This file includes my code from working with a sports collaborator from CU Boulder. The original data was taken by the sports collaborator using Data Volley. This data set had over 200,000 points of data and over 120 variables. Due to the data set including sensitive information including players' names, this code will not produce any results when running. This data was collected for the women's volleyball seasons of 2017, 2018, 2019, 2021, and 2022. This project was in collaboration with Arnaldo Bayona (PhD candidate at CU Boulder), and domain expert Eduardo Fiallos (Sports Collaborator at CU Boulder).
Python Code: The Python code shown was used to clean the data efficiently. Part of this was to handle missing data, identify non-CU Boulder players, and estimate percentages of their teams' receptions and attacks that specific players are responsible for.
R Code: The R code was used for data analysis and hypothesis testing. One of the positions we were testing for was the Outside Hitter position, labeled OH. There were three questions that we set out to answer.
- Does the ability to win a set or match depend on how often the outside hitter receives the ball?
- Does the load imposed on outside hitters make it more difficult for them to do their jobs effectively?
- What kinds of groupings will the data show based on two contacts: receptions and attacks?
In summary, the results of this work showed no evidence that the physical load imposed on outside hitters influences the team's ability to win a volleyball set or match. We also found that the distribution of outside hitters encountered helps to compare the performance of each outside hitter with the overall performance of all outside hitters (regardless of team) that were examined.