Research and Statistical Modeling Projects

We're researching ways to use social network analysis (SNA) to model and analyze passing networks to uncover team dynamics and player contributions. This approach enhances our understanding of in-game strategies and player interactions. We see this research as a way to understand and improve team performance and cohesion. Our goal is to elevate strategic analysis and decision-making in lacrosse, contributing to deeper strategic and tactical decisions.

Ranking systems play a critical role in the seeding and selection processes for seeded tournaments (like the NCAA Championship), informing team selections for postseason play. Our research explores cutting-edge ranking methodologies that synthesize a variety of factors, offering a comprehensive assessment of a team's quality. We are also actively researching ensemble strategies that blend advanced and traditional ranking systems. This approach is designed to amplify the strengths and mitigate the limitations of individual models, delivering superior predictions and indices tailored for high-stakes decisions in collegiate athletics. Our commitment lies in enhancing the precision and reliability of rankings, setting a new standard for competitive fairness and integrity in NCAA sports.

Our research is looking into the development profiles for teams, units, and individual players through latent class models (LCMs). These probabilistic models are designed to reveal underlying "patterns of play" by analyzing a blend of traditional and advanced lacrosse metrics. The utility of these models lies in their ability to segment complex data into distinct categories, offering insights into strategic behaviors and performance tendencies not immediately visible through standard analysis. This approach not only enriches player and team evaluations but also informs coaching strategies, player development, and game preparation. By leveraging LCMs, we're unlocking a deeper, data-driven understanding of lacrosse dynamics, setting a new standard for analytical precision in sports analytics.

We are using the latest advances in survey design to develop instruments for better understanding affective factors associated with team and individual performance. By integrating psychological insights with analytics, we're uncovering the less visible, yet potent, emotional elements that flow through the heart of athletic performance. This research not only enhances our grasp of the affective dimensions within lacrosse but also informs tailored interventions to drive success.

Our current work in performance factors leverages latent variable modeling, particularly factor analysis, to identify hidden factors that are pivotal to team success and failure in lacrosse. This methodology allows us to isolate and understand the latent structures that govern game outcomes, providing a strategic edge in evaluating and enhancing team performance. With this data-driven approach, we aim to redefine the benchmarks of success in lacrosse, translating statistical patterns into actionable strategies. 

This research aims to explore and apply the principles of decision theory to the dynamic and fast-paced sport of lacrosse, with the goal of enhancing understanding and execution in both individual and team contexts. By integrating decision-theoretic models with empirical data from lacrosse games, the study seeks to uncover optimal strategies for key decision points within the sport, such as shot selection, offensive/defensive formations, and player positioning. The research employs a combination of approaches to evaluate decisions under different game states.