Relationships Between the External and Internal Training Load in Professional Soccer: What Can We Learn From Machine Learning?

in International Journal of Sports Physiology and Performance
Restricted access

Purchase article

USD $24.95

Student 1 year subscription

USD $107.00

1 year subscription

USD $142.00

Student 2 year subscription

USD $203.00

2 year subscription

USD $265.00

Purpose: Machine learning may contribute to understanding the relationship between the external load and internal load in professional soccer. Therefore, the relationship between external load indicators (ELIs) and the rating of perceived exertion (RPE) was examined using machine learning techniques on a group and individual level. Methods: Training data were collected from 38 professional soccer players over 2 seasons. The external load was measured using global positioning system technology and accelerometry. The internal load was obtained using the RPE. Predictive models were constructed using 2 machine learning techniques, artificial neural networks and least absolute shrinkage and selection operator (LASSO) models, and 1 naive baseline method. The predictions were based on a large set of ELIs. Using each technique, 1 group model involving all players and 1 individual model for each player were constructed. These models’ performance on predicting the reported RPE values for future training sessions was compared with the naive baseline’s performance. Results: Both the artificial neural network and LASSO models outperformed the baseline. In addition, the LASSO model made more accurate predictions for the RPE than did the artificial neural network model. Furthermore, decelerations were identified as important ELIs. Regardless of the applied machine learning technique, the group models resulted in equivalent or better predictions for the reported RPE values than the individual models. Conclusions: Machine learning techniques may have added value in predicting RPE for future sessions to optimize training design and evaluation. These techniques may also be used in conjunction with expert knowledge to select key ELIs for load monitoring.

Jaspers and Helsen are with the Movement Control & Neuroplasticity Research Group, Dept of Movement Sciences; De Beéck and Davis, the Dept of Computer Science; and Staes, the Musculoskeletal Rehabilitation Research Group, Dept of Rehabilitation Sciences, KU Leuven, Leuven, Belgium. Brink and Frencken are with the Center for Human Movement Sciences, University Medical Center, University of Groningen, Groningen, the Netherlands. Jaspers and De Beéck share first authorship, and Davis and Helsen share last authorship.

Jaspers (arne.jaspers@kuleuven.be) is corresponding author.
International Journal of Sports Physiology and Performance
Article Sections
References
  • 1.

    Akenhead RNassis GP. Training load and player monitoring in high-level football: current practice and perceptions. Int J Sports Physiol Perform. 2016;11(5):587593. PubMed doi:10.1123/ijspp.2015-0331

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2.

    Impellizzeri FMRampinini EMarcora SM. Physiological assessment of aerobic training in soccer. J Sports Sci. 2005;23(6):583592. PubMed doi:10.1080/02640410400021278

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3.

    Gabbett TJ. The training-injury prevention paradox: should athletes be training smarter and harder? Br J Sports Med. 2016;50(5):273280. PubMed doi:10.1136/bjsports-2015-095788

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4.

    Impellizzeri FMRampinini ECoutts AJSassi AMarcora SM. Use of RPE-based training load in soccer. Med Sci Sports Exerc. 2004;36(6):10421047. PubMed doi:10.1249/01.MSS.0000128199.23901.2F

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5.

    Drew MKCook JFinch CF. Sports-related workload and injury risk: simply knowing the risks will not prevent injuries. Br J Sports Med. 2016;50:13061308. doi:10.1136/bjsports-2015-095871

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6.

    Gaudino PIaia FMStrudwick AJet al. Factors influencing perception of effort (session rating of perceived exertion) during elite soccer training. Int J Sports Physiol Perform. 2015;10(7):860864. PubMed doi:10.1123/ijspp.2014-0518

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7.

    Lovell TWSirotic ACImpellizzeri FMCoutts AJ. Factors affecting perception of effort (session rating of perceived exertion) during rugby league training. Int J Sports Physiol Perform. 2013;8(1):6269. PubMed doi:10.1123/ijspp.8.1.62

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8.

    Scott BRLockie RGKnight TJClark ACJanse de Jonge XA. A comparison of methods to quantify the in-season training load of professional soccer players. Int J Sports Physiol Perform. 2013;8(2):195202. PubMed doi:10.1123/ijspp.8.2.195

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9.

    Bartlett JDO’Connor FPitchford NTorres-Ronda LRobertson SJ. Relationships between internal and external training load in team sports athletes: evidence for an individualized approach. Int J Sports Physiol Perform. 2017;12(2):230234. PubMed doi:10.1123/ijspp.2015-0791

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10.

    Bishop C. Pattern Recognition and Machine Learning (Information Science and Statistics). 1st ed. New York, NY: Springer; 2007.

  • 11.

    Malone JJLovell RVarley MCCoutts AJ. Unpacking the black box: applications and considerations for using GPS devices in sport. Int J Sports Physiol Perform. 2017;12(suppl 2):218226. PubMed doi:10.1123/ijspp.2016-0236

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Varley MCGabbett TAughey RJ. Activity profiles of professional soccer, rugby league and Australian football match play. J Sports Sci. 2014;32(20):18581866. PubMed doi:10.1080/02640414.2013.823227

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13.

    Foster CFlorhaug JAFranklin Jet al. A new approach to monitoring exercise training. J Strength Cond Res. 2001;15(1):109115. PubMed

  • 14.

    Malone JJDi Michele RMorgans RBurgess DMorton JPDrust B. Seasonal training-load quantification in elite English premier league soccer players. Int J Sports Physiol Perform. 2015;10(4):489497. PubMed doi:10.1123/ijspp.2014-0352

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15.

    Barrett SMidgley ALovell R. PlayerLoad™: reliability, convergent validity, and influence of unit position during treadmill running. Int J Sports Physiol Perform. 2014;9(6):945952. PubMed doi:10.1123/ijspp.2013-0418

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16.

    Spencer MLawrence SRechichi CBishop DDawson BGoodman C. Time–motion analysis of elite field hockey, with special reference to repeated-sprint activity. J Sports Sci. 2004;22(9):843850. PubMed doi:10.1080/02640410410001716715

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 17.

    Austin DJGabbett TJJenkins DJ. Repeated high-intensity exercise in a professional rugby league. J Strength Cond Res. 2011;25(7):18981904. PubMed doi:10.1519/JSC.0b013e3181e83a5b

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Tibshirani R. Regression shrinkage and selection via the lasso. J R Stat Soc B. 1996;58:267288.

  • 19.

    Meinshausen NBühlmann P. Stability selection. J R Stat Soc B. 2010;72(4):417473. doi:10.1111/j.1467-9868.2010.00740.x

  • 20.

    Hopkins WGMarshall SWBatterham AMHanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc. 2009;41(1):313. PubMed doi:10.1249/MSS.0b013e31818cb278

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21.

    Hopkins WG. A scale of magnitudes for effect statistics. In: A New View of Statistics. 2002. . Accessed November 13 2017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    McKinney W. Data structures for statistical computing in Python. Proceedings of the 9th Python in Science Conference; June 28 2010–July 03 2010; Austin TX. . Accessed November 13 2017.

    • Crossref
    • Export Citation
  • 23.

    Pedregosa FVaroquaux GGramfort Aet al. Scikit-learn: machine learning in Python. J Mach Learn Res. 2011;12:28252830.

  • 24.

    Nédélec MMcCall ACarling CLegall FBerthoin SDupont G. Recovery in soccer: part I—post-match fatigue and time course of recovery. Sports Med. 2012;42(12):9971015.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25.

    Lindstedt SLLaStayo PCReich TE. When active muscles lengthen: properties and consequences of eccentric contractions. News Physiol Sci. 2001;16(6):256261.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26.

    Coutts AJKempton TSullivan CBilsborough JCordy JRampinini E. Metabolic power and energetic costs of professional Australian football match-play. J Sci Med Sport. 2015;18(2):219224. PubMed doi:10.1016/j.jsams.2014.02.003

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27.

    Brito JHertzog MNassis GP. Do match-related contextual variables influence training load in highly trained soccer players? J Strength Cond Res. 2016;30(2):393399. PubMed doi:10.1519/JSC.0000000000001113

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 28.

    McLaren SJSmith ASpears IRWeston M. A detailed quantification of differential ratings of perceived exertion during team-sport training. J Sci Med Sport. 2017;20(3):290295. PubMed doi:10.1016/j.jsams.2016.06.011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29.

    Weston MSiegler JBahnert AMcBrien JLovell R. The application of differential ratings of perceived exertion to Australian football league matches. J Sci Med Sport. 2015;18(6):704708. PubMed doi:10.1016/j.jsams.2014.09.001

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 30.

    Arcos ALYanci JMendiguchia JGorostiaga EM. Rating of muscular and respiratory perceived exertion in professional soccer players. J Strength Cond Res. 2014;28(11):32803288. PubMed doi:10.1519/JSC.0000000000000540

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 31.

    Vanrenterghem JNedergaard NJRobinson MADrust B. Training load monitoring in team sports: a novel framework separating physiological and biomechanical load-adaptation pathways. Sports Med. 2017;47(11):21352142. PubMed doi:10.1007/s40279-017-0714-2

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 32.

    Gallo TFCormack SJGabbett TJLorenzen CHPre-training perceived wellness impacts training output in Australian football players. J Sports Sci. 2016;34(15):14451451. PubMed doi:10.1080/02640414.2015.1119295

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 33.

    Saw AEMain LCGastin PB. Monitoring the athlete training response: subjective self-reported measures trump commonly used objective measures: a systematic review. Br J Sports Med. 2016;50(5):281291. PubMed doi:10.1136/bjsports-2015-094758

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34.

    Akubat IBarrett SAbt G. Integrating the internal and external training loads in soccer. Int J Sports Physiol Perform. 2014;9(3):457462. PubMed doi:10.1123/ijspp.2012-0347

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 35.

    Buchheit MCholley YLambert P. Psychometric and physiological responses to a pre-season competitive camp in the heat with a 6-hour time difference in elite soccer players. Int J Sports Physiol Perform. 2015;11(2):176181. PubMed doi:10.1123/ijspp.2015-0135

    • Crossref
    • Search Google Scholar
    • Export Citation
Article Metrics
All Time Past Year Past 30 Days
Abstract Views 244 244 56
Full Text Views 19 19 6
PDF Downloads 15 15 6
Altmetric Badge
PubMed
Google Scholar
Cited By