The Relationship Between Managerial Factors and Performance and Motivation in the NBA: A Data-Driven Analysis Using Artificial Intelligence and Advanced Algorithms

Introduction: Sports management in the digital age is undergoing fundamental transformations, necessitating innovative, evidence-based, and data-driven approaches. The National Basketball Association (NBA), as one of the world's most advanced, lucrative, and complex sports leagues, serves as an ideal case study for examining the interplay of managerial factors on team performance and fan motivation (Vivovinco, ۲۰۲۴). Despite a substantial body of research in sports management, a significant gap exists in studies that integrate extensive quantitative data (Big Data) with strategic qualitative analysis for high-level decision-making. This research aims to bridge this gap and provide an operational framework for sports managers. The primary research questions are: First, what is the quantitative relationship between managerial indicators (such as team ranking, financial strategy, and marketing) and success metrics (attendance, revenue)? Second, with what accuracy can artificial intelligence algorithms predict the complex behavior of fans and sports outcomes? And third, how can data-driven findings be formulated into an integrated and executable strategy for sports clubs? Methodology: This study adopted a mixed-methods approach utilizing three complementary and rich data sources. The first source is a comprehensive historical NBA database for ۲۰۰۱-۲۰۱۵, comprising ۴۵۰ observations from ۳۰ teams. Key variables include home game average attendance (Home Avg_Attendance), final seasonal team rank (Rank), stadium capacity percentage (Home_Percent_Capacity), and away game attendance statistics. The second source is a set of attendance prediction data containing actual values (Actual), values predicted by a preliminary model (Predicted), and prediction error (Error, Absolute_Error) for ۲۱۰ observations across different teams and years. This dataset allows for evaluating prediction model accuracy and analyzing error patterns. The third source is a qualitative strategic document from the Golden State Warriors, outlining an integrated framework consisting of four pillars-sports, finance, marketing, and technology-with specific quantitative goals (e.g., improving rank from ۱۲th to ۷th and increasing revenue by ۲۵%.) Analysis of quantitative data was performed using a suite of advanced artificial intelligence and machine learning algorithms in Python (Horvat & Job, ۲۰۲۰). For modeling the nonlinear and complex relationship between various factors and attendance, Multi-layer Artificial Neural Networks (MLPs) with an ۸-۱۶-۸-۱ architecture (input, two hidden, and output layer neurons, respectively) and ReLU activation function were employed. For classifying teams into "successful" and "needs improvement" groups based on composite indicators, the Support Vector Machine (SVM) algorithm with a Radial Basis Function (RBF) kernel was used. To identify the most influential factors, a Random Forest model comprising ۱۰۰ decision trees was applied(Breiman, ۲۰۰۱). Furthermore, for simulating and optimizing dynamic strategies like pricing, Reinforcement Learning algorithms based on Q-Learning were utilized. All models were evaluated using ۵-Fold Cross-Validation, and metrics such as MAE, RMSE, R² (for regression) and accuracy, precision, recall (for classification) were used to report performance. Findings: Quantitative analyses revealed significant structural relationships. A strong and significant negative correlation of -۰.۷۲ was identified between team rank (where ۱ indicates the best team) and average attendance. This implies that each step of improvement in team rank is associated with an average increase of approximately ۱۱۲ in attendance. The artificial neural network model, trained on ۸۰% of the data, achieved a Mean Absolute Error (MAE) of ۲۴۷ persons and a coefficient of determination (R۲) of ۰.۸۹۳ on the ۲۰% test data, indicating high model accuracy in attendance prediction. Feature importance analysis by the Random Forest model clearly showed the hierarchy of influencing factors: Current season team rank with a ۲۸.۳% share was the most important predictor. This was followed by the team's average rank over the past five years (۱۹.۷%), indicating historical stability and credibility; stadium capacity percentage from the previous season (۱۵.۲%), reflecting core fan enthusiasm; and presence of star players on the team (۱۲.۸%). Other factors like day of the week, ticket price, and opponent's reputation had lesser shares. An in-depth case study of the Golden State Warriors, which transformed from a mediocre team (rank ۲۴ in ۲۰۰۱) to a dominant powerhouse (rank ۷ in ۲۰۱۵) during the research period, demonstrated a successful pattern of strategic integration. Aligning quantitative data with the qualitative strategy document revealed that the staggering ۴۲.۱% growth in average attendance (from ۱۴,۴۶۷ to ۱۹,۵۹۶) and the ۱۷-position improvement in rank were not accidental but the result of systematically implementing four pillars: targeted investment in talent (sports pillar), implementation of dynamic pricing (financial pillar), execution of smart digital campaigns (marketing pillar), and development of an integrated data analytics platform (technology pillar). Additionally, prediction error analysis showed that model accuracy was very high for stable teams (like the Bulls and Lakers, with error below ۱%) and lower for volatile teams (like the Nets). Conclusions: The findings of this research decisively confirm that the sustainable and competitive success of modern sports clubs hinges on the systematic and intelligent integration of four key domains: ۱) Athletic Performance, which is the cornerstone of appeal and requires advanced talent identification and game analysis systems; ۲) Financial Management, which ensures economic sustainability and can be optimized with tools like dynamic pricing; ۳) Smart Marketing, built upon precise fan segmentation and two-way engagement; and ۴) Advanced Technology, which provides the fuel for evidence-based decision-making through integrated data platforms. This study demonstrated that the use of artificial intelligence algorithms-not as a replacement for managerial judgment but as a powerful complement-can significantly enhance prediction accuracy, analytical depth, and ultimately, the quality of managerial decisions. Sports club managers are advised to establish or strengthen specialized data analytics and AI units within their organizational structure, institutionalize a data-driven decision-making culture, and invest in the continuous development of predictive models tailored to the unique characteristics of their team and market. For future research, investigating the impact of socio-cultural factors on predictive models and adapting the proposed framework to other sports leagues (e.g., soccer) is recommended.

کلیدواژه ها:

Sports Management ، Artificial Intelligence ، Big Data Analytics ، NBA ، Integrated Strategy ، Performance Prediction ، Fan Motivation