Identifying performance patterns in professional mixed martial arts: An exploratory data approach

Background: Mixed martial arts (MMA) performance is shaped by a complex interaction of physical, technical, and tactical variables. With the sport’s rapid professionalization, data-driven approaches have become increasingly important in evaluating fighter performance and informing strategy. Research Objectives: This study aims to explore performance patterns among Ultimate Fighting Championship™ (UFC) fighters using Exploratory Data Analysis (EDA), uncovering trends and relationships among key performance attributes. Methods: A dataset of 4.111 UFC fighters was analyzed across 18 variables, including physical characteristics (e.g., height, weight, reach) and performance metrics (e.g., striking accuracy, takedown success, submission attempts). EDA techniques, including visualizations, descriptive statistics, and correlation analysis, were applied to detect patterns and summarize key characteristics. Finding/Results: The analysis revealed that takedown accuracy and takedown defense are moderately related, suggesting a possible interdependence of offensive and defensive grappling skills. However, most correlations between physical traits and outcomes, such as between height and win totals, were weak, indicating that physical attributes alone are insufficient to predict success. Observations on stance effectiveness and high-performing athletes provided illustrative, but non-generalizable, insights. Conclusion: This study demonstrates the utility of EDA as a foundational tool in MMA performance analysis. While limited in inferential scope, the findings offer preliminary guidance for coaches and sports analysts seeking to design targeted training interventions. Future research should integrate psychological, contextual, and opponent-based data to enhance the accuracy and depth of predictive modeling in combat sports.