This post discusses our new open access research article on data analytics practices and reporting strategies in football worldwide.
I'm excited to share our latest open-access research, published in Science and Medicine in Football. Led by Antonio Dello Iacono, this project with Naomi Datson, Mathieu Lacome, Adam Sullivan, Tzlil Shushan, and myself, explores how football clubs worldwide use data analytics. Interestingly, this study stemmed from a simple coffee chat at a conference—highlighting the value of informal discussions!
This post summarises key findings, but I encourage you to read the full article here.
Despite the surge in data collection, football still lags in data integration, with expertise in analytics described as ‘embryonic’ (Ward et al., 2019; Gregson et al., 2022). To better understand the landscape, we conducted a cross-sectional survey of practitioners in senior, professional, and semi-professional football, identifying trends, disparities, and opportunities.
We explored six key domains:
Who – Who is applying data analytics?
What – What evidence, metrics, and approaches are used?
Where – Where is data collected and stored?
Content – What data is being analysed?
Target – How is data reported and shared?
Live – Is live data used in decision-making?
For detailed methodology and survey instructions, see the full text.
Who: Who is Applying Data Analytics in Football?
A total of 206 respondents were included in the final analysis. Of these, 176 (86%) were male practitioners, 26 (13%) were female, and 2 (1%) preferred not to disclose their gender. Respondents represented all six of FIFA’s confederations, with the majority (68%) working with football organisations based in Europe (UEFA).
Most held postgraduate qualifications and had 10+ years of experience, primarily in top-tier leagues. Most respondents worked in the performance department (73%), while data (18%) and medical (9%) departments were less represented.
While the industry is evolving, women remain underrepresented in football analytics. Only 13% of respondents were female, and few worked exclusively with female players. While the sex gap disparities are expected to narrow in the future due to substantial investments made by governing bodies in women’s football (Beissel et al. 2024), the sport remains a male-dominated industry, with significantly lower funding available in women’s football to date and fewer female employees in medical, performance, and data roles compared to their male counterparts (Luteberget et al. 2021; Bryan 2022). This is something I’m personally passionate about addressing through the female practitioner network.
What: What Evidence and Metrics Inform Decision-Making?
Interestingly, scientific literature was the least preferred source of guidance among practitioners, with many relying instead on peer discussions, professional experiences, and in-house projects. Although it should be noted that scientific research can and probably does, influence such discussions and projects as well. Nonetheless, this disconnect suggests that research findings often fail to translate directly into applied settings.
While the majority of performance and medical staff reported seeking guidance from the professional or industry community, respondents in data-related roles were more reliant on in-house projects. With that in mind, collaboration between performance and medical staff with data departments can support the influence of both industry views and in-house projects.
When assessing training load, locomotive and perceptual measures were the most commonly used metrics, regardless of role. Performance testing primarily occurred in field- and gym-based environments, offering pragmatic, non-invasive, and time-efficient solutions compared to lab-based assessments. We discussed submaximal fitness tests as an illustration of such practical solutions.
For analytical approaches, exploratory data analysis was the most frequently used method, whereas modelling, forecasting, and prediction were least utilised. These findings are consistent with two other surveys on load monitoring practices in elite men’s and women’s football (Luteberget et al. 2021; Houtmeyers et al. 2021), which revealed that only a minority of surveyed practitioners reported using machine learning techniques for data analysis.
I've often discussed on the blog the rapid evolution of data and technology in sports science. Within this context, the data analytics literacy required has also rapidly increased. Therefore, enhancing analytics literacy remains a key development area.
Where and What Data (Content) is Collected?
Data collection strategies varied based on resources and organisational structures. Performance staff typically led analytics, with medical and data departments playing a lesser role, particularly in lower-tier clubs.
Most practitioners (73%) used multiple storage solutions, with cloud-based systems and in-house storage being the most common, and often used in tandem. Elite clubs and national teams often outsourced data analytics or relied on commercial software (right), prioritising ease of use over fully customised solutions.
Word clouds published in the article's supplementary files illustrate the choices across i) cloud-based systems (Supplementary File 6), analytical company services (Supplementary File 7: above right), software used for aggregating and analysing (Supplementary File 8), software used for reporting and visualisation tasks (Supplementary File 9: right), and visualisation methods (Supplementary File 10).
While spreadsheets were the most widely used tool for reporting and visualisation (n = 157, 76%), the data highlights the prominence of tools such as Power BI, Tableau, R, and Python. This is precisely why I've collaborated with Sport Horizon on a course combining load monitoring with Power BI visualisation. For more information, including an early bird discount, sign up here.
Summary tables were the most common choice for reports and visualisation strategies (n = 189; 92%). The bar plot below illustrates the frequency of each visualisation method.
Target: How is Data Reported and Shared?
Data reporting was most frequently conducted daily or weekly, supporting training planning, monitoring, evaluation, and player availability. However, some differences emerged. National teams were more likely to use monthly reports, presumably due to less frequent service provision (Buchheit & Dupont, 2018). In lower-tier clubs, financial and staffing constraints likely contribute to reduced reporting frequency.
Live: Is Live Data Reported?
Most respondents from top-tier clubs and national teams (76–79%) reported using live data to adjust training at both squad and individual levels, compared to 54% in third-tier or lower clubs. These differences likely stem from resource availability. National teams and higher-tier clubs have the staff and technology to enable real-time adjustments, whereas lower-tier clubs, often semi-professional or developmental, operate with fewer resources, limiting their capacity for live monitoring.
Barriers and Future Directions?
Several barriers to effective data analytics in football were identified:
Financial constraints, particularly in women’s football, limiting access to advanced tools and expertise.
Organisational structures that restrict collaboration between performance, medical, and data teams.
Limited analytics literacy, reducing the adoption of predictive and inferential statistical approaches.
To advance data analytics in football, these findings suggest clubs and national teams should consider prioritising:
Investment in staff education to improve data literacy.
Integration of interdisciplinary teams, combining performance, medical, and data expertise.
Bridging the gap between research and practice, ensuring that scientific findings inform real-world decision-making.
It's important to note that despite the survey’s broad reach, the use of a convenience sample may not fully represent football practitioners across different roles and regions. While the sample was large (n = 206) and geographically diverse (all six FIFA confederations), most respondents were male (86%) and based in Europe (68%).
Final Thoughts
We hope this study offers valuable insights into the current state of data analytics in elite football. While clubs and national teams are increasingly adopting data-informed approaches, there remains significant variability in practices, methodologies, and reporting strategies. Addressing these challenges through education, collaboration, and research translation will be key to maximising the impact of data analytics in football.
A big thank you to everyone who participated in the survey. I also want to acknowledge my co-authors and especially our lead researcher, Antonio Dello Iacono.
You can find the full text and its supplementary files here.
Komentarze