Recent advances of artificial intelligence, data science, sensors and wearable technologies have fostered the emergence of an increasing number of tools, services and approaches for collecting and processing sports data by promising information that can be used to improve athletes’ performance in a variety of contexts. Recent sports research publications indicate a trend on topics related to artificial intelligence (AI) and computer vision for data acquisition and analysis. However, this trend also raises questions from practitioners and researchers that are not familiar with the underlying concepts and technologies of such tools and approaches since their experience and education are mostly focused on the sports domain. Some of such questions are: what is the difference between these aforementioned technologies and techniques? What are common mistakes and caveats when trying to apply them in practice? Which of them better fits to overcome the challenges related to sports science from data collection to its analysis? With those questions in mind, this paper aims to provide an overview of the fundamental concepts of artificial intelligence and statistical data analysis, with a particular focus on their application to badminton-related challenges. We cover from classic statistics-based analysis to state-of-the-art deep learning models, addressing issues such as data acquisition and processing, and the limitations of each technique. We also discuss the interpretation of complex data outputs, since users must be aware of the limitations and potential biases of the algorithms to ensure that the insights provided by the results are relevant. We aim this knowledge can empower sports professionals to make informed decisions and effectively leverage technology to improve athletic performance and sports organizations.
Dorini et al. (Sun,) studied this question.