Purpose: This study aims to address methodological challenges in evaluating biclustering algorithms under simultaneous collinearity and overlap, which often co-occur in real world multivariate data but are rarely analyzed simultaneously. This research highlights the importance of understanding how these structural challenges affect local pattern detection in data mining applications. Methods: A simulation study was conducted using synthetic matrices embedded with two constant biclusters under 15 combinations of collinearity levels (ρ = 0.3,0.6,0.9) and overlap degrees (none, small, large). Each scenario was replicated 100 times. Performance was assessed using the Liu and Wang Index (ILW), while a three-way ANOVA tested the effects of algorithm type, collinearity, and overlap. Result: Spectral Biclustering maintained stable ILW scores despite increasing collinearity, while CC performed better in low-overlap scenarios but was more sensitive to collinearity. Under high collinearity and large overlap, both algorithms experienced notable degradation. The ANOVA confirmed all main effects and interactions were significant (p < 0.001). Novelty: This study contributes empirical evidence regarding the influence of interacting structural characteristics on biclustering performance. The results deliver practical insights for selecting suitable algorithms and emphasize the potential advantages of hybrid approaches that integrate the stability of spectral methods with the adaptability of residual-based techniques.