also see it as a strength that we describe the “recovery” from peak periods in 1-min periods to better understand the kinetics of the change in locomotor behaviour. Finally, we apply rolling averages instead of pre-defined fixed time-intervals. However, the study also has some limitations. For example, we do not in this study perform direct assessments of neuromuscular fatigue; however, we try in the discussion to link our findings to other studies where muscle function and high intensity performance has been assessed directly. Thus, we cannot rule out that the changes in intensity in conjunction with the different peak periods are related to changed pacing or locomotion behaviour, due to other factors than physiologically mediated fatigue. Moreover, the large data set allows for detecting small differences, which however may not be meaningful (e.g., significant different HSRD and SpD of 0% in the recovery period following 2-min peak periods) and therefore should be interpreted accordingly.
Conclusion
To our knowledge, this is the first study that provides a more detailed analysis of the recovery period after a peak intensity period in a large cohort of elite male football players. The data analysis provides insight into what impact a peak intensity period can exert on the physical performance in the match and the recovery kinetics from these intense game intervals. Furthermore, our data indicates that the duration of a peak intensity period impacts the recovery period as prolonged recovery was observed after peak 5-min both on SpD and HSRD compared to 1-min and 2-min peak. In contrast, physical performance after AccD peak periods were not negatively impacted on this performance metric as no decrements in the associated recovery period for all peak period epochs were observed for AccD.
Acknowledgments
The authors would like to thank the Danish League for providing access to the tracking data of three full seasons. Moreover, we would like to thank Mikkel Keldmann, Brøndby IF for helping with the data extraction. The authors declare no conflicts of interests.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The data for this study is owned by the Danish League (Divisionsforeningen) and the author of this paper have gained access to the data from the Danish League. The author cannot share the data but refers to the data owner for access to the data.
Additional information
Funding
References
- Akenhead, R., Hayes, P. R., Thompson, K. G., & French, D. (2013). Diminutions of acceleration and deceleration output during professional football match play. Journal of Science and Medicine in Sport, 16(6), 556–561. https://doi.org/10.1016/j.jsams.2012.12.005
- Allen, T., Taberner, M., Zhilkin, M., & Rhodes, D. (2024). Running more than before? The evolution of running load demands in the English premier league. International Journal of Sports Science & Coaching, 19(2), 779–787. https://doi.org/10.1177/17479541231164507
- Bangsbo, J., Norregaard, L., & Thorso, F. (1991). Activity profile of competition soccer. Canadian Journal of Sport Sciences = Journal Canadien Des Sciences Du Sport, 16(2), 110–116.
- Baptista, I., Johansen, D., Figueiredo, P., Rebelo, A., & Pettersen, S. A. (2019). Positional differences in peak-and accumulated-training load relative to match load in elite football. Sports, 8(1), 1. https://doi.org/10.3390/sports8010001
- Barnes, C., Archer, D. T., Hogg, B., Bush, M., & Bradley, P. S. (2014). The evolution of physical and technical performance parameters in the English premier league. International Journal of Sports Medicine, 35(13), 1095–1100. https://doi.org/10.1055/s-0034-1375695
- Bowtell, J. L., Jackman, S. R., Scott, S., Connolly, L. J., Mohr, M., Ermidis, G., Julian, R., Yousefian, F., Helge, E. W., Jørgensen, N. R., Fulford, J., Knapp, K. M., & Krustrup P. (2016). Short Duration Small Sided Football and to a Lesser Extent Whole Body Vibration Exercise Induce Acute Changes in Markers of Bone Turnover. BioMed research international. 2016, 3574258. https://doi.org/10.1155/2016/3574258
- Bradley, P. S., Archer, D. T., Hogg, B., Schuth, G., Bush, M., Carling, C., & Barnes, C. (2016). Tier-specific evolution of match performance characteristics in the English premier league: It’s getting tougher at the top. Journal of Sports Sciences, 34(10), 980–987. https://doi.org/10.1080/02640414.2015.1082614
- Bradley, P. S., Sheldon, W., Blake Wooster, B., Olsen, P., Boanas, P., & Krustrup, P. (2009). High-intensity running in English FA Premier League soccer matches. Journal of Sports Sciences, 27(2), 159–168. https://doi.org/10.1080/02640410802512775
- Bush, M., Barnes, C., Archer, D. T., Hogg, B., & Bradley, P. S. (2015). Evolution of match performance parameters for various playing positions in the English premier league. Human Movement Science, 39, 1–11. https://doi.org/10.1016/j.humov.2014.10.003
- Dalen, T., Sandmæl, S., Stevens, T. G., Hjelde, G. H., Kjøsnes, T. N., & Wisløff, U. (2021). Differences in acceleration and high-intensity activities between small-sided games and peak periods of official matches in elite soccer players. The Journal of Strength & Conditioning Research, 35(7), 2018–2024. https://doi.org/10.1519/JSC.0000000000003081
- Delaney, J. A., Thornton, H. R., Rowell, A. E., Dascombe, B. J., Aughey, R. J., & Duthie, G. M. (2018). Modelling the decrement in running intensity within professional soccer players. Science and Medicine in Football, 2(2), 86–92. https://doi.org/10.1080/24733938.2017.1383623
