Loading Score Dashboard for High-Performance Managers
Project Background
With the abundance of GPS metrics available for analysis, sports scientists and high-performance staff face the challenge of accurately prescribing, delivering, and modifying training sessions. This dashboard was designed to provide clarity, offering a snapshot of the squad's current loading status while enabling staff to pinpoint its source. It allows for both squad and individual viewing, with the flexibility to analyse single days or track trends over a custom time period.
A key component of this dashboard is the “Loading Score”, a measure used to assess whether an athlete is experiencing an optimal level of training load to improve performance while reducing the risk of injury. A lower loading score indicates that an athlete has a well-balanced and stable workload, whereas a higher loading score suggests instability in training load, placing the athlete at an increased risk of injury if they were to compete. Risk appetite varies depending on the time of year. For example, in a Grand Final, the team may be willing to select a key player with a high loading score due to their importance to the squad. However, in Round 1, a more conservative approach would likely be taken to prioritise seasson availability.
This dashboard is designed for the High-Performance Manager, in collaboration with the medical and physiotherapy team, to manage training plans effectively and identify key contributors to high loading scores at both the squad and individual levels.
Stakeholders
High-Performance Manager
Strength & Conditioning coaches
Physiotherapists
Medical Team (Club Doctor, Welfare)
Selection Committee (Head Coach, Assistant Coaches, Head of Football)
Data Structure
This is a real-world dataset consisting of 132 GPS metrics and has been randomised to protect athlete privacy. Only 10 columns were utilised for this project, with an additional 24 columns created using SQL to create the “Loading Score” logic. The final dataset, ‘gps_cleaned’ contains 1,965 rows × 34 columns. To protect intellectual property, the full list of columns, and logic behind them, is not disclosed. The dataset follows the structure outlined below:
Data Transformation & Manipulation
This process used window functions, conditional logic, and rolling aggregations to compute custom "Loading Score" metrics. Speed thresholds were established as percentage-based targets, with new columns identifying whether the current training session met these thresholds, the most recent date they were achieved, and the number of days since the last achievement.
Velocity thresholds and workload ratios were then assigned a loading Score on a scale from 0 to 3, with the sum of these representing the athlete's total “Loading Score”.
Data Visualization
Using Tableau, action filters, dynamic parameters, and custom calculations create an interactive timeline with adjustable date ranges and data smoothing options. Stakeholders can click and drag to select data points, refine the date range with a slider, and adjust smoothing levels via a drop-down menu. Embedded charts in tooltips are included to highlight the specific sources of loading scores.
For this project, the scenario question is used:
"During which time periods were our athletes best and least prepared in terms of loading, and what sources of training contributed to this?"
Overview of findings
Athletes were best prepared between February 10-22, where the Loading Score was at its lowest (1.46) due to the controlled pre-season phase, with no match demands and a more controlled training environment. In contrast, April 20 until May 3 was the least prepared period, with the highest Loading Score (2.6) driven by match unpredictability and reconditioning of injured players. The peak single day Loading Score (3.67) on April 26 was influenced by 95% Max Velocity, Decelerations, High-Speed Running, and Sprint Efforts.
Lowest Loading Score Period
Between February 10 and 22, the squad’s 7-day moving average Loading Score reached its lowest point of the season at 1.46.
Margot Robbie and Nick Daicos were the only two players who recorded individual Loading Scores exceeding 7.6. Their elevated scores were primarily driven by 85%, 90%, and 95% Max Velocity efforts.
All other GPS metrics contributing to Loading Scores remained minimal during this period
This period fell within the pre-season phase, where training loads were highly controlled due to the absence of matches. Match simulation and drills were adjustable, ensuring that planned and executed loads remained closely aligned using historical drill data.
No players were injured during this phase, meaning there were no heightened reloading periods or spikes in workload.
Highest Loading Score Period
Between April 20 and May 3, the squad’s 7-day moving average Loading Score peaked at 2.6, signifying the highest loading period of the season.
This period falls within the competition phase, where match loads are unpredictable, and external factors such as injuries, soreness, and recovery constraints limit the ability to execute controlled training at specific intensities.
Several players during this period were returning from injury and underwent increased training loads to meet match demands, increasing the overall squad loading scores.
The squad’s highest single day Loading Score was 3.67, recorded on April 26.
The most significant contributing factor was 95% Max Velocity, with additional factors including 90% Max Velocity, Decelerations, High-Speed Running, and Sprint Efforts.
On this day, Lewis Hamilton, J.K. Rowling, Mitchell Starc, and Nick Daicos recorded the highest individual Loading Scores, all exceeding 7.6, signifying a heightened injury risk.
Recommendations
Based on the uncovered insights, the following recommendations have been provided:
Live GPS tracking and load-monitoring during training sessions should prioritise 95% Max Velocity, Decelerations, High-Speed Running, and Sprint Efforts, as these metrics contribute the most to loading scores.
Players returning from injury should receive increased support from sports scientists to ensure they are reloading within safe ranges to minimise reinjury risk.
Athletes should be educated on the impact of high velocity efforts on loading scores, as the primary concern is not excessive exposure but rather insufficient exposure, which increases the risk of soft-tissue injuries.