In learning management systems, like Moodle, log files containing data on students' learning paths offer insights into how students interact with their tutors, peers, and the provided educational material. This study focused on 152 postgraduate students using Moodle in an Open University, engaging with a specific module in the 2021-2022 academic year. Analysis of their log files aimed to correlate learning path states with final grades, providing insights for optimizing the learning process, tutor guidance, and course structure. Markov Chain Models were used to analyze students' navigation and transitions between seven states in Moodle. Results indicated a predominant engagement with the Course and Forum states, suggesting the university platform emphasizes content delivery and collaborative learning. Measures such as Entropy, Frobenius Norm, Euclidean Distance, and Cosine Similarity were used to assess data patterns and relationships. Variations in probabilities among states made this study challenging to differentiate students' academic performance in the specific module, based on eigenvectors and the above metrics alone. The diverse range of features, activities, and interactions within Moodle results in data patterns that may not always have straightforward interpretations.