Journal of Science &Technology Metrics R. Sebastiyan, Pit Pichappan 7 2 2026 https://doi.org/10.6025/jstm/2026/7/2/88-108 https://www.dline.info/jstm/fulltext/v7n2/jstmv7n2_3.pdf This longitudinal study examines the dynamics of global university rankings using data from the Academic Ranking of World Universities (ARWU) for the top 25 institutions from 2021 to 2025. Employing a multifaceted quantitative approach including descriptive statistics, rank volatility analysis, correlation and multiple regression modeling, Principal Component Analysis (PCA), K-means clustering, and trajectory tracking this research investigates temporal changes in ranking indicators, institutional stability, the relative influence of performance metrics, and latent strategic profiles among elite universities. Results reveal substantial heterogeneity in indicator dynamics, with the Award (Nobel/Fields Medal staff) and Nature & Science (N&S) scores exhibiting the largest absolute changes. Regression analysis demonstrated that prestige related indicators, particularly Award ( = 0.402), exert the strongest influence on Total Score, collectively explaining over 92% of the variance. PCA identified two dominant dimensions Academic Prestige and Research Excellence (PC1, 71.8% variance) and Publication Productivity and Efficiency (PC2) highlighting the primacy of impact and reputation over publication volume. K-means clustering (validated via elbow, silhouette, and Davies Bouldin metrics) delineated three strategic archetypes: Elite Prestige Leaders (e.g., Harvard, Stanford), Research Driven Institutions (e.g., MIT, UC Berkeley), and Balanced High Performers.Elite institutions displayed remarkable positional stability, while competitive pressures drove volatility in other tiers. These findings underscore how scientific prestige, high impact research, and cumulative advantage sustain stratification in global higher education. The study offers data driven insights for institutional strategy and policy to enhance international competitiveness.