Δείκτες για την αξιολόγηση της απόδοσης παικτών και ομάδων σε αγώνες μπάσκετ και παράγοντες που τους επηρεάζουν
Indices for evaluating player and team performance in basketball games, and factors affecting these indices
View/
Keywords
Μπάσκετ ; Στατιστική ανάλυση ; Δείκτης αξιολόγησης ; Λογιστική παλινδρόμηση ; Μηχανική μάθησηAbstract
In a data-driven era, the field of data analytics has emerged as a powerful tool in various industries, including sports. With the exponential growth of technology and the availability of vast amounts of data, teams and organizations are now able to extract valuable insights and make informed decisions to gain a competitive advantage over their opponents. Data analytics in sports involves the collection, interpretation and visualization of complex data sets, allowing teams to optimize their performance, enhance player development strategies, improve game strategies and ultimately achieve greater success on the field. From player statistics to game performance metrics, data analytics has revolutionized the way sports are played and analyzed, opening up new possibilities for teams and athletes to unlock their full potential. In this master thesis, using real data from the EuroLeague, which is considered the top club basketball competition in Europe, we use statistical techniques in order to analyze, by examining all the seasons since its creation, what are the important, predictive factors in the qualification of teams to the Playoffs and the Final Four, as well as what is the contribution of the five best players based on PIR to the progress of their teams. Additionally, we present a descriptive analysis of our variables and illustrate the results through graphs, charts, and tables. Then, we apply logistic regression models to find the key variables that affect a team's qualification (or not) in the two phases we examine. Finally, through appropriate machine learning techniques, we will investigate whether there are already existing clusters in our data with similar characteristics (clustering), and we try to build efficient classification models, to examine the qualification of teams in the individual phases of the competition.
Postgraduate Studies Programme
Εφαρμοσμένη ΣτατιστικήDepartment
Σχολή Χρηματοοικονομικής και Στατιστικής. Τμήμα Στατιστικής και Ασφαλιστικής ΕπιστήμηςNumber of pages
219Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα