Analysis of Valpo Women\u27s Soccer Heart Rate Data

Prompted by the head coach, we would like to help our NCAA Division One Women’s Soccer Team use the data gathered from their heart rate monitors. We want to know how types of information recorded correlate with each other to aid in developing optimal rest and recovery. The questions we wish to answer include: “What variables contribute to trimp score?”, “Is QRT score predictive of performance?”, and “Are previous training sessions predictive of performance?” Data are provided by head coach Marovich and span several years. As technology improved over the years, the data collected became more complex, so only data from the fall season of 2017 was retained to ensure completeness of records for all variables of interest. The cleaned dataset contained 3629 instances and 43 variables, and analysis was conducted in Excel and R. From analysis it was found that there is a strong positive correlation between average heart rate and trimp score (r = .9744), total calories burned and trimp score (r = .9510), and total minutes played and trimp score (r = .8372). These results are fairly intuitive; A higher average heart rate, as well as the other independent variables, imply strenuous activity, so the workload would also be higher. Further results show it was difficult to predict game outcome and performance from previous training sessions and QRT scores. From this data, we recommend that coaches focus mainly on average heart rate when determining workload to maximize player performance throughout the season. Keywords: soccer, modeling, trimp score, QRT, heart rate, training effect, heart rate monitor