<![CDATA[Data analysis has fundamentally altered contemporary football performance assessment, with Expected Goals (xG) emerging as a key metric. Despite its widespread use, comprehensive documentation of its modeling methodologies remains scarce. This study aims to map the global xG research landscape using bibliometric analysis. Scopus data was analyzed using VOSviewer to identify trends, prevalent methodologies, and implementation domains. Findings indicate a significant rise in publications from 2020 to 2025, signaling growing scholarly interest. Random Forest is identified as the most widely used technique, though recent trends suggest a resurgence of Logistic Regression. The primary application domain is Performance Analysis, followed by Tactical & Strategic Analysis. Keyword analysis reveals three main clusters: machine learning, regression models, and deep learning. It is concluded that current xG research trends are moving towards a balance between algorithmic complexity and model interpretability, while expanding beyond mere shot evaluation to more holistic performance metrics.]]>
Copyrights © 2026
