Data Science for Sports Injuries Using R, Python, and Weka

Sports analytics is a new field in data science which promises to revolutionise the world of sports. The use of data to study and predict injuries has come into the front of research in the last few years and can completely change the game for team and individual sports alike. This course deals with the use of data and advanced techniques in injury prevention and treatment. Some of the issues discussed are: 1) How can we assess the importance of a factor towards injury? 2) How can we predict injuries before they take place? 3) How can we predict recovery after an injury has taken place? 4) How should data be recorded in order to analyse the relationship with injuries? 5) How can we deliver these models in a way that can aid decision making within a club? This course focuses on football (soccer), but the lessons taught also apply for other team (and individual sports). The course targeted towards sports scientists, data scientists and medical practitioners. The tools used are R, Python (the most popular computer languages for data science) and Weka (a GUI tool for machine learning, useful for those who do not want to delve in coding). The course explains all the most important concepts in statistics and machine learning and how these relate to injury prediction and exposes different use cases based on real-world examples, where data is analysed in order to aid injury-related decision making within a soccer team. The course outlines each single step in the solving the problem, from defining the problem, to the analysis, to presenting results. Professionals from a non-technical background (sports scientists and medical practitioners) will benefit the most from the high-level explanation of the technical concepts and the tutorial on the Weka software, which does not require any coding skills. Data scientists will benefit the most from the exposition and analysis of the use cases. The use cases replicate a real-world setting, where the data scientist has to deal with trade-offs, conflicting results and multiple stakeholders. This course should also interest any individual or sports club that wants to learn how to use data in order to reduce injury incidence. The course will also be of wider interest to anyone who is willing to learn how to use data science in a real world setting, as it exposes a complete overview of the basic concepts in data science, the most popular tools for data science (R, Python and Weka) and real-world scenarios, similar to those that are met in practice in many organisations. Skills learned: Data collection, data wrangling and manipulation, statistical analysis, survival analysis, predictive modelling, machine learning Tools used: R, Python, Weka Soft skills: Communication skills, learn about the intricacies of sports data and working with sports clubs Syllabus 1. Outline a. Audience b. Objectives c. Tools 2. Data in sports clubs a. Segregation, subjective opinion and lack of standards b. Culture and organization c. Data protocols and spurious relationships d. Data problems 3. Aiding decision making a. Basic principles b. Data limitations and concept limitations c. Aiding decision making: basic principles d. Statistics vs Machine Learning e. Presenting results 4. Model testing and metrics a. Metrics overview b. Statistical metrics c. Machine learning model validation 5. Use case 1: Survival analysis a. A quick introduction to R b. Survival analysis i. Current uses in sports ii. Types of models c. Dataset i. Exploratory analysis d. Analysis e. Results f. Summary and exercises 6. Use case 2: Injury prediction based on exposure records a. Problem introduction b. Introduction to Weka i. Using Weka ii. Feature selection c. Analysis d. Results e. Summary and exercises 7. Use case 3: Predicting the recovery time after an injury a. A quick introduction to Python for data analysis b. Problem introduction c. Analysis d. Results e. Interpreting results i. Understanding the results ii. Communicating the results to the club f. Summary and exercises 8. Summary