- DiMascio, M., & Bradley, P. S. (2013). Evaluation of the most intense high-intensity running period in English FA premier league soccer matches. The Journal of Strength & Conditioning Research, 27(4), 909–915. https://doi.org/10.1519/JSC.0b013e31825ff099
- Fereday, K., Hills, S. P., Russell, M., Smith, J., Cunningham, D. J., Shearer, D., McNarry, M., & Kilduff, L. P. (2020). A comparison of rolling averages versus discrete time epochs for assessing the worst-case scenario locomotor demands of professional soccer match-play. Journal of Science & Medicine in Sport / Sports Medicine Australia, 23(8), 764–769. https://doi.org/10.1016/j.jsams.2020.01.002
- Fransson, D., Krustrup, P., & Mohr, M. (2017). Running intensity fluctuations indicate temporary performance decrement in top-class football. Science and Medicine in Football, 1(1), 10–17. https://doi.org/10.1080/02640414.2016.1254808
- Fritz, C. O., Morris, P. E., & Richler, J. J. (2012). Effect size estimates: Current use, calculations, and interpretation. Journal of Experimental Psychology: General, 141(1), 2–18. https://doi.org/10.1037/a0024338
- Ju, W., Doran, D., Hawkins, R., Gómez-Díaz, A., Martin-Garcia, A., Ade, J., Laws, A., Evans, M., & Bradley, P. (2021). Contextualised peak periods of play in English premier league matches. Biology of Sport, 39(4), 973–983. https://doi.org/10.5114/biolsport.2022.112083
- Juel, C., Pilegaard, H., Nielsen, J. J., & Bangsbo, J. (2000). Interstitial K(+) in human skeletal muscle during and after dynamic graded exercise determined by microdialysis. American Journal of Physiology - Regulatory, Integrative, and Comparative Physiology. 278(2). https://doi.org/10.1152/ajpregu.2000.278.2.R400
- Lago- Peñpeñas, C., Lorenzo-Martinez, M., López-Del Campo, R., Resta, R., & Rey, E. (2023). Evolution of physical and technical parameters in the Spanish LaLiga 2012–2019. Science and Medicine in Football, 7(1), 41–46. https://doi.org/10.1080/24733938.2022.2049980
- Martin-Garcia, A., Casamichana, D., Diaz, A. G., Cos, F., & Gabbett, T. J. (2018). Positional differences in the most demanding passages of play in football competition. Journal of Sports Science and Medicine, 17(4), 563–570.
- McKenna, M. J., Bangsbo, J., & Renaud, J. M. (1985). K+, Na+, and Cl disturbances and Na+-K+ pump inactivation: implications for fatigue. Journal of applied physiology. 104(1). 288–295. https://doi.org/10.1152/japplphysiol.01037.2007
- Mohr, M., Fatouros, I. G., Jamurtas, A. Z., Draganidis, D., Thomassen, M., Ørntoft, C., Ermidis, G., Loules, G., Batsilas, D., Poulios, A., Papanikolaou, K., Randers, M. B., Krustrup, P., & Nybo, L. (2022). Skeletal muscle phenotype and game performance in elite women football players. Journal of medicine & science in sports. 32(1), 39–53. https://doi.org/10.1111/sms.14022
- Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match performance of high-standard soccer players with special reference to development of fatigue. Journal of Sports Sciences, 21(7), 519–528. https://doi.org/10.1080/0264041031000071182
- Mohr, M., Krustrup, P., & Bangsbo, J. (2005). Fatigue in soccer: A brief review. Journal of Sports Sciences, 23(6), 593–599. https://doi.org/10.1080/02640410400021286
- Mohr, M., Krustrup, P., Nielsen, J. J., Nybo, L., Rasmussen, M. K., Juel, C., & Bangsbo, J. (2007). Effect of two different intense training regimens on skeletal muscle ion transport proteins and fatigue development. American Journal of Physiology - Regulatory, Integrative, and Comparative Physiology. 292(4). https://doi.org/10.1152/ajpregu.00251.2006
- Mohr M., Thomassen M., Girard O., Racinais S., & Nybo L. (2016). Muscle variables of importance for physiological performance in competitive football. European journal of applied physiology, 116(2), 251–62. https://doi.org/10.1007/s00421-015-3274-x.
- Novak, A. R., Impellizzeri, F. M., Trivedi, A., Coutts, A. J., & McCall, A. (2021). Analysis of the worst-case scenarios in an elite football team: Towards a better understanding and application. Journal of Sports Sciences, 39(16), 1850–1859. https://doi.org/10.1080/02640414.2021.1902138
- Oliva-Lozano, J. M., Fortes, V., & Muyor, J. M. (2021). The first, second, and third most demanding passages of play in professional soccer: A longitudinal study. Biology of Sport, 38(2), 165–174. https://doi.org/10.5114/biolsport.2020.97674
- Oliva-Lozano, J. M., Martin-Fuentes, I., Fortes, V., & Muyor, J. M. (2021). Differences in worst-case scenarios calculated by fixed length and rolling average methods in professional soccer match-play. Biology of Sport, 38(3), 325–331. https://doi.org/10.5114/biolsport.2021.99706
- Oliva-Lozano, J. M., Rojas-Valverde, D., Gomez-Carmona, C. D., Fortes, V., & Pino-Ortega, J. (2020). Worst case scenario match analysis and contextual variables in professional soccer players: A longitudinal study. Biology of Sport, 37(4), 429–436. https://doi.org/10.5114/biolsport.2020.97067
- Schimpchen, J., Gopaladesikan, S., & Meyer, T. (2021). The intermittent nature of player physical output in professional football matches: An analysis of sequences of peak intensity and associated fatigue responses. European Journal of Sport Science, 21(6), 793–802. https://doi.org/10.1080/17461391.2020.1776400
- Varley, M. C., & Aughey, R. J. (2013). Acceleration profiles in elite Australian soccer. International Journal of Sports Medicine, 34(1), 34–39. https://doi.org/10.1055/s-0032-1316315