Determinants of coach dismissals in the Italian soccer league Serie A: a competing risks approach
In this talk, we investigate the determinants of head coach turnover in Italian Serie A over the period 2010-2019, accounting for multiple exit mechanisms within a competing risks framework. We consider four mutually exclusive exit types: end of contract, mutual consent, resignation, and sacking. To identify robust predictors of turnover, we use classical cause-specific Cox model and a variable selection procedure based on forward selection, LASSO regression, and RSFs. Our findings show that performance-related variables are the primary drivers of coach turnover, while personal characteristics play a limited role. Short-term performance indicators, such as points per match, strongly influence dismissal risk, and cumulative poor performance substantially increases this risk.
By considering four exit types and using a conservative variable selection strategy, our study provides a robust assessment of managerial turnover dynamics in the most important Italian professional soccer league.
Palabras clave: Random survival forest LASSO Competing risks model Sports analytics