Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon

PurposeThis study introduces two-dimensional (2D) Kernel Density Estimation (KDE) plots as a novel tool for visualising Training Intensity Distribution (TID) in biathlon. The goal was to assess how KDE plots, alongside traditional training metrics, might provide a more detailed understanding of hear...

Full description

Saved in:
Bibliographic Details
Main Authors: Craig A. Staunton, Andreas Kårström, Hannes Kock, Marko S. Laaksonen, Glenn Björklund
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-06-01
Series:Frontiers in Sports and Active Living
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fspor.2025.1546909/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849421419757174784
author Craig A. Staunton
Craig A. Staunton
Andreas Kårström
Andreas Kårström
Hannes Kock
Hannes Kock
Marko S. Laaksonen
Glenn Björklund
author_facet Craig A. Staunton
Craig A. Staunton
Andreas Kårström
Andreas Kårström
Hannes Kock
Hannes Kock
Marko S. Laaksonen
Glenn Björklund
author_sort Craig A. Staunton
collection DOAJ
description PurposeThis study introduces two-dimensional (2D) Kernel Density Estimation (KDE) plots as a novel tool for visualising Training Intensity Distribution (TID) in biathlon. The goal was to assess how KDE plots, alongside traditional training metrics, might provide a more detailed understanding of heart rate (HR) intensity patterns, aiding in the evaluation of training quality and compliance.MethodsFifteen elite-level youth biathletes from two national academy programmes were monitored over 5–6 weeks using HR monitors. Training sessions were measured via time-in-zone (TIZ) within a five-zone HR model with any time accumulated below the threshold for Zone 1, considered Zone 0. Sessions were dichotomised into those planned as low-intensity training (LIT) or those planned with high-intensity training (HIT). KDE analyses were conducted in MATLAB (Version R2020b) using the “ksdensity” function to create 2D KDE plots that visualise HR intensity accumulation across each programme, session type (e.g., Low-intensity training: LIT; High-intensity training: HIT), and individual athlete responses. Traditional histogram plots and grouped bar charts were also used for comparison.ResultsFor LIT sessions, athletes performed less time in Zone 1 than planned, while performed time exceeded planned time in Zone 2. For HIT sessions, performed time in Zone 5 was lower than planned. All sessions contained unplanned time in Zone 0. The 2D KDE plots provided a continuous and detailed representation of HR intensity accumulation throughout training sessions, revealing patterns and intensity fluctuations that complement traditional TIZ analyses.Conclusions2D KDE plots might serve as a valuable complementary tool for assessing TID in biathlon, offering a more nuanced and continuous view of HR intensity. By identifying discrepancies between planned and performed training intensity, coaches can refine strategies and provide individualised feedback. Incorporating KDE plots into training monitoring could improve training alignment, helping reduce overtraining or undertraining risks and optimising athlete development.
format Article
id doaj-art-f6772ee4babc418c8e80dcaf04dc82c7
institution Kabale University
issn 2624-9367
language English
publishDate 2025-06-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Sports and Active Living
spelling doaj-art-f6772ee4babc418c8e80dcaf04dc82c72025-08-20T03:31:27ZengFrontiers Media S.A.Frontiers in Sports and Active Living2624-93672025-06-01710.3389/fspor.2025.15469091546909Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlonCraig A. Staunton0Craig A. Staunton1Andreas Kårström2Andreas Kårström3Hannes Kock4Hannes Kock5Marko S. Laaksonen6Glenn Björklund7Department of Health Sciences, Swedish Winter Sports Research Centre, Mid Sweden University, Östersund, SwedenDepartment of Environmental and Bioscience, School of Business, Innovation and Sustainability, Halmstad University, Halmstad, SwedenDepartment of Health Sciences, Swedish Winter Sports Research Centre, Mid Sweden University, Östersund, SwedenSwedish Biathlon Federation, Östersund, SwedenDepartment of Health Sciences, Swedish Winter Sports Research Centre, Mid Sweden University, Östersund, SwedenDepartment of Endurance Sports, Institute for Applied Training Science, Leipzig, GermanyDepartment of Health Sciences, Swedish Winter Sports Research Centre, Mid Sweden University, Östersund, SwedenDepartment of Health Sciences, Swedish Winter Sports Research Centre, Mid Sweden University, Östersund, SwedenPurposeThis study introduces two-dimensional (2D) Kernel Density Estimation (KDE) plots as a novel tool for visualising Training Intensity Distribution (TID) in biathlon. The goal was to assess how KDE plots, alongside traditional training metrics, might provide a more detailed understanding of heart rate (HR) intensity patterns, aiding in the evaluation of training quality and compliance.MethodsFifteen elite-level youth biathletes from two national academy programmes were monitored over 5–6 weeks using HR monitors. Training sessions were measured via time-in-zone (TIZ) within a five-zone HR model with any time accumulated below the threshold for Zone 1, considered Zone 0. Sessions were dichotomised into those planned as low-intensity training (LIT) or those planned with high-intensity training (HIT). KDE analyses were conducted in MATLAB (Version R2020b) using the “ksdensity” function to create 2D KDE plots that visualise HR intensity accumulation across each programme, session type (e.g., Low-intensity training: LIT; High-intensity training: HIT), and individual athlete responses. Traditional histogram plots and grouped bar charts were also used for comparison.ResultsFor LIT sessions, athletes performed less time in Zone 1 than planned, while performed time exceeded planned time in Zone 2. For HIT sessions, performed time in Zone 5 was lower than planned. All sessions contained unplanned time in Zone 0. The 2D KDE plots provided a continuous and detailed representation of HR intensity accumulation throughout training sessions, revealing patterns and intensity fluctuations that complement traditional TIZ analyses.Conclusions2D KDE plots might serve as a valuable complementary tool for assessing TID in biathlon, offering a more nuanced and continuous view of HR intensity. By identifying discrepancies between planned and performed training intensity, coaches can refine strategies and provide individualised feedback. Incorporating KDE plots into training monitoring could improve training alignment, helping reduce overtraining or undertraining risks and optimising athlete development.https://www.frontiersin.org/articles/10.3389/fspor.2025.1546909/fullbig datadata sciencedensity estimationheart rateNordic skiingtraining load
spellingShingle Craig A. Staunton
Craig A. Staunton
Andreas Kårström
Andreas Kårström
Hannes Kock
Hannes Kock
Marko S. Laaksonen
Glenn Björklund
Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon
Frontiers in Sports and Active Living
big data
data science
density estimation
heart rate
Nordic skiing
training load
title Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon
title_full Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon
title_fullStr Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon
title_full_unstemmed Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon
title_short Kernel Density Estimation: a novel tool for visualising training intensity distribution in biathlon
title_sort kernel density estimation a novel tool for visualising training intensity distribution in biathlon
topic big data
data science
density estimation
heart rate
Nordic skiing
training load
url https://www.frontiersin.org/articles/10.3389/fspor.2025.1546909/full
work_keys_str_mv AT craigastaunton kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT craigastaunton kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT andreaskarstrom kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT andreaskarstrom kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT hanneskock kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT hanneskock kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT markoslaaksonen kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon
AT glennbjorklund kerneldensityestimationanoveltoolforvisualisingtrainingintensitydistributioninbiathlon