Performance Metrics And Game Outcomes In Division I Women's Soccer: A Sport Analytics Case Study

Soccer is the world’s most popular sport; however, collegiate women’s soccer is underrepresented in sport analytics. Analysis of sport data, including players’ external load data and match outcome, is crucial when evaluating performance, planning practices/periodization, and preventing injuries. Common Key Performance Indicators (KPIs) in soccer include Total Distance (TD; meters), Sprint Distance (SD; meters), Top Speed (TS; meters per second), and Total Player Load (TPL; unitless composite score). PURPOSE: To determine if a relationship exists between wins/losses and team TD, SD, TS, and TPL data in a women’s division I soccer team for the 2022 season. METHODS: An NCAA division I female soccer team (n = 27) participated. Data on TD, SD, TS, and TPL were collected via Global Positioning System (GPS) monitors (Vector S7; Catapult Sports) during the 2022 season (17 matches; six wins, eight losses). The statistical analysis, facilitated by Python libraries (Pandas, SciPy, and Matplotlib), involved calculating descriptive statistics (means and standard deviations), conducting independent two-sample t-tests for comparing means between wins and losses, and determining effect sizes using Cohen's d. Data visualization was done using histograms and Kernel Density Estimation (KDE) plots. RESULTS: A) Descriptives: a) Wins: MTD = 6700, SD = 3809; MSD = 356, SD = 249; MTS = 6.69, SD = 0.97; MTPL = 740, SD = 400 and b) Losses: MTD = 6713, SD = 3807; MSD = 356, SD = 249; MTS = 6.7, SD = 0.97; MTPL = 742, SD = 400. B) T-tests: a) tTD (332) = -0.05, p = .960, d = -0.004; b) tSD (332) = -0.050, p = .960, d = -0.004; c) tTS (332) = -0.117, p = .907, d = -0.009; and d) tTPL (332) = -0.049, p = .961, d = -0.004. CONCLUSIONS: The analysis indicates no statistical or practical significance in any of the KPIs between wins and losses. Coaches along with sport scientists should contextualize these findings, considering alternative KPIs that might more distinctly reflect performance outcomes. Future research should focus on position-specific KPIs, include a broader team sample, and span different divisions, acknowledging limitations like a 12.3% data gap and small sample size.