This project is a design study of an analytical tool for lead climbing performances, created in collaboration with the French Federation of Climbing (FFME). Its goal is to facilitate the exploration of a dataset to identify new performance indicators and better train French athletes. Regular interviews enabled us to identify the analysts’ needs and design a robust system they can access constantly. A thematic analysis of transcripts from two evaluation days with the national coach highlighted the importance of accessing performance videos to view key passages, the need to navigate between several levels of information and the value of relying on qualitative data to better understand athletes’ choices. We detail our design choices and outline the solutions implemented to inform the design of future sports performance analysis systems.
2024
SportsUser-Centered DesignDesign Study
See also Climbing Annotations
The tool uses quantitative and qualitative data produced manually using a video annotation tool that relies on videos captured during training and competition sessions. All the data visualizations rely on this data.
Types of data analyses available Through interviews with analysts and coaches, we identified several types of analyses they were interested in. They wanted to focus on 3 primary types, with others uncovered later in the project by actually using the tool. Analysts need to compare athletes' profiles over routes and competitions which is challenging because in lead climbing the routes are always unique. Then need to compare performances of an athlete over multiple competitions. They need to compare athletes' performances over a single route (e.g., final round of a competition).
Interactive analytical tool From these interviews, and after having implemented early prototypes experts could play with, we conducted a thematic analysis on the transcripts and identified significant themes in expert data analyses, e.g., on the critical functionalities to implement, on how they analyze data, on the data set they deal with and on the potential impact of the tool. We built and refined the analytical tool using user-centered design methods over a year, and present an overview below.
Design implications We formulated design implications based on our observation of experts using the tool. One implication was that experts need to have access to overviews and details on the data analyzed, and interaction means should support them in navigating between various levels of information. To support this, we enabled fast access to specific sequences in the videos of the performances by clicking on data points (see image on the right). All details are available in the paper (written in French).
Etude de conception d'un outil d'analyse de données d'escalade de difficulté Bruno Fruchard, Timo Maszewski, Cécile Avezou IHM'24: 35e conférence internationale francophone sur l'Interaction Humain-Machine